Managerial Economics and Organizational Architecture - Brickley 5th Ed - Online - No Password
391 Pages

Managerial Economics and Organizational Architecture - Brickley 5th Ed - Online - No Password

Course Number: FINA 6387 6387, Spring 2012

College/University: U. Houston

Word Count: 210617

Rating:

Document Preview

Economics Course: ECO 305 Managerial Economics Instructor: Dr. Bruce Hartman Cal Maritime Business Administration =>? McGraw-Hill/Irwin McGrawHill Primis ISBN10: 0390670014 ISBN13: 9780390670014 Text: Managerial Economics and Organizational Architecture, Fifth Edition BrickleySmithZimmerman This book was printed on recycled paper. Economics http://www.primisonline.com Copyright 2008 by The McGrawHill...

Unformatted Document Excerpt
Coursehero >> Texas >> U. Houston >> FINA 6387 6387

Course Hero has millions of student submitted documents similar to the one
below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.

Course Hero has millions of student submitted documents similar to the one below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.

305 Economics Course: ECO Managerial Economics Instructor: Dr. Bruce Hartman Cal Maritime Business Administration =>? McGraw-Hill/Irwin McGrawHill Primis ISBN10: 0390670014 ISBN13: 9780390670014 Text: Managerial Economics and Organizational Architecture, Fifth Edition BrickleySmithZimmerman This book was printed on recycled paper. Economics http://www.primisonline.com Copyright 2008 by The McGrawHill Companies, Inc. All rights reserved. Printed in the United States of America. Except as permitted under the United States Copyright Act of 1976, no part of this publication may be reproduced or distributed in any form or by any means, or stored in a database or retrieval system, without prior written permission of the publisher. This McGrawHill Primis text may include materials submitted to McGrawHill for publication by the instructor of this course. The instructor is solely responsible for the editorial content of such materials. 111 ECONGEN ISBN10: 0390670014 ISBN13: 9780390670014 Economics Contents BrickleySmithZimmerman Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 2 2. Economists View of Behavior 3. Markets, Organizations, and the Role of Knowledge 2 46 II. Managerial Economics 90 4. Demand 5. Production and Cost 6. Market Structure 7. Pricing with Market Power 8. Economics of Strategy: Creating and Capturing Value 9. Economics of Strategy: Game Theory 10. Incentive Conflicts and Contracts 90 1 26 1 62 1 90 2 24 2 65 2 97 III. Designing Organizational Architecture 324 17. Divisional Performance Evaluation 3 24 IV. Applications of Organizational Architecture 358 21. Understanding the Business Environment: The Economics of Regulation 3 58 iii Notes 1 2 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Economists View of Behavior CHAPTER 2 CHAPTER OUTLINE Economic Behavior: An Overview Economic Choice Marginal Analysis Opportunity Costs Creativity of Individuals Graphical Tools Individual Objectives Indifference Curves Opportunities and Constraints Individual Choice Changes in Choice Motivating Honesty at Merrill Lynch Managerial Implications Alternative Models of Behavior Only-Money-Matters Model Happy-Is-Productive Model Good-Citizen Model Product-of-the-Environment Model Which Model Should Managers Use? Decision Making under Uncertainty Risk Aversion Certainty Equivalent and Risk Premium Risk Aversion and Compensation Summary Appendix: Consumer Choice I n May 2002, Merrill Lynch agreed to pay $100 million to settle charges that its analysts had recommended stocks to clients that they privately thought were poor investments. Internal e-mails provided strong support for this claim leveled by the New York State attorney general. For example, InfoSpace, an Internet services company, was rated highly by analysts, yet privately the analysts suggested that it was a powder keg and a piece of junk. Although InfoSpaces share price dropped from $261 to $14, Merrill analysts never recommended selling the stock. Merrill analysts BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 3 17 rated Excite@Home accumulate or buy, while privately the investment team called it a piece of crap. This episode at Merrill sent shock waves through other major investment houses indeed through the entire investment community. Other investment rms publicly stated that they were taking strong steps to make sure that the situation at Merrill would not be repeated within their organizations. Fortune magazine ran a cover story entitled, In Search of the Last Honest Analyst.1 The scandal generated signicant concerns throughout the world among both the general public and government regulators. For example, the New York attorney general began a sweeping investigation of analysts at Salomon Smith Barney and other investment rms that had recommended WorldCom to investors. In July 2002, WorldCom became the biggest company ever to le for bankruptcy in U.S. history. In December 2002, the nations 10 top investment banks agreed to a $1.2 billion settlement with regulators aimed at protecting investors from brokerages conicts of interest. Managers at Merrill, Salomon Smith Barney, and other investment companies had to act quickly to address this potential problem. As a rst step, management had to understand what motivated the Merrill analysts to mislead their investment clients. Only then could they choose a policy to redress the situation. If management thought this problem was caused by a few dishonest employees, the appropriate response would have been to try to identify and re those employees. If, instead, management believed the problem was caused by disgruntled employees taking out their frustrations on customers, a potential response would have been to adopt a job-enrichment program to increase employee satisfaction and, it would be hoped, analyst honesty. Alternatively, Merrill Lynch might have created incentives through its compensation plan that caused its analysts to issue misleading investment reports. If so, the appropriate response would be to restructure its compensation plan. Many other assumptions and responses are possible. The example of Merrill Lynch illustrates a general point: Managers responses to problems are likely to depend on their understanding of peoples motives and their forecast of peoples reactionstheir responses thus depend on their underlying model of behavior. Most managerial actions attempt to change the behavior of individuals, such as employees, customers, union ofcials, or subcontractors. Managers with different understandings (or models) of what motivates behavior are likely to make different decisions and take different actions. We begin this chapter by briey summarizing the general framework economists use to examine individual behavior. Selected graphical tools are introduced to aid our analysis. Next, we use this economic framework to analyze the problem at Merrill Lynch. The managerial implications of this analysis are discussed. We contrast this economic view of behavior with alternative views and explore why the economic framework is particularly useful in managerial decision making. Finally, we analyze decision making under uncertainty. Economic Behavior: An Overview Individuals have unlimited wants. People generally want greater wealth, more attentive service, larger houses, more luxurious cars, and additional personal material items. They want more time for leisure activities. Most also want to improve the plight of othersstarving children, the homeless, and disaster victims. People are concerned about vitality, religion, integrity, and gaining the respect and affection of others. 1 June 10, 2002, issue. 4 18 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts In contrast to wants, resources are limited. Households face limited incomes that preclude all the purchases and expenditures that household members might like to make. The available amount of land, trees, and other natural resources is nite. There are only 24 hours in the day. People become ill; death is inevitable. Economic Choice Economic analysis is based on the notion that individuals assign priorities to their wants and choose their most preferred options from among the available alternatives. If Kathy Measer is confronted with a choice between a laptop or a desktop computer, she can tell you whether she prefers one over the other or whether she is indifferent between the two. Depending on the relative prices of the two products, she purchases her preferred alternative. If Kathy has a weekly budget of $1,000, she considers the many ways she might spend the money and then chooses the package of goods and services that will maximize her personal happiness. She cannot make all desired purchases on her limited budget. However, this choice is optimal for Kathy, given her limited resources. Economists do not assert that people are selsh in the sense that they care only about their own personal consumption. Within the economic paradigm, people also care about such things as charity, family, religion, and society. For instance, Kathy will donate $100 to her church, as long as the donation provides greater satisfaction than alternative uses of the money. Neither do economists contend that individuals are supercomputers that make infallible decisions. Individuals are not endowed with perfect knowledge and foresight, nor is additional information costless to acquire and process.2 For example, Kathy might order an item from a restaurant menu only to nd that she dislikes what she is served. Within this economic paradigm, she simply does the best she can in the face of her imperfect knowledge. But she learns from her experience and does not repeat the same mistakes in judgment time after time.3 Marginal Analysis Marginal costs and benets are the incremental costs and benets that are associated with making a decision.4 It is the marginal costs and benets that are important in economic decision making. An action should be taken whenever the marginal benets of that action exceed its marginal costs. Mary ODwyer has a contract to help sell products for an ofce 2 Economists sometimes use the idea of bounded rationality. Under this concept, individuals act in a purposeful and intendedly rational manner. However, they have cognitive limitations in storing, processing, and communicating information. It is these limitations which make the question of how to organize economic activity particularly interesting. H. Simon (1957), Models of Man ( John Wiley & Sons: New York). 3 At least this learning appears to occur outside the comics. For decades, Charlie Brown from Peanuts continued to try to kick the football held by Lucy van Pelt. Yet Lucy always pulled the ball at the last second. Few individuals are as incurably optimistic as Charlie Brownthey learn. 4 Technical note: Marginal costs and benets are typically dened as changes in costs and benets associated with very small changes in a decision variable. For instance, the marginal costs of production are the additional costs from producing a small additional amount of the product (for instance, one more unit). Often decisions involve discrete choices, such as whether or not to build a new plant. In these cases, it is not possible to dene a small change in the decision variable. Incremental costs and benets are those costs and benets which vary with such a decision. For our present discussion, the technical distinction between marginal and incremental is not important. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 5 19 MANAGERIAL APPLICATIONS Marginal Analysis of Customer Protability First Union Corp. (since acquired by Wachovia) used a computer program it called Einstein to rank customers based on their protability. Protable customers keep several thousand dollars in their accounts, use tellers less than once a month, and rarely make calls to the banks customer call center. Unprotable customers make frequent branch visits, keep less than a thousand dollars in the bank, and call frequently. When a customer requests a lower credit card interest rate or a waiver of the banks $28 bounced-check fee, the operator pulls up the customers account. The computer system displays the customers name with a red, yellow, or green box next to it. A green box signals the call operator to keep this protable customer happy by granting the request (within the limits of their authority). Customers with red boxes rarely get what they request, in hopes they will go to another bank. This system is an example of how First Union used marginal analysis to decide the level of service supplied to individual customersEinstein helped the operator identify the marginal costs and benets of satisfying the banks customer demands. Source: R. Brooks (1999), Unequal Treatment, The Wall Street Journal ( January 7), A1. supply company. She is paid $50 for every sales call that she makes to customers. Thus, Marys marginal benet for making each additional sales call is $50. Mary enjoys playing tennis more than selling. If she places a marginal value of more than $50 on the tennis that she would forgo by making an extra call, she should not make any more sales calls that daythe marginal costs would have exceeded the marginal benets. She continues to make additional sales calls as long as the reduction in tennis playing is valued at less than $50.5 Marginal analysis is a cornerstone of modern economic analysis. In economic decision making, bygones are forever bygones. Costs and benets that have already been incurred are sunk (assuming they are nonrecoverable) and hence are irrelevant to the current economic decision. Mary paid $5,000 to join a tennis club last month. This fee does not affect her current decision of whether to play tennis or make an extra sales call. That expenditure is ancient history and does not affect Marys current trade-offs. As another example, consider Ludger Hellweg who owns a company that installs wood oors. He is offered $20,000 to install a new oor. The cost of his labor and other operating expenses (excluding the wood) are $15,000. He has wood for the job in inventory. It originally cost him $2,000. Price increases have raised the market value of the wood to $6,000, and this value is not expected to change in the near future. Should he accept the contract? He should compare the marginal costs and benets from the project. The marginal benet is $20,000. The marginal cost is $21,000$15,000 for the labor and operating expenses and $6,000 for the wood. The historic cost for the wood of $2,000 is not relevant to the decision. To replace the wood used on this job costs $6,000. Since the marginal costs exceed the marginal benets, Ludger would be better off rejecting the contract than accepting it. This example illustrates that in calculating marginal costs, it is important to use the opportunity cost of the incremental resources, not their historic (accounting) cost. 5 To keep this example simple, we abstract from several issues. We ignore any pleasure Mary receives from the process of selling. Also, selling effort today is likely to have some effect on her future professional progress. Finally, if Mary values a tennis game at 9 A.M. and one at 7 P.M. equally, she will sell during the business day and postpone tennis to the evening. 6 20 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Opportunity Costs and V-8 The Campbell Soup Company used the idea of an opportunity cost to create a successful ad campaign for its V-8 vegetable juice. Upon nishing a soft drink, the fellow in the ad would look into the camera, slap his forehead, and exclaim: Wow I coulda had a V-8. Since one is unlikely to drink both a soft drink and a V-8, the opportunity cost of the soft drink is the forgone V-8a cost that these commercials sought to convince the viewing audience is quite high. Opportunity Costs Because resources are constrained, individuals face trade-offs. Using limited resources for one purpose precludes their use for something else. For example, if Larry Matteson takes four hours to play golf, he cannot use that same four hours to paint his house. The opportunity cost of using a resource for a given purpose is its value in its best alternative use. The opportunity cost of using four hours to play golf is the value of using the four hours in Larrys next best alternative use. Marginal analysis frequently involves a careful consideration of the relevant opportunity costs. If Larry starts a new pizza parlor and hires a manager at $30,000 per year, the $30,000 is an explicit cost (a direct dollar expenditure). Is he better off managing the restaurant himself, since he can avoid the explicit cost of $30,000 by not paying himself a salary? The answer to this question depends (at least in part) on the opportunity cost of his time. If he can earn exactly $30,000 in his best alternative job, the implicit cost of self-management is the same as the explicit cost of hiring an outside manager: He forgoes $30,000 worth of income if he manages the parlor himself. Both explicit and implicit costs are opportunity costs that should be considered in the analysis. Suppose that Larrys gross prot from the pizza parlor, before paying the manager a salary, is $35,000 and that he can earn $40,000 in an outside job. Hiring a manager for $30,000 yields a net prot of $5,000 from the pizza parlor. He also earns $40,000 from the outside job, for total earnings of $45,000. If he manages the pizza parlor himself, he earns only $35,000. In this example, it is better for him to work at the outside job and hire a manager to run the restaurant.6 Creativity of Individuals7 Within this economic framework, individuals maximize their personal satisfaction given resource constraints. Indeed, people are quite creative and resourceful in minimizing the effects of constraints. For instance, when the government adopts new taxes, almost immediately accountants and financial planners begin developing clever ways to reduce their impact. Some self-employed individuals were able to reduce the impact of recent tax increases by changing their status from a proprietorship to a corporation. 6 Again, to keep the example simple, we assume there is no difference in personal satisfaction between Larrys outside job and managing the pizza parlor. We also postpone the discussion of consequences for the success of the pizza parlor from hiring a manager versus self-management until Chapter 10. 7 This section draws on W. Meckling (1976), Values and the Choice of the Model of the Individual in the Social Sciences, Schweizerische Zeitschrift fr Volkswirtschaft und Statistik 112, 545560. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 7 21 ANALYZING MANAGERIAL DECISIONS: Marginal Analysis You own a business that services trucks. A customer would like to rent a truck from you for one week, while you service his truck. You must decide whether or not to do this. You have an extra truck that you will not use for any other purpose during this week. This truck is leased for a full year from another company for $300/week plus $.50 for every mile driven. You also have paid an annual insurance premium, which costs $50/week to insure the truck. The truck has a full 100-gallon fuel tank. The customer has offered you $600 to rent the truck for a week. This price includes the 100 gallons of fuel that is in the tank. It also includes up to 500 miles of driving. The customer will pay $.50 for each additional mile that he drives above the 500 miles. You anticipate that the customer will bring back the truck with an empty fuel tank and will have driven more than 500 miles. You sell fuel to truckers at a retail price of $4.00/gallon. Any fuel you sell or use can be replaced at a wholesale price of $3.25/gallon. The customer will rent a truck from another company if you do not accept the proposed deal. In either case, you will service his truck. You know the customer and are condent that he will pay all charges incurred under the agreement. 1. Should you accept or reject the proposed deal? 2. Would your answer change if your fuel supplier limited the amount of fuel that you could purchase from him at the wholesale price? Explain. As another example, a 33-year-old Brazilian farm hand recently retired with full social security benets after he satised social security auditors that he had been working since he was three years old. Because Brazil doesnt specify a minimum retirement age, the average Brazilian retires at age 49.8 Similarly, when hackers and corporate spies continue to develop more sophisticated schemes to steal information from Web sites or networks, software tools that detect break-ins also have grown in popularity and sophistication. This intrusion-detection software was about a $100 million industry in 1999 and is now a billion dollar industry.9 Understanding this creative nature of individuals has important managerial implications that we discuss later in this chapter, as well as throughout the book. MANAGERIAL APPLICATIONS Creative Gaming of the System An MIS manager bought computers for his company one at a time, charging them on his personal credit card. He then led for reimbursement on his expense account. Although this process imposed delay costs on the rmit required almost a year for the company to acquire 20 computersthe manager received frequent yer miles given by his credit card company. Source: S. Adams (1996), The Dilbert Principle (Harper Business: New York), 326. 8 P. Fritsch (1999), In Brazil Retirement Has Become a Benet Nearly All Can Enjoy, The Wall Street Journal (September 9), A1. 9 J. DAllegro (1999), Intrusion Detection Matures, National Underwriter (March 8), 9. 8 22 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts Graphical Tools Economists often employ a set of graphical tools to illustrate how individuals make choices. These tools distinguish between the preferences (level of satisfaction) that the individual associates with each potential opportunity and the set of feasible opportunities that an individual faces. We use these tools throughout this book. They also are used in other courses within the typical business school curriculum, such as in nance, human relations, and marketing courses. Our intent is to introduce these tools so that the reader is comfortable using them in basic business applications. We subsequently apply the tools to analyze the problems at Merrill Lynch. The appendix to this chapter provides a more detailed development of the economic theory of individual choice (commonly called the Theory of Consumer Choice). Individual Objectives Goods are things that people value. Goods include standard products like food and clothing, services like haircuts and education, as well as less tangible emotions such as love of family and charity. The economic model of behavior posits that people acquire goods that maximize their personal satisfaction, given their resource constraints (such as a limited income). Economists traditionally use the term utility in referring to personal satisfaction. To provide a more detailed analysis of how people make choices, economists represent an individuals preferences by a utility function. This function expresses the relation between total utility and the level of goods consumed. The individuals objective is to maximize this function, given the resource constraints.10 This concept can be illustrated most conveniently through a simple example where an individual cares about only two goods. The insights from this two-good analysis can be extended readily to the case of additional goods such as food, housing, clothing, respect, and charity. Suppose that Dominique Lalisse values only food and clothing. In general form, his utility function can be written as follows: Utility F (Food, Clothing) (2.1) Dom prefers more of each goodthus, his utility rises with both food and clothing. In Doms case, his specic utility function is Utility Food1 2 Clothing1 2 (2.2) For instance, if Dom has 16 units of food and 25 units of clothing, his total utility is 20 (that is, utility 161 2 25 1 2 4 5 20). Dom is better off with 25 units of both food and clothing. Here, his utility is 25 (utility 251 2 251 2 5 5 25). Utility functions rank alternative bundles of food and clothing in the order of most preferred to least preferred, but they do not indicate how much one bundle is preferred to another. If the utility index is 100 for one combination of food and clothing and 200 10 Clearly, most individuals do not actually consider maximizing a mathematical function when they make these choices. However, this formulation can provide useful insights into actual behavior to the extent that it approximates how individuals make choices. Mathematicians have shown that if an individuals behavior is consistent with some basic axioms of choice (comparability, transitivity, nonsatiation, and willingness to substitute), the individual will make choices as if he or she were trying to maximize a mathematical function. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 23 Indifference Curves These indifference curves picture all combinations of food and clothing that yield the same amount of utility. The specic utility function in this example is U F 1 2 C1 2, where F is food and C is clothing. Northeast movements are utility-increasing. Indifference curve 2 represents all combinations of food and clothing that yield 20 units of utility, whereas curve 1 pictures all combinations that yield 8 units of utility. Other indifference curves could be drawn for different levels of utility. Increasing utility F 25 Quantity of food Figure 2.1 Economists View of Behavior 9 16 2: U = 20 4 1: U = 8 4 16 25 C Quantity of clothing for another, Dom will prefer the second combination. The second bundle does not necessarily make him twice as well off as the rst bundle.11 Neither does this formulation allow one persons utility of a bundle to be compared to another persons utility. Indifference Curves Preferences implied by the utility function can be illustrated graphically through indifference curves. An indifference curve pictures all combinations of goods that yield the same utility. Given his utility function in Equation (2.2), Dom is indifferent between either 16 units of food and 25 units of clothing or 25 units of food and 16 units of clothing. Both combinations yield 20 units of utility, and hence are on the same indifference curve. Figure 2.1 shows two of Doms indifference curves. For example, if given a choice between any two points on curve 1, Dom would say that he does not care which one is selectedin either case, he obtains 8 units of utility. The slope at any point along one of Doms indifference curves indicates how much food he would be willing to give up for a small increase in clothing (his utility remains unchanged by this exchange).12 Standard indifference curves that illustrate trade-offs 11 This is like rankings on a testan individual who scores in the 80th percentile is not twice as smart as one from the 40th. 12 Recall that the slope of a line is a measure of steepness, dened as the increase or decrease in height per unit of distance along the horizontal axis. Slopes of curves are found geometrically by drawing a line tangent to the curve at the point of interest and determining the slope of this tangent line. The slope at a point along one of Doms indifference curves indicates how the quantity of food changes for small changes in the amount of clothing in order to hold utility constant. Since by denition Dom is indifferent to this exchange (he remains on the same indifference curve), he is willing to make the exchange. 10 24 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts between two goods have negative slopes. If Dom obtains a smaller amount of one good such as clothing, the only way he can be equally as well off is to obtain more of another good like food. If at a point along an indifference curve the slope is 2, Dom is willing to give up 2 units of food to obtain 1 unit of clothing. Alternatively he is willing to give up 1 2 unit of clothing to obtain 1 unit of food. This willingness to substitute has important implications, which we discuss below. North and east movements in graphs like Figure 2.1 are utility-increasing. Holding the amount of food constant, utility increases by increasing clothing (an eastward movement). Holding the amount of clothing constant, utility increases by increasing the amount of food (a northward movement). Thus, in Figure 2.1, Dom would rather be on indifference curve 2 than on 1. He obtains 20 units of utility rather than 8. Economists typically picture indifference curves as convex to the origin (they bow in, as in Figure 2.1). Convexity implies that if Dom has a relatively large amount of food, he would willingly exchange a relatively large quantity of food for a small amount of additional clothing. Thus, the indifference curves in Figure 2.1 are steep when the level of food is high relative to the level of clothing. In contrast, if he has a relatively large amount of clothing, he would be willing to substitute only a small amount of food for additional clothing. Correspondingly, the indifference curves in Figure 2.1 atten as Dom has less food and more clothing. The behavior implied by the convexity of indifference curves is consistent with the observed behavior of many individualsmost people purchase balanced combinations of food and clothing. Opportunities and Constraints Dom would like more of both food and clothing. Unfortunately, he faces a budget constraint that limits his purchases. Suppose that he has an income of I and the prices per unit of food and clothing are Pf and Pc , respectively. Since he cannot spend more than I, his consumption opportunities are limited by the following constraint: I Pf F PcC (2.3) where F and C represent the units of food and clothing purchased. This budget constraint indicates that only combinations of food and clothing that cost no more than I are feasible. Rearranging terms, this constraint can be written as F I Pf (Pc Pf )C (2.4) Figure 2.2 depicts these consumption opportunitiesfrequently called a budget line. All combinations of food and clothing on or below the line are attainable. Combinations above the line are infeasible given an income of I. The F intercept (on the vertical axis) of the line I Pf indicates how much food Dom can purchase if his entire income is spent on food (no clothing is purchased). The C intercept is correspondingly I Pc . The slope of the line Pc Pf is 1 times the ratio of the two prices. The ratio Pc Pf is the relative price of clothing in terms of food. It represents how many units of food he must forgo to acquire a unit of clothing: It is the opportunity cost of clothing. For example, if the price of clothing is $8 and the price of food is $2, the relative price of clothing is 4. To keep total expenditures constant, 4 units of food must be given up for every unit of clothing purchased. The relative price of food is Pf Pc (in this example, 0.25); 1 4 unit of clothing must be given up for each unit of food purchased. The constraint changes with changes in Doms income and the relative prices of the two goods. As shown in Figure 2.3, changes in income result in parallel shifts of the constraint: Its slope is unaffected. An increase in income shifts the constraint outward BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 2 Figure 2.2 Opportunities and Constraints Economists View of Behavior 25 F Combinations above line are unaffordable I Pf Quantity of food The constraint reects the feasible combinations of food and clothing that are attainable given the persons income (I ). The vertical and horizontal intercepts, respectively, show the amounts of food and clothing that can be purchased if no income is spent on the other good. The slope of the constraint is equal to 1 times the ratio of the prices of the two goods. For instance, if the price of clothing is $8 and the price of food is $2, the slope will be 4. This slope implies that 4 units of food must be given up for 1 unit of clothing. If both goods have the same price, the slope will be 1. 11 The McGrawHill Companies, 2009 2. Economists View of Behavior Combinations on/below line are affordable Pc Pf C I Pc Quantity of clothing (up and to the right), while a decrease in income shifts the constraint inward. The slope of the constraint changes with the relative prices of the two goods. As shown in Figure 2.4 (page 26), if the price of clothing increases relative to the price of food, the constraint becomes steeper. If the price of clothing falls relative to the price of food, the constraint becomes atter. Figure 2.3 Income Changes F This gure shows that there is a parallel shift in the budget constraint when income changes. The slope of the constraint does not change because there is no change in the prices of the two goods. The slope is 1 times the ratio of the prices. Quantity of food I Pf Higher income Original consumption opportunities Lower income I LO Pc Quantity of clothing I Pc I HI Pc C 26 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 Figure 2.4 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts Price Changes This gure shows how his consumption opportunities change with changes in the price of clothing. The slope of the line is (Pc Pf). Thus, an increase in the price of HI clothing (from Pc to Pc ) produces a steeper line, while a decrease (from LO Pc to Pc ) produces a atter line. Changes in the price of food also affect the slope of the line. F I Pf Increase in the price of clothing Quantity of food 12 Original consumption opportunities Decrease in the price of clothing I PcHI I Pc I C PcLO Quantity of clothing Individual Choice Within this economic framework, Doms goal is to maximize utility given his opportunities. Utility is maximized at the point of tangency between the constraint and an indifference curve.13 Figure 2.5 portrays the optimal choice. Dom could choose points like b and c on indifference curve 1. However, point a on curve 2 yields greater satisfaction and thus is preferred. Dom would prefer to be at any point on curve 3. Yet, these points are unattainable given his income. This graphical solution to Doms choice problem has a simple intuitive interpretation. At the point of tangency, the indifference curve and the constraint have equal slopes. Recall that the slope of the indifference curve represents Doms willingness to trade food for clothing, whereas the slope of the constraint represents the terms of trade available in the marketplace. At the optimal choice, the willingness and ability to trade are equal. At other feasible combinations of food and clothing, Doms utility could be increased by making substitutions. For instance, if Dom were at a point where he was willing to trade 5 units of food for 1 unit of clothing and if the relative price of clothing were 4 (the slope of the indifference curve is steeper than the constraint), Dom would be better off purchasing less food and more clothing. (He is willing to trade 5 units of food for one unit of clothing, but only must forgo 4 units of food to obtain 1 unit of clothing in the marketplace.) Alternatively, if Dom were at a point where he was only willing to forgo 1 unit of food for 1 unit of clothing (the slope of the indifference curve is atter than the constraint), he would be better off purchasing more food and less clothingsince he receives 5 units of food for each unit of clothing forgone. 13 For simplicity, we ignore the possibility of corner solutionsthe points where the budget constraint intersects the axes. With corner solutions, the individual spends all income on only one good. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 2 Optimal Choice The individual is best off by choosing point a where the constraint is tangent to indifference curve 2. This optimal combination of food and clothing, F* and C*, yields higher utility than other feasible alternatives (for example, points b and c). The individual would prefer points on indifference curve 3, but these points are infeasible given his consumption opportunities. Economists View of Behavior 27 F b Quantity of food Figure 2.5 13 The McGrawHill Companies, 2009 2. Economists View of Behavior F* 3 a 2 c 1 C C* Quantity of clothing Earlier in this chapter, we discussed how marginal analysis is the cornerstone of modern economics. It is important to understand that the graphical tools presented in this section depict marginal analysis. In marginal analysis, individuals take actions as long as their incremental benets are greater than their incremental costs. Our graphical analysis of individual choice corresponds to this decision rule. The relative price ratio, PC PF, is the marginal cost of a unit of clothing, expressed as units of foodthe units of food that are forgone is the opportunity cost of an additional unit of clothing. Similarly, the opportunity cost of an additional unit of food is PF PC units of clothing. The slope of the indifference curve reects the marginal benet of an additional unit of clothing expressed as units of food. For example, if Dom is willing to trade 5 units of food for 1 unit of clothing (slope of the indifference curve 5), his marginal benet of one additional unit of clothing must equal the utility from 5 units of food. Similarly, his marginal benet of a unit of food is equivalent to .2 units (1 5) of clothing. If Dom is not at the point of tangency between the indifference curve and the budget line, the marginal benet of trading one good for the other must be greater than the marginal cost. Suppose Dom is willing to trade 5 units of food for 1 unit of clothing, but only has to trade 2 units of food for 1 unit of clothing in the marketplace. In this case, Dom should trade food for clothing since the marginal (incremental) benet is greater than the marginal (incremental) cost. At the optimum (point of tangency) the marginal benet of consuming 1 more unit of either good is equal to the marginal cost and there is no reason to make additional trades. Changes in Choice Doms consumption opportunities will change whenever prices or income change. Consequently, he will make different choices. Recall that changes in relative prices alter the slope of the constraint. When the relative price of a good increases, individuals typically 28 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 Figure 2.6 Changes I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts Optimal Choice and Price This gure shows how the optimal choice changes with an increase in the price of food. In this example, the individual chooses less food (F1 rather than F0 This * *). is the typical caseusually, an individual will purchase less of a good when its price increases. Due to the particular utility function used in this example, the amount of clothing purchased remains unchanged (C*). More generally, the amount of clothing purchased can go either up or down. It depends on the location of the new tangency point. F Original consumption opportunities Consumption opportunities after increase in price of food Quantity of food 14 * F0 * F1 C* C Quantity of clothing choose less of that good.14 Figure 2.6 shows how Dom will purchase less food as its relative price increasesfood is more expensive and so less attractive than it was at a lower price. Generally the amount of clothing purchased can go either up or down; it depends on the location of the new tangency point. (Given the particular utility function assumed in this example, the amount of clothing purchased remains unchanged.) Even though the price of clothing is relatively more attractive, the increase in food prices can limit available income so as to reduce the amount purchased of both goods. Changes in Doms income cause parallel shifts in the constraint and will change his optimal choice. In Chapter 4, we examine in more detail how changes in income and prices affect consumption choices. The appendix to this chapter contains a more detailed analysis of the effects of price changes on individual choice. Choices also change if preferences change. Now changes in preferences undoubtedly occur. (Do you really believe that Toys R Us will have any difculty satisfying the demands for toys that were highly popular in past years, such as Teenage Mutant Ninja Turtle action gures, Tomaguchi virtual pets, Tickle-Me-Elmo dolls, or Pokemon Cards next Christmas?) Yet, economists rarely focus on such explanations. Economics has little theory to explain what might cause preferences to change. And since a large premium is placed on operationalism in managerial economics, preference-based explanations generally are appealed to only after other potential explanations are exhausted. In a sense, these preference-based explanations are too easythey work too well. Virtually any observed behavior could be explained by appealing to preferences: Why did the consumption of frozen yogurt increase relative to that of ice cream? Peoples preferences changed so that more frozen yogurt and less ice cream was demanded. But an observed 14 Although in principle some individuals might purchase more of a good if its price increases, this outcome is rarely observed. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 15 The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 29 ANALYZING MANAGERIAL DECISIONS: Consumer Choice and Graphical Tools You are a manager for a company that bottles and sells wine in two different countries. You charge the same price for a bottle of wine in both countries. Yet, your wine sales are much higher in one country than the other. Your boss asks you to develop an explanation for the differences in wine sales between the two countries and to develop a plan to sell more wine in the country with low wine consumption. Population sizes and family incomes in the two countries are very similar. You also know that each country imposes a per bottle tax on wine. Begin by providing a plausible economic explanation (focusing on constraints) for the differences in wine sales in the two countries. Illustrate your explanation by using indifference curves and budget lines for representative consumers from the two countries. What data would you want to determine if your explanation is likely to be correct? Are there other plausible explanations for the differences in wine consumption? Are there ways to determine which of these explanations is most likely to be driving the differences in consumption? 1. Suppose that your economic explanation is likely to be correct and that your company will not allow you to lower the price per liter that you charge for wine in the two countries. Discuss at least two potential actions that you might take to sell more wine in the country with low demand. 2. Now provide a potential preference-based explanation for the differences in wine sales. Suppose that this explanation is correct. Discuss whether there are likely to be feasible policies that you could use to increase wine sales in the country with the low demand. reduction in consumption could have been explained just as readily. Without a deeper understanding of why preferences change, one is left explaining everything but with an analysis that allows you to predict nothing. Ultimately, the managerial usefulness of this analysis comes from its power to identify policy instruments that have a predictable impact on the problems at hand. Across a broad array of problems, assuming that underlying preferences are reasonably stable and analyzing the impact of changes in opportunities and constraints regularly will yield important managerial insights and identify productive managerial tools. Motivating Honesty at Merrill Lynch Often, economists focus on consumption goods such as food and clothing. This focus is natural given the interests economists have in understanding consumer behavior. Yet this analysis can be extended easily to consider other goods that people care about, such as love and respect.15 Such an extension can be used to analyze the problem at Merrill Lynch. Suppose that Susan Chen, like other analysts at Merrill Lynch, values two goods money and integrity. Her utility function is Utility F (Money, Integrity) (2.5) Money is meant to symbolize general purchasing power; it allows the purchase of goods such as food, clothing, and housing. Integrity is something Sue values for its own sake 15 G. Becker (1993), Nobel Lecture: The Economic Way of Looking at Behavior, Journal of Political Economy 101, 385409. 30 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts Figure 2.7 Nature of Opportunities Facing an Analyst at Merrill Lynch The constraint depicts the maximum amounts of money and integrity that are possible for the analyst given the bonus plan and conditions at the company. If the analyst sacrices all integrity and recommends stocks even if they are poor investments, the employee earns a maximum of $max a year. Investment banking business is lost if the analyst gives objective advice and rates certain stocks as poor investments (selects a higher level of integrity). Income is lower since the analyst is paid a bonus based on investment banking revenues. Ic represents complete honesty. $ $max Income (in dollars) 16 $min I Ic Quantity of integrity being honest in her dealings with other people provides Sue with satisfaction and she values it for that reason. Suppose that integrity can be measured on a numerical scale with Sue preferring higher values. For example, 5 units of integrity provide more utility than 4 units of integrity. (In actuality, measuring a good like integrity on a numerical scale might be quite difcult. Yet this complication does not limit the qualitative insights that we can derive from the analysis.) Merrill paid its stock analysts an annual bonus that was based partly on the analysts contribution to the investment banking side of the business (e.g., the rms underwriting activities). If Sue were completely honest and rated a company as a poor investment, the management of that company might take its investment banking business to another rm. The resulting loss in Merrills investment banking revenue would reduce Sues annual bonus. This bonus scheme thus confronts Sue with a trade-off. She can be honest and derive satisfaction from maintaining her integrity, or she can be dishonest in her rating of the stock and obtain a higher bonus. (She also might consider the future effects on her income from developing a good or bad reputation as an investment analyst. However, the analysis in this chapter is framed in a simple one-period context and does not consider monetary returns from developing a good reputation. In subsequent chapters we extend the analysis and consider such multiperiod effects.) Figure 2.7 depicts Sues implied opportunities. This constraint shows the maximum combinations of income and integrity that are feasible given the compensation plan and conditions at the company.16 If Sue sacrices all integrity, she earns $max a year. If 16 For simplicity, we draw the constraint as linear. Linearity is not necessary for our analysis. Also, we want to emphasize that we put dollars on the vertical axis only because it is a convenient general indication of value, not because money is more important than other things. We could illustrate Sues willingness to trade integrity against anything else Sue values, such as Big Macs, pianos, or pairs of jeans. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 2 Figure 2.8 Optimal Choices of an Analyst at Merrill Lynch under Two Different Compensation Plans Economists View of Behavior 31 $ Income (in dollars) Case 1 reects the original compensation plan. In this case, compensation includes a high bonus based on investment banking revenues and the constraint is relatively steep. In Case 2, the rm reduces the emphasis on investment banking revenues in compensating analysts. The slope of the constraint is atter. The result is that the individual chooses a higher level of integrity in Case 2 than in Case 1. 17 The McGrawHill Companies, 2009 2. Economists View of Behavior $* 1 $* 2 Case 2 Case 1 * I1 * I2 I Quantity of integrity she is scrupulously honest in her investment recommendations, she earns less (there is a positive oor on her income, $min, since her base salary does depend on the amount of investment banking business and her analysis undoubtedly will suggest recommending some of Merrills clients stocks). Intermediate options along the constraint are possible. While Sue would like to earn more than $max, higher income is not feasible in this job. Sue chooses the combination of integrity and income that places her on the highest attainable indifference curve. This choice occurs at the point of tangency between her indifference curve and the constraint. Sue ends up selecting relatively low amounts of integrity because the bonus plan adopted by Merrills management has made integrity expensive. If Sue chooses more integrity, she must forfeit a relatively large amount of income. Management can alter the opportunities Sue and her colleagues face by changing its compensation plan. In the Merrill case, reducing the emphasis of investment banking revenue in determining the annual bonus reduces the gains from dishonest advice and thus attens the constraint. Changes in the slope of the constraint result in a different tangency point and hence a different choice. Figure 2.8 shows how Sues optimal choice changes when the emphasis on investment banking revenue is decreased.17 The result is more honest behavior. In essence, Sue purchases more integrity because it 17 We have altered the compensation scheme in a manner that places Sue on the same indifference curve. Our rationale for doing this is as follows: Merrill Lynch must provide Sue with sufcient utility to retain her at the rm. Below this level of utility, Sue will quit and work elsewhere. Merrill Lynch is unlikely to want to pay Sue more than this minimum utility because it reduces rm prots. Thus, Merrill Lynch has an incentive to adjust compensation in a manner that keeps her on the same indifference curve. Sues indifference curve in Figure 2.8 can be viewed as this reservation utility. These issues are covered in more detail in Chapter 14. 18 32 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Money and Job Satisfaction A former investment banker related that: I recently decided to leave a promising career in an investment bank. The salary was lavish but the working hours were inhuman, social or family life was nonexistent and the level of stress was intolerable. Ive given it up for a job with a measly salary but one thats interesting and personally satisfying andguess whatIve never been happier in my life. In surveys of 400 executives by the Young Presidents Organization, they admitted that the pursuit of money consumed them, yet played down the importance of money in career choices. Some equated wealth with self-worth and others said, Theres never enough. However in attitude surveys, career-development programs consistently top employees list of wants. Managers often insist they would never make career decisions primarily based on money. One manager ranks his familys well-being and happiness as most important, but still appreciates the monetary value of the job. One manager who received a large bonus said he will use the money for his childrens education, Its given me a sense of relief. Now I can redirect some money to things we havent done or divert it into a retirement fund. Thus, these executives clearly value things other than money and regularly make choices that trade off money for other things that they value. Source: H. Lancaster (1998), Needy or Greedy? The Wall Street Journal ( June 30), B1; L. Kellaway (2007), Do I Go for Money or Stay Put in a Job I Like? Financial Times (March 21), 12. now is less expensive. Consistent with this analysis, Merrill, in its settlement with the State of New York, agreed to change the way it evaluated and compensated its analysts. Bonuses now are based on the quality of investment advicenot tied to investment banking business. Managerial Implications This analysis illustrates how the economic framework can be used to analyze and address management problems. Managers are interested in affecting the behavior of individuals such as employees, customers, union leaders, or subcontractors. Understanding what motivates individuals is critical. The economic approach views individual actions as the outcomes of maximizing personal utility. People are willing to make substitutions (for example, less leisure time for more income) so long as the terms of trade are advantageous. Managers can affect behavior by appropriately designing the opportunities facing individuals. The design of the opportunities affects the trade-offs that individuals face and hence their choices. For example, management can motivate employees through the structure of compensation plans or customers through pricing decisions. The outcome of individuals making economic choices is a function of both opportunities and preferences. Individuals try to achieve their highest level of satisfaction given the constraints they face. Our discussion of management implications, however, intentionally focuses on opportunities and constraints, not preferences. As a management tool, the usefulness of focusing on personal preferences often is limited. It is difcult to change what a person likes or does not like. Moreover, preferences rarely are observable, and (as we noted earlier) virtually any observed change in choice can be explained as simply a matter of a change in personal tastes. For instance, a preferencebased explanation as to why employees were dishonest at Merrill Lynch is that these employees gained personal utility from being dishonest (or compared to employees at other rms, Merrill Lynch employees were willing to trade large amounts of personal integrity for small nancial rewards). This explanation is not very helpful in giving management guidance on how to address the problem. It suggests that Merrill Lynch might try to re BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 19 33 MANAGERIAL APPLICATIONS Medicare Creates Perverse Incentives for Doctors Doctors do not care about money but are motivated by concerns about providing the best care for patientsright? Apparently many doctors and the major drug company employees do not think so. Perverse incentives among physicians arguably have contributed to the problem of spiraling health care costs in the United States. For years, Medicare (federal health program for the elderly) reimbursement policies allowed individual doctors to make hundreds of thousands of dollars a year in extra prots from the drugs they administered to patients in their ofces (the doctors would buy the drugs themselves and bill Medicare, rather than having the patients get them directly from pharmacies). For example, many cancer doctors earned over $1 million per year on drug sales alone. Because the prots on different drugs varied enormously, doctors had incentives to prescribe medications with the highest prot margins. Some physicians have acknowledged that they performed treatments that got them the best reimbursements, whether or not the treatments beneted patients. Drug companies were well aware of the Medicare policies and calculated the prots that doctors received from prescribing specic drugs down to the penny. For example, in 1998 Schering-Plough told its sales representatives that its drug for the treatment of bladder cancer could produce a prot for a physician of $2,373.84 per patient. Sales representatives in turn made sure that doctors were well aware when their drugs were in the high prot category. For instance, a sales representative for AstraZeneca wrote in a letter to Arizona urologists, DO THE MATH. Medicare changed its reimbursement policies in 2005 and reduced the prots that physicians could make on drug sales. At least some physicians have responded by shifting from drug intensive treatments to other treatments that have higher prot margins. To quote one doctor, People go where the money is, and youd like to believe its different in medicine, but its really no different . . . as long as oncologists continue to be paid by the procedure instead of spending time with patients, they will nd ways to game the system. Source: Alex Berensen (2007), Incentives to Limit Any Savings in Treating Cancer, nytimes.com (June 12). dishonest employees and replace them with employees who care more about personal integrity. But the difculty of observing personal preferences limits the viability of this approach. It would be difcult for Merrill Lynch to know if, as a group, its new hires would be any less dishonest than the old employees. You cannot just ask applicants if they are honestif they are not, they will have no qualms about claiming that they are. The fact that individuals are clever and creative in limiting the effects of constraints greatly complicates management problems. Changing incentives will affect employee behavior, though sometimes in a perverse and unintended manner. Consider two of the Soviet Unions early attempts to adopt incentive compensation to motivate employees. To discourage taxi drivers from simply parking their cabs, they were rewarded for total miles traveled; to encourage additional production, chandelier manufacturers were rewarded on total volume of productionmeasured in kilograms. In response to these incentive plans, Moscow taxi drivers began driving empty cabs at high speeds on highways outside the city and chandelier manufacturers started producing such massive xtures that they literally would collapse ceilings. (It is less costly to make one 100-kilo chandelier than ve 20-kilo chandeliers; manufacturers also substituted lead for lighter-weight inputs.) Merrill Lynch initially adopted bonuses to motivate analysts to work harder and cooperate across business units. The dishonest behavior was a side effect that potentially was unanticipated when the plan was adopted. In summary, the economic approach to behavior has important managerial implications. The framework suggests that a manager can motivate desired actions by establishing appropriate incentives. However, managers must be careful because setting improper incentives can motivate perverse behavior. 20 34 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts It is worth noting that economic analysis is limited in its ability to forecast the precise choices of a given individual because individual preferences are largely unobservable. The focus is on aggregate behavior or on what the typical person tends to do. For example, an economist might not be very good at predicting the responses of individual employees to a new incentive plan. An economist will be successful in predicting that the typical employee will work harderand thus output for the group will risewhen compensation is tied to output, than when a xed salary independent of performance is paid. Managers typically are interested in structuring an organizational architecture that will work well and does not depend on specic people lling particular jobs. Individuals come and go, and the manager wants an organization that will work well as these changes occur. In this context, the economic framework is likely to be useful. To solve management problems where the characteristics of a specic individual are more important, other frameworks may be more valuable. For example, if the board is interviewing a potential new CEO, insights into that individuals behavior derived from psychology might be extremely useful. Alternative Models of Behavior18 We have shown how the economic view of behavior can be used in managerial decision making. We now discuss four other models that are commonly used by managers (either explicitly or implicitly) to explain behavior. Our discussion of each of these models is simplied. The intent, however, is to capture the essence of a few of the more prominent views that managers have about behavior and to illustrate how managerial decision MANAGERIAL APPLICATIONS Happy-Is-Productive versus Economic Explanations of the Hawthorne Experiments Seven productivity studies were conducted at Western Electrics Hawthorne plant over the period 19241932. All seven studies focused on the response of assembly workers productivity when different aspects of the work environment were manipulated (for example, length of break times and workday). Surprisingly, productivity rose virtually regardless of the particular manipulation. For example, it is claimed that productivity increased whenever illumination of the work area was changed, regardless of the direction of the change. When the lights were turned up, productivity increased, and when they were turned down, productivity increased, as well. This result is known as the Hawthorne Effect and is among the most discussed ndings in psychology; it often is taken as support for the happy-is-productive model. The workers in the experiment were given special attention and nonauthoritarian supervision relative to other workers at the plant. Also, the affected workers views on the experiments were solicited by management, and the workers were given more responsibility. These actions, it has been argued, increased job satisfaction and performance. Parsons (1974) presents evidence that the ndings of the Hawthorne experiments also can be explained by accompanying changes in the compensation system. Prior to the experiment, all workers were paid based on the output of a group of about 100 workers. During the experiment, the compensation plan was changed to base pay on the output of only ve workers. In this case, a given workers output more directly affects her own pay, and economic theory predicts increased output. Interestingly, the last of the original Hawthorne experiments observed workers where the compensation system was not changed. In that seventh experiment, there was no change in output. Source: H. Parsons (1974), What Happened at Hawthorne? Science 183, 922932. 18 This section draws on W. Meckling (1976). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 21 35 making is affected by the particular view. We contrast these alternative views with the economic view and argue why the economic framework is a particularly useful tool for managers. Only-Money-Matters Model Some people believe that the only important component of the job is the level of monetary compensation. But as we have already suggested, people have an incredibly broad range of interests, extending substantially beyond money. And these interests are reected in a diverse array of activities. As examples, much of the work through the Red Cross is undertaken by unpaid volunteers; people frequently choose early retirement, forgoing a regular paycheck to enjoy additional leisure time; riskier occupations command higher pay in order to attract people into those jobs. Some of this confusion can result from a misinterpretation of standard economic analysis. Central to economics is the study of trade-offs (recall our discussion of indifference curves illustrating trade-offs between food and clothing). Economists frequently use money as one of the goods being considered. But in these cases, money is merely a convenient unit of value: It simply represents general purchasing power. Its use does not suggest that only money matters. Happy-Is-Productive Model Managers sometimes assert that happy employees are more productive than unhappy employees. Managers following this happy-is-productive model see as their goal the designing of work environments that satisfy employees. Psychological theories, such as Maslows and Herzbergs, are frequently used as guides in efforts to increase job satisfaction.19 A manager adhering to the happy-is-productive model might suggest that the problem at Merrill Lynch was motivated by disgruntled employees who took out their frustrations on customers. This view implies that Merrill Lynch could reduce the problem by promoting employee satisfaction through such actions as designing more interesting jobs, increasing the rates of pay, and improving the work environment. Happier employees would be expected to provide customers with better investment advice. The economic and happy-is-productive models do not differ based on what people care about. The economic model allows individuals to value love, esteem, interesting work, and pleasant work environments, as well as more standard economic goods such as food, clothing, and shelter. The primary difference in the models is what motivates individual actions. In the happy-is-productive model, employees exert high effort when they are happy. In the economic model, employees exert effort because of the rewards. To contrast the two models, consider offering an employee guaranteed lifetime employment plus a large salary, which will be paid independent of performance. The happy-is-productive model suggests that the employee will be more productive, because the high salary and job security are likely to increase job satisfaction. The economic model suggests that the employee would exert less effortsince the employee receives no additional rewards for working harder and will not be red for exerting low effort. 19 F. Herzberg, B. Mausner, and B. Snyderman (1959), The Motivation to Work ( John Wiley & Sons: New York); and A. Maslow (1970), Motivation and Personality (Harper & Row: New York). 22 36 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Culture and Behavior In Tokyo, lost cell phones, umbrellas, and cash regularly nd their way to the Tokyo Metropolitan Police Lost and Found Centerthe Japanese are scrupulous about turning in found articles. In 2002 the center handled $23 million in cash and 330,000 umbrellas. The system traces its roots to a code written in 718. Lost goods, animals, and servants had to be handed over to a government ofcial within ve days of being found. Not handing over found objects was severely punished. In 1733 two ofcials who kept a parcel of clothing were led around town and executed. Current law gives the nder seven days to turn in found goods. If the item is reclaimed, the nder is entitled to a reward (5 to 20 percent). If the item is not reclaimed within six months, the nder can claim it. The most commonly reclaimed item is a cell phoneabout 75 percent are returned. The least reclaimed are umbrellas at 0.3 percent. Source: N. Onishi (2004), Never Lost, but Found Daily, New York Times (January 8), A1. Good-Citizen Model Some managers subscribe to the good-citizen model. The basic assumption is that employees have a strong personal desire to do a good job; they take pride in their work and want to excel. Under this view, managers have three primary roles. First, they need to communicate the goals and objectives of the organization to employees. Second, they must help employees discover how to achieve these goals and objectives. Finally, managers should provide feedback on performance so that employees can continue to improve their efforts. There is no reason to have incentive pay, since individuals are interested intrinsically in doing a good job. This view suggests that the problems at Merrill Lynch occurred because employees misunderstood what was good for the company. Employees might have thought that increasing investment banking revenues was in the companys best interests, even if it required a certain amount of dishonesty. Under the good-citizen view, the management of Merrill Lynch could motivate employee honesty by clearly communicating to its analysts that Merrill Lynch would be better off in the long run if they did not deceive their customers. Managers might be instructed to hold a series of analyst meetings to stress the value of honesty and objective investment advice. In the good-citizen model, employees place the interests of the company rst. There is never a conict between an employees personal interest and the interest of the company. In contrast, the economic model posits that employees maximize their own utility. Potential conicts of interest often arise. The economic view predicts that pleas from Merrill Lynch management that analysts be more honest would have little effect on behavior unless they also changed the reward system to make it in the interests of analysts to be more honest. Product-of-the-Environment Model The product-of-the-environment model argues that the behaviors of individuals are largely determined by their upbringings. Some cultures and households encourage positive values in individuals, such as industry and integrity, whereas others promote negative traits, such as laziness and dishonesty. This model suggests that Merrill Lynch had dishonest analysts. A response would have been to re these employees and replace them with honest analysts from better backgrounds. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 23 37 Which Model Should Managers Use? Behavior is a complex topic. No behavioral model is likely to be useful in all contexts. For example, the economic model is unlikely to be helpful in predicting whether a given individual will prefer a red shirt to a blue shirt (selling at the same price). But our focus is on managerial decision making. In this context, there are reasons to believe that the economic model is particularly useful. Managers are frequently interested in fostering changes in behavior. For example, managers want consumers to buy more of their products, employees to exert more effort, and labor unions to accept smaller wage increases. In contrast to other models, the economic framework provides managers with concrete guidance on how to alter behavior. Desired behavior can be encouraged by changing the feasible opportunities facing the decision maker. For example, incentive compensation can be used to motivate employees, and price changes can be used to affect consumer behavior. There is ample evidence to support the hypothesis that this economic framework is useful in explaining changes in behavior. The most common example is that consumers tend to buy fewer products at higher prices. The evidence suggests that the model is also useful in explaining aspects of behavior in many other contexts, including voting; the formation, dissolution, and structure of families; drug addiction; and the incidence of crime.20 The good-citizen model appears less successful in predicting behavior in business settings. Management would be an easy task if employees would work harder and produce higher-quality products simply on request. The happy-is-productive model also has material limitations. Most importantly, the existing evidence suggests that there is little relation between job satisfaction and performance (see Scotts Criticisms of the Happy-Is-Productive Model in the accompanying box). Happy employees are not necessarily more productive. Sometimes, managers might want to follow the implications of the product-of-the-environment model and re employees with undesirable traits. Yet, this approach is of limited use in solving most managerial problems. Also, given laws that limit discrimination, this approach can subject the rm to potentially serious legal sanctions. A CADEMIC APPLICATIONS The Economic Framework and Criminal Behavior Criminals often are viewed as psychologically disturbed. Evidence, however, suggests that criminal behavior can be explained, at least in part, by the economic framework. This framework predicts that a criminal will consider the marginal costs and benets of a crime and will commit the crime only when the benets exceed the costs. Under this view, increasing the likelihood of detection and/or the severity of punishment will reduce crimes. In a pioneering study, Issac Ehrlich examined whether the incidence of major felonies varied across states with the expected punishment. He found that the incidence of robberies decreased about 1.3 percent in response to each 1 percent increase in the proportionate likelihood of punishment. The incidence of crime also decreased with the severity of the punishment. Since Ehrlichs study, scholars have conducted extensive research on this topic. In general, the results support the conclusion that the economic model plays a useful role in explaining criminal activity. Source: I. Ehrlich (1973), Participation in Illegitimate Activities: A Theoretical and Empirical Investigation, Journal of Political Economy 81, 521565. 20 G. Becker (1993). 24 38 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts ACADEMIC APPLICATIONS Criticisms of the Happy-Is-Productive Model W. Richard Scott summarizes some of the major concerns about the happy-is-productive model (sometimes referred to as the human-relations movement): Virtually all of these applications of the human-relations movement have come under severe criticism on both ideological and empirical grounds. Paradoxically, the human-relations movement, ostensibly developed to humanize the cold and calculating rationality of the factory and shop, rapidly came under attack on the grounds that it represented simply a more subtle and rened form of exploitation. Critics charged that workers legitimate economic interests were being inappropriately deemphasized; actual conicts of interest were denied and therapeutically managed; and the roles attributed to managers represented a new brand of elitism. The entire movement was branded as cow sociology just as contented cows were alleged to produce more milk, satised workers were expected to produce more output. The ideological criticisms were the rst to erupt, but reservations raised by researchers on the basis of empirical evidence may in the long run prove to be more devastating. Several decades of research have documented no clear relation between worker satisfaction and productivity. Source: W. Scott (1981), Organizations: Rational, Natural and Open Systems (Prentice Hall: Englewood Cliffs, NJ), 8990. ANALYZING MANAGERIAL DECISIONS: Interwest Healthcare Corp. Interwest Healthcare is a nonprot organization that owns 10 hospitals located in three western states. Cynthia Manzoni is Interwests chief executive ofcer. Vijay Singh, Interwests chief nancial ofcer, and the administrators of the 10 hospitals report to Manzoni. Singh is deeply concerned because the hospital staffs are not being careful when entering data into the rms management information system. This data involves information on patient intake, treatment, and release. The information system is used to compile management reports such as those relating to the costs of various treatments. Also, the system is used to compile reports that are required by the federal government under various grant programs. Singh reasons that without good information, the management and government reports are less useful and potentially misleading. Moreover, the federal government periodically audits Interwest and might discontinue aid if the reports are deemed inaccurate. Thus, Singh is worried about the managerial implications and the potential loss of federal funds. Singh has convinced Manzoni that a problem exists. She also realizes the importance of an accurate system for both management planning and maintaining federal aid. Six months ago, she invited the hospital administrators and staff members from the corporate nancial ofce to a retreat at a resort. The purpose was to communicate to the hospital administrators the problems with the data entry and to stress the importance of doing a better job. The meeting was acrimonious. The hospital people accused Singh of being a bureaucrat who did not care about patient services. Singh accused the hospital staffs of not understanding the importance of accurate reporting. By the end of the meeting, Manzoni thought that she had a commitment by the hospital administrators to increase the accuracy of data entry at their hospitals. However, six months later, Singh claims that the problem is as bad as ever. Manzoni has hired you as a consultant to analyze the problem and to make recommendations that might improve the situation. 1. What are the potential sources of the problem? 2. What information would you want to analyze? 3. What actions might you recommend to increase the accuracy of the data entry? 4. How does your view of behavior affect how you might address this consulting assignment? BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 25 39 Decision Making under Uncertainty Throughout this chapter, we have considered cases where the decision maker has complete certainty about the items of choice. For instance, Dom Lalisse knew the exact prices of food and clothing, and Sue Chen knew the precise trade-off between integrity and compensation at Merrill Lynch. Decision makers, however, often face uncertainty. For instance, in choosing among risky investment alternatives (such as stocks and bonds), an individual must forecast the likely payoffs. Even so, there can be signicant uncertainty about the eventual outcomes. The analysis presented in this chapter can be extended readily to incorporate decision making under uncertainty.21 A detailed analysis of decision making under uncertainty is beyond the scope of this book. This section introduces a few key concepts that we will use later in this book. Expected Value Taylor McClure sells real estate for RealCo. He receives a sales commission from his employer. For simplicity, suppose that Taylor has three possible incomes for the year. In a good year, he sells many houses and earns $200,000, whereas in a bad year he earns nothing. In other years, he receives $100,000. Probability refers to the likelihood that an outcome will occur. In this example, each outcome is equally likely, and thus has a probability of 1 3 of occurring. The expected value of an uncertain payoff is dened as the weighted average of all possible outcomes, where the probability of each outcome is used as the weights. The expected value is a measure of central tendencythe payoff that will occur on average. In our example, the expected value is:22 Expected value (1 3 0) $100,000 (1 3 100,000) (1 3 200,000) (2.6) Variability Although Taylor can expect average earnings of $100,000, his income is not certain. The variance is a measure of the variability of the payoff. It is dened as the expected value of the squared difference between each possible payoff and the expected value. In this example, the variance is Variance 1 3(0 100,000)2 1 3(100,000 1 3(200,000 100,000)2 6.7 billion 100,000)2 (2.7) The standard deviation is the square root of the variance: Standard deviation (6.7 billion)1 2 $81,650 (2.8) Variances and standard deviations are used as measures of risk. It does not really matter which we use, since one is a simple transformation of the other (higher standard deviations correspond to higher variances). In this example, we focus on the standard deviationin 21 For example, E. Fama and M. Miller (1972), The Theory of Finance (Dryden Press: New York), Chapter 5. Note that the expected value need not equal one of the possible outcomes. As a weighted average, it can be a value between outcomes. In this example, it happens to correspond to one of the possible outcomes, $100,000. 22 40 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts Figure 2.9 Indifference Curves for Expected Value and Standard Deviation This gure displays three indifference curves for a riskaverse individual. The individual prefers higher expected value but lower standard deviation. Standard deviation is a measure of risk. Since risk is a bad, the indifference curves are positively sloped. Northwest moves are utility-increasing. Currently, the individual has a compensation package that has an expected value of $100,000 and a standard deviation of $81,650. The certainty equivalent of this package is $80,000. The risk premium is $20,000. $ Increasing utility 3 Expected value (in dollars) 26 2 1 100,000 Risk premium = $20,000 80,000 $ 81,650 Standard deviation (in dollars) part because the standard deviation is expressed in the same units as the mean, dollars (the units for the variance would be dollars squared). Higher standard deviations reect more risk. An event with a denite outcome has a standard deviation of zero. Risk Aversion Like most people, Taylor is risk-averse: Holding the expected payoff xed, he prefers a lower standard deviation. He therefore gains utility from an increase in expected value, but he experiences a reduction in utility from increases in standard deviation. Figure 2.9 shows three of Taylors indifference curves. Each curve shows all combinations of expected value and standard deviation that give Taylor equal utility. In contrast to our previous analysis, here one of the objects of choice is a badTaylor dislikes risk. Thus, in this gure, the indifference curves have positive slopes, and northwest movements are utility-increasing (recall in the standard analysis that the curves have negative slopes, and northeast movements are utility-increasing). The slopes of the indifference curves indicate Taylors degree of risk aversion. Steeper slopes translate into higher risk aversion. (If the slopes of the indifference curves are steep, Taylor must receive a relatively large increase in expected value for each additional unit of risk to maintain a constant level of utility.) If his indifference curves were totally at, he would be risk-neutral. A risk-neutral person cares only about expected value and is indifferent to the amount of risk. Indifference curve 3 is associated with the highest level of utility, whereas curve 1 is associated with the lowest utility. Taylor is currently on curve 2. Given a choice among compensation plans with different expected payoffs and risk, Taylor will choose the combination that places him on the highest attainable indifference curve. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 27 41 ANALYZING MANAGERIAL DECISIONS: Risk Aversion versus Risk Taking Lauren Arbittier decides to bet $2,000 on number 35 of the roulette wheel in a Las Vegas casino. Almost immediately she starts to question her decision. Lauren normally is a risk avoider who hardly ever gambles. But she works at Trilogy Software where the CEO understands that taking risks and suffering the consequences are critical to the rms success. The CEO wants to develop people who take chances. You dont win points . . . for trying. Lauren is participating in Trilogys three-month training program for all new recruits. It educates employees about, among other things, how to evaluate risky projects, not just to immediately accept or reject the project because it is risky. The program also suggests to employees that they will not be rewarded at Trilogy unless they take risks. Thus, although Lauren does not like taking risks, working for Trilogy, she has economic incentives to do so. There are at least three ways in which the Trilogy training program might be effective: (1) It changes employees preferences regarding risk bearing. (2) It more effectively identies individuals with the risk to tolerances that Trilogy desires. (3) It better communicates the consequences to Trilogy employees of undertaking risky ventures. Discuss the likely importance of these three mechanisms. SOURCE: E. Ramstad (1998), High Rollers, How Trilogy Software Trains Its Raw Recruits to Be Risk Takers, The Wall Street Journal (September 21), A1. Certainty Equivalent and Risk Premium Figure 2.9 indicates that Taylor is indifferent between the risky commission scheme, which has an expected payoff of $100,000, and a certain income of $80,000. The $80,000 is Taylors certainty equivalent for the risky income streamhe is willing to trade the uncertain income of $100,000 for a certain income of $80,000. The difference between the expected value of the risky income stream and the certainty equivalent is called the risk premium. This $20,000 premium, which comes in the form of a higher expected payoff, must be paid to keep Taylor indifferent between the risky income stream and his certainty equivalent. Suppose that Taylor receives a job offer from a competing real estate company that would pay him a xed salary of $90,000 per year. Taylor considers the new job to be the same as his current job in all dimensions other than the compensation plan. Taylors current compensation plan will not be sufcient to motivate him to continue to work for RealCo. Even though his current plan has a higher expected payoff, he would prefer the certain $90,000 to RealCos risky commission plan. If RealCo wants to retain Taylor, it must offer him a compensation package that provides the same level of utility as the $90,000 for certain. Risk Aversion and Compensation Diversied shareholders, who invest in portfolios of companies, own much of the stock of large rms. Managers are often ill-diversied, having much of their human and nancial capital invested in one rm. As we will discuss later in this book, this difference in diversication can lead to managers being overly risk averse in their investment decisions relative to those shareholders would prefer. Shareholders can induce managers to 28 42 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts undertake more risky investment by adopting compensation plans that reward good outcomes, but that do not penalize bad outcomes heavily. The top management of Enron (see Chapter 1) arguably went too far in this direction, inducing their managers to take too much risk (effectively transforming their behavior from risk averse to risk loving).23 We expand on this issue later in the book. Summary In this chapter we summarize the way economists view behavior. In the economic model, individuals are seen as having unlimited wants but limited resources. They rank alternative uses of limited resources in terms of preference and choose the most preferred alternative. Individuals are clever in guring out ways of maximizing their satisfaction (utility) in the face of resource constraints. Individuals are not necessarily selsh in the sense that they care only about their personal wealth: They also care about charity, family, religion, and society. They are not infallible supercomputers. The opportunity cost of using a resource is the value of the resource in its best alternative use. For example, the cost of having a manager use ve hours to work on a project is the value of the managers time in working on the next best alternative project. Economic decision making requires careful consideration of the relevant opportunity costs. Marginal costs and benets are the incremental costs and benets that are associated with a decision. In calculating marginal costs, it is important to use the opportunity costs of the incremental resources. For example, in deciding whether to purchase a new laptop computer, the marginal cost is its price and the marginal benet is the value that the person places on the new computer. It is the marginal costs and benets that are important in economic decision making. Action should be taken when the marginal benets are greater than the marginal costs. Sunk costs that are not affected by the decision (for example, unrecoverable funds previously spent on computers) are not relevant. A utility function is a mathematical function that relates total utility to the amounts that an individual has of whatever items the individual cares about (goods). Preferences implied by a utility function are pictured graphically by indifference curves. Indifference curves picture all combinations of goods that yield the same level of utility. Individual choice involves maximizing utility given resource constraints. Graphically, the constraint depicts all combinations of goods that are feasible to acquire; it denes the feasible consumption opportunities. The optimal choice is where the indifference curve is tangent to the constraint. At this point, the individual is at the highest level of utility possible given the feasible opportunities. Changes in opportunities result in changes in the optimal choice. An important implication is that managers can affect behavior by affecting constraints and opportunities. Managers, however, have to be careful. Individuals are clever at maximizing their utility, and establishing disfunctional incentives can have perverse consequences. We contrast the economic model with other models of human behavior that managers often use. We argue that the economic model is often more useful than alternative models in managerial decision making. The analysis in this chapter can be extended to the case where the decision maker faces uncertainty about the items of choice. An example of decision making under uncertainty is choosing among risky investment alternatives. One concept that we will rely 23 Most managers have risk-averse preferences (utility functions). Managerial actions, however, are a function of both preferences and constraints. Thus risk-averse preferences can be offset or reinforced by the design of the compensation plan. Compensation plans that limit the upside potential but not the downside induce less risky choices, whereas plans that limit the downside but not the upside induce more risky choices. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 29 43 on later in this book is risk aversion. When confronted with both a risky and a certain alternative having the same expected (or average) payoffs, a risk-averse person always will choose the certain outcome. A risk premium must be offered to entice the person to choose the risky alternative. Throughout this chapter, we focus primarily on how managers might use the economic view to analyze and inuence the behavior of employees. As we will see, the economic view is quite powerful and is useful in explaining behavior in a variety of different contexts. Appendix Consumer Choice The main text of this chapter provided a simple graphical analysis of individual choice. It introduced utility functions, indifference curves, and budget lines. It depicted the optimal choice as the tangency between an indifference curve and the budget line where the willingness and ability to trade are equal. This simple analysis explained how the optimal choice changes with changes in relative prices or income. This economic framework has important implications since managers often want to inuence and/or predict the behavior of individuals, such as customers and employees. This appendix extends the economic framework of individual choice (commonly called consumer choice). Its intent is to provide a deeper and richer understanding of this important model of behavior. This appendix also discusses how this model relates to an important topic that is covered in more detail later in this bookdemand functions. Marginal Utility A utility function expresses the relation between a persons total utility and the level of goods consumed. Utility functions can take many forms. For illustration, suppose that Tom Morrell values only food and clothing and that his utility function is U FC (2.9) where F is the units of food and C is the units of clothing that Tom consumes for the period. Notice, this simple utility function is multiplicative in both food and clothing. If Tom has no clothing, then no matter how much food he has, he has utility of zero. And likewise, if he has no food, then no matter has much clothing he has, his utility again is zero. If Toms consumption bundle consists of 20 units of both food and clothing his utility is 400, while his utility is only 100 if he has 10 units of both goods. Tom prefers the rst bundle, but he is not necessarily four times happier when he has 20 units, rather than 10 units, of each good. The utility function provides an ordinal ranking of consumption bundlesnot a cardinal ranking where absolute comparisons can be made. Marginal utility measures the additional utility that is obtained by consuming one additional unit of a good, while holding all other goods constant. Marginal utility is an important concept in economic analysis since optimizing individuals focus on the marginal (incremental) benets and costs in making consumption choices. Figure 2.10 (page 44) graphs Toms utility as a function of food, while holding clothing constant at 10 units. The equation for this graph is U 10F (2.10) The marginal utility of food in this example is 10for each additional unit of food that Tom consumes he receives 10 additional units of utility. More generally for any given 44 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts Figure 2.10 Toms Utility as a Function of Food with Clothing Held Constant at 10 Units (U 10F ) This gure displays Toms utility as a function of food with clothing held constant at 10 units. Marginal utility measures the additional utility that is obtained by consuming one additional unit of a good, while holding all other goods constant. The marginal utility of F in this example is 10 (the slope of the line)for each additional unit of F that Tom consumes he receives 10 additional units of utility. 400 350 300 Utility 30 Slope = 100/10 = 10 250 200 150 10 100 1 50 0 1 3 5 7 9 11 13 15 17 19 21 Food (F ) 23 25 27 29 31 33 35 quantity of clothing, C, the marginal utility of food is C (given Toms utility function). Similar logic implies that the marginal utility of clothing is F.24 We denote the marginal utilities for food and clothing by MUF and MUC , respectively. Slope of an Indifference Curve Now that we have dened marginal utility, we can derive the slope of an indifference curve. Consider Toms indifference curve for U 100. The equation for this indifference curve, which is pictured in Figure 2.11, is F 100 C.25 Because all points on the indifference curve generate 100 units of utility, the total gain in utility associated with an increase in C must be balanced by an offsetting decline in utility from reduced consumption of F (as Tom moves along the curve): MUC ( C ) MUF ( F ) 0, (2.11) where F and C represent the changes in food and clothing. The rst term in Equation (2.11) represents the change in utility from changing the amount of clothing, while the second term represents the change in utility from changing the amount of food. These terms are equal in magnitude and of opposite sign along an indifference curve, and so total utility remains unchanged. The slope of the indifference curve for a small change in C is given by ( F C ). Rearranging Equation (2.11): ( F C) Slope of the indifference curve (MUC MUF) (MUC MUF) (2.12) 24 Note for the mathematically inclined: The marginal utility of any good X is equal to the partial derivative of the utility function with respect to X. In this example, the partial derivative with respect to F is C, and with respect to C is F. In this example, the marginal utility of each good is constant. More realistically, the marginal utility of a good will eventually decline as the consumer continues to receive more of the good. We chose the utility function in this example to simplify the presentation. 25 The decision to place food on the y-axis is arbitrary. If we had placed clothing on the y-axis the equation for the indifference curve would be C 100 F. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Figure 2.11 Slope of One of Toms Indifference Curves: (MUC/MUF) 45 Indifference curve for 100 units of utility Quantity of food This gure displays an indifference curve for 100 units of utility from Toms utility function: U FC. The equation for the indifference curve is F 100 C. The slope of an indifference curve at any point is (MUC MUF). The slope at a point is dened as the slope of the tangency line at that point. The tangency lines at points A and C are two examples. In this example, the slope at any point is (F C). The absolute value of the slope, which is called the Marginal Rate of Substitution (MRS), declines continuously along the curve. This property implies that Tom becomes less willing to trade F for C as C increases relative to F. Economists View of Behavior 31 A (5, 20): MRS = 4 20 B (10, 10): MRS = 1 10 C (20, 5): MRS = .25 5 5 10 20 Quantity of clothing In the example where U FC, the slope of the indifference curve is (F C ) since MUC F and MUF C. For example, at the point where F 5 and C 20 the slope is .25. The intuition for this result is as follows: If a marginal unit of clothing yields ve units of utility, while a marginal unit of food yields 20 units of utility, food can be traded for clothing at a rate of .25 units for one unit, and utility will remain the same (for small changes in the two goods). The absolute value of the slope of an indifference curve is called the Marginal Rate of Substitution (MRS). The MRS reects the individuals willingness to trade at a point on an indifference curve (in this example, trading food for clothing). Consider Toms willingness to trade when he has a consumption bundle of 10 units of both goods. The slope at this point is (10 10) 1. The MRS, which is equal to one, implies that Tom is willing to give up a small quantity of food to receive an equal number of units of clothing. (If Tom were to increase his consumption of C by one unit he would have to reduce his consumption of F by approximately one unit to keep his utility the same.) The MRS declines along the convex curve indicating that Tom becomes less willing to trade food for clothing as the amount of clothing increases relative to food.26 Individual Choice Recall that the equation for the budget line is: F (I PF) (PC PF)C (2.13) The absolute value of the slope, (PC PF), reects the consumers ability to trade in the market place. For example, when the price of clothing is $2/unit and the price of food is $1/unit, two units of food must be given up to consume one additional unit of clothing 26 The slope of a curve is dened at a point on a curve and the slope changes along the curve. The slope of 1 at the point [10,10] reects Toms willingness to trade for very small changes in C. For a full unit change in C the decline in F is only approximately equal to one. A one unit increase in C and a one unit decrease in F produce a consumption bundle with 99 units of utility. This value is approximately equal to the starting point of 100. 32 46 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts (PC PF 2). The intercept (I PF) indicates how many units of food could be purchased if the entire budget is spent on food. At the optimal consumption bundle the budget line is tangent to an indifference curve (which is the highest attainable indifference curve given the budget constraint). This condition implies that the MRS is equal to the ratio of the prices at the optimum. Since the MRS is equal to the ratio of the marginal utilities for the two goods: MUC MUF PC PF (2.14) The left side of Equation (2.14) represents the willingness to trade, while the right side reects the ability to trade. At the point where the consumer is maximizing his utility (the optimum), the two are equal. We can rearrange Equation (2.14) as follows: MUC PC MUF PF (2.15) Equation (2.15) is an important and familiar result in consumer theory. It says that the consumers utility is maximized when the budget is allocated among goods so that the marginal utility per dollar of expenditure is the same for each good. At any combination where this condition does not hold, the consumer can be made better off by making feasible changes in the consumption bundle. For example, suppose Tom has an initial bundle where the marginal utility per dollar for clothing is 10 and for food is 20. Since he is getting more utility per dollar from food, he should spend less money on clothing and more on food. As he trades clothing for food, his marginal utility of clothing increases while his marginal utility of food decreases. Tom will eventually reach the optimal consumption bundle where the marginal utility-to-price ratios are equal. Equation 2.15 reects a condition known as the equal marginal principle (the marginal utility per dollar is the same for all goods at the optimum).27 This principle reappears in various forms in the economic analysis of both consumer and producer behavior. Solving for the Optimal Consumption Bundle Suppose that Tom has a budget of $100 and the prices of food and clothing are $1 and $2, respectively. How much of each good will he buy? This problem is straightforward since it involves two unknown variables (F and C ) and two independent equations. One equation is the optimality condition Equation (2.14); the second is the budget line Equation (2.13). At Toms optimal choice, the MRS must equal the price ratio (i.e., MUC MUF PC PF). Substituting the values for Toms marginal utilities yields: FC C 2 F2 (2.16) Tom must also satisfy his budget constraint: F 100 2C. (2.17) We can solve for the amount of food that Tom will purchase by substituting Equation (2.16) into Equation (2.17): F F* C* 25 is found by substituting F * 27 100 50. 2 (F 2) 50 into Equation 2.16. This condition is also referred to as the equimarginal principle. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 33 47 At the optimal consumption bundle, Tom obtains 1,250 units of utility (25 50). He can increase his consumption of clothing by purchasing 26 units of clothing and 48 units of food (26 $2 48 $1 $100). However, this bundle would yield only 1,248 units of utility (26 48). Alternatively, he could decrease his consumption of clothing by one unit and increase his consumption of food by two units (C 24; F 52). This bundle also would produce only 1,248 units of utility. Indeed any feasible alternative bundle would yield less than 1,250 units of utility. The equal marginal principle holds at Toms optimal consumption bundlethe marginal utility per dollar of expenditure (MUi Pi) is 25 for both goods. This condition implies that Tom has no incentive to shift expenditures from one good to the other since both goods yield the same marginal utility per dollar of spending. The marginal utility of income is dened as the additional utility that the consumer receives from one additional dollar of income. It can be shown that at the optimum, the marginal utility-to-price ratio for all goods is equal to the marginal utility of income. For example, if Toms income increases by $1, he could increase his utility by 25 units by purchasing additional quantities of either good. Demand Functions A demand function expresses the mathematical relation between the quantity demanded for a product (how many units consumers will purchase) and the factors that determine consumer choice (such as prices and income). In a more general setting than our simple example, the demand for clothing is likely to be affected by the price of clothing, consumer income, the prices of other products, and other variables (such as advertising expenditures). Managers care about consumer choice since a good understanding of the demands for their products is important for making productive investment, pricing, advertising, and other decisions. In subsequent chapters, we focus on aggregate demand for a product (total demand across all consumers in the market) without directly tying the analysis back to individual consumer behavior as analyzed in this chapter. Nevertheless, it is useful to recognize that aggregate demand for a given product can conceptually be derived from the framework presented in this appendix. The derivation of Toms demand function for either food or clothing is particularly easy (given his utility function). Toms optimal consumption bundle is where his MRS equals the price ratio. In Toms case, this condition is: FC PC PF (2.18) Toms total expenditures on either food or clothing is equal to the quantity purchased of the good times its price. By cross-multiplying Equation (2.18) we see that Toms expenditures on food and clothing are always equal: (F PF) (C PC) (2.19) Equation (2.19) implies that Tom will always spend half his income on each good (this result is driven by his utility function). It follows that Toms total expenditures on clothing are (C PC) I 2. Solving for C produces Toms demand function for clothing: C I (2PC) (2.20) This demand function implies that Tom will purchase more clothing as his income rises and less clothing as his income falls. His clothing purchases vary inversely with the price of clothing. $2/unit. Consistent with In our example, Tom had an income of $100 and PC Equation (2.20), we found that he consumed 25 units of clothing. The demand function 34 48 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts implies that if Toms income were to increase to $200 (holding price constant) he would purchase 50 units of clothing. In contrast, his clothing purchases would fall to 12.5 units if PC increased to $4 (holding income constant at $100). Tom is only one consumer who purchases clothing. The total (aggregate) quantity demanded for clothing at a given price is equal to the sum of the purchases made by all consumers in the market. Since Tom always spends half his income on clothing, the amount of clothing that he purchases is not affected by the price of food. This is a special case, which does not hold for many other utility functions. Consider Anne George whose utility function is U 2 C .5 F .5. Annes demand for clothing is C I [(P C PF) PC ].28 This function indicates that Annes clothing purchases increase with income and the price of food, but decrease with the price of clothing. For example, at the initial prices and income (PF $1, PC $2, and I $100) Anne purchases 16.67 units of clothing. If the price of food were to increase from $1 to $2, her demand for clothing increases to 25 units. When the prices for food and clothing are equal she spends half her income on each good. As relative prices change, Anne spends a higher percentage of her income on the relatively less-expensive good. Income and Substitution Effects Equation (2.20) indicates that Toms demand for clothing decreases with the price of clothing. Figure 2.12 displays the example where the price of clothing increases from $2 to $4 (holding income constant at $100 and the price of food at $1). Remember, Toms utility function is U CF. The price increase causes the budget line to rotate inward. The new budget line, B 2 is steeper than the original line, B1 (slopes of 4 and 2, respectively). The inward shift of the budget line implies that Tom has less purchasing power than he had prior to the price increase. The area between the two budget lines contains consumption bundles that he could have purchased when PC $2 that he can no longer afford. The reduced consumption possibilities imply that Tom has effectively less income than he had prior to the price increase. Thus, an increase in the price of clothing has two effects. One is to increase the price of clothing relative to the price of food (i.e., PC PF increases); the other is to reduce Toms effective income (purchasing power). As we will see, both effects inuence Toms response to the price increase. Tom purchased 25 units of clothing and 50 units of food when PC $2, PF 1, and I $100. This choice, which is at the point of tangency between the original indifference curve, I1, and budget line, B1, is labeled as t1 in Figure 2.12. His optimal consumption bundle following the price increase, consisting of 12.5 units of clothing and 50 units of food, is pictured by the point of tangency, t2, between the indifference curve, I2, and the new budget line, B2. The decline in the quantity demanded for clothing from 25 units to 12.5 units represents the total effect of the price change (a decrease of 12.5 units). The total effect can be decomposed into a substitution effect and an income effect. The substitution effect is the change in the quantity demanded of a good when its price changes, holding the prices of other goods and utility constant. To hold utility constant, Tom must be compensated for the price increase by receiving enough additional income to maintain his previous level of utility of 1,250 units (U FC 25 50 1,250). Without this increase in income, he could not afford any of the bundles on the original indifference curve at the new prices. To focus on the effect of changes in relative prices (versus changes in effective income), we examine what Tom would do if he actually 28 For practice, derive the demand function from Annes utility function (for this utility function: MUC and MUF .5F ( .5)). .5C ( .5) BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Figure 2.12 Income and Substitution Effects 49 100 t 70.6 Food This gure illustrates income and substitution effects. Toms original budget line and indifference curve are denoted by B1 and l1; here he chooses 25 units of clothing and 50 units of food as denoted by t1. An increase in the price of clothing from $2 to $4 causes the budget line to rotate inward as pictured by B2. At the new optimum, t2, Tom purchases 12.5 units of clothing and 50 units of food. The 12.5 unit decline in the demand for clothing is the total effect of the price change, which is the sum of the substitution and income effects. The substitution effect is 7.3 units. It is pictured by Toms optimal choice, t , which assumes that Tom has received additional income to keep him on the original indifference curve. The income effect of 5.2 units is the additional decline in demand due to the fact that Tom does not actually receive the hypothetical increase in income. The hypothetical increase is used to isolate the pure price effect from the effect of reduced purchasing power due to the price increase. Economists View of Behavior 35 t1 50 t2 I1 (utility = 1,250 units) I2 (utility = 625 units) B2 B B1 12.5 17.7 25 50 Clothing Substitution effect: 25 17.7 = 7.3 Income effect: 17.7 12.5 = 5.2 Total effect = (7.3 + 5.2) = 12.5 received this hypothetical income increase. This hypothetical situation is depicted by an imaginary budget line, B in Figure 2.12which is parallel to the new budget line, B2, and tangent to the original indifference curve, I1, at t . The resulting combination of 70.6 units of food and 17.7 units of clothing is the least expensive bundle that Tom can purchase at the new prices that yields 1,250 units of utility.29 Since the cost is $141.40, Toms income would have to increase by $41.40 to afford this combination. Thus, if he were to receive enough additional income to compensate for the price increase, he would respond by purchasing 7.3 units less of clothing and 20.6 units more of food than when the price of clothing was only $2. This substitution between clothing and food occurs because clothing is relatively more expensive. Figure 2.12 depicts the 7.3 unit decline in Toms clothing purchases with an arrow labeled substitution effect. The convexity of the indifference curves implies that the substitution effect is positive. The hypothetical $41.40 increase in income is used to isolate the pure effect of the change in relative prices. Since Tom does not actually receive this extra income, he will 29 This consumption bundle is found by solving two equations simultaneously. One equation is for the indifference curve containing bundles that yield 1,250 units of utility (F 1,250 C); the second equation is that the slope of the indifference curve and the new budget line are equal at the point of tangency (F C 4). 36 50 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts not be able to purchase 70.6 units of food and 17.7 units of clothing (his actual income is still $100). The decline in purchasing power from the price increase has an additional effect on Toms demands for clothing and food. The income effect is the change in the quantity demanded of a good because of a change in income, holding prices constant. The parallel shift in the budget constraint from B to B2 in Figure 2.12 captures Toms effective decrease in income. As Toms budget decreases from the hypothetical level of $141.40 to the actual level of $100, he consumes 5.2 fewer units of clothing (17.7 12.5 5.2). The 5.2 unit reduction in quantity demanded is depicted in Figure 2.12 with an arrow labeled income effect. The total effect is that Toms quantity demanded for clothing drops by 12.5 units (25 12.5 12.5) due to the price increase. The total effect, which is the sum of the substitution and income effects, is pictured by an arrow labeled total effect. The nal result is that Tom purchases 12.5 units of clothing and 50 units of food and obtains 625 units of utility (50 12.5). Total Effect 12.5 Substitution Effect 7.3 Income Effect 5.2 The substitution effect is always positivechanges in relative prices motivate substitutions toward the relatively less-expensive good. The income effect for a normal good is also positive. As income decreases (increases) total consumption must decrease (increase); thus, on average the demand for goods must move in the same direction as the income change. Nonetheless, for some goods the income effect is negative. For example, in contrast to a normal good the demand for canned meat products is likely to vary inversely with income (richer people are likely to shun canned meat and purchase fresh meat, such as steak). We call goods with demands that vary inversely with income inferior goods. A positive income effect reinforces the substitution effect and increases the magnitude of the response, while a negative income effect mitigates the substitution effect and reduces the magnitude of the response. In Toms case the 7.3 unit substitution effect is reinforced by the 5.2 unit income effect. For most goods, the income effect is small relative to the substitution effect, and thus the total effect usually is in the same direction as the substitution effect. The income effect in Toms case is relatively large (42 percent of the total effect). This is due to the assumption that Tom can only purchase two goods. Since Tom spends half his budget on clothing, he experiences a large drop in purchasing power when the price of clothing doubles. In contrast to this simple example, most consumers purchase many goods and spend a relatively small percentage of their budgets on any one good (e.g., salt, toothpaste, apples, and so on). Thus, a change in the prices of the typical good does not have an important effect on the purchasing power of the consumer. This observation implies that for many products the substitution effect is much more important than the income effect. For example, suppose that your nearby grocery store raises the price of cucumbers by $1/pound. Conceptually, your income (purchasing power) is lower than it was before since you can no longer purchase as many potential consumption bundles. This small decline in effective income, however, is not likely to be the driving force behind your response to the price change. The relative increase in the price of cucumbers might motivate you to use more tomatoes and fewer cucumbers in your next salad. However, this decision is driven by the change in relative prices of cucumbers and tomatoesnot by the small change in your effective income. Some goods, such as housing or transportation, constitute a relatively large proportion of the typical consumers budget. For these goods, income effects can be more important. Since these goods are the exception rather than the rule, we can effectively ignore income BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Figure 2.13 Income Effects in the Supply of Labor Leisure time (hr) This gure displays Ralph Kramdens choice between work and leisure. Ralph has a total of 100 hours per week that he divides between work and leisure activities (the remaining hours are used for sleeping, etc.). At a wage rate of $10/hr., Ralph works 60 hours/week and has a total income of $600. At a wage rate of $20/hr., he chooses to work fewer hours (40) and to consume more leisure time (60 hours). While the increase in the wage rate increases the opportunity cost of leisure time, the income effect is larger than the substitution effect. At the higher income level ($800), Ralph places greater value on leisure time and works 20 fewer hours than when the wage rate was $10/hr. 100 Economists View of Behavior 37 51 Budget line for wage = $10/hr. 60 Budget line for wage = $20/hr. 40 $600 $1,000 Total income $2,000 effects in many applications. Correspondingly, we tend to concentrate on substitution effects in analyzing the effects of changes in relative prices. The reader, however, should be aware that other applications exist where income effects are important. One prominent case where income effects can be important is the supply of labor. Figure 2.13 depicts Ralph Kramden who is choosing between work and leisure time. Ralph is a bus driver whose employer allows him to choose the number of hours he works each week. Ralph has a total of 100 hours per week that he divides between work and leisure activities (the remaining hours are used for sleeping, etc.). At a wage rate of $10 per hour, Ralph chooses to work 60 hours per week and has a total income of $600; the other 40 hours are used for leisure activities. The budget line rotates outward when the wage rate is increased to $20/hr. The new budget line is atter than the original line (slope is 1 20 versus 1 10). The reduced slope captures the increase in the opportunity cost of leisureleisure now costs Ralph $20/hr. The substitution effect works in the direction of motivating Ralph to reduce his leisure time and to work more hours. The substitution effect in this example, however, is outweighed by the income effect. At the higher wage rate, Ralph chooses to work only 40 hours per week; his total income is $800, which is $200 more than he made working 60 hours at $10/hr. At an income level of $800, Ralph values an additional hour of leisure time at more than the $20 he could make from using the hour for work. At the lower level of income ($600) he placed a smaller value on an extra hour of leisure time (he had to work more hours to provide basic support for his family). Magnitude of the Substitution Effect Economists typically assume that indifference curves are convex to the origin. Convexity is consistent with the behavioral observation that a persons willingness to trade one good for the other generally declines as the relative amount of the second good increases (i.e., the MRS declines as the good on the horizontal axis increases). While Tom is Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts Figure 2.14 Convexity of Indifference Curves Good Y This gure compares the typical indifference curve with the two extremes. The rst extreme is the case of perfect complements where the indifference curve is shaped as a right angle. In this case, the two goods are used in xed proportions. An individual receives no additional utility from receiving more units of just one of the goods. The other extreme case is perfect substitutes, where the indifference curve is a straight line. In this case, the MRS does not change as the person receives more of one good relative to the other. The substitution effect in response to a change in relative prices is larger when the two goods are close substitutes than when they are close complements. Good Y 52 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Good Y 38 Good X Good X Good X Perfect Complements Normal Case Perfect Substitutes willing to trade a large amount of food for a unit of clothing when he has lots of food and little clothing, his willingness to trade food for clothing declines as he reduces his food stock relative to his supply of clothing. The substitution effect is always positive with convex indifference curves. An increase in the relative price of one good motivates substitution away from that good toward other goods (holding utility constant). The magnitude of the substitution effect, however, varies depending on the convexity (curvature) of the indifference curve. Figure 2.14 compares the typical indifference curve with the two extremes. The rst extreme is the case of perfect complements where the indifference curve is shaped as a right angle. In this case, the two goods are used in xed proportions. An individual receives no additional utility from receiving more units of just one of the goods. For utility to increase, the quantity of both goods must increase. An example is shoes. The typical individual requires both a left and right shoe. Utility is not increased if the individual receives a right shoe unless it is matched with a left shoe. The other extreme is perfect substitutes, where the indifference curve is a straight line. In this case, the MRS does not change as the person receives more of one good relative to the other. For example, a persons willingness to trade $10 bills for $20 bills remains at 2 for 1 regardless of the relative supply of the two goods. While most goods are neither perfect complements nor perfect substitutes, the convexity of indifference curves varies among products. Some indifference curves have signicant curvature (tend to be close to right angles), while others are relatively straight. The substitution effect is smaller when the indifference curve is more convex (closer to perfect complements). For example, a small increase in the price of left shoes will not motivate consumers to purchase fewer left shoes and more right shoes. In contrast, a small price change can motivate large shifts from one good to another when they are close substitutes. For instance, a small price change can motivate a consumer to switch from one brand of orange juice to another if the consumer is largely indifferent between the two brands (i.e., they are viewed as close substitutes). Geometrically, as the convexity of an indifference curve increases, BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 39 53 the consumer does not have to move as far from the initial optimum to reach the new optimum as relative prices change (higher convexity implies that the slope of the indifference curve is changing more rapidly along the curve). Additional Considerations Our analysis has focused exclusively on interior solutions where the consumer optimally purchases positive quantities of both goods. This focus is justied because it is the usual case with convex indifference curves. Nevertheless, there are special cases where it is optimal for the consumer to spend the entire income on only one of the goods. This outcome is known as a corner solution. For example, in the case of perfect substitutes it is optimal for the consumer to purchase only one of the goods unless the budget line and straightline indifference curve have the same slopes, in which case the consumer is indifferent between purchasing either of the two goods (and thus might purchase positive quantities of both goods). For instance, you might be relatively indifferent between holding your cash as $5 bills or $10 bills at an exchange rate of two for one. However, you will hold only one type of bill at other exchange rates. If you have the ability to trade one $10 bill in the marketplace for three $5 bills you should clearly do so. Similarly, while you might be indifferent between purchasing two similar brands of orange juice when they sell for identical prices, you will quickly shift to buying only one brand if that one goes on sale. For simplicity we have focused on an example based on only two goods. When there are more than two goods in the marketplace, an increase in the price of one good can motivate a reduction in the demand for other complementary goods. For example, an increase in the price for playing golf can reduce the demand for golf equipment. The substitution effect constitutes movement away from golf-related goods to other goods, which are now relatively less expensive. We examine complementarity and substitutability of products in greater detail in Chapter 4. Calculus Derivation of Equal Marginal Principle The equal marginal principle states that the marginal utility to price ratio is equal for all goods at the consumers optimal consumption bundle. At any combination where this condition does not hold, the consumer can be made better off by making feasible changes in the consumption bundle. This section provides a calculus-based derivation of this principle. The utility function for the two-good case takes the following general form: U f (xi, xj) (2.21) To nd the slope of an indifference curve, we totally differentiate Equation (2.21). We set this differential equal to zero, since utility does not change along an indifference curve: dU [ U xidxi] [ U xjdxj] 0 (2.22) The slope of the indifference curve is dened by dxi dxj (when good i is placed on the y-axis). Thus the slope of the indifference curve ( U xj) ( U xi) MUj MUi (2.23) (2.24) This expression has a straightforward interpretation. For illustration, assume that at some xed combination of xi and xj, the marginal utility of good i is 1 and the marginal utility of good j is 2. This means that 2 units of i can be given up for 1 unit of j and utility will stay the same. This is true by denition, since j has twice the marginal utility of i. 40 54 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Basic Concepts ANALYZING MANAGERIAL DECISIONS: Consumer Choice 1. Dene the following terms: marginal utility, ordinal utility, marginal rate of substitution, equal marginal principle, demand function, substitution effect, income effect, normal good, inferior good, perfect complement, and perfect substitute. 2. Susan Pettits preferences for coffee (by the pound) and doughnuts (by the dozen) can be characterized as follows: MUcoffee MUdoughnuts a. b. c. d. e. f. MUx MUy y2 2xy If the ratio of relative prices is (Px Py) 6 3 2, and Susans income is $90 per period, what combination of pounds of coffee and dozens of doughnuts will she choose? Now let the ratio of coffee to doughnut prices decline to unity ( 1), holding the price of doughnuts constant. How does Susan respond to the reduction in the relative price of coffee? Redo parts (a) and (b) for the case of income of $60 per period. Derive Susans demand function for coffee. Is coffee a normal or inferior good for this consumer? Does Susan consider coffee and doughnuts to be either perfect complements or perfect substitutes? Explain. 3. Susans demand function for coffee in the previous problem includes only the price of coffee and income. Thus, changes in the price of doughnuts do not affect the demand for coffee. Does this imply that there is no substitution effect between the two goods? Explain. 4. (More challenging problem) Mario Casali is a TV newscaster who gets an annual clothing allowance to buy suits that he must wear during his televised forecasts. He allocates the allowance each year between expensive Italian suits and cheap American suits. Marios utility function for suits is SA.5 where S is the number of Italian suits bought and A is the number of American suits bought. Last year, Mario bought two Italian suits and four American suits. [Note: MUS A.5 and MUA .5SA ( .5)] a. If Mario was maximizing his utility last year, what was the ratio of the price of an Italian suit to the price of an American suit (PS PA)? b. What was Marios clothing allowance last year if the price of an Italian suit was $1,000? c. If Mario has the same allowance this year as last year, and American suit prices have not changed, how high would the price of Italian suits have to rise in order for Mario to want to buy exactly one Italian suit this year? At a consumers optimum the slope of the budget line (Pj Pi) is equal to the slope of the indifference curve: MUj MUi Pj Pi (2.25) Rearranging this expression yields the Equal Marginal Principle: MUj Pj MUi Pi (2.26) This principle immediately generalizes to utility functions with more than two goods. Suggested Readings G. Becker (1993), Nobel Lecture: The Economic Way of Looking at Behavior, Journal of Political Economy 101, 385409. M. Jensen and W. Meckling (1994), The Nature of Man, Journal of Applied Corporate Finance 7, 419. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 2 Economists View of Behavior 55 21. Suppose there are only two goods that Bob cares about(1) material welfare and (2) leisure time that he buys from the outside world at $40 per unit and $20 per hour, respectively. He currently lives and works in Atlanta, has a budget totaling $1,000 per week for these two goods, and consumes 11 units of material welfare and 28 hours of leisure time. a. Show Bobs consumption choice on a graph using the actual budget line and a hypothetical indifference curve. (Label the axes and show the Y and X intercepts.) b. How will Bobs optimal consumption choice change if the government imposes a $10/unit tax on the material welfare good? c. Suppose Bob has a chance to move (at zero cost) to Saint Louis where material welfare and leisure time cost $50 and $10, respectively. His budget remains the same as before and the government has decided not to impose a tax on the material good. Would Bob move to Saint Louis? Why or why not? Explain. 22. Amiko is an investor in the stock market. She cares about both the expected value and standard deviation of her investment. Currently she is invested in a security that has an expected value of $25,000 and a standard deviation of $10,000. This places her on an indifference curve with the following formula: Expected Value $15,000 Standard Deviation a. Is Amiko risk-averse? Explain. b. What is Amikos certainty equivalent for her current investment? What does this mean? c. What is the risk premium on Amikos current investment? 23. You have won a free ticket to see an Eric Clapton concert (which has no resale value). Bob Dylan is performing on the same night and is your next best alternative activity. Tickets to see Dylan cost $40. On any given day, you would be willing to pay up to $50 to see Dylan. Assume there are no other costs of seeing either performer. Based on this information, what is the opportunity cost of seeing Eric Clapton? (a) $0, (b) $10, (c) $40, or (d) $50?30 Solutions to Self-Evaluation Problems 21. Individual Choice a. 25 Material welfare units Self-Evaluation Problems 41 The McGrawHill Companies, 2009 2. Economists View of Behavior u1 11 B1 50 28 Leisure time (hours) b. The $10 tax increases the price of material welfare from $40 to $50. Refer to Figure 2.6 to see the effects of a price increase on optimal consumption (replace food with material welfare on the vertical axis and clothing with leisure time on the horizontal axis). Bob is likely to reduce his consumption of material welfare due to the price increase. His consumption of leisure time could either go up, down, or stay the same depending on the exact nature of his adjustment. The change in relative prices (reected in the atter budget line) will work in the direction of motivating Bob to substitute units of material welfare for more leisure time. 30 Suggested by R. Frank (2005), The Opportunity Costs of Economics Education, New York Times (September 1), C2. 42 56 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts However, Bobs new budget line will not allow him to stay on the same indifference curve as before (it rotates inward). His consumption of leisure time will depend on the exact location of the new tangency point on a lower indifference curve in the graph. Bob was spending $560 to purchase 28 hours of leisure time. Whether Bob consumes more or less leisure time after the price increase depends on how much money he has left over after buying fewer units of material welfare at the new higher price. For example, if he only reduces his consumption of material welfare from 11 to 10 units, he will have less money to spend on leisure time than before ($500 versus $560). Alternatively, if he reduces his consumption of material welfare to 8 or fewer units he will have more money than he had before to spend on leisure time. Which option he chooses depends on his specic utility function. c. Bob would want to move to Saint Louis. In Saint Louis, his current consumption bundle only costs $830. This leaves $170 to spend on other goods. Since more is better than less, he can always do better in Saint Louis than he is currently doing in Atlanta. 22. Decision Making Under Uncertainty a. Yes, Amiko is risk-averse. She is willing to take on more risk only if it is associated with a sufciently higher expected return. b. Amikos certainty equivalent is $15,000. She would be willing to accept a certain return of $15,000 (the vertical intercept of her indifference curve) in lieu of her current risky investment that has an expected return of $25,000 and a standard deviation of $10,000. c. The risk premium on Amikos current investment is $10,000. This is the difference in the expected return of her risky investment and the risk-free investment (certainty equivalent). The $10,000 risk premium is what it takes in expected return to make her indifferent between the risk and risk-free investments. 23. Opportunity cost is a subtle concept that requires careful analysis to implement. Even trained economists can make mistakes if they are not careful to include all relevant costs in the analysis. Two researchers from Georgia State University (P. Ferraro and L. Taylor) posed the question to 200 professional economists at an annual meeting. A careful application of the denition of opportunity costs yields a clear answer$10. The next best alternative use of your time, going to the Bob Dylan concert, produces a net benet of $10 (the $50 value you place on the Dylan concert minus the $40 to purchase the ticket). Marginal analysis implies that you should go to the Clapton concert as long as you obtain at least $10 worth of happiness from the concert. For example, if you value the Clapton concert at $15, you are $5 better off going to the Clapton concert than the Dylan concert, which yields only $10 of net value. Interestingly, only 21.6 percent of the professional economists surveyed chose the correct Register to View Answersmaller percentage than if they had chosen randomly. Additional surveys revealed that the incorrect answers were driven by faulty analysis and not by the specic wording of the question. College students who had taken a course in economics did even worse. The lesson is that managers, students, and even economists should be careful to include all of the relevant explicit and implicit opportunity costs in their analyses. Missing a hypothetical question on opportunity costs is inconsequential. Managers can destroy signicant value if they make mistakes in evaluating opportunity costs in their decision making. Review Questions 21. Which costs are pertinent to economic decision making? Which costs are not relevant? 22. A noted economist was asked what he did with his free time. He responded by saying that time is not free. Explain this response. 23. The Solace Company has an inventory of steel that it originally purchased for $20,000. It currently has an offer to sell the steel for $30,000. Should Solaces management agree to sell? Explain. 24. Suppose that you have $900 and want to invest the money for one year. There are three existing options. a. The city of Rochester is selling bonds at $90 per unit. The bonds pay $100 at the end of one year when they mature (no other cash ows). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 43 The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 57 b. Put the money under your mattress. c. The one-year interest rate of saving in the Chase Bank is 7 percent. Which one will you choose? What is the opportunity cost of your choice? Explain. 25. Suppose Juans utility function is given by U FC, where F and C are the two goods available for purchase: food and clothing. a. Graph Juans indifference curves for the following levels of utility: 100, 200, and 300. b. Are these curves convex or concave to the origin? What does this shape imply about Juans willingness to trade food for clothing? c. Suppose Juans budget is $100 and the prices of F and C are both $5. Graph the budget constraint. d. How many units of food and clothing will Juan purchase at these prices and income? Show graphically. What is his corresponding level of utility? e. The Johnson Company is the sole producer of clothing. What can the company do to induce Juan to purchase more clothing? Show graphically. (The graph does not have to be exact.) 26. Suppose that Bobs indifference curves are straight lines (as opposed to being convex to the origin). What does this imply about Bobs willingness to trade one good for the other? Give examples of goods where this type of behavior might be expected? 27. Suppose that Bobs indifference curves are perfectly L-shaped with the right angle occurring when Bob has equal amounts of both goods. What does this imply about Bobs willingness to trade one good for the other? Give examples of goods where this type of behavior might be expected? 28. a. Briey describe the ve models of behavior presented in this chapter. b. What are the implications of these models for managers attempting to inuence their employees behavior? 29. Employees in a plant in Minnesota are observed to be industrious and very productive. Employees in a similar plant in southern California are observed to be lazy and unproductive. Discuss how alternative views of human behavior and motivation might suggest different explanations for this observed behavior. 210. Employees at a department store are observed engaging in the following behavior: (a) they hide items that are on sale from the customers, and (b) they exert little effort in designing merchandise displays. They are also uncooperative with one another. What do you think might be causing this behavior, and what might you do to improve the situation? 211. One of the main tenets of economic analysis is that people act in their own narrow self-interest. Why then do people leave tips in restaurants? If a study were to compare the size of tips earned by servers in restaurants on interstate highways with those in restaurants near residential neighborhoods, what would you expect to nd? Why? 212. Several school districts have attempted to increase teacher productivity by paying teachers based on the scores their students achieve on standardized tests (administered by outside testing agencies). The goal is to produce higher-quality classroom instruction. Do you think that this type of compensation scheme will produce the desired outcome? Explain. 213. A company recently raised the pay of employees by 20 percent. Employee productivity remained the same. The CEO of the company was quoted as saying, It just goes to show that money does not motivate people. Provide a critical evaluation of this statement. 214. One physician who worked for a large health maintenance organization was quoted as saying: One day I was listening to a patients heart and realized there was an abnormal rhythm. My rst thought was that I hoped that I did not have to refer the patient to a specialist. Indeed, HMO physicians have been criticized for not making referrals when they are warranted. How do you think the physician was compensated by the HMO? Explain. 215. Insurance companies have to generate enough revenue to cover their costs and make a normal prototherwise, they will go out of business. This implies that the premiums charged for insurance policies must be greater than the expected payouts to the policyholders. Why would a person ever buy insurance, knowing that the price is greater than the expected payout? 44 58 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 2. Economists View of Behavior The McGrawHill Companies, 2009 Basic Concepts 216. Critically evaluate the following statement: Risk-averse people never take gambles. 217. Suppose that an investment can yield three possible cash ows: $5,000; $1,000; or $0. The probability of each outcome is 1 3. a. What is the expected value and standard deviation of the investment? b. How much would a risk-neutral person be willing to pay for the investment? c. How much would a risk-averse person be willing to pay for the investment? 218. In order to spur consumer spending in 1998, the Japanese government considered an $85 billion voucher system whereby every Japanese consumer would receive a shopping voucher that could be used to purchase Japanese products. For simplicity, assume the following: Each consumer has wealth of 1 million yen, consumers must allocate this wealth between consumption now (c1) and consumption later (c2), the interest rate is zero, the voucher is worth 100,000 yen, and it can be spent only in the current period. If it is not spent, it is lost. a. Plot a budget line for a representative consumer both before and after the voucher program (c1 and c2 are on the axes). b. Do you expect that current consumption of a typical consumer will increase by the full 100,000 yen of the voucher? Explain. c. How does the impact of this 100,000-yen voucher differ from simply giving the individual 100,000 yen? 219. People give to charity. a. Is this action consistent with the economic view of behavior? Explain. b. Suppose there is a big drop in charitable giving. At the same time there has been no decline in per capita income or total employment. Using the economic model, what potential factors might have led to this decline in giving? c. How might the decline in giving be explained by the product-of-the environment model? 220. The Japanese are very good at returning lost property to local police stations. If you lose a wallet lled with cash in Japan it is likely to be turned into the police. This is true even though the person nding it could keep it without anyone else knowing. This behavior is not what you would nd in New York City. a. Does this observation about Japan imply that the economic model does not explain behavior in Japan? Explain. b. Police stations in Japan are lled with lost umbrellas. It used to be that the typical Japanese would make a trip to the local police station to search for a lost umbrella. Now they dont. Explain this behavior using the Economic Model. c. Do you think that the typical Japanese is more likely to come to a police station to nd a lost cell phone or a lost umbrella? Explain using the Economic Model. 221. Some states in the United States allow citizens to carry handguns. Citizens can protect themselves in the case of robberies by using these guns. Other states do not allow citizens to carry handguns. Criminals, however, tend to have handguns in all states. Use economic analysis to predict the effects of handgun laws on the behavior of the typical criminal. In particular: (1) Do you think criminals will commit more or fewer robberies in the states with the laws? (2) How do you think the laws will affect the types of robberies criminals commit? Be sure to explain your economic reasoning. 222. Discuss the following statement: Sunk costs matter. People who pay $20,000 to join a golf club play golf more frequently than people who play on public golf courses. 223. Jenny is an investor in the stock market. She cares about both the expected value and standard deviation of her investment. Currently she is invested in a security that has an expected value of $15,000 and a standard deviation of $5,000. This places her on an indifference curve with the following formula: Expected Value $10,000 Standard Deviation. a. Is Jenny risk averse? Explain. b. What is Jennys certainty equivalent for her current investment? What does this mean? c. What is the risk premium on her current investment? 224. Accounting problems at Enron ultimately led to the collapse of the large accounting rm Arthur Andersen. When the Enron scandal rst became public, Andersens top management BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 2. Economists View of Behavior Chapter 2 Economists View of Behavior 45 59 blamed one rogue partner in the Houston ofce who they claimed was less honest than other partners at the rm. They red the partner and asked that people not hold the remaining partners accountable for one bad apple. What model of behavior was Andersens management using when it analyzed the source of the problem? According to the economic view of behavior, what was the more likely cause of the problem? 225. According to a recent article in The Atlanta Journal-Constitution ( January 29, 2004), materialism, not necessity, gave birth to dual-income families. In supporting the argument, the author cites the following gures from the Department of Commerce: In 1970 the average wage per job was $6,900, which in 2001 dollars (adjusting for ination) amounts to $31,500. In 2001, the average wage per job was $35,500. The main thesis of the article is that dual-income families are a result of a shift in consumer preferences toward consumption as opposed to leisure time/time spent with the family. a. Assume the average person worked 250 days during a year both in 1970 and 2001, and that, as reported in the article, only one person worked in the average family in 1970, while both parents did in 2001. Provide a graphical analysis of the typical familys choice between family income and combined parent leisure time that supports the authors argument, relying on the tools presented in class. Be careful in labeling your graph(s), and provide a clear and concise explanation for your graph(s). Note that there are 365 days in a year so that the total parent leisure time that is possible is 730 days (assuming neither spouse works). Assume it is possible for each family member to work anywhere from 0 to 365 days a year (at the going salary rate) if they choose to do so. b. Assume that in 1971 the average single person worked 220 days per year, while the same person worked 260 days per year in 2001. Moreover, suppose the average daily wage in 2001 dollars was $125 in 1970 and $140 in 2001. Show graphically how the authors argument would not necessarily apply to the average single person (i.e., assume preferences are unchanged). Explain clearly and concisely why the average worker may be choosing to work more in 2001 and carefully label your graph. 46 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge CHAPTER 3 CHAPTER OUTLINE Goals of Economic Systems Property Rights and Exchange in a Market Economy Dimensions of Property Rights Gains from Trade Basics of Supply and Demand The Price Mechanism Linear Supply and Demand Prices as Social Coordinators Measuring the Gains from Trade Government Intervention Externalities and the Coase Theorem Markets versus Central Planning General versus Specic Knowledge Knowledge Creation Specic Knowledge and the Economic System Incentives in Markets Contracting Costs and Existence of Firms Contracting Costs in Markets Contracting Costs within Firms Managerial Decisions Summary Appendix: Shareholder Value and Market Efciency M uch of the worlds economic activity occurs within free markets where individual decisions are coordinated through the price mechanism. For example, the three countries with the largest gross national products (GNP) in 2005 (the United States, Japan, and Germany) all have developed market systems. Other more centrally planned economies, such as China with the fourth largest GNP in 2005, are making increased use of markets. On closer inspection, however, it is evident that a substantial amount if not most of the production in modern economies takes place inside rms, where multimillion dollar resource allocation decisions (e.g., on what to produce and how to produce it) are made by managers without the use of market transactions. The monetary size of the worlds BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 47 61 largest rms exceeds that of many economies. For instance, the 2005 GNPs of Peru, Kenya, and Portugal were $73 billion, $18 billion, and $171 billion, respectively; the 2005 net sales of Exxon Mobil, General Motors, and Wal-Mart were $359 billion, $192 billion, and $258 billion, respectively. To be effective, managers must have a working understanding of both markets and rms. In this chapter, we contribute to this understanding by examining three important questions: How do market systems work? What are the relevant advantages of using market systems compared to central planning in large economies? Why do we observe so much economic activity conducted within rms in market economies? Answers to these questions are particularly important to managers for two reasons. First, an understanding of how markets work helps managers make appropriate strategic and operational decisions (e.g., input, output, and pricing decisions). The supply and demand analysis that we introduce in this chapter is especially useful in many management applications. Second, understanding the relative advantages and disadvantages of markets, central planning, and rms is directly relevant to understanding rm-level issues such as which decision rights to decentralize to employees and whether to make or buy each of the rms inputs. The basic tools and concepts introduced in this chapter are used to analyze these specic management decisions in more depth in subsequent chapters. Goals of Economic Systems Every economic entitybe it a national economy, rm, or householdis confronted with three basic issues: What to produce How to produce it How to allocate the nal output Economic entities can be organized in alternative ways to address these issues. For instance, national economies can rely on either central planning or free markets. Similarly, rms and households can use centralized decision making, where the CEO or head of household makes all major decisions. Alternatively, other people in the rm or household can be granted substantial decision-making authority. Given the alternatives, what is the best way to organize economic activities? To answer this question, we need some criterion for comparing alternative systems. Unfortunately, uniform agreement over such a criterion is unlikely. For instance, you might argue that an ideal system would produce your preferred mix of products and give them all to youalthough your neighbor is certain to disagree. Given these differences in opinion, economists generally focus on a relatively uncontroversial but narrow criterion for comparing the effectiveness of economic systems: Pareto efciency.1 A distribution of resources is said to be Pareto-efcient if there exists no alternative allocation that keeps all individuals at least as well off but makes even one person better off. If an economic system is not producing an efcient allocation of resources, it is possible to make its members better off by adopting Pareto-improving changes (thus beneting some members without hurting others). As an example, assume that the economy produces 1,000 laptop computers and no DVDs. This distribution of resources is not Pareto-efcient if there is a subset of 1 The term is named after Vilfredo Pareto, 18481923, an Italian economist and sociologist. 48 62 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts individuals who prefer to substitute a DVD for a computer and could make this substitution without reducing the utility of other individuals in the society. Based on the concept of Pareto efciency, its citizens would be better off if the economy produced more DVDs and fewer computers. If this action adversely affects even one person, the move would not be Pareto-improving and an economist would have little formal basis to conclude whether the move would be good or bad for society at large.2 Pareto efciency also requires that goods be produced in an efcient manner. The resource allocation would not be Pareto-efcient if production could be rearranged so that more DVDs were produced without lowering computer production (and vice versa). Within centrally planned economies, government ofcials decide what to produce, how to produce it, and who obtains the nal output. In free markets, these decisions are decentralized to individuals within the economy. At least in concept, a central planner could order any feasible production and distribution of goods. Thus, any allocation of resources that could be achieved by a market economy also could be achieved by a centrally planned economyat least in principle. We begin by discussing how market systems work and how they can produce a Pareto-efcient allocation of resources. We then discuss why in large economies a market is more likely to produce an efcient resource allocation than central planning. Property Rights and Exchange in a Market Economy A property right is a legally enforced right to select the uses of an economic good. A property right is private when it is assigned to a specic person. Private property rights are alienable in that they can be transferred (sold or given) to another individual. For example, if Valerie Fong owns an automobile, she can use the automobile as she sees t (within limits set by trafc laws). Valerie can restrict others from using her vehicle. She also can sell the automobile (transfer to another person whatever property rights her ownership confers in the vehicle). The government maintains police and a court system to help enforce these property rights. An important feature of a market economy is the use of private property rights. Owners of land and other resources have the legal rights to decide how to use these resources and frequently trade these rights to other individuals. They are free to start new businesses and to close existing businesses. In contrast, in centrally planned economies, property tends to be owned by the state; government ofcials decide how to use these resources. Dimensions of Property Rights Ownership involves two general dimensions: use rights and alienability rights. These aspects of ownership are not always bundled together. You own your body in the sense that you can decide what activities to pursue. Yet, there are signicant legal restrictions on alienability. For instance, you cannot enter a legally enforceable contract to sell one of your kidneys, despite the fact that you have two, can live comfortably with one, and might value your second kidney much less than a wealthy individual who is dying from 2 Therefore, economics does not address the question of which of the many possible efcient resource allocations is best for a society. Producing your preferred set of products and giving them all to you is efcient (the allocation cannot be changed without making you worse off ). However, others will argue that the allocation is not fair or equitable. Economists have no special training in resolving these fairness or equity issues and thus rarely attempt to settle these types of debates. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 49 63 MANAGERIAL APPLICATIONS Patent for Priceline.com Government-enforced patents better-dene property rights in new inventions. Patents in the United States are awarded for processes, machines, manufacturers, or compositions of matter that are considered useful, novel, and unobvious. Patents protect the intellectual property rights of the inventor and thus protect the common good by providing incentives to innovate novel and unobvious inventions. Priceline.com received a patent for the worlds rst buyer-driven e-commerce system where users can go to the Internet to name their price for goods and services. Expedia.com challenged whether Priceline.coms process is really novel and unobvious. In 2001, the parties settled. Internet businesses where consumers can name their own price have to pay Priceline.com a royalty. This royalty is a tax on all Internet consumers. Awarding a patent for something that is obvious lowers incentives for future innovations that use this process. Source: J. Gurley (1999), The Trouble with Internet Patents, Fortune ( July 19), 118; L. Flynn (2002), The Web World Watches Closely as British Telecommunications Stakes a Patent Claim on a Now-Ubiquitous Function: Hyperlinking, New York Times (March 11). kidney failure. This restriction eliminates the possibility of a free market in kidneys. In some transactions, it is possible to sell use rights while retaining alienability rights. For instance, in a rental contract, the renter obtains the rights to use an apartment, but does not own or have the right to sell the unit. Conversely, the landlord has the right to sell the apartment, but does not have the right to use it while the lease is in force. (Rental, lease, and franchise agreements separate alienability and use rights; we examine these contracts in Chapter 19.) Gains from Trade To understand how a market economy works, we must understand the motives for trading property rights. Why do people buy and sell? The basic answer is to make themselves better off. Within the economic framework, people order their preferences and take actions that maximize their level of satisfaction (utility). Trade takes place because the buyer places a higher value on the item than the seller. The corresponding gains from trade make both parties better offvoluntary trade is mutually advantageous. For example, if Jos Coronas is willing to pay up to $26,000 for a particular automobile and Rochester Motors is MANAGERIAL APPLICATIONS Property Rights Insecurity in Colombia Colombia has a continuing stream of impoverished farmers who are leaving the countryside and migrating to cities. They live in shantytowns that breed crime and violence. Yet Colombia has substantial arable landan area equivalent in size to North Dakota. And only about 20 percent is used for agriculture. These seemingly inconsistent facts are both by-products of Colombias 40-year conict between the government and a paramilitary force that is deeply involved in drugs. The resulting violence induces many to ee. But others are forced off their land or intimidated into selling at bargain-basement prices. Source: J. Forero (2004), Colombias Landed Gentry: Coca Lords and Other Bullies, New York Times ( January 21), A4. 50 64 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS While Animosity between the Governments of Venezuela and the United States Grows, So Does Trade Venezuelan President Hugo Chavez has predicted that capitalism will lead to the destruction of humanity. In turn he has worked hard to redirect his nations trade away from the United States to what he considers more like-minded nations, such as China and Iran. Washington has also taken steps to reduce trade with Venezuela, such as halting American weapon sales to Venezuela. The potential gains from trade between Venezuela and the United States, however, are large. For example, Venezuela is a leading producer of oil that is in high demand in the United States, while U.S. manufacturers produce automobiles and other products that are in high demand in Venezuela. Meanwhile the typical Venezuelan places lower value on many products produced by the like-minded nations, such as Chinese cars. Despite the animosity between the Venezuelan and U.S. governments, trade between the two countries continues to soar. Venezuela is the fourth largest oil supplier to the United States, while non-oil exports to the United States increased 116 percent during the rst three months of 2006. Meanwhile, General Motors and Ford have been striving to meet soaring demand in Venezuela, with automobile sales up over 28 percent between July 2005 and 2006. General Motors, as Venezuelas largest car manufacturer, indicated that it planned to invest $20 million to expand its output in the country by 30 percent, adding 600 new workers. Trading partners are made better off through exchange. There are strong incentives to engage in trade. This example illustrates that these incentives are not easily thwarted by political rhetoric. Trade continues to thrive even though companies and individuals face potential government actions that could affect their trading relationships and corresponding investments. Source: Simon Romero (2006), For Venezuela, as Distaste for U.S. Grows So Does Trade, nytimes.com (August 16). willing to sell the automobile for as little as $20,000, the potential gains from trade are $6,000 ($26,000 $20,000). If the automobile trades at $23,000, both parties are $3,000 better off. Jos gives up $23,000 to buy something that he values at $26,000, while Rochester Motors obtains $23,000 for something it values at only $20,000. At other prices between $20,000 and $26,000, the total gains are still $6,000 but they are not split evenly. For example, at a price of $25,000, Jos gains $1,000 in value, while Rochester Motors gains $5,000.3 From where do these gains from trade come? One source is differences in preferences. The buyer and seller simply may place different values on the traded item. For example, some people value new automobiles more than other people do. Another important source of gains is that the seller may be able to produce the item more cheaply than the buyer and thus has a comparative advantage in its production. In advanced economies, individuals specialize in producing goods where they have a comparative advantage; they then trade to acquire other goods. Specialization greatly enhances the standard of living of a society. Imagine that you had to be completely self-sufcient, making your own clothing, growing your own food, building your own house, and producing your own vehicles for transportation. Your overall standard of living would be dramatically lower than it is living in a modern, specialized economy. 3 Sometimes, individuals regret a trade after the fact. For instance, Jos might be unhappy after he purchased a particular automobile from Rochester Motors. But given the information he had at the time of the transaction, he must have expected it to be advantageous to purchase automobiles or else he would not have done so (at least from Rochester Motors). Joss ability to say no limits the extent to which he can be exploited in any voluntary trade. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Chapter 3 Markets, Organizations, and the Role of Knowledge 51 65 A. Time it takes for Donna and Mario to produce meat and beer Meat (1 lb) Beer (1 quart) 1 hour 6 hours 2 hours 3 hours Donna Mario B. Allocation of time (30 hours/week) and output prior to specialization and trading Meat 18 hours; 18 lbs 18 hours; 3 lbs 21 lbs 12 hours; 6 quarts 12 hours; 4 quarts 10 quarts Meat (lbs) Beer (quarts) 30 0 30 lbs 0 10 10 quarts Meat (lbs) Donna Mario Total production Beer Beer (quarts) 23 7 6 4 C. Production with specialization Donna Mario Total production D. One possible allocation after trading Donna Mario Table 3.1 Comparative Advantage This table provides an example of comparative advantage. Panel A shows how many hours it takes for Donna and Mario to produce 1 pound of meat and 1 quart of beer. Donna and Mario each work 30 hours/week. Panel B shows their allocation of time and resulting output prior to meeting and trading. While Mario is less productive than Donna in an absolute sense for both goods, he has a comparative advantage in making beer (opportunity cost of 1 2 pound of meat for 1 quart of beer compared to Donnas opportunity cost of 2 pounds of meat). Donna has a comparative advantage in producing meat. Panel C illustrates how total production can be increased by having both people specialize in the activity where they have a comparative advantage. Panel D displays a possible nal allocation after Donna and Mario trade. Specializing and trading produce real gains for both people. Table 3.1 presents a numerical example of comparative advantage. Donna Meyers and Mario Santini each produce their own food and drink through hunting and brewing beer. Panel A shows how many hours it takes for them to produce 1 pound of meat and 1 quart of beer. Panel B shows their allocation of time and resulting output working independently prior to their meeting and trading. Both work 30 hours/week. Donna spends 18 hours/week hunting and 12 hours/week making beer, producing a total of 18 pounds of meat and 6 quarts of beer. Mario spends 18 hours hunting and 12 hours making beer, producing a total of 3 pounds of meat and 4 quarts of beer. Their total production prior to meeting is 21 pounds of meat (18 3) and 10 quarts of beer (6 4). Donna has an absolute advantage over Mario in making both goodsit takes her fewer hours to produce either a pound of meat or a quart of beer. Mario, however, has a comparative advantage (lower opportunity cost) for producing beer. Marios opportunity 52 66 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Strategic Business Planning: Ignoring Economics of Trade During the 1970s, many rms adopted a particular form of strategic business planning. The idea behind this process is to treat the projects of a rm like stocks in a portfolio. Through systematic analysis, winners are to be kept and losers sold. Specically, all projects of the rm are ranked based on growth potential and market share. Projects with high growth potential and high market share are called stars, while projects with low growth potential and market share are referred to as dogs. Dogs are sold, while stars are kept. Funding for the stars comes from cash cows, projects with high market share and low growth potential. Thus, money is invested in the winners to enhance the rms competitive advantage. Although the idea might sound intriguing, its underpinnings are inconsistent with the basic economics of tradesell if, and only if, you can get a price that exceeds the value of keeping the item yourself. This principle implies that, contrary to the process, dogs should be kept unless they can be sold at sufciently high prices. Moreover, stars should be sold if the price is sufciently high. By the 1980s, many rms found that violating the basic economics of trade had led them to accumulate suboptimal collections of projects. Large increases in stock prices were observed as these rms reshufed plants, divisions, and subsidiaries through sell-offs, spin-offs, and divestitures. Source: The New Breed of Strategic Planner (1984), BusinessWeek (September 17), 6268. cost for producing 1 quart of beer is 12 pound of meat (he could have produced 12 pound of meat with the 3 hours he uses to produce a quart of beer), while Donnas opportunity cost is 2 pounds of meat. Conversely, Donna has a comparative advantage in hunting. Donnas opportunity cost for producing 1 pound of meat is 12 quart of beer, while Mario forgoes 2 quarts of beer to produce a pound of meat. Panel C illustrates how total production can be increased by having each person specialize in producing the product for which they have a comparative advantage. Donna can produce 30 pounds of meat by spending all 30 hours on hunting, while Mario can produce 10 quarts of beer by focusing exclusively on beer production. This specialization maintains total beer production at 10 quarts and increases the production of meat by 9 pounds.4 By specializing and trading, both parties can be made better offthere are gains from trade. The nal allocation depends on the specic bargain reached by Donna and Mario. One possible outcome is presented in Panel D, where both parties have the same amount of beer as before but more meat. Specializing and trading results in a Pareto improvement relative to working in isolation. While it is possible to have an absolute advantage in producing all goods, it is impossible to have a comparative advantage in all activities.5 Specialization and trading are common features in economies throughout the world. Comparative advantage also 4 In this example, Donna and Mario completely specialize and produce only one product. More generally, at least one of the two people will specialize in producing one product. The other person might allocate some time to producing the same product (the one for which he does not have a comparative advantage) if additional gains are derived from producing more of the product than can be produced by the rst person. 5 Note for the mathematically inclined: The opportunity cost for producing a product is the ratio of the time it takes to make that product to the time it takes to make the other product. For example, Donnas opportunity cost for producing 1 pound of meat is 12 quart of beer (1 hr 2 hr). The reciprocal of this ratio, 2 (2hr 1hr) is Donnas opportunity cost for producing beer expressed in pounds of meat. If Donnas ratio is smaller than Marios ratio for one product, the reciprocal of Marios ratio must be smaller than the reciprocal of Donnas ratio. Thus Donna has a comparative advantage in producing the rst product, while Mario has a comparative advantage in producing the second product. It is a mathematical impossibility for one person to have a comparative advantage in producing all products. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 53 67 ACADEMIC APPLICATIONS Gains from Trade In 1880, the United States was about to become the worlds most efcient economy. Yet labor productivity varied substantially among states. North Carolina, the least productive state, was only 18 percent of Nevada, the most productive. (In 1880, Nevadas productivity was high because many had migrated there to work in the mines.) In 2002, New Mexico, the poorest state, had a per capita income that was almost 60 percent of Connecticut, the richest state. As a giant free-trade zone, incomes in the United States have converged to similar standards. Although there are still differences, those differences have fallen substantiallyand not at the expense of the rich states. Source: V. Postrel (2004), A Case Study in Free Trade: American Incomes Converge, but Not at the Bottom, New York Times (February 24), C2. arises in many management situations. For example, while a top-level manager might be able to perform many activities more effectively and in less time than a lower-level employee, the manager should not try to do all activities himself (make sales calls, work on the manufacturing line, type papers, answer phones, and so on). More value will be created if managers concentrate on activities for which they have a comparative advantage. A common misconception is that trade takes place because people have too much of some goodspeople sell to others what they cannot use themselves. This view, however, does not explain why individuals sell houses, cars, jewelry, land, and other resources that they value highly and have in short supply. The economic explanation for trade argues that trade takes place not because people have too little or too much of a good. Rather, trade takes place because a person is willing to pay a higher price for a good than it is worth to its current owner. While you might love your new sports car, you would still sell it if someone offered you a high-enough price. And winning bidders of collectibles auctioned on eBay are frequently individuals with collections of related items. ANALYZING MANAGERIAL DECISIONS: Comparative Advantage in the Workplace You are a manager of a division of a company that is responsible for the nal assembly of two computer products, modems and keyboards. You manage two employees, Julio and Chenyu, who each work 8 hours per day. Currently you have assigned both Julio and Chenyu to spend the rst 7 hours of the day assembling keyboards and the last hour assembling modems. Julio can assemble 2 modems per hour and 14 keyboards per hour. Chenyu is more highly skilled in both activities. She averages 3 modems per hour and 15 keyboards per hour. 1. How many modems and keyboards are being assembled under the current work assignments? 2. What are Julios opportunity costs for assembling modems and keyboards? What are Chenyus? Does either employee have a comparative advantage in assembling one of the products? 3. Devise a way of reassigning the work activities between the two employees that keeps the number of modems being assembled the same as before but increases the number of keyboards. 4. What are potential reasons why you might not want to change the work assignments (assume that more assembly of either or both products is desirable)? 54 68 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Basic Concepts It is important to recognize that trade is an important form of value creation. Trading produces value that makes individuals better off. Gains from trade also provide important incentives to move resources to more productive uses. If George Nichols can make the most productive use of a piece of land, he will be willing to pay a higher price for the land than other potential users. The current owner, Jody Crowe, has the incentive to sell the land to George, because she gets to keep the proceeds from the sale. It is these incentives that help to promote a Pareto-efcient allocation of resources in a market economy. After all mutually advantageous trades are completed, it is impossible to change this allocation without making someone worse off. Basics of Supply and Demand Gains from trade explain why individuals buy and sell. But what coordinates the separate decisions of millions of individuals in a market economy to prevent chaos? Why are there not massive surpluses of some goods and huge shortages of other goods? What restricts the amounts demanded by the public to the amounts supplied? Answers to these questions come from an understanding of the market price system. The Price Mechanism The basic economics of a price system can be illustrated through standard supply-anddemand diagrams. Figure 3.1 displays a supply-and-demand diagram for a particular model of personal computerfor example, a Pentium dual-core machine with standard quality and features. The vertical axis on the graph shows the price for a PC, and the horizontal axis shows the total quantity of PCs demanded and supplied in the market for the period (for example, a month). Figure 3.1 Supply and Demand in the PC Industry Supply Price (in dollars) The demand curve shows the number of PCs that consumers want to purchase at each price. The supply curve shows the number of PCs that producers want to sell at each price. Equilibrium occurs where the two curves intersect. Here, the quantity supplied equals the quantity demanded. If the price is above the market-clearing price of P*, say at PHI, there is a surplus of PCs. Producers supply more PCs than consumers want to purchase, and inventories build. If the price is below the market-clearing price, say at PLO, there is a shortage. Producers supply fewer PCs than consumers want to purchase and inventories shrink. Surpluses and shortages put pressure on prices and quantities to move to equilibrium levels of P*. $ P HI Surplus P* P LO Shortage Demand Q Q* Quantity of PCs BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 55 69 MANAGERIAL APPLICATIONS Shifts in Demand, Quantity, and Price at the PGA Tournament The PGA Tournament features competition among the worlds top golfers. In 2003, the PGA was held at Oak Hill Country Club in Rochester, New York. The event attracted over 30,000 spectators a day. Many of these spectators were from outside the Rochester area. A signicant number of these visitors were avid golfers who wanted to play while they were in Rochester. Rochester has several courses that are open to the public. However, many courses in the area are private (only members and their guests can play). Facing this dramatic temporary increase in the demand for public golf courses, several of the private courses decided to become public during the week of the PGA. These courses charged high fees ranging from $100 to $250 per round (their normal guest fees were approximately $50). This example highlights that shifts in demand motivate increases in the quantity supplied and the price of a product (in this case, golf times). The market includes all potential buyers and sellers of this type of PC. Suppose that in this market there are many buyers and sellers and that individual transactions are so small in relation to the overall market that the price is unaffected by any single sale or purchase. In this case, no buyer or seller has market power: All trades are made at the going market price. We label this type of market as competitive. (In Chapter 6, we extend our analysis of competitive markets; we also examine noncompetitive market structures.) The demand curve depicts how many total PCs consumers are willing to buy at each price (holding all other factors constant). The demand curve slopes downward because consumers typically buy more if the price is lower. For example, consumers are likely to buy more PCs if the price is P LO (say, $900) than if the price is P HI (say, $1,500). The supply curve depicts how many PCs producers are willing to sell at each price (holding all other factors constant). The curve slopes upward: At higher prices, producers are able and willing to produce and sell more units. For example, at a price of $900, many potential producers cannot cover their costs, and thus they refrain from entering production. At a price of $1,500, more units are manufactured and brought to market. The two curves cross at the market-clearing price P * and quantity Q *. At the marketclearing price, the quantity of PCs demanded exactly equals the quantity supplied. Here, at a price of $1,200, the market is said to be in equilibrium. There are strong pressures within markets that push prices and quantities toward their equilibrium levels. To see why, suppose that the market price is above the equilibrium price, such as P HI in Figure 3.1. At this higher price, there is a surplus of PCssuppliers produce more PCs than consumers are willing to purchase. As inventories of unsold PCs build, this surplus places downward pressure on prices as suppliers compete to try to sell their products. As prices fall, fewer PCs will be produced and more will be demanded, thus reducing the surplus. In contrast, if the price is below the market-clearing price, such as P LO in Figure 3.1, inventories dwindle and back orders accumulatethere is a shortage of computers. Here, consumers will bid up the price of PCs as they compete for the limited supply. As prices rise, producers increase their output and consumers demand fewer PCs, thus reducing the shortage. When the market is in equilibrium, there is no pressure on prices and quantitiesthe quantity demanded exactly equals the quantity supplied. Inventories are stable at their desired levels, and the market price is stable at this point. Supply-and-demand diagrams like that in Figure 3.1 are snapshots at a point in time. As time passes, both the supply and the demand curves are likely to change. Figure 3.2 (page 70) shows the effects of a shift in the demand curve in the PC market. The left panel depicts an increase in demand. Here, there is a shift in the demand curve to the right, since 70 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Basic Concepts Figure 3.2 The Effects of a Shift in Demand on the Equilibrium Price and Quantity of PCs The initial equilibrium is where the demand curve, labeled D0, intersects the supply curve, labeled S0. The left panel shows the effects of an increase in demand. The result is a higher equilibrium price and quantity. The right panel shows the effects of a decrease in demand. The result is a lower equilibrium price and quantity. $ $ Initial demand S0 Price (in dollars) 56 S0 * P1 * P0 * P0 Initial demand P* 2 D1 D0 * Q0 * Q1 Q D2 D0 * * Q2 Q0 Quantity of PCs Quantity of PCs Increase in demand Q Decrease in demand at each price, consumers demand more PCs. Demand for PCs might increase for a variety of reasons, including an increase in the purchasing power of consumers or a decline in the prices of supporting software. These types of changes motivate consumers to purchase more PCs at any given price. After the demand shift at the old equilibrium price, inventories shrink and there is a shortage of PCs. This shortage places upward pressure on prices; higher prices in turn stimulate more production. The end result is a higher equilibrium price and quantity. The right panel shows that the opposite effect occurs with a reduction in demand. This shift to the left in the demand curve also can be caused by a variety of factors (for example, a recession that causes businesses to reduce their purchases of PCs or an increase in personal tax rates that reduces consumers purchasing power). Figure 3.3 depicts the effects of a shift in supply in the PC market. The left panel displays a shift in the supply curve to the right. A rightward shift implies an increase in supply, because at each price producers make and offer more PCs. Many factors might cause an increase in supply. For example, a decline in the prices of labor and other inputs used for manufacturing PCs will make PC production more protable and increase supply. Supply also might increase because of changes in technology that allow for less expensive, more efcient production. After the supply shift at the old equilibrium price, inventories accumulate and there is a surplus of PCs. This surplus places downward pressure on prices, which in turn stimulates more demand. The end result is a lower equilibrium price and higher equilibrium quantity. The right panel shows that the opposite effect occurs when supply shifts to the left. Linear Supply and Demand Throughout this book we use linear demand and supply curves (as pictured in Figures 3.1 to 3.3). Linearity simplies the analysis and is often a reasonable approximation in actual applications (at least over the range of actions being considered). This section provides a numerical example of supply and demand analysis using linear supply and demand curves. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 3 Figure 3.3 The Effects of a Shift in Supply on the Equilibrium Price and Quantity of PCs Markets, Organizations, and the Role of Knowledge $ 71 $ Initial supply S2 S0 S0 S1 Price (in dollars) The initial equilibrium is where the demand curve, labeled D0, intersects the supply curve, labeled S0. The left panel shows the effects of an increase in supply. The result is a lower equilibrium price and an increase in equilibrium quantity. The right panel shows the effects of a decrease in supply. The result is a higher equilibrium price and a lower equilibrium quantity. 57 The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge P2 * P0 * P0 * Initial supply P1 * D0 D0 Q Q Q0 Q1 ** Q2 Q0 ** Quantity of PCs Quantity of PCs Increase in supply Decrease in supply Suppose that the supply function for apples is: Qs 30 0.2Pa 3W (3.1) where Qs is the quantity supplied of apples in millions of pounds, Pa is the market price for apples in cents, and W is the hourly wage rate paid to agricultural workers. The supply function indicates that farmers will produce more apples as either the market price of ANALYZING MANAGERIAL DECISIONS: Ethanol and Pork Prices Over the past few years the federal government has taken signicant steps to encourage the development of ethanol and other fuels made from plants as a partial replacement for gasoline. These actions have been undertaken by politicians in the midst of public concerns about the dependence on foreign oil, war in the Middle East, and global warming. The primary input for ethanol production is corn. In 2006, the 5 billion gallons of ethanol produced in the United States consumed about 20 percent of the domestic corn supply. Under new legislation passed in late 2007, the production of corn ethanol would reach 15 billion gallons by 2015 and 21 billion gallons by 2022. You manage the Hog Heaven restaurant chain. Your restaurant chain, which has about 300 outlets throughout the United States, specializes in barbecue pork dishes but also offers chicken, beef, and vegetarian meals. Currently about 80 percent of your revenue comes from your pork dishes. The price of pork has a major impact on your costs. You are concerned that the federal promotion of ethanol might have an impact on pork prices and the protability of your restaurant chain. Feed cost is typically about 5060 percent of the total cost of production of pork producers. About 80 percent of the feed that hogs consume is corn. 1. Use basic supply and demand analysis to illustrate the likely effect of the governments mandated increase of ethanol production on (1) corn prices and (2) pork prices. 2. What actions might you consider given the results of your analysis? 58 72 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Basic Concepts apples rises or the wage rate for workers falls. Farmers produce more apples when the wage rate falls since production costs are lower. Currently the wage rate is $10. Substituting this value in Equation (3.1) and solving for Pa produces the current supply curve:6 Pa 5Qs (3.2) Supply curves show the relation between price and quantity supplied holding all other factors constant (in this case the wage rate for agricultural workers). Suppose that the demand function for apples is: Qd 20 1 3Pa 0.002I (3.3) where I per capita income. The demand function indicates that consumers will purchase more apples as the price falls and/or as income increases. Currently income is $10,000. Substituting this value into the demand function and solving for Pa produces the current demand curve: Pa 120 3Qd (3.4) In equilibrium, the quantity supplied equals the quantity demanded: Qs Qd Q * where Q * denotes the equilibrium quantity. Substituting Q * into the supply and demand curves (Equations [3.2] and [3.4]) and setting them equal (since there is one equilibrium price) allows us to nd the equilibrium quantity, Q *: 5Q * Q* 120 15 3Q * (3.5) The equilibrium price of 75 cents is found by substituting the equilibrium quantity of 15 into either the demand or supply curve equations ([3.2] or [3.4]).7 Note that changes in the wage rate shift the supply curve, while changes in income shift the demand curve. See if you can nd the new equilibrium price and quantity if income increases to $20,000. Answer the problem before looking in this footnote for the answer.8 Prices as Social Coordinators The equilibrium of supply and demand highlights the crucial role that prices play in coordinating the consumption and production decisions of individuals. For example, if too few PCs are being produced, inventories will shrink and dealers will raise prices. High prices signal would-be producers to shift from producing lower-valued products to producing computers. Because property rights are private, individuals reap the reward from redirecting their efforts and therefore have strong incentives to shift production. Higher prices also motivate consumers to reduce the quantity of PCs demanded. The end result is that the quantity demanded equals the quantity supplied. This is what Adam Smith referred to as the invisible hand. 6 Recall that when graphing the supply and demand curves, the convention is to place price on the vertical axis. We could have found the equilibrium price by setting the demand and supply functions (Equations [3.1] and [3.3]) equal after substituting for the current values of W and I. The equilibrium quantity then could be found by substituting the equilibrium price into either (3.1) or (3.3). We took the extra steps of solving for the demand and supply curves to illustrate how they are derived from the underlying demand and supply functions. We elaborate on this derivation in the case of the demand curve in the next chapter. 8 An increase in income in this example shifts the demand curve to the right, resulting in both a higher equilibrium price and quantity (see Figure 3.2). More specically, shifting the income from $10,000 to $20,000 results in a equilibrium quantity of 22.5 million pounds and an equilibrium price of $1.125 (112.5 cents). 7 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 59 73 MANAGERIAL APPLICATIONS Supply of Online Rsums Bogs Down Employers The Internet has reduced signicantly the cost of submitting rsums to would-be employers. Job seekers no longer must print their rsums on high-quality paper, address, stamp, and mail an envelope. A click of the mouse and the rsum is gone. Some companies have thousands of rsums dumped into their e-mail boxes each day. During 1999 there were almost 5 million rsums on the Internet200 times more than in 1994. When the cost of a good (like submitting a rsum) falls, the quantity supplied increases. Source: S. Armour (1999), Online Rsums Bogging Down Employers, Democrat and Chronicle ( July 19), 1F. If everyone trades in the marketplace and all mutually advantageous trades are completed, the price system results in a Pareto-efcient resource allocation.9 No government intervention or central planning is required. Rather, consumers and producers, acting in their own self-interest, react to price signals in a manner that produces an efcient resource allocation. Prices act to control and coordinate the many individual decisions made in the economy. After trading is completed, the output mix and nal distribution of products cannot be changed without making someone worse off. Also, suppliers engage in efcient production. The basic logic for efciency in a competitive economy is straightforward. At equilibrium prices, the quantity supplied equals the quantity demanded for all goods and there are no shortages or surpluses. Everyone who wants to make trades has done so, and all gains from trade have been exhausted. In making supply decisions, rms have strong private incentives to adopt the most efcient production methods and the value-maximizing output mix (these production choices maximize their prots). No changes in either production or distribution can be made without making someone worse off. Measuring the Gains from Trade In some applications it is useful to have measures of the gains from trade that are in units, such as dollars, that are independent of individuals subjective utilities. Consumer surplus and producer surplus display this property and are commonly used to measure the gains from trade and the effects of specic actions and events on consumers, producers, and society as a whole. Figure 3.4 (page 74) displays supply and demand curves for a market in which the equilibrium price and quantity are $10 and 10 units. The demand curve indicates that some consumer would be willing to pay $19 to obtain the rst unit of the product; however, the consumer only has to pay the market price of $10. The $9 surplus is a measure of the consumers gains from trade from purchasing the rst unit. The sum of the surplus at all points along the demand curve up to the equilibrium quantity of 10 units represents the aggregate difference between what consumers would be willing to pay for the product and what they have to pay given the market price. We call this difference, consumer surplus and display it graphically in Figure 3.4 by Triangle A. In this example, the consumer surplus is $50 (1 2 10 10). Note that consumer would be willing to 9 These conditions will be met in a competitive market when trading costs are sufciently low. Later, we will discuss factors that can motivate inefciency in a market economy. 60 74 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 Figure 3.4 Surplus I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Basic Concepts Consumer and Producer In this gure, consumers pay $100 to obtain 10 units of the product ($10 10 units). They would have been willing to pay an additional $50 to obtain the 10 units. This $50 (Triangle A) represents the gains from trade to consumers and is called consumer surplus. Producers receive $100 of revenue, but would have been willing to supply the product if they had covered their incremental production costs of $50 (Triangle C). The extra $50 (Triangle B) is the gains from trade to producers and is called producer surplus. P $20 Consumer surplus (Triangle A) Supply A Producer surplus (Triangle B) $10 B Incremental production costs (Triangle C) C Demand 10 Q pay up to $150 dollars to obtain the 10 units (Triangle A Triangle B Triangle C), but only have to pay $100 (Triangle B Triangle C). Their gains from trade from participating in this market are $50. The same idea is used for measuring the net gains to producers. Later in this book, we show that the area under the supply curve represents the incremental costs that producers incur to produce the output. In this example, the incremental costs are $50 (Triangle C). Producers are willing to supply 10 units as long as the incremental revenue is at least equal to the incremental costs of $50. Producers, however, receive $100 (Triangle B Triangle C). The extra $50 (Triangle B) represents the gains to trade for producers and is called producer surplus. The total gains from trade produced in the market are measured by the sum of consumer and producer surplus. In this example, the total gains from trade (surplus) are $100 (Triangle A Triangle B). We use the concepts of producer, consumer, and total surplus in several places in this book to measure the effects of various actions on consumers, producers, and social welfare. Government Intervention We have discussed how a well-functioning price system can produce an efcient allocation of resources without government intervention or central planning. Nonetheless, governments sometimes intervene to establish caps or oors on prices. This section examines the economic effects of these actions (in an otherwise well-functioning market). Price Controls The average retail price for gasoline in the United States was about $1.15 per gallon in 2003. Gasoline prices increased dramatically over the period 20042008 due to factors such as the increased demand for oil in China and India, the war in Iraq, and Hurricane Katrina. Gasoline prices, which averaged $1.60 per gallon in 2004, broke the $2.00 per gallon barrier in July of 2005. In August 2005, gasoline supply was disrupted by BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 3 Figure 3.5 Economic Effects of a Government Price Cap on Gasoline The free market equilibrium price in this gure is $3 per gallon. The government does not allow stations to charge more than $2 per gallon. At the $2 price, the quantity demanded is greater than the quantity supplied there is a shortage of gasoline. The excess demand implies that gasoline must be allocated through nonprice mechanisms, such as waiting in line. Triangles A and B represent the lost gains from trade to consumers and producers (consumer and producer surpluses), respectively, induced by the price cap. Rectangle C represents a transfer of surplus from producers to consumers (ignoring other costs imposed on consumers). The price cap can also distort incentives of consumers to reduce consumption of gasoline, for example, by moving closer to work or buying smaller automobiles. 61 The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Markets, Organizations, and the Role of Knowledge 75 $/Gallon Supply Lost gains from trade A $3.00 B C Excess demand (shortage) for gasoline $2.00 $2.00 price Demand QS QD Q Katrina, the devastating storm that hit the Gulf Coast. Gasoline prices surged to over $3.00 per gallon. In subsequent months gasoline prices were volatile, rising from a low of $2.10 per gallon in November 2005 to a high of $3.20 per gallon in May 2007. In 2008 the price was above $4.00 per gallon. Consumers expressed fear and outrage over the high gasoline prices. Some groups asked the U.S. government to implement price controls to protect consumers from unfair gasoline prices. Figure 3.5 displays the economic effects of a cap on the price of gasoline. The free market equilibrium price is $3.00 per gallon. However, suppose that the government passes a law that does not allow stations to charge more than $2.00 per gallon. At the $2.00 price, the quantity demanded, QD, is greater than the quantity supplied, QS there is a shortage of gasoline. The excess demand implies that gasoline must be allocated through nonprice mechanisms. One mechanism is to serve customers in the order that they arrive until the supply is exhausted. This mechanism is likely to produce long lines and waits for gasoline. Also, customers who place the highest value on the gasoline (e.g., due to the importance of their travel) do not necessarily receive the product. The quantity supplied falls as a result of the price cap, resulting in lost gains from trade (total surplus). The reduction in consumer surplus and producer surplus is pictured by Triangles A and B, respectively. The consumers who obtain the gasoline for $2.00 potentially benet from the cap at the expense of the gasoline dealers, the gasoline distributors, the gasoline reners, and the individuals who own the mineral rights and receive a lower price for the quantity sold (unless the gains are offset by costs such as having to wait in line). Rectangle C pictures the transfer from producers to consumers (ignoring other costs). 76 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Basic Concepts Figure 3.6 Economic Effects of Minimum Wage Laws This gure displays the supply and demand for unskilled labor. The free market equilibrium is a $4 wage rate and Q* people being hired. The government has imposed a minimum wage of $5.85 that results in an excess supply of labor (unemployment). The QD people who are employed at the minimum wage of $5.85 benet from the law (unless they incur offsetting costs to obtain and keep their jobs). Rectangle C represents the transfer of surplus from employers to employed workers. In contrast, the people between QD and Q* are hurt by the law. These people are willing to work for as little as $4.00 per hour and would obtain jobs in a free market. However, they are unemployed due to the minimum wage law. Overall there is a reduction in the total gains from trade in the labor market. The lost surpluses from reduced trade for rms and labor are pictured by Triangles A and B, respectively. $ Unemployment (excess supply of labor) Labor supply Minimum wage $5.85 Wage ($/hr.) 62 C A $4.00 Lost gains from trade (surplus) B Labor demand QD Q* QS Q Quantity of labor During the 1970s, there was a severe shortage of gasoline in the United States due to an oil embargo and price controls. People still remember the long gas lines and substantial inconveniences experienced during that period. Some gas stations served customers based on their license plate numbersodd numbers one day and even numbers the next. Customers also were limited in the number of gallons that they could purchase. These inconveniences have rarely been observed since the elimination of price controls in the early 1980s. Consumers responded to higher gasoline prices during the 20042007 period in a variety of ways. Some reduced their travel plans, while others shifted to less expensive forms of transportation (e.g., carpooling, buses, and bikes). Some moved closer to work, while others purchased more fuel-efcient automobiles. Price controls not only produce shortages, but also distort incentives to take actions that reduce the consumption of the product. Price Floors A prominent example of a price oor is a minimum wage law. In the United States, the Fair Labor Standards Act (FLSA) requires employers to pay employees at least a minimum wage (currently $5.85 per hour) for all hours they work. If a state has a minimum wage that is higher than the federal minimum, employers are obligated to pay the higher rate to employees working in that state. For example, the minimum wage in New York State was raised to $7.15 per hour on January 1, 2007. Figure 3.6 displays the economic effects of a minimum wage law. The market clearing price for unskilled labor in this example is $4.00 per hour. However, employers are BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 63 77 ACADEMIC APPLICATIONS Labor Unions and Minimum Wage Laws The Federal minimum wage was increased to $5.15 per hour in 1997 during the Clinton Administration. Labor unions provided strong political support for this action. In signing the legislation, President Clinton stated: There are very few unions in America that have minimum wage workers. Most of these unions supported the minimum wage law because they thought it was the right thing to do. They spent their time and their money and their energy trying to help other people who do not belong to their organization, and I thank you for that. President Clintons statement is consistent with the Good Citizen Model discussed in Chapter 2unions spent time and money, not because it beneted them, but because it was the right thing to do. The Economic Model suggests another possibility. Higher wages for low-skilled workers increase the demand for higher-skilled union workers (especially union workers at the low end of the wage scale). For example, rms are more likely to contract with unionized companies for services as the cost differential between unionized and nonunionized labor declines. Labor unions have also exerted signicant political pressure to limit free trade with countries, such as China. Labor union leaders argue that these actions are motivated by concerns about human rights and the poor treatment of workers in developing countries. Another possibility is that they do not want to compete with a large supply of low-paid workers from other countries. One concern is that rms will shift work away from high-price unionized labor in the United States to offshore companies with low-paid workers. not allowed to pay wages below $5.85 per hour. At $5.85, the quantity supplied of labor, QS, is greater than the quantity demanded, QD there is unemployment. Firms would hire more workers at $4.00 and fewer people would enter the labor market; at $4.00, the quantity supplied equals the quantity demanded, Q *, and there is no unemployment.10 The QD people who are employed at the minimum wage of $5.85 benet from the law at the expense of their employers who have to pay higher wages (unless the employees incur offsetting costs to obtain and keep their jobs). Rectangle C pictures the transfer of surplus from employers to employed workers. In contrast, the people between QD and Q * are hurt by the law. These people are willing to work for as little as $4.00 per hour and would obtain jobs in a free market. However, they are unemployed due to the minimum wage law. Overall, there is a reduction in the total gains from trade in the labor market. The lost surpluses for rms and labor are pictured by Triangles A and B, respectively. These triangles depict the lost surplus from not allowing companies and unemployed workers to enter into mutually benecial employment relations below the minimum wage. In November 2007, the unemployment rate among teenagers in the United States was 16.3 percent compared to 4.1 percent for adults. One potential reason for the high unemployment rate among teenagers is the minimum wage.11 10 In reality, measured unemployment would not be zero absent minimum wage regulation. There are always going to be individuals changing jobs or searching for better jobs. We abstract from these considerations in this example. 11 The U.S. government has tried to reduce the effects of the minimum wage law on teenagers by exempting them from the law for their rst 90 days of employment. During this period, teenagers must be paid at least $4.25 per hour. Note that the amount of unemployment caused by the minimum wage laws depends on the slopes of the supply and demand curves (see Figure 3.6). Although most economists would agree with the direction of the effect, there is disagreement as to its magnitude. Most economists believe this effect is large, but some disagree. For example, See D. Card and A. Krueger (1995), Myth and Measurement: The New Economics of the Minimum Wage (Princeton University Press: New Jersey). 64 78 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts Minimum wage laws can also affect how people are paid. For example, suppose that prior to a new minimum wage law unskilled workers are paid $4.00 per hour plus health benets. One likely affect of forcing employers to pay higher cash wages is a reduction in health benets. Employees who prefer the health benets to higher cash wages are worse off after the law. We examine the composition of the pay package in more detail in Chapter 12. Externalities and the Coase Theorem12 Externalities exist when the actions of one party affect the well-being or production possibilities of another party outside an exchange relationship. Externalities can prevent a free market from being efcient. If a rm emits pollution into the air, it can adversely affect the welfare of the rms neighbors. If the rm does not bear these costs, it is likely to select an inefcient level of pollution (that is, to overpollute). In choosing how much to invest in pollution control equipment, the rms managers will consider only its own costs and benets. Efcient investment would require them also to consider costs and benets imposed on neighbors (the efcient level of investment is where the total marginal costs of additional investment equal the total marginal benetsnot just those incurred privately by the rm). Economists used to think that externalities surely would prevent a market system from producing an efcient allocation of resources. Government intervention seemed to be required to enhance efciency. For example, the traditional recommendation was to ACADEMIC APPLICATIONS Property Rights in Russia An exchange transaction is an agreement among individuals on property rights to goods. Exchange is limited dramatically if property rights are not enforceable. Within well-functioning economies, the legal system is an important institution for enforcing property rights and increases trade in the economy. Former Communist countries, such as Russia, have had difculty converting to a market system because they do not have established legal systems for enforcing property rights. Although there are court systems, the inefciencies of these systems signicantly limited their usefulness to private companies and individuals (the courts are quite slow, expensive, and sometimes corrupt). The lack of a good legal system offers prot opportunities for rms to create their own mechanisms for enforcing property rights and facilitating trade. Large rms in Russia have established their own security forces. If another party does not honor a contract, the security force uses coercive power to force compliance (much like the Maa). In turn, these large companies have incentives to honor contracts, because it is important for them to maintain good reputations to encourage other parties to deal with them in the future (see Chapters 10 and 22). Allegedly, the ability of large rms to enforce contracts in Russia serves as a source of competitive advantage (prots). Small rms have difculty competing with large rms because they do not have these security forces or the reputations to ensure contract compliance. Utilizing legal systems can be expensive, even in developed economies. Managers, throughout the world, have the potential to create value if they can devise more efcient methods for contract enforcement. Many rms invest substantial resources to develop reputations as honest trading partners. Source: A. Grief and E. Kandel (1995), Contract Enforcement Institutions: Historical Perspective and Current Status in Russia, in Economic Transition in Eastern Europe and Russia (Hoover Institution: Palo Alto, CA), 291321. 12 This section draws on R. Coase (1960), The Problem of Social Cost, Journal of Law and Economics 3, 144. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 65 79 ACADEMIC APPLICATIONS The Coase Theorem and the Fable of the Bees A prominently discussed case of externalities is the so-called Fable of the Bees. Beekeepers provide pollination services for the surrounding fruit growers, and the growers, in turn, provide nectar for the bees. Many economists would consider this to be a classic case of externalities. If beekeepers and growers do not receive compensation for the benets they bestow on other parties, they will underinvest in their activities (from a social standpoint). The Coase Theorem suggests that beekeepers and growers can privately negotiate to overcome this externality problem. This is exactly what is done. Beekeepers and growers often enter into contracts. Fruit growers hire beekeepers to supply hives of bees for pollination of those trees that give little suitable nectar, while the beekeepers pay growers for the privilege of grazing their bees on high nectar-producing trees. Given these payments, beekeepers and growers have incentives to consider the effects on the other party when they make their investment decisions. Through this process, beekeepers and growers can reach efcient levels of investment without help from the government. Source: S. Cheung (1973), The Fable of the Bees: An Economic Investigation, Journal of Law and Economics 16, 1134. tax rms based on their levels of pollution. Such a tax would give rms incentives to reduce pollution. In 1960, Nobel Prizewinner Ronald Coase presented a convincing argument that exchange in a free market is more powerful in producing efcient results than had been thought previously. As long as property rights can be traded, there is an incentive to rearrange these rights to enhance economic efciency. The often-recommended government intervention might be unnecessary and in many cases undesirable. Suppose that a rm has the legal right to pollute as much as it wants. Its neighbors always can offer to pay the rm to reduce its pollution level. Thus, the rm faces a cost for polluting (if the rm pollutes, there is an opportunity cost of not receiving compensation from its neighbors). The rm will pollute only if the pollution is more valuable to the rm than the costs it imposes on its neighbors. This efcient solution is obtained without a pollution tax. The same level of pollution can occur even if the neighbors have the legal right to stop the rm from emitting any pollution as opposed to the rms having the legal right to pollute as much as it wants. In this case, the rm can pay its neighbors for the right to pollute. Regardless of whether the rm or the neighbors have the legal right, the gains from trade are exhausted when the marginal benet to the rm of polluting is equal to the sum of the marginal costs imposed on its neighbors plus those that the rm bears. Coases argument convinced most economists that externalities were less of a problem than previously thought. It also implied that the distribution of property (legal) rights might have less of an effect on the ultimate use of resources than it has on the distribution of incomeas long as these rights can be exchanged. In our example, the rm might emit the same amount of pollution regardless of who initially is assigned the property right. However, the party with the property right obtains more wealth (since it is the one receiving payments). Nonetheless, as Coase points out, market exchange will not always solve the problem of externalities. The transactions that are necessary to overcome this problem are not free: There are contracting costs. These costs include search and information costs, bargaining and decision costs, and drafting, policing, and enforcement costs.13 These 13 C. Dahlman (1979), The Problem of Externality, The Journal of Law and Economics 22, 148162. 66 80 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Property Rights Help Make Niger Greener Niger in northern Africa historically has been known for being a barren, dust-choked country with many starving citizens. Most observers saw no hope in Nigers continued battle against desertication. The end result would be increased poverty and starvation. Recent studies of vegetation patterns, however, show that Niger recently has added millions of new trees and is far greener than it was 30 years ago. Interestingly some of the vegetation is densest in some of the most densely populated regions of the country. This nding runs counter to the conventional claim that population growth leads to the loss of trees and accelerates land degradation. One important factor for why Niger has become greener is a change in property rights. From colonial times, all trees in Niger were property of the state. State ownership gave farmers little incentive to grow and protect trees. Trees were cut down by residents for construction and chopped for rewood with little regard for the environmental costs. Government foresters were supposed to foster the growth of trees and to protect them from illegal destruction, but there were not enough foresters for an area nearly twice the size of Texas. Now the government allows individuals to own treesthere is private rather than public ownership. Farmers make money from their trees by selling branches, pods, fruits, and bark. Because these sales are more lucrative over time than cutting trees for rewood, the farmers preserve and protect them. Niger remains in a fragile position since it is located in a drought-prone area. Recently acquired private property rights, however, have helped to foster the growth and protection of trees. According to experts, more trees will help Nigers people to withstand whatever changes the climate might bring. Observers note that the improved situation in Niger was accomplished without having to spend a lot of money. Source: Lydia Polgreen (2007), In Niger, Trees and Crops Turn Back the Desert, nytimes.com (February 11). costs can prevent a preferred outcome from occurring. In our example, the rm might limit its pollution for a payment that is far lower than the The ultimate resource allocation will collective damage imposed on its neighbors. Nonetheless, the costs of be efcient, regardless of the initial bargaining with the rm and the costs of reaching agreement on how the assignment of property rights, as neighbors should split the payment can prevent this mutually benecial long as contracting costs are sufciently low and the property agreement from being reached. Generally, the costs of reaching an agreerights are assigned clearly, are well ment increase with the number of bargainers. In our example, the likelienforced, and can be exchanged hood of reaching an efcient agreement is highest if the rm only has to readily. bargain with a single neighbor who owns all the surrounding property. It also is important that property rights be clearly assigned, enforced, and exchangeable. Suppose there were no legal system to enforce property rights. Neighbors would be reluctant to pay a rm not to pollutethey do not obtain an enforceable property right to prevent the rm from polluting. After collecting the payment, the rm could renege on its promise to reduce pollution and the neighbors would have no recourse. This discussion suggests that market economies will tend to produce an efcient resource allocation whenever property rights are clearly assigned and contracting costs of exchanging them are sufciently low. When these conditions are met, efciency will occur regardless of the initial distribution of property rights. This general principle is often referred to as the Coase Theorem. The driving force behind the Coase Theorem is gains from trade: Individuals have incentives to search out and undertake mutually advantageous trades. This principle has important managerial implications. Even if a manager does not have all the property rights necessary to undertake a particular project, it does not mean that the project cannot be undertaken. If the proposed project creates enough value, the manager often can The Coase Theorem BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 3 Figure 3.7 The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Markets, Organizations, and the Role of Knowledge 67 81 The Cost of Transferring Knowledge The costs of transferring knowledge can be displayed on a continuum. At one end is general knowledge, which essentially is free to transfer. As the costs of information transfer increase, the information is said to become more specic. We use the term specic knowledge to denote knowledge that is relatively expensive to transfer. General knowledge Specific knowledge Information transfer costs acquire the necessary property rights from their current owners. Suppose the Watts Construction Company can create substantial value by developing a shopping center on a site that currently is zoned for residential housing. Surrounding property owners might support a change in the zoning requirement, as long as they share in the value creation. Watts might be able to increase this support by offering to develop a new neighborhood park near the shopping mall. The Coase Theorem also suggests that contracting costs are central to the study of organizations. In the absence of contracting costs, efcient outcomes will occur independent of the way decision rights are assigned. From an efciency standpoint, it does not matter whether decision rights are centralized or decentralized. It is contracting costs that make these organizational considerations important. We elaborate on this issue in the section that follows. Markets versus Central Planning History suggests that the price system is more efcient at controlling and coordinating production and consumption decisions in large economies than is central planning. Without the aid of government planners, market economies have produced products that are highly valued by consumers while avoiding large shortages or surpluses. In planned economies such as the former Soviet Union, shortages, surpluses, and other production mistakes are common. There are at least two reasons why markets have been more successful than central planning in large economies. First, the price system motivates better use of knowledge and information in economic decisions. Second, it provides stronger incentives for individuals to make productive decisions. As we will discuss later, an understanding of these advantages can be useful for managers in making rm-level decisions, such as what decision-making authority to delegate to employees and whether to make or buy a rms inputs. General versus Specic Knowledge14 Figure 3.7 shows how the costs of transferring knowledge can be displayed on a continuum. At one end of this continuum is general knowledge. General knowledge essentially is free to transfer. Examples of general knowledge are prices and quantitiesa 14 This section draws on M. Jensen and W. Meckling (1995), Specic and General Knowledge, and Organizational Structure, Journal of Applied Corporate Finance 8:2, 418. 68 82 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Topic-Specic Search Engines GlobalSpec.com is competing with Yahoo and Googleand winning. Its search engine for engineers has 3.5 million users and adds 20,000 more each week. They own that market, says Charlene Li of Forrester. GlobalSpec has a well-dened customer base and detailed understanding of its users; this sets it apart from the generalist search engines. These features allow its vertical site to provide search results from a select group of topic-specic Web sites and precisely target advertising at particular audiences. Source: Specic Knowledge about Your Customer, The Economist (July 14, 2007), 75. storekeeper easily can tell you that the price of sugar is $1 per pound. As the costs of information transfer increase, the information is said to become more specic. We use the term specic knowledge to denote knowledge that is relatively high on this scale: It is expensive to transfer. At least three factors inuence the costs of transferring information. First are the characteristics of the sender and receiver. Generally, it is less expensive for people of similar training, language, and culture to communicate than for people from different backgrounds. Second is the technology available for communication. For example, the development of electronic mail (e-mail) has lowered the costs of transferring information. Third is the nature of the knowledge itself. Some knowledge is difcult to summarize, comprehend, or transfer in a timely fashion. Depending on the exact setting, the following types of knowledge often are specic in nature: Idiosyncratic knowledge of particular circumstances. The employee on the spot is most likely to know if a particular truck has room for additional cargo or if a certain customer wants to purchase a specic product. If this information is not used immediately, it may become useless. For example, by the time the information about the truck is transferred to another person (such as a central planner), the opportunity to load the truck with additional cargo can be lost (for instance, if the truck has left). MANAGERIAL APPLICATIONS The Dynamic Nature of Specic Knowledge Historically, economies of scale have motivated rms in retailing to concentrate on standardized production and distribution. Knowledge about the idiosyncratic demands of people in particular neighborhoods tended to be ignored in stocking individual stores within a large retail chain: The information simply was too expensive to collect and process. This limited their ability to compete with small local stores that catered to the specic demands of local customers. But the development of computers and electronic scanners has made information about idiosyncratic demands of individuals less specic. As a result, retail companies have begun to engage in more micromarketing. For instance, the Sears outlet in the North Hollywood section of Los Angeles is tailor-made to suit the neighborhoods Hispanic population. Signs are in Spanish. The store is stocked with ethnic items, such as a broad selection of compact discs and tapes by Latin American artists. A few hundred miles to the north, the Sears store in San Jose offers a large number of clothing items in extra-small sizes to attract the areas Asian population. On the other hand, Sears stores in Florida carry large, roomy clothes that appeal to the large population of elderly residents. Source: Customers on Target, Financial Times (August 18, 1995). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 69 83 MANAGERIAL APPLICATIONS Use of Specic Knowledge at Apple Computer Apples rst portable Mac had so many bells and whistles that it weighed 17 pounds. It did poorly in the market. In 1990, Apple began completely reworking the design of the computer from the customers viewpoint. The entire productdevelopment team of software designers, industrial engineers, marketing people, and industrial designers were sent into the eld to observe potential customers using other products. The team discovered that people used laptops on airplanes, in cars, and at home in bed. People did not want just small computers but mobile computers. In response, Apple designed two distinctive features for its PowerBook computerthe TrackBall pointer and the palm rest in the front of the keyboard. The new product was easy to use and distinctive. Sales improved. The knowledge of what customers really wanted in a laptop computer was acquired by a team of employees who interacted closely with customers. The team members also had important scientic and assembled knowledge that allowed them to take this new information and use it to design a marketable product. Finally, they had the authority to modify the product based on their ndings. It is less likely that such specic knowledge would be incorporated in product design within a large centrally planned economywhere a central ofce is in charge of making decisions on literally millions of products. Source: Hot Products, Smart Design Is the Common Thread, BusinessWeek ( June 7, 1993), 5457. Scientic knowledge. Knowledge of how recombinant DNA works is not easily transferred to nonscientists. Assembled knowledge. An accountant who has completed a clients tax returns for several years is likely to have assembled important knowledge about the relevant parts of the tax code and the idiosyncrasies of the individuals income and deductions. Another example is learning to operate a complex machine. In neither case is this information easily transferred to others. Specic knowledge is critical in properly allocating resources. Many economic opportunities are short-lived and must be acted on quickly by the person on the spot (who has the specic information of the opportunity) or lost. Not incorporating the proper scientic or assembled knowledge into economic decisions can have costly implications. For an economic system to be successful, it must promote the use of relevant specic knowledge in economic decisions. Knowledge Creation Figure 3.7 displays knowledge on a continuum at a point in time. It is important to realize that knowledge is dynamic. There are at least two factors that can motivate changes in the costs of transferring knowledge. The rst is technology: Improved communications and computer technology have greatly lowered the costs of transferring certain types of information, making it more general. Second, individuals can take actions to convert specic knowledge to more general knowledge, for example, by drafting an operating manual. Nonaka and Takeuchi argue that converting hunches, perceptions, mental models, beliefs, experiences, and other types of specic knowledge into a form that can be communicated and transmitted in formal and systematic language is a key aspect of 70 84 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts ACADEMIC APPLICATIONS Markets versus Central Planning in Russia Since 1994 the Republic of Georgia moved toward a market economy by privatizing many businesses and creating labor and product markets where prices are unregulated. The Republic of Uzbekistan privatized only a small part of its economy, and government monopolies and central planners still control various economic activities tightly. From 1995 to 1998 gross domestic product has been growing between 7 and 11 percent in Georgia, whereas the Uzbekistan economy is stagnant. Nonetheless, most ex-Soviet republics are reluctant to adopt a decentralized economy. Free markets benet the young and energetic and often hurt politically inuential groups who benet from the old system: pensioners, employees at inefcient state enterprises, and government ofcials. Source: G. Becker (1998), A Free-Market Winner vs. a Soviet Style Loser, BusinessWeek (August 3), 22. successful new product innovation.15 As one example, consider Matsushitas development of an automated fresh bread maker in the 1980s. Specic knowledge of how to knead dough to produce tasty bread was held by master bakers. This knowledge was not easily transferred to others, and past attempts to produce fully automated bread makers had failed because they produced poor-quality bread. Yet specic knowledge about how to manufacture automated bread machines was held by engineers. To produce a successful bread machine, relevant specic knowledge had to be transferred between bakers and engineers. To accomplish this transfer, managers from Matsushita took bread-making lessons from a master baker at an Osaka hotel. Eventually, the managers discovered that the key to good bread making is to twist and stretch the dough during the kneading process. This concept was general knowledge that could be passed along to design engineers. Matsushitas Home Baker was the rst fully automatic bread-making machine for home use and has become a quite successful product. MANAGERIAL APPLICATIONS Converting IT Wetware into Software Charles Belford, president of the Canadian-based management consulting rm Managements Smarts, Inc., advises clients to get more value out of their existing information technology (IT) without buying new software or hardware. He argues that rms should revise their current delegation of authority, governance, and planning and management practices to identify lowcost, low-risk IT enhancements. Managers should be given incentives to ensure that successful pilot tests are properly identied and then disseminated throughout the organization. He encourages clients to reorganize so that a senior manager has enterprise-wide oversight to exploit IT packages to ensure their full potential for the company. For an example, Mr. Belford says, By revising the current delegation of authority for managing Web site content in your company, you may be able to turn your obese Web site or your bulimic internal intranet network into healthy assets that actually serve their respective constituencies. Mr. Belford maintains that the only way to convert IT wetware (successful local IT applications) into recipes or software and then leverage this newly created software throughout the rm is by changing the companys organizational architecture. Source: C. Belford, Add Value to Tech Assets without Breaking the Bank, The Globe and Mail (March 28, 2002), B16. 15 I. Nonaka and H. Takeuchi (1995), The Knowledge-Creating Company (Oxford University Press: New York). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 71 85 MANAGERIAL APPLICATIONS FiatUsing Specic Knowledge about Developing Markets Traditionally, leading car manufacturers simply have adapted their existing models for sales in developing countries, in spite of the fact that unique conditions prevail in these markets. Fiat, however, has invested about $2 billion in developing a new car, the A178, to meet the particular demands of people in developing countries. The car is attractively styled and inexpensive but built tough to withstand poor road conditions. Specic knowledge about the appropriate design features was obtained from a task force of engineers from Brazil, Italy, Turkey, Poland, Argentina, India, South Africa, and Morocco. Fiat forecasts that it will sell up to 900,000 A178s a year once production goes into full swing. Fiat has the decision rights to produce this product and does not have to convince some central planner of its merits. Source: Fiat Steers New Model towards Developing Markets, Financial Times (August, 24, 1995). In 1778, economistclergyman Thomas Malthus predicted that population would grow more rapidly than the food supply resulting in mass starvation.16 His argument was straightforward: Because land and other natural resources are nite, the growth in population (fueled by the passion of the sexes) would eventually exceed the available food supply. But this predictionwhich prompted economics to be labeled the dismal sciencehas not come to pass. For instance, population in the United States increased from 76 million in 1900 to 296 million in 2005 (an increase of nearly 300 percent), yet the amount of land and the number of workers devoted to agricultural production over the same period fell dramatically. Today less than 3 percent of the U.S. population works in agriculture, while the per capita food supply is at an all-time high. And this increase in the food supply is not just a U.S. phenomenon. From 1951 to 1992, world food production per capita increased 34 percent.17 So why was Malthus wrong? Malthus, as well as more recent pessimists, has underestimated the importance of improvements in technologyguring out better ways to use existing resources.18 For example, todays computers are far more powerful than those of a decade ago, yet they take fewer resources to produce. Moreover, computer designers have discovered ways to make the materials used in computers more recyclable. This process of discovering better ways to use existing resources occurs not only in manufacturing but also in service-related industries as well. For example, consider the implications of McDonalds innovations in the 1950s in the delivery of fast food. Economist Paul Romer argues that the opportunities for this type of discovery and growth essentially are unlimited.19 People are constantly taking ideas and knowledge that are in their wetware (brains) and converting them into software (recipes and formulas) that can be employed to produce new products and services. Matsushitas conversion of the specic knowledge held by the master baker into more general knowledge that could be used by engineers is a good example. As we will discuss in Chapter 8, good managers understand this mechanism for creating knowledge and value and foster it within their rms. 16 Thomas Malthus (1778), An Essay on the Principle of Population (printed for J. Johnson, in St. Pauls Church Yard, London). 17 J. Perloff (2001), Microeconomics (Addison Wesley: Boston), 154. 18 For a more recent example of concern about the implications of nite resources, see D. Meadows (1977), Limits to Growth: A Report for the Club of Romes Project on the Predicament of Mankind (New American Library). 19 For nontechnical discussions of Romers theory of economic growth, see P. Romer (1993), Economic Growth, Fortune Encyclopedia of Economics (Time Warner Books: NY); and P. Romer (1995), Beyond the Knowledge Worker, World Link, January/February, 5560. 72 86 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts Specic Knowledge and the Economic System20 Nobel Prizewinner Friedrich Hayek offered a convincing argument that market economies are more likely than centrally planned economies to incorporate relevant specic knowledge in economic decision making. He argued that the relevant specic knowledge for economic decision making is not given to any one individual; instead, it is distributed among many people in the economy. This knowledge, by denition, does not lend itself to statistical aggregation; it is costly to transfer. A central planner invariably lacks the mental or computing ability to process large volumes of this sort of information. Hayek thus concluded that central planners often ignore important specic knowledge in economic decisions. In contrast, economic decisions in a market system are decentralized to individuals who are likely to have the relevant specic knowledge. Technical and marketing geniuses, like William Gates at Microsoft and Michael Dell at Dell Computer, are free to start new businesses and to market products of their choosing. The information that motivates these decisions does not have to be transferred to some central ofce in Washington where centralized production decisions are made. Thus, the information is more likely to be used effectively. The activities of decentralized decision makers are coordinated by prices. For instance, an increase in market-determined wage rates (the price of labor) signals to producers that labor is in short supply and should be conserved. Higher wages, in turn, motivate producers to conserve labor. An important advantage of the price system that is stressed by Hayek is that prices economize on the costs of transferring information to coordinate decisions. Companies normally do not have to know all the details of why labor costs have increased. The simple fact that wages have increased tells them most of the things they need to know to make value-maximizing decisions.21 Incentives in Markets Private property rights are critical for making a market economy work because they provide strong incentives for decentralized decision makers to act on their specic informationthe wealth effects of economic decisions are borne directly by the resource owners. If Alice Chan owns a piece of property, she has incentives to use the land productively because she gets to keep the prots. If Jamal Hammoud can make more productive use of the land, Alice will sell the land to Jamal (there are gains from trade). Property rights are rearranged so that decision rights over resources are linked with the relevant specic knowledge. In contrast, decision makers in centrally planned economies have limited incentives to make productive use of information (even if they have it) since they do not own the resources under their control. Further, lower-level bureaucrats have limited incentives to carry out decisions made by the central authority. The best use of a particular automobile might be to transport tourists from a local airport. A central planner, however, might give the car to his brother because he is more concerned about making his brother happy than about making the economy more productive. After all, he does not keep the prots from transporting touriststhey go to the state. 20 This section draws on F. Hayek (1945), The Use of Knowledge in Society, American Economic Review 35, 519530. 21 Producers also might want to know the expected future prices of labor. For instance, if the price increase is expected to be transitory, the company might want to avoid making layoffs. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 73 87 ANALYZING MANAGERIAL DECISIONS: Nobel PrizeWinner F. A. Hayek on the Miracle of the Price System It is worth contemplating for a moment a very simple and commonplace instance of the action of the price system to see what precisely it accomplishes. Assume that somewhere in the world a new opportunity for the use of some raw material, say, tin, has arisen, or that one of the sources of supply of tin has been eliminated. It does not matter for our purpose and it is signicant that it does not matterwhich of these two causes has made tin more scarce. All that the users of tin need to know is that some of the tin they used to consume is now more protably employed elsewhere and that, in consequence, they must economize tin. There is no need for the great majority of them even to know where the more urgent need has arisen, or in favor of what other needs they ought to husband the supply. If only some of them know directly of the new demand and switch resources over to it, and if the people who are aware of the new gap thus created in turn ll it from still other sources, the effect will rapidly spread throughout the entire economic system. This inuences not only all the uses of tin but also those of its substitutes and the substitutes of these substitutes, the supply of all things made of tin, and their substitutes, and so on. All this takes place without the great majority of those instrumental in bringing about these substitutions knowing anything at all about the original cause of these changes. The whole acts as one market, not because any of its members surveys the whole eld, but because their limited individual elds of vision sufciently overlap so that through many intermediaries the relevant information is communicated to all. The mere fact that there is one price for any commodityor rather that local prices are connected in a manner determined by the cost of transport, etc.brings about the solution which (if conceptually possible) might have been arrived at by one single mind possessing all the information which is in fact dispersed among all the people involved in the process. Some people (for example, Hayek) argue that decentralization of economic decisions in the economy leads to an efcient resource allocation. What differences exist within the rm that make the link between decentralization and efciency less clear? SOURCE: F Hayek (1945), The Use of Knowledge in Society, . American Economic Review 35, 519530. Contracting Costs and Existence of Firms Hayeks argument suggests that markets are better than central planning. Why, then, is so much activity conducted within rms, where resource allocation decisions are made by managers in a manner that is closely akin to central planning?22 Conceptually, rms do not have to exist. All production and exchange could be carried out through market transactions. In the case of the PC, each consumer could buy all the parts that make up the PC in separate market transactions and then pay someone to assemble them. In reality, of course, most computers are made by rms and only the nal products are sold to the consumer. Ronald Coase provides an answer to the question as to why resources are allocated by both markets and rms.23 His basic argument is that economic transactions involve contracting costs, including search and information costs, bargaining and decision costs, 22 Within a rm, resources often are transferred from one division to another by an administrative order from management. For example, managers often are transferred among divisions by administrative decisions. Prices are not used to make these decisionsthe divisions typically do not bid for the managers. 23 R. Coase (1937), Economica, The Nature of the Firm, New Series, IV, 386405. 74 88 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Japan, Computers, and Industrial Policy In the early 1990s, a group of prominent policy advocates argued that Japans output of computers and computer-related products would pass that of the United States during the decade due to the alleged power and success of Japanese government planners. It was argued that the United States was at a competitive disadvantage because it did not rely on central economic planning. By 1995, it is not Fujitsu or an NEC or a Hitachi that you stare at every day at the ofce, but a Compaq or IBM or Appleall American designed and manufactured. Japanese companies have also been crushed in software. In 1994, U.S. computer companies invaded the Japanese market for personal computers in a serious manner. In 1994, U.S. companies doubled their share of the Japanese computer market to more than 30 percent. Source: Japan, Computers, and Industrial Policy, The American Enterprise ( July/August, 1995), 86. and policing and enforcement costs. There is also an opportunity cost if the transaction results in an inefcient resource allocation (we discuss this in detail in Chapter 10). The optimal method of organizing a given economic transaction is the one that minimizes contracting costs.24 In some cases, the method will be market exchange. In other cases, the method will involve rms. Contracting Costs in Markets A primary set of costs of using markets for exchange involves the discovery and negotiation of prices.25 For example, rms have the following two potential advantages: Fewer transactions. If there are N customers and M factors of production, a rm can hire the M factors and sell to the N customers. The total transactions are N M. In contrast, if each customer contracts separately with each factor of production, there are N M transactions. For example, 10 workers might be required to assemble a computer. If there are 1,000 customers and each customer negotiates with each worker, there are a total of 10,000 transactions. If a rm hires the 10 workers and sells computers to the 1,000 customers, there are 1,010 transactions. Informational specialization. Think of buying a PC. How much do you know about buying each separate part? PC producers, on the other hand, specialize in this knowledge. The consumer buying from a rm only has to be concerned with the quality of the end product. In Chapter 19, we shall elaborate on one particularly important set of contracting costs that motivates the existence of rms, those associated with specic assets. Assets are specic when they are worth more in their current use than in alternative uses. An 24 It is not always possible to separate contracting costs from the basic costs of production. The optimal method of production can depend on the way the transaction is organized. Therefore, it is more precise to say that the optimal method of organization is the one that minimizes total costs (production and contracting costs). The basic arguments are easier to explain if we focus on contracting costs. 25 Economists generally agree that contracting costs motivate the existence of rms. There is disagreement concerning which contracting costs are most important. Our intent in this chapter is to give the reader a general sampling of the kinds of costs that can be important. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 75 89 ACADEMIC APPLICATIONS Herbert Simon on Organizations and Markets The United States often is referred to as a market economy. In reality, much of the economic activity in the United States, as well as in other market economies, is conducted within rms. To quote Herbert Simon, a former Nobel Prize winner, Suppose a visitor from Mars approaches the earth from space, equipped with a telescope that reveals social structures. The rms reveal themselves, say, as solid green areas with faint interior contours marking out divisions and departments. Market contracting costs show as red lines connecting rms, forming a network in the spaces between them. Within the rms the approaching visitor also sees pale blue lines, the lines of authority connecting bosses with various levels of workers. . . . No matter whether the visitor approached the United States or the Soviet Union, urban China or the European Community, the greater part of the space below would be within the green areas, for almost all the inhabitants would be employees, within rm boundaries. Organizations would be the dominant feature on the landscape. A message sent back home, describing the scene, would speak of large green areas interconnected by red lines. It would not likely speak of a network of red lines connecting green spots. Source: H. Simon (1991), Organizations and Markets, Journal of Economic Perspectives 5, 2544. extreme example is a machine that is used to produce parts that can be used only by one particular producer. The machine is valuable in producing parts for the particular buyer but is essentially worthless in alternative uses. In this case, independent suppliers are reluctant to purchase the machine since they do not want to be at the mercy of a single buyer. For instance, suppliers might worry that the buyer will try to force a reduction in future prices, make unreasonable quality or quantity demands, or curtail purchases. It is these concerns that make simple market transactions between buyers and sellers unlikely when the relevant assets are highly specic. A potential response to this problem is for the producer to own the machine and make the input parts within a single larger rm. Another potential advantage of rms is that in some cases they can reduce contracting costs through established reputations. Individuals are likely to have condence in trading with parties who are expected to continue to participate in the marketplace over a long time. They understand that these parties have incentives to be honorable in order to enhance their reputation and future business opportunities. Organizations tend to have longer lives than individuals and thus might be expected to be more likely to honor agreements than unknown individuals (some major corporations date back to the nineteenth century). This increased trust can motivate lower expenditures on negotiating and policing agreements. We discuss this issue in greater detail in Chapters 10 and 22. Government regulation also helps explain the existence of some rms. Sometimes rms can produce more cheaply because they avoid taxes at various stages of production compared to market transactions. Contracting Costs within Firms We have discussed several contracting costs that can motivate the existence of rms. Given these costs, why isnt the economy just one big rm? The answer is that resource allocation by rms also involves contracting costs. For example, as rms become larger, it becomes increasingly difcult for managers to make efcient and timely decisions. They are more likely to make errors and to be less responsive to changing circumstances. 76 90 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts ACADEMIC APPLICATIONS Corporate Focus and Stock Returns Ronald Coase argues that the use of markets involves contracting costs and that sometimes these costs can be reduced by including transactions within rms. However, rms also involve contracting costs. In the 1990s, many companies concluded that they had become too large and diversied. These companies, in turn, decided to refocus on their core businesses and to shed unrelated activities (for example, through asset sales). Evidence suggests that on average, these rms increased their stock market values by increasing their focus on core activities. Source: R. Comment and G. Jarrell (1995), Corporate Focus and Stock Returns, Journal of Financial Economics 37, 6787. As a rm grows, important decisions must be delegated to employees who are not owners of the rm, thereby generating costs to motivate these nonowners to work in the interests of the owners. Chapters 10 and 19 contain more detailed discussions of these contracting costs within rms. Efcient Organization Individuals involved in trade and production have incentives to implement cost-reducing methods of organization because there are greater gains to be shared. For example, at a given price, more prots can be generated if costs are reduced.26 In competitive markets, individuals will constantly search for new and better ways to reduce costs to improve their competitive advantage and prots. The bottom line is that rms will be used to organize economic activities whenever their cost is lower than that of using markets, and vice versa. Also, as we will see, this same process has important implications for the internal design of organizations. Managerial Objectives Our discussion to this point has treated decision makers within rms as owners. Owners have a strong interest in increasing the prots of the rm, since they get to keep the proceeds. In public corporations managers are rarely major owners of the rm. Nonetheless, in Part 2 of the book, we assume that managers strive to maximize rm protsor more precisely rm value, the present value of the rms prot stream: They make input, output, and pricing decisions with value maximization as their sole objective.27 This perspective is a reasonable starting point because if rms fail to make prots over time, they cease to exist. Most managers are under constant pressure to create value. There also are other mechanisms, such as incentive compensation, that work to align the interests of managers and owners. These mechanisms help make prot maximization a reasonable rst approximation of the managers objective function. Prot maximization is the basic premise used in most economics textbooks. Starting in Chapter 10, however, we shall present a richer characterization of the rm and analyze management/owner conicts in greater detail. The appendix to this chapter provides a more detailed discussion of managerial objectives, focusing on the topics of shareholder value and stock-market efciency. 26 A rms prot ( ) is the difference between its total revenues (TR) and total costs (TC): TR TC. If a company has sales of $1 million and costs of $750,000, it earns a prot of $250,000. 27 Much of our basic analysis focuses on maximizing prot in a single period. This approach yields useful managerial insights without overly complicating the analysis. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 77 91 ACADEMIC APPLICATIONS CEO Turnover and Firm Prots A standard assumption in microeconomics is that managers strive to maximize prots. One reason that managers are likely to be concerned about prots is that poor prots and stock price performance increase the likelihood that they will be red. For instance, research suggests that the worst performing rms are about 1.5 times as likely to have a management change as the best performers. Source: For a review of this evidence, see J. Brickley (2003), Empirical Research on CEO Turnover and Firm Performance: A Discussion, Journal of Accounting and Economics 36, 227233. Managerial Decisions We began this chapter with an overview of how market economies operate. An understanding of this topic is critical if managers are to make productive economic decisions. It is important to understand how a shift in either supply or demand affects product prices. (In Part 2, we shall extend this analysis and examine in more detail how managers might make optimal input, output, and pricing decisions.) We also discussed the role of knowledge and incentives in determining the effectiveness of alternative economic systems and the importance of contracting costs in determining whether or not economic transactions are conducted within markets or organizations. ACADEMIC APPLICATIONS Firms versus Markets: When Markets Ruled Economic theory argues that activities are organized within rms when the cost is lower than using markets, and vice versa. Today, much of the economic activity in the world is conducted within rms. It is hard to envision a world where large rms do not play an important role in the production and distribution of products. The importance of rms, however, is a relatively recent phenomenon. Prior to the middle of the nineteenth century, there were virtually no large rms. Most production was conducted by small, owner-managed operations. The activities of these operations were coordinated almost entirely through market transactions and prices. To quote Alfred Chandler in describing business organization before 1850, The traditional American business was a single-unit business enterprise. In such an enterprise an individual or a small number of owners operated a shop, factory, bank, or transportation line out of a single ofce. Normally this type of rm handled only a single economic function, dealt in a single product line, and operated in one geographic area. Before the rise of the modern rm, the activities of one of these small, personally owned and managed enterprises were coordinated and monitored by market and price mechanisms. The large rm became feasible only with the development of improved energy sources, transportation, and communications. Coal-red steam power generators provided a source of energy that made it possible for the factory to replace artisans and small mill owners, and railroads enabled rms to ship production in large quantities to newly emerging urban centers. The telegraph allowed rms to coordinate activities of workers over larger geographic areas. These developments tended to make it less expensive to coordinate production and distribution using administrative controls, rather than to rely on numerous market transactions among all the intermediaries in the system. Source: A. Chandler (1977), The Visible Hand: The Managerial Revolution in American Business (Harvard University Press: Cambridge, MA). 78 92 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts MANAGERIAL APPLICATIONS Hewlett-Packard and Corporate Focus Many companies often split themselves into several companies when their operations have grown too large and unwieldy. In 1996 AT&T spun off its equipment arm to Lucent Technologies and its computer-manufacturing division to NCR. In 1999, Hewlett-Packard announced it would spin off as a separate company its $7.6 billion a year test and measurement operations, leaving the remaining $39.5 billion a year computers and printers business. This was intended to allow H-P to become more focused on making these products. H-P faced slowing revenue growth in 19981999, partly because it moved slower than its rivals to capitalize on new developments like the Internet. One analyst commented, H-P needs to be more agile and faster-moving, if its going to keep pace with the Dells and Sun Microsystems. Source: D. Hamilton and S. Thurm (1999), H-P to Spin Off Its Measurement Operations, The Wall Street Journal (March 3), A3. ANALYZING MANAGERIAL DECISIONS: Property Right Security in Russian Deprivatization Since 1992, approximately 70,000 state-owned enterprises in Russia were privatized. Many of the private buyers were foreign companies and investors, for example from the United States and Western Europe. The idea was to move from a centrally planned economy to a market system. Yet in the late 1990s a weak economy caused great concern among Russian voters. Politicians, such as Moscows Mayor Yuri M. Luzhkov, began promoting deprivatization or, as the locals put it, deprivatizatsia. Under this policy certain past privatizations would be declared illegal and the transactions would be reversed. The company then would be either run as a state-owned enterprise or sold to another party. For example, in October 1999, a court stripped Wall Streets Kohlberg Kravis Roberts and the U.S.-Russia Investment Fund of their majority interest in the Lomonosov Porcelain Factory in St. Petersburg. These companies had purchased the factory in 1998, but the courts ruled that the companys initial privatization ve years earlier was illegal. Sources suggested that the company was likely to be resold to Soviet-era managers who were set to lose their jobs when the new investors entered the picture. Politicians, such as Luzhkov, vow that not all privatizations will be reversedonly the illegal ones. But one current problem is that privatization legislation is nebulous about what could be termed a violation. Anything from a missing piece of paper in the original tender offer to investment requirements not being met might be ruled a violation. And virtually anyone could le a complaint to trigger an inquiry into a past deal. 1. What impact will the prospect of deprivatization have on investment by managers of privatized rms? 2. What effect will deprivatization have on foreign investment in Russia? 3. Do you think that mass deprivatization is in the long-run best interests of Russia? 4. Who gains from deprivatization? Who loses? 5. Assuming more people are hurt by deprivatization than helped, why would a local politician support such a policy? SOURCE: M. Coker (1999). That Russian Company You Bought? Maybe You Didnt, BusinessWeek (December 13), 70. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 79 93 Although we have focused our discussion at the economic-system level, these issues are directly relevant to understanding rm-level decisions on organizational architecture. If rms are to be productive, they must be structured in ways that promote the use of the relevant specic knowledge and economize on the costs of organization. They also must establish appropriate incentives, so that their employees act in a productive manner. Starting in Chapter 10, we shall extend the concepts introduced in this chapter to questions of organizational architecture. Summary There are many different ways of organizing economic activities. Economists focus on Pareto efciency in evaluating the effectiveness of alternative economic systems. An allocation is Pareto-efcient if there is no alternative that keeps all individuals at least as well off but makes at least one person better off. Pareto-improving changes in a resource allocation are viewed as welfare-increasing. An important feature of a market economy is the use of private property rights. A property right is a legally enforced right to select the uses of an economic good. A property right is private when it is assigned to a specic person. Private property rights are alienable in that they can be transferred (sold or gifted) to other individuals. In free markets, property rights frequently are exchanged. Trade occurs because it is mutually advantageous. The buyer values the good more than the seller, and there are gains from trade. Trade is an important form of value creation. Trading produces value that makes individuals better off. Gains from trade also motivate the movement of resources to more productive users. Total output and standards of living often increase when individuals specialize in production activities for which they have a comparative advantage (lower opportunity costs). Prices coordinate the individual actions in a market economy. If too little of a good is being produced, inventories will shrink, prices will rise, and producers will have incentives to increase output to exploit the prot opportunity. If too much of a good is being produced, prices will fall, inventories will build, and producers will have incentives to cut production. The market is in equilibrium when the quantity supplied of a product equals the quantity demanded. There are strong pressures in competitive economies that move the market toward equilibrium. In equilibrium, there are no shortages or surpluses and inventories are stable at their desired levels. Equilibrium prices and quantities change with changes in the supply and demand for products. Consumer surplus and producer surplus are measures of the gains from trade to consumers and producers from participating in a market. Government-imposed price caps or oors result in market imbalances and lost surplus (in an otherwise well-functioning market). Externalities exist when the actions of one party affect the consumption or production possibilities of another party outside an exchange relationship. Externalities can cause markets to fail to produce an efcient resource allocation. Competitive markets will produce a Pareto-efcient allocation of resources if the costs of making mutually advantageous trades are sufciently low. The Coase Theorem indicates that the ultimate resource allocation will be efcient, regardless of the initial assignment of property rights, as long as contracting costs are sufciently low and property rights are clearly assigned, well enforced, and readily exchangeable. General knowledge is inexpensive to transfer, whereas specic knowledge is expensive to transfer. Specic knowledge is quite important in economic decisions. Central planning often fails because important specic knowledge is not incorporated in the planning 80 94 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts process. Within market systems, economic decisions are decentralized to individuals with the relevant specic knowledge. Prices convey general knowledge that coordinates the decisions of individuals. Private property rights provide important incentives to individuals to act productively, since they bear the wealth effects of their decisions. In principle, all economic activity could be conducted through market transactions. However, even in market economies, much economic activity occurs within rms, where administrative decisions rather than market prices are used to allocate resources. Firms exist because of the contracting costs of using markets. However, organizing transactions within rms also involves costs. Individuals have incentives to organize transactions in the most efcient mannerto increase the gains from trade. Economic activities tend to be organized within rms when the cost is lower than that of using markets, and vice versa. This chapter provides important background information on both markets and organizations. In Part 2, we shall extend the analysis of markets and study important managerial decisions such as output, inputs, pricing, and strategy. In these next six chapters, we assume that managers strive to maximize rm prots. In the remainder of the book, we shall extend the analysis of organizations and cover a variety of important topics about organizational design. A reader interested primarily in organizational design can move directly to Chapter 10 without loss of continuity. Appendix Shareholder Value and Market Efciency In the United States, top managers (ofcers and directors) have a duciary duty to act in the interests of the corporation and its shareholders. Consistent with this legal obligation, executives constantly profess to the media, stock analysts, and other constituencies a fundamental allegiance to increasing shareholder value, as reected by the price of the common stock. Some critics, however, contend that managers ignore this duty and make decisions that benet themselves at the expense of shareholders. Theory and evidence suggests that there are a variety of internal and external control mechanisms that provide incentives to managers to be concerned about shareholder value. We examine these mechanisms beginning in Chapter 10. Analyzing managerial decisions based on the assumption of shareholder wealth maximization has two benets. First, it suggests what managers should do to meet their duciary responsibilities. Second, it describes what good managers actually do when they have sufcient incentives to focus on shareholder value. Finance courses analyze investment and nancing decisions where it is crucial to consider intertemporal tradeoffs among cash ows. For example, should a manager invest $1,000,000 to build a plant today that has the potential to yield $100,000 per year prot in the future? In contrast, managerial economics largely focuses on operational decisions where intertemporal tradeoffs in cash ows are relatively less important. For example, what current price should the manager charge to maximize prots? These decisions can be analyzed under the simplifying assumption that managers seek to maximize single-period prots. Little is gained from the added complexity of assuming that managers seek to maximize shareholder wealth, which involves the valuation of multiperiod cash ows.28 In Chapters 47, we present an economic analysis of demand, production, cost, market structure, and pricing. Here we follow the standard approach in managerial economics and assume that managers strive to maximize single-period prots. In subsequent chapters, 28 In stationary settings where the same action is optimal in each period, the two objective functions are equivalent. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Chapter 3 Markets, Organizations, and the Role of Knowledge 81 95 where intertemporal considerations are more important, we assume that mangers seek to maximize shareholder wealth. The main text does not require a detailed understanding of stock market valuation. It is sufcient simply to understand that share price incorporates the effects of managerial actions on both current and future protsappropriately adjusted for the timing and risk of the cash ows. In this appendix we go beyond what is necessary for the main text by providing a more detailed analysis of shareholder value. We begin by introducing the concept of present value and deriving an expression for the current value of a share of common stock. We then discuss the concept of stock market efciency and its resulting managerial implications. Present Value Is it better to receive a dollar today or a dollar a year from now? The obvious answer is that it is better to receive the dollar today, since it can be invested to yield more than a dollar in the future. Suppose that you invest a dollar in a risk-free asset (such as a U.S. Treasury security) with a 5 percent interest rate. At the end of one year, your investment will be worth $1.05. The investment, which promises $1.05 for certain in one year, and a dollar today have equivalent value since the dollar can be invested to produce the same future cash ow.29 The future value of a dollar invested in the risk-free asset for one year is $1 (1.05) $1.05. Conversely, the present value of $1.05 received for certain in one year is $1.05 1.05 $1. The concept of present value extends to multiple periods. One dollar invested at 5 percent for n periods is worth $(1.05)n at the end of the nth period. The present value of this future value is $(1.05)n (1.05)n $1.00. More generally the present value of W dollars received for certain at the end of n periods is W (1 r)n, where r is the risk-free interest rate. For example, suppose that a risk-free investment promises a cash ow of $1.50 in ve years and the annual interest rate is 5 percent. The present value of the investment is $1.50 (1.05)5 $1.175 (investing $1.175 for ve years at 5 percent yields a terminal value of $1.50, assuming any intermediate cash ows are reinvested). Investments often generate multiple cash ows over time. For example, a 10-year U.S. Treasury bond pays a xed rate of interest every 6 months until the security matures at which time a nal principal payment is made along with the nal interest payment (20 total payments). The present value of a stream of cash ows is equal to the sum of the present values of the cash ows for each period. The general formula for the present value of a risk-free investment is: Present Value CFt (1 r)t (3.6) where CFt is the cash ow that occurs in period t (t 1 to n, the terminal period). For example, the present value of a 3-year bond that pays $100 at the end of each year and a principal payment of $1,000 at maturity is: 100 (1.05) 100 (1.05)2 1,100 (1.05)3 $1,136.16 (assuming a 5 percent risk-free rate). The concept of present value is extremely useful when comparing alternative investments with different time-series patterns of cash ows. Present value allows an apples-toapples comparison of the alternatives since all are expressed in a common dimension their present values. The best investment is the one with the highest net present valuethe difference between the present value of its benets and costs. 29 This equivalence of the risk-free investment and the dollar today holds even if the investor wants to consume a dollar today as long the investment can be sold in the marketplace for its present value. 82 96 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Basic Concepts Share Value Our analysis of present value has focused on risk-free investments, where promised future cash ows were known with certainty and correspondingly discounted by the riskfree rate to obtain present values. Common stocks, however, are risky investmentsthe ultimate payouts to shareholders are not certain, but depend on the fortunes of the rm. While a rm might be expected to pay a given stream of dividends over time, investors might receive higher or lower payments depending on the fortunes of the rm. The values of stocks and other risky investments are determined by discounting expected cash ows (see Chapter 2) by risk-adjusted discount rates. The discount rate used for valuing an investment increases with the risk of the expected cash ows. (Finance courses teach that the relevant risk is its systematic risk, that which cannot be eliminated through holding a diversied portfolio of investments.) The intuition of stock valuation can be illustrated using a simple example. Suppose that a stock has the potential to pay a liquidating dividend to shareholders of either $50 or $150 at the end of the year. Each outcome has a probability of .5, and thus the expected cash ow is (.5 $50) (.5 $150) $100. If we discount the $100 expected cash ow at the risk-free rate of 5 percent we obtain $100 (1.05) $95.24. Purchasing the stock at a price of $95.24 by denition yields an expected return of 5 percent ((100 95.24) 95.24). Risk averse investors (see Chapter 2), however, will not purchase the stock at this price since they can earn the same return for certain by investing the $95.24 in a risk-free asset. For the stock to appeal to a risk-averse investor, it must sell for a lower price and thus a higher expected return. Suppose investors would purchase this stock if it yielded an expected rate of return of 10 percent. The resulting price of $90.91 is found by discounting the expected cash ow of $100 by the risk-adjusted discount rate of 10 percent ($100 1.10 $90.91). More generally, companies are expected to last more than one period and may pay dividends in multiple periods. The value of a stock is equal to the sum of the present values of each of the expected future dividends. Expressed in equation form, the current value of a share of stock, P0, is: P0 D1 (1 k) D2 (1 k)2 . . . Dq (1 k)q (3.7) where Dt is the expected dividend paid to the investor at each time t , and k is the riskadjusted discount rate (expected return).30 Cash ows that are expected to occur further into the future have less impact on the valuation due to this discounting process. Thus, analysts generally exert most of their effort predicting cash ows over the rst 5 to 10 years of the investment. Simplifying assumptions typically are used for estimating the present value of the remaining cash ows. In special cases, Equation (3.7) reduces to simple expressions. One prominent example is the constant growth model. This model assumes that investors expect the rms dividends will grow each period at a constant rate g k, and that the rm will last forever. It is easy to show that Equation (3.7) reduces in this case to: P D1 (k g) (3.8) Suppose that investors expect that the HG Corporation will pay a $5 dividend at the end of the year and that the dividend will grow at 5 percent annually forever. The 30 Most investors receive part of their returns in the form of capital losses or gains when they sell the stock. This observation does not invalidate Equation (3.7) since the price of the stock at the time of the sale reects the discounted value of the remaining expected cash ows. Also, the equation holds even when some shareholders receive payouts in the form of share repurchases. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 83 97 price of the stock at the beginning of the year, assuming a risk-adjusted discount rate of .10 percent, is P0 $5 (.10 .05) $100. Equation (3.8) illustrates that the value of the stock is not determined by the current dividend alone. Growth rms often pay few dividends in their early years so that internally generated cash can be reinvested in the business. Nonetheless, they can sell at high prices if the market anticipates that they will make large payouts after their growth slows. Consider Microsoft Corporation as its growth rate slowed in 2004. In July 2004, Microsoft announced that it would double its annual dividend to shareholders by $3.5 billion per year, pay a one-time special dividend of $32 billion, and repurchase $30 billion of company stock over the following four years. The special dividend payment was the largest in S&P 500 history and quickly turned Microsoft from a small dividend payer to the tenth highest on the S&P 500 Index. Stock Market Efciency Investments with identical cash ows and risk sell for the same price in a well-functioning stock market.31 Since rational investors always will purchase the lower-priced of two identical securities, identical securities must sell for the same price for the market to clear (quantity supplied equal to quantity demanded for each security). In equilibrium the expected returns on identical securities are equal. A stock market is efcient if it responds quickly and rationally to new information (i.e., share prices fully reect available information; each stock is priced to yield a competitive return given its risk; stocks are not systematically under- or overvalued). In an efcient capital market, the market values of securities reect the present values of the expected future net cash ows to shareholders, including expected cash ows from future investments. If an event occurs that changes expected cash ows or risk of a rm, the share price will adjust quickly to reect the new information. As a result, investors should expect to receive competitive returns from purchasing stocks at current prices (i.e., the market-determined expected return for securities with similar characteristics). Depending on the fortunes of the rm, investors may end up earning more or less than the expected return. However, they should not expect to beat the market on a systematic basis. The efcient markets hypothesis is perhaps the most extensively tested hypothesis in all the social sciences. The evidence is consistent with the view that stock markets are at least reasonably efcient with respect to public information. Major stock markets react quickly to new information, and investors generally can expect that they will not earn abnormal returns from trading stocks based on publicly available information. The research on efcient markets has important implications for corporate managers. First, there is no ambiguity about the rms objective functionmanagers should maximize the current market value of the rm. Hence, management does not have to choose between maximizing the rms current value or its future value, and there is no reason for management to have a time horizon that is too short. Second, management decisions that increase reported earnings, but do not affect current or future cash ows, represent wasted effort. Third, if new securities are issued at market prices, which reect an unbiased assessment of future payoffs, then concern about dilution or the sharing of positive net present value projects with new security holders is eliminated. Fourth, security returns are meaningful measures of rm performance. This allows scholars, 31 Other characteristics such as liquidity and tax implications can be valued by the market. For simplicity we concentrate on risk. The same principle holds in the more general case; investments with identical characteristics that are valued by the market should sell at the same price. 84 98 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts ANALYZING MANAGERIAL DECISIONS: Shareholder Value and Market Efciency 1. Suppose that you purchase a newly issued 10-year U.S. Treasury bond for $10,000. The bond has a promised interest rate of 5 percent ($250 every six months). The stated interest rate of 5 percent (annual payment of $500 divided by the initial face value of $10,000) does not change over the life of the bond. Do you expect that the market value of the bond will be constant or variable over the life of the bond? Explain. 2. Calculate the present value of an investment with the following expected cash ows at a discount rate of 10 percent: year 1 $500, year 2 $600, and year 3 $650. Recalculate the present value at discount rates of 15 percent and 5 percent. 3. Is the discount rate used by investors to value a given stock necessarily constant over time? Explain. 4. Find an event reported in todays business press that is likely to have an important effect on the cash ows for a given rm. Use Yahoos nance Web site to produce a chart of the companys stock price around the time or day of the announcement of the event (http://nance. yahoo.com/). Explain why the market reacted the way it did. management, and analysts to use security returns to estimate the effects of various corporate policies and events on the market value of the corporation. For example, soon after Hurricane Katrina wreaked havoc on the Gulf Coast in summer 2005, the stocks of property insurance companies fell dramatically, while the stocks of companies such as Home Depot (which sells lumber and other building supplies) increased. According to the efcient markets hypothesis these reactions reect the stock markets unbiased estimate of the valuation effects of the storm. The evidence that security returns are a meaningful measure of rm performance also provides support for using equity-based compensation to provide incentives to top management. (Chapter 15 presents an analysis of incentive compensation.) Suggested Readings Self-Evaluation Problems R. Coase (1988), The Firm, the Market, and the Law (The University of Chicago Press: Chicago). J. Earwell, M. Milgate, and P. Newman (1989a), Allocation, Information, and Markets (W.W. Norton: New York). (1989b), The Invisible Hand (W.W. Norton: New York). F. Hayek (1945), The Use of Knowledge in Society, American Economic Review 35, 519530. M. Jensen and W. Meckling (1995), Specic and General Knowledge, and Organizational Structure, Journal of Applied Corporate Finance 8:2, 418. O. Williamson (1985), The Economic Institutions of Capitalism (Free Press: New York). 31. Two men, Tom Hanks and Forest Gump, have been marooned separately on the same deserted island. There are two activities each man can undertake to obtain food: shing and gathering coconuts. Tom can catch 40 sh per hour or gather 10 coconuts per hour. Forest can catch 10 sh per hour or gather 8 coconuts per hour. Answer the following questions: a. Does Tom have a comparative advantage in producing both products? Explain. b. Tom and Forest have not yet met. Tom is working 2 hours a day and producing (and consuming) 48 sh and 8 coconuts (note: the sh are very small). Forest is also working BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts Chapter 3 85 The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Markets, Organizations, and the Role of Knowledge 99 2 hours a day, but he is producing and consuming 15 sh and 4 coconuts. Now assume that Tom and Forest meet and develop a trading relationship. Come up with a production and trading scheme such that they can each work the same amount per day as before, but each is better off than before. Provide specic numbers to show how they are better off. 32. a. Suppose sugar has the demand curve P 50 5Q and the supply curve P 5Q. Compute the equilibrium price and quantity and show graphically. Calculate the consumer surplus and producer surplus associated with this outcome. b. What factors might cause the equilibrium price and quantity of sugar to change? Solutions to Self-Evaluation Problems 31. a. No, while Tom has an absolute advantage in producing both products, he only has a comparative advantage in shing. Forest has a comparative advantage in gathering coconuts. His marginal cost for gathering one coconut is 1.25 sh (10 8), while Tom has a marginal cost of 4 sh (40 10). b. One specific example is as follows: Say Tom produces only fish and Forest produces only coconuts. There will be a total of 80 fish and 16 coconuts. Now suppose they set a price of 2 fish per coconut and Tom buys 10 coconuts (for a price equal to 20 fish). Then Tom consumes 60 fish and 10 coconuts and Forest consumes 20 fish and 6 coconuts. They each consume more of both commodities, so they are each better off even though they are working the same amount as before. Many other examples could be constructed. 32. a. Set demand equal to supply: 50 5Q 5Q and solve for the equilibrium quantity, Q * 5. Place Q * into either the supply or demand equation and solve for the equilibrium price, P * $25. Graphically, the picture is: P $50 Supply C $25 P Demand 5 Q Triangle C pictures the consumer surplus. The area of a triangle is .5(Base Height). Therefore the consumer surplus is .5(5 25) $62.5. The producer surplus pictured by Triangle P is also $62.5. b. Changes in the equilibrium price and quantity are induced by shifts in either the supply or demand curves. Factors that affect demand include such things as consumer income and the prices of other goods. For example, if an increase in consumer income caused the demand curve for sugar to shift to the right (an increase in demand) both the equilibrium price and quantity would increase. Factors that affect supply of sugar include such things as the prices of inputs used in the production process (e.g., land, labor, and fuel prices). If an increase in the price of labor caused the supply curve to shift to the left (a decrease in supply), the equilibrium price of sugar would increase and the equilibrium quantity would decrease. 86 100 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 Review Questions I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge The McGrawHill Companies, 2009 Basic Concepts 31. What is Pareto efciency? Why do economists use this criterion for comparing alternative economic systems? 32. What is a property right? What role do property rights play in a market economy? 33. Twin brothers, Tom and Bill, constantly ght over toys. For instance, Tom will argue it is his turn to play with a toy, while Bill argues it is his turn. Their parents frequently have to intervene in these disputes. Their mom has conceived an idea that might reduce these conicts. In particular, every toy in the house would be owned by one of the boys. The owner would have complete authority over the use of the toy. The mom reasons that ownership would cut down on disputes. Any time there is an argument over a toy, the owner gets the nal and immediate say. The boys dad is concerned that this idea will prevent the boys from learning to share. He envisions that under the new system, Tom will not allow Bill to play with his toys and Bill will not allow Tom to play with his toys. The old system forces them to gure out a way to share the toys. Do you think that their dads concerns are valid? Explain. 34. Many economists favor free trade between nations. They argue that free trade will increase total world output and make people of trading nations better off. Discuss how this argument relates to concepts presented in the chapter. 35. What do you think will happen to the price and quantity of DVD players if a. The availability of good movies to play on DVD players increases? b. Personal income increases? c. The price of inputs used to produce DVD players decreases? d. Ticket prices at local movie theaters decline substantially? 36. Suppose that the U.S. government caps the price of milk at $1.00 per gallon. Prior to the cap milk sold for $1.00 per gallon. Picture the effects of the price cap using a supply and demand graph. Explain how the cap affects consumers and producers. 37. a. What is an externality? b. Why might externalities lead a rm to discharge too much pollution into a river? c. Congress has passed a law that limits the level of cotton dust within textile factories. Why might a textile rm allow too much cotton dust within its workplace? 38. What is the difference between general and specic knowledge? How can specic knowledge motivate the use of decentralized decision making? 39. Evaluate the following statement: Using free markets and the price system always results in a more efcient resource allocation than central planning. Just look at what happened in Eastern Europe. 310. a. What are contracting costs? b. Give a few examples of contracting costs. c. What effect does the existence of contracting costs have on market economies? 311. If markets are so wonderful, why do rms exist? 312. In certain professional sports, team owners own the players. Owners can sell or trade players to another team. However, players are not free to negotiate with other team owners on their own behalf. The team owners initially obtain the rights to players through an annual draft that is used to allocate new players among the teams in the league. They also can obtain the rights to players by purchasing them from another team. Players do not like this process and often argue that they should be free to negotiate with all teams in the sporting league. In this case, they would be free to play for the team that offers the most desirable contract. Owners argue that this change in rights would have a negative effect on the distribution of talent across teams. In particular, they argue that all the good players would end up on rich, media-center teams such as New York or Los Angeles (because these teams could afford to pay higher salaries). The inequity of players across teams would make the sport less interesting to fans and thus destroy the league. Do you think the owners argument is correct? Explain. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 87 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 101 313. The guide at the Washington Monument tells your 10-year-old nephew, Enjoy the monument. As a citizen, you are one of its owners. Your nephew asks you if that is true. What do you say? 314. Locust Hill Golf Club is a private country club. It charges an initiation fee of $23,000. When members quit the club, they receive no refund on their initiation fees. They simply lose their membership. Salt Lake Country Club is also a private golf course. At this club, members join by buying a membership certicate from a member who is leaving the club. The price of the membership is determined by supply and demand. Suppose that both clubs are considering installing a watering system. In each case, the watering system is expected to enhance the quality of the golf course signicantly. To nance these systems, members would pay a special assessment of $2,000 per year for the next 3 years. The proposals will be voted on by the memberships. Do you think that the membership is more likely to vote in favor of the proposal at Locust Hill or for the one at Salt Lake Country Club? Explain. 315. Critically evaluate the advice of the Providence Consulting Group, which recommended to your company, That you analyze all the business divisions in your company. Rank them on growth potential. Sell all the low-growth units and invest the money in the high-growth units. Make sure not to sell the high-growth units. 316. Suppose that the U.S. government begins charging a $1 sales tax to all consumers for each dress shirt they buy. a. What is likely to happen to the price (not including the tax) and quantity demanded of dress shirts? Show using supply and demand graphs. b. What is likely to happen to the demand for sport shirts (not taxed) and undershirts (which are worn primarily with dress shirts)? Explain. 317. Title-loan rms offer high-interest loans (the interest rate can exceed 200 percent per year) to high-risk customers. The title of a car often is used as collateral. If the borrower defaults on the loan, the company can repossess the car. Recently, the nancial press has reported stories of poor people who have had their cars repossessed by title lending companies. Legislation is being proposed in some states to make this lending practice illegal. A proponent of the law made the following argument. The market for loans is very competitive given all of the banks, savings and loans, and nance companies. Outlawing title lending will make poor people better off. It will motivate the lending companies to provide loans with less onerous terms. Thus low income people and people with bad credit histories will be able to obtain credit on more favorable terms. Do you agree with this argument? Explain. 318. Suppose that annual demand in the U.S. market for ice cream cones can be expressed as QD 800 0.2I 100P, where QD is the number of cones demanded in millions of cones, I equals average monthly income in dollars, and P is price in dollars per cone. Supply can be expressed as QS 200 150P (with the same units for quantity and price). a. Graph the demand and supply curves for ice cream cones, assuming that average monthly income is $2,000, and solve for the equilibrium price and quantity. b. Now assume that average monthly income drops to $750 and supply is unchanged. Draw the new demand curve on the same graph as used in (a) above and solve for the new equilibrium price and quantity. How would you describe the shift in demand intuitively? 319. The rent control agency of Rochester has found that aggregate demand is P 500 5Q D. Quantity, Q D, is measured in thousands of apartments. Price, P, equals the monthly rental rate in dollars. The citys board of realtors acknowledges that this is a good demand estimate and has shown that supply can be expressed as P 5Q S. a. If the agency and the board are right about demand and supply, respectively, what is the free-market price? How many apartments are rented? b. If we assume an average of 3 persons per apartment, what is the expected change in city population if the agency sets a maximum average monthly rent of $100 and all those who cannot nd an apartment leave the city? 88 102 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 1 I. Basic Concepts The McGrawHill Companies, 2009 3. Markets, Organizations, and the Role of Knowledge Basic Concepts 320. Assume that before the ice storm of 2003, the weekly demand and supply for ice in the Rochester Metro Area was given by the following equations: Dpre: P Spre: S 100 Q 5 0.5Q a. Draw a graph representing the Rochester ice market before the storm and label it carefully. What was the equilibrium price for the Rochester ice market before the storm? And the total quantity of ice traded? b. As a result of the ice storm, electricity went out in the Rochester area. The demand for ice increased due to the lack of electricity to power refrigerators. The lack of power also caused the supply to decrease. Ice producers were still able to produce some ice using electric generators. Other ice had to be imported from other areas with power. The relevant poststorm equations are the following: Dpost: P Spost: P 110 Q 10 2Q Draw a graph representing the Rochester ice market after the storm and label it carefully. What is the new equilibrium price? What is the quantity? c. An open-ad in a local newspaper, commenting on the dramatic increase in price of ice following the storm, stated: Obviously, avarice and greed won out over decency and morality as ice-vendors took advantage of the ice storm to increase prices and gouge their loyal customers. Do you agree with this statement? Explain. 321. Suppose the supply and demand for wheat is given by: Supply: Qs Demand: Qd 1,800 2,550 240P 10I 266P Where P the price per bushel of wheat and I income. The current value of I is 100. a. Find the current equilibrium price and quantity of wheat sold in the marketplace. b. Find the equilibrium price and quantity if income increases to 150. c. Show the change in equilibrium using a standard supply and demand graph. Make sure to label the axes and the curves. The graph does not need to be to scale. Just illustrate in a general way what is going on. 322. Assume that the demand curve for sporting guns is described by Q D 100 2p and the supply is described by Q S 20 p (Q D and Q S are in millions, p is in $). a. Compute the competitive equilibrium price and quantity. Draw a graph of a supply and demand curve and label it correctly. Compute the total value created in the market for sporting guns (hint: total value consumer surplus producer surplus). b. Suppose that the government views sporting guns as a luxury product and taxes the consumers $6 for each sporting gun they buy. Solve the new competitive equilibrium. What losses do consumers of sporting guns incur as a result of the tax? What losses, if any, do the producers of sporting guns incur? 323. Suppose there has been a storm in Nebraska that has destroyed part of the corn crop in the eld. The demand curve for corn has not changed. As a result, the market clearing prices and quantities before and after the storm are: Pb 50, Qb 2,000; Pa 100, Qa 1,500. (The subscripts a and b refer to after the storm and before the storm.) a. Assume a linear demand curve for corn; that is P Q. Calculate , with the provided information, and draw the demand curve with P on the y-axis and Q on the x-axis. Label the intercept and the slope on the graph. b. The supply curve for the period after the storm is P (1 15)Q, and it is parallel to the supply curve before the storm. Is the supply curve before the storm above or below that BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition I. Basic Concepts 3. Markets, Organizations, and the Role of Knowledge Chapter 3 The McGrawHill Companies, 2009 Markets, Organizations, and the Role of Knowledge 89 103 after the storm? Calculate the slope and the intercept of the supply curve before the storm. Draw both supply curves on a new graph with P on the y-axis and Q on the x-axis. Add the demand curve (calculated in part a) to the graph. c. Suppose consumers care only about corn consumption and apple consumption (they live in a two-good world). How would the change in the price of corn affect the budget constraint of the typical consumer? Show graphically. How would the change in relative prices affect the typical consumers consumption of corn versus apples? Is this result consistent with your observation from the demand and supply framework (i.e., an increase in the price of corn is associated with a decrease in the equilibrium quantity)? Explain. 90 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand The McGrawHill Companies, 2009 Demand CHAPTER 4 CHAPTER OUTLINE Demand Functions Demand Curves Law of Demand Elasticity of Demand Linear Demand Curves Other Factors That Inuence Demand Prices of Related Products Income Other Variables Industry versus Firm Demand Network Effects Product Attributes Product Life Cycles Demand Estimation Interviews Price Experimentation Statistical Analysis Implications Summary Appendix: Demand T he Players Theater Company is a regional repertory theater in the Midwest. Each year, it produces six plays, ranging from Shakespeare to contemporary musicals. PTC has priced its tickets at $30. On a typical night, approximately 200 of the theaters 500 seats are lled. The PTC board met recently to discuss a possible price decrease to $25 for next season. Advocates of the proposal argued that the decrease in ticket prices would increase the theaters customer base, the number of tickets sold, and revenues for the company. At the meeting, the PTC board engaged in a heated debate over the proposal. It soon became evident that the board had insufcient information to make a sound decision. For instance, nearby restaurants, which serve PTC customers, have indicated that they are planning to implement substantial price increases before the beginning of the next season. Would this increase affect the demand faced by PTC and thus the appropriate ticket price? Although customers might buy more tickets at lower prices, would total BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Chapter 4 Demand 91 107 revenue or prots necessarily increase? Would it be better to attract additional customers by lowering price or by improving the quality of PTC plays? After a lively discussion, the proposed decrease in price was tabled for further study. This discussion at the PTC board meeting highlights the fact that managers require a detailed understanding of product demand to make sound pricing decisions. Understanding product demand also is important for decisions on advertising, production levels, new product development, and capital investment projects. Chapter 3 offered a brief introduction to supply and demand analysis. In that chapter, we introduced the notion of a demand curve and briey discussed some of the factors that might cause a demand curve to shift. In this chapter we provide a more extensive analysis of demand. Important topics include demand functions, demand curves, factors affecting demand, industry versus rm-level demand, network effects, demand for product attributes, product life cycles, and demand estimation. In the technical appendix to this chapter, we derive point elasticities, analyze marginal revenue for a linear demand curve, and examine a special (log-linear) demand function. Demand Functions Managers require a fundamental knowledge of the factors that affect the demand for their product. Only by understanding these factors can they make sound decisions on pricing, output, capital expenditures, and other strategic issues. In fact, poor pricing decisions can destroy rm value and damage executive careers. Kraft Foods demoted the head of its North American business unit, Betsy Holden, after she raised prices and Kraft lost market share. Over this period Krafts stock price fell 20 percent while other food companies stock prices rose 9 percent.1 A demand function is a mathematical representation of the relation between the quantity demanded of a product and all factors that inuence this demand. In its most general form, a demand function can be written as Q f (X1, X2 , . . . Xn ) (4.1) where the Xi s are those factors that affect the demand for this product. The quantity demanded Q is the dependent variable in the demand function, since its value depends on the variables on the right-hand side of the equation. The Xi s are the independent variables. In this chapter, we focus on three particularly important independent variables: the price of the product, the prices of related products, and the incomes of potential customers. This analysis can be extended to include other variables, such as advertising expenditures, tastes and preferences, and consumer expectations (for example, about future prices). For concreteness, we continue to focus on PTC as an example. We assume that PTC faces a demand function for tickets on any given night that can be expressed by the following function2: Q 1 117 6.6P 1.66Ps 3.3Pr 0.0066I (4.2) S. Ellison and V. OConnell (2003), Kraft Removes Holden as Co-Chief, The Wall Street Journal (December 17), A3 and A8; and E. Herman (2005), Former Co-CEO Holden Leaves Kraft, Chicago Sun-Times (June 25). 2 Note that this function assumes that PTC can sell fractional tickets. This assumption does not have a material effect on our analysis. However, it allows us to draw continuous demand curves. One way to think of quantity in this example is as the average number of tickets sold for a performance. In this case, fractional tickets are possible. Note also that we assume that demand is constant for each performance by PTC. PTC performances are all scheduled for Friday and Saturday nights. If they expand their schedule to include weeknights or matinees, it is likely that demand conditions for these performances will differ and hence so should prices. These issues are discussed in Chapter 7. 92 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 108 Part 2 Figure 4.1 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics Demand Curves The left panel shows the demand curve for the Players Theater Company tickets. By convention, price is placed on the vertical axis, while quantity is placed on the horizontal axis. The equation for PTCs demand curve is: P 60 0.15Q. The curve indicates that, for example, 200 tickets are purchased at $30 and 133 tickets are purchased at $40. The right panel indicates that the demand curve shifts to the right as income increases from $50,000 to $51,000at each price, consumers buy more tickets. Movements along a demand curve are motivated by changes in price and are called changes in the quantity demanded. Movements of the entire demand curve are motivated by other factors, such as changes in income, and are referred to as changes in demand. Ticket price (in dollars) $ $ 61 60 Income = $51,000 60 Income = $50,000 D1 D D0 Q 400 Quantity of PTC tickets Q 400 406.6 Quantity of PTC tickets where P is the price of PTC tickets, Ps is the ticket price at a nearby symphony hall, Pr is the average meal price at nearby restaurants, and I is the average household income of area residents. As our starting point, we assume PTC tickets are currently priced at $30; symphony tickets and meals are priced at $50 and $40, respectively; income is $50,000. Given these values, the demand function implies that PTC sells 200 tickets per night. We now examine each of the independent variables in this demand function in more detail. Demand Curves The price of the product is particularly important in demand analysis for two reasons: First, prices are among the most important variables that customers consider in making purchasing decisions. Second, managers choose the price of their products; variables such as the prices of other products and income levels largely are beyond their control. Given its special importance, economic analysis traditionally singles out the effects of price from other independent variables in the demand function. A demand curve for a product displays for a particular period of time how many units will be purchased at each possible price, holding all other factors xed.3 The left panel of Figure 4.1 depicts the demand curve for PTC tickets. By convention, price is placed 3 Technical note: We derive an individuals demand curve from the indifference curve/budget line analysis in the appendix to Chapter 2. The price of one goodsay, foodis varied, holding the price of other goods and income xed. The persons optimal choices are recorded. The individuals demand curve simply plots the optimal choices of the good (in this case, food) against the associated prices. The rm-level demand curve, in turn, is the sum of the demands of all individuals at each price. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 93 The McGrawHill Companies, 2009 4. Demand Chapter 4 Demand 109 MANAGERIAL APPLICATIONS Learning the Law of Demand the Hard Way Mercury One-2-One is a British mobile-phone company. In a promotion to attract new customers, the company offered free telephone calls on Christmas to customers who signed on between November 8th and Christmas Eve. The company never dreamed its customers would be so generous in spreading the holiday cheer. The promotion generated more than 33,000 hours of calls, jamming the network and prompting hundreds of complaints from people who couldnt get through to place their calls. The volume on Christmas was about 10 times the daily average. Many people placed overseas calls and simply left the phone line open, logging free international calls of up to 12 hours. The average call was about 11 hours long; the typical caller rang up about $60 worth of callsequal to the average monthly bill of a cellular 2 company in the United States. The promotion ended up costing the rm millions of dollars. One member of Parliament vowed to le a complaint with Britains Board of Trade. To quote one executive of the company, Theres certainly been insatiable demand. Source: K. Pope (1994), Phone Companys Gift of Gab Jams Its Lines, The Wall Street Journal (December 28), B1. on the vertical axis, while quantity is placed on the horizontal axis.4 The equation for PTCs demand curve is: P 60 0.15Q (4.3) This expression is obtained by substituting the current values of the other variables into Equation (4.2) and solving for P. The equation indicates that, for example, 200 tickets are purchased at $30 and 133 tickets are purchased at $40.5 Demand curves hold other factors xed. Changes in income or the prices of symphony tickets or restaurant meals will cause shifts in the position of the demand curve (the intercept changes). For instance, the right panel of Figure 4.1 indicates that the demand curve shifts to the right as income increases from $50,000 to $51,000at each price, patrons purchase 6.6 more tickets. Movements along a demand curve reect changes in price and conventionally are called changes in the quantity demanded. Movements of the entire demand curve are caused by other factors (such as this change in income) and are referred to as changes in demand. Law of Demand As we discussed in Chapter 3, demand curves generally slope downwardindividuals purchase less (or certainly no more) of a product as the price increases. PTCs demand curve has a slope of 0.15. Although it is conceptually possible that individuals might purchase more of a product as the price rises, as a practical consideration, managers are quite safe in assuming that the quantity demanded for their products varies inversely with price.6 It would be foolish for PTC board members to think that they would sell more tickets if they raised the price. The negative slope of demand curves has become known as the law of demand. 4 In subsequent chapters, we consider costs which are a function of quantity produced. Placing P on the vertical axis allows us to display both demand (revenue) and costs on the same graph in a convenient fashion. 5 Rounded to the nearest dollar. 6 Goods for which the income effect swamps the substitution effect so that consumers purchase more at higher prices are called Gifn Goods. We ignore this possibility throughout the rest of this book. 94 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 110 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics ANALYZING MANAGERIAL DECISIONS: Setting Tuition and Financial Aid The Board of Ursinus College in Pennsylvania raised its tuition and fees 17.6 percent to $23,460 in 2000. It subsequently received 200 more applications than the year before. The president of the college surmised that applicants had apparently concluded that if the college cost more, it must be better. Other colleges that raised tuition to match rival colleges in recent years include University of Notre Dame, Bryn Mawr College, Rice University, and the University of Richmond. They also experienced an increase in applications. In contrast, North Carolina Wesleyan College lowered their tuition and fees about 10 years ago by 22 percent and attracted fewer students. The college president concluded that it didnt work out the way it had been hoped. People dont want cheap. You are hired as a consultant to a President of a liberal arts college in the East. You are asked to evaluate a recommendation by the colleges Admissions Director, Susan Hansen, to increase tuition and to reduce nancial aid to students. Susan argues that the data from competing colleges suggest that the demand curves for colleges slope upward the quantity demanded increases with price. Susan projects that the increase in tuition and reduction in nancial aid will solve the schools nancial problems. Last year, the college enrolled 400 new students who each paid an effective tuition of $15,000 (after nancial aid), totaling $6,000,000. She projects that with the increased demand from charging an effective tuition of $25,000, the college will be able to enroll 600 new students (of equal or better quality), totaling $15,000,000. Evaluate Susans analysis and recommendation. SOURCE: Jonathan D. Glater and Alan Finder (2006), In Twist on Tuition Game, Popularity Rises with Price, nytimes.com (December 12). Elasticity of Demand Demand curves vary in their sensitivity of quantity demanded to price. In some cases, a small change in price leads to a big change in quantity demanded, whereas in other cases a big price change leads to only a small change in quantity demanded. Information on this sensitivity is critically important for managerial decision making. For instance, the board would not want to lower ticket prices to $25 if it could ll the theater by reducing prices only to $28. One measure of the responsiveness of quantity demanded to price is simply the slope of the demand curve. But this measure is of limited usefulness, in part because it depends MANAGERIAL APPLICATIONS Increased Foreign Competition and Demand Elasticities Price elasticities for products usually increase with available substitutes. In recent years, there has been a dramatic increase in the amount of foreign competition facing many American companies. One result has been an increase in the demand elasticities for many American products. A specic example is lm produced by Eastman Kodak. For years, Kodak had a virtual worldwide monopoly in the production of lm. Correspondingly, consumers were relatively insensitive to the price of Kodak lmthey had no alternative sources. Kodak now faces intense pressure from Japans Fuji Corporation. Competition also comes from producers of store-brand lm, such as the 3-M Corporation in the United States (store-brand lm is sold under the store name at large discount drug, retail, and grocery stores). As a result, the demand for Kodak lm is much more price-elasticthe quantity sold is more sensitive to price changes. This change in price elasticities has motivated Kodak to change its pricing and product development strategies: It can no longer focus exclusively on selling high-quality lm at high prices. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Chapter 4 This gure displays two points on PTCs demand curve for theater tickets. As displayed in the gure, the arc elasticity between these two points is 1.4. Thus, over this range, for every 1 percent increase in price there is approximately a 1.4 percent reduction in the quantity of tickets purchased. 111 $ Arc Elasticity = [Q/(Q1 + Q2)/2] [P/(P1 + P2)/2] = [67/(200 + 133)/2] [10/(30 + 40)/2] 60 Ticket price (in dollars) Figure 4.2 Demand 95 = 1.4 (133, $40) P2 (200, $30) P1 D Q Q1 400 Q2 Quantity of PTC tickets on the particular dimensions in which quantities are quoted. For instance, if the slope of a demand curve is 2 when the quantity is expressed in tons, it is only 0.001 when the quantity is stated in pounds. Using the magnitude of the slope coefcient to derive insights into the sensitivity of quantity demanded to price requires additional computation. Economists more frequently use a dimensionless measure of this sensitivity known as the price elasticity of demand, . (Frequently, this elasticity is simply referred to as the elasticity of demand.) Demand elasticity measures the percentage change in quantity demanded given a percentage change in its price. The law of demand indicates that price elasticities are negative; convention, however, dictates that we state this elasticity as a positive number. Higher price elasticities mean greater price sensitivity. The elasticity of demand, , thus is given by (% change in Q ) (% change in P ) (4.4) Calculating Price Elasticities This elasticity can be approximated between any two points using the concept of arc elasticity.7 The formula for an arc elasticity is [ Q (Q1 Q2) 2] [ P (P1 P2) 2] (4.5) 8 where represents the change between the two points. Figure 4.2 displays two points on PTCs demand curve for theater tickets. As shown in the gure, the arc elasticity between 7 Price elasticity can be measured at a point on the demand curve. The concept of point elasticity requires elementary knowledge of calculus and, more importantly, a smooth mathematical demand curve. While our example assumes such a curve, data on demand often is available for only a few price-quantity combinations. We show how to calculate point elasticities in the appendix to this chapter. 8 Equation (4.4) can be expressed as QQ P P. When calculating the elasticity between two points, the question arises as to which Q and P to use in this expression, the starting or ending values. Equation (4.5) uses the average of these two valuesthe initial plus the ending values divided by 2. 96 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 112 Part 2 Figure 4.3 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics Range of Price Elasticities Price elasticities lie between zero and innity. If the price elasticity is zero, quantity demanded is unaffected by price. In this case, as depicted in the left panel of the gure, the demand curve is vertical. If the price elasticity is innite, as in the right panel, a small increase in price will cause people to purchase none of the product, and the demand curve is a horizontal line. $ $ Price (in dollars) Price (in dollars) D ( = 0) D ( = ) Q Quantity Q Quantity these two points is 1.4. Hence, over this region, for every 1 percent increase in price, patrons reduce the quantity of tickets purchased by approximately 1.4 percent. Price elasticities lie between zero and innity. If the price elasticity is zero, quantity demanded is unaffected by price. In this case, as depicted in the left panel of Figure 4.3, the demand curve is vertical. If the price elasticity is innite, a small increase in price will lead customers to purchase none of the product. In this case, as displayed in the right panel of Figure 4.3, the demand curve is a horizontal line. For instance, a small farmer might not be able to sell any soybeans if they were priced above the prevailing market price. Demand is elastic if the price elasticity is greater than one, unitary if equal to one, and inelastic if less than one. Elasticity varies along most demand curves. For instance, with a linear demand curve, elasticity will be high when quantities are low and approach zero as the quantities become large. (Try calculating some arc elasticities along PTCs demand curve.) We discuss this topic in greater detail below. (In the appendix to this chapter, we present a special demand curve that has constant elasticityit does not vary along the curve.) Price Changes and Total Revenue One of the boards concerns is how total revenue changes if it lowers ticket prices. We now demonstrate that the relation between revenue and price depends on the demand elasticity. Total revenue is calculated by multiplying the quantity purchased times the price (that is, P Q ). If price elasticity is inelastic (less than one), then the quantity demanded is less responsive to a change in price; a 1 percent increase in price results in less than a 1 percent decrease in quantity. Thus total revenue increases. Conversely, a price decrease results in a decrease in revenue. In contrast, if demand elasticity is unitary (equal to one), a 1 percent change in price results in an offsetting 1 percent change in quantity and hence total revenue is unchanged. Finally, if demand is elastic (value greater than one), a small increase in price results in a decline in revenue, whereas a small BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Chapter 4 Figure 4.4 Revenue Price Elasticities, Price Changes, and Total How total expenditures on a product change with price depends directly on the price elasticity. This gure displays the relation between small price changes, total revenue, and price elasticities. Demand 97 113 Inelastic demand ( < 1) P Total revenue P Total revenue Unitary elasticity ( = 1) P = No change in total revenue Elastic demand ( > 1) P Total revenue P Total revenue decrease in price results in an increase in revenue. These relations are summarized in Figure 4.4. We discuss these relations in greater detail below. Determinants of Price Elasticities The elasticity of demand tends to be high when there are good substitutes for the product. For instance, if a ight is overbooked airlines have little trouble nding volunteers to surrender their seats when alternate ights are available that involve no material delay in arrival. Conversly, if the overbooked ight is the last of the day, to elicit volunteers might require several free tickets. The elasticity of demand for PTC tickets is likely to increase with the number of competing events in the city. With many entertainment options, a small increase in the price of PTC tickets might be sufcient to induce a substantial number of potential consumers to attend other events. When alternatives are more limited, additional customers will decide to pay the higher price for PTC tickets rather than just stay at home. Demand elasticities also can depend on the importance of the goods within consumers budgets. Goods such as salt and pepper, which consume a relatively small proportion of a persons income, tend to be relatively price-insensitive, or inelastic. On the other hand, goods such as major appliances and automobiles represent more substantial MANAGERIAL APPLICATIONS Price Elasticities Economists have estimated the price elasticities of various products, such as Sugar 0.31 Potatoes 0.31 Tires 1.20 Electricity 1.20 Haddock 2.20 Movies 3.70 These estimates indicate that sugar and potatoes have relatively low price elasticities. This might be expected given that these products represent a small portion of most peoples budgets. Also, sugar has few close substitutes. Haddock and movies have high elasticities. Haddock is a narrowly dened product (as opposed to sh) and has many close substitutes. Movies are a luxury item for many people; higher prices cause individuals to consume other forms of entertainment. Source: E. Manseld and G. Yohe (2004), Microeconomics (W.W. Norton: New York), 135. 98 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 114 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics MANAGERIAL APPLICATIONS Demand Elasticities and Airline Pricing Round-trip airfares are substantially lower if the traveler stays over a Saturday night. Airline companies offer this discount to increase revenues (and prots). The typical traveler who stays over a Saturday night is a tourist. Tourists have relatively high price elasticities for air travel. Lowering the price from the standard fare correspondingly increases revenue: The price decrease is more than offset by the increase in tickets sold. Airline companies do not offer comparable discounts to travelers who complete the round-trip midweek. These customers are primarily business travelers who have relatively inelastic demands. Lowering price would decrease revenue because the decrease in price would not be offset by an increase in tickets sold. Airline companies also offer fewer discounts during peak periods, such as the period around the Thanksgiving holiday. During these periods, demand is relatively inelastic and they can ll the planes without offering substantial discounts. (Chapter 7 presents a detailed discussion of product pricing.) purchases. Customers are more likely to comparison-shop to collect product information and thus are more likely to be price-sensitive. A third determinant of price elasticity is the length of the period to which the demand curve pertains. Demand tends to be more elastic or responsive to price changes over a longer period than within a shorter period. An increase in PTC ticket prices is likely to result in an immediate decline in tickets sold. Long-run effects will be even larger as consumers identify other entertainment options or fail to renew season tickets (these effects will cause the demand curve to shift to the left). Similarly, a large increase in the price of oil will result in a near-term decline in the quantity of oil demanded people will set their thermostats to a lower temperature and drive less. Over time, the effect will be larger as consumers insulate their homes better, buy smaller, more fuelefcient cars, and shift to alternative energy sources. Linear Demand Curves The PTC boards decision on whether or not to lower prices depends on the relation between price and total revenue and thus its demand elasticity: It would make little sense for the board to lower prices if a price reduction would lower total revenue. We now provide a more in-depth analysis of the relation between price and revenue and discuss the PTC boards optimal pricing policy. Through this analysis, we illustrate the properties of linear demand curves. Knowing these properties is useful for understanding the subsequent analysis in this book.9 Total Revenue PTCs total revenue (TR) for any given performance is equal to the quantity of tickets sold times the price. Price is given by the demand curve in Equation (4.3). Thus, total revenue can be expressed as TR 9 PQ (60 0.15Q )Q 60Q 0.15Q 2 (4.6) Technical note: This result also can be applied to nonlinear demand curves. At a specic point, construct the tangent to the demand curve. Now, if the tangency point is at the midpoint of the tangent, the elasticity is unitary, and so forth. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 4 Ticket price (in dollars) This gure displays PTCs demand and total revenue curves in the upper and lower panels. Total revenue increases as price decreases up to the midpoint of the demand curve. Thus, over this range, demand is elastic: The percentage decline in price is smaller than the percentage increase in quantity demanded. The elasticity is unitary at the midpoint. Past the midpoint, price declines result in reduced total revenue; and thus, demand is inelastic over this range. An important concept in economics is marginal revenue, which is dened as the change in total revenue given a unitary change in quantity. In the appendix, we show that marginal revenue (MR) for a linear demand curve is a line with the same intercept as the demand curve but with twice the negative slope. The marginal revenue curve for PTC is pictured in the gure. Demand 115 $ 60 Elastic demand ( > 1) =1 30 Inelastic demand ( < 1) D MR Q $ Total revenue (in dollars) Figure 4.5 Demand, Total Revenue, and Marginal Revenue for Linear Demand Curves 99 The McGrawHill Companies, 2009 4. Demand 6,000 Q 200 Quantity of PTC tickets Figure 4.5 displays PTCs demand and total revenue curves. Total revenue increases as price decreases up to the midpoint of the demand curve. Over this range, demand is elastic: The percentage decline in price is smaller than the percentage increase in quantity demanded. The elasticity is unitary at the midpoint. Past the midpoint, price declines result in reduced total revenue; thus, demand is inelastic over this range. These are general properties of linear demand curves. Marginal Revenue An important concept in economics is marginal revenue, which is dened as the change in total revenue given a one-unit change in quantity. Intuitively, marginal revenue for the rst unit is just its price. Thus, the intercepts of the demand and marginal revenue curves are the same. As quantity increases, marginal revenue is below priceto sell an extra unit, the price charged for all units must decrease. Marginal revenue is positive up to the midpoint of the demand curve (total revenue is increasing over this interval). At the midpoint, demand elasticity is unitary and marginal revenue is zero. Beyond the midpoint, marginal revenue is negativethe increase in revenue from selling another unit is less than the decline in revenue from lowering price (see the appendix). Hence, marginal 100 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 116 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics ANALYZING MANAGERIAL DECISIONS: Demand Curve for an Electronics Product You work for a company in India that manufactures and exports batteries and other charge storage devices. You are the sales manager for a DC-DC converter that is used to step up or step down the voltage in various industrial applications. You currently price the product at 4,000 Indian Rupees (INR) and sell 100,000 units. You estimate that if you priced the product at 3,000 INR you would sell 150,000 units. You think it is reasonable to assume that your demand curve is linear. 1. Derive the equation for your demand curve from the two price and sales points discussed above. 2. Are you currently operating in the elastic or inelastic portion of your demand curve? 3. You are paid a sales commission based on your total sales. What price would you charge to maximize your bonus? 4. Is this price likely to be optimal from your rms standpoint, which has prot maximization as a goal? revenue (MR) for a linear demand curve is a line with the same intercept as the demand curve but with twice the negative slope (see Figure 4.5). The equation for PTCs marginal revenue is MR 60 0.3Q (4.7) Prot Maximization All of PTCs costs are xed and do not depend on the quantity of tickets sold on a given eveningactors and utilities have to be paid regardless of how many people attend the performance. Thus, the PTC boards objective is to maximize total revenue (for PTC, with costs xed, maximizing total revenue is equivalent to maximizing total prot). Figure 4.5 indicates that revenues are maximized at a price of $30. Hence under current conditions, the PTC board should not lower the ticket price to $25. Currently, the company is collecting $30 200 $6,000 in revenue per night. If the price is decreased to $25, total revenue would be $25 233 $5,825 per night.10 (The upcoming increase in restaurant prices will change the optimal pricing policy. A practice problem at the end of this chapter explores this policy change.) Note that, in contrast to this example, most rms do not want to maximize total revenue. PTC, with only xed costs, is a special case. In most rms, both costs and revenues vary with output. A prot-maximizing rm must consider both effects. We discuss these considerations in greater detail in Chapters 5 through 7. Other Factors That Inuence Demand In addition to a products own price, the prices of related products and incomes of potential customers are among the more important factors that influence product demand. 10 From the demand curve: $25 60 0.15Q. Therefore, Q 233. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand 101 The McGrawHill Companies, 2009 Chapter 4 Demand 117 MANAGERIAL APPLICATIONS Complementarity between Computer Hardware and Software Over the past decade, there has been a dramatic decrease in the price of personal computers. Not only has the price of PCs decreased, but their quality and computing power have improved substantially as well. This decrease in the price of personal computers has increased the quantity of PCs demanded enormously. In addition, it also has increased the demand for software products. Today, some of the largest companies in the world (for example, the Microsoft Corporation) specialize in the production of software for PCs. Computer hardware and software are complements and thus have negative cross elasticities. Prices of Related Products Complements versus Substitutes The demand for a product can be affected by the prices of related products. For instance, if the local symphony raises its ticket prices, arts patrons will be less likely to attend the symphony and more likely to attend the PTC. Thus, there is a positive relation in Equation (4.3) between the demand for PTC tickets and the price of symphony tickets. Goods that compete with each other in this manner are referred to as substitutes. In contrast, if local restaurants raise their prices, the demand for PTC tickets falls (note the negative sign in the demand function). For instance, some potential PTC customers will choose to stay home because the total cost of an evening on the town has increased. Products like theater tickets and meals at restaurants, which tend to be consumed together, are complements. Another example of complements is digital video disk players and DVD movies. Between 1997 and 1999, the price of DVD players fell from $600 to $299. Sales of DVD players rose from under 50,000 to 600,000 over these two years. And as consumers experience the better sound quality and video, they also are buying bigger TVs and better sound systems. Big-screen TVs were up 12 percent and audio sales 11 percent. The biggest market driver is DVD, said Terry Shimek, owner of Shimeks Audio Video in Anchorage, Alaska. Moreover, the demand for DVD movies increased as well. Initially caught at-footed, Hollywood started making more movies available on DVD. The number jumped from 1,800 in 1988 to 5,000 in 1999. (That is compared to 18,000 on VHS tape.)11 MANAGERIAL APPLICATIONS Derived Demand Some products are demanded, not because individuals receive pleasure from consuming them, but rather because they are useful in the consumption of other products. Demands for these products are derived from the demands from other products. Take motor oil, for example. Few people derive satisfaction from purchasing oil for their automobiles. Rather, it is a derived demand from consuming transportation services provided by their cars. Procter & Gamble (P&G) discovered that spraying a bit of their Clean Shower bathroom cleaning product on a razor each day can extend the razors life three or four times. They are formulating a product targeted to this use. Thus, this new products demand is derived from the demand for razor blades. 11 E. Ramsted (1999), As Prices Tumble, Sales of DVD Players Explode for the Holidays, The Wall Street Journal (December 9), 31. 102 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 118 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics MANAGERIAL APPLICATIONS Estimates of Cross Elasticities Economists have estimated the cross elasticities for various commodities. Below are a few of these estimates: Electricity and natural gas 0.20 Beef and pork 0.28 Natural gas and fuel oil 0.44 Margarine and butter 0.81 All the pairs of commodities listed above are substitutes. Complements such as DVD players and DVD movies have negative cross elasticities. Natural gas apparently is not a very strong substitute for electricity. Although people can use either gas or electricity for heating, natural gas is not generally used for lighting. On the other hand, natural gas and fuel oil are closer substitutes (both tend to be used for heating). Margarine and butter are strong substitutes. Source: E. Manseld and G. Yohe (2004), Microeconomics (W. W. Norton: New York), 135. Cross Elasticities One frequently used measure of substitution between two products is the cross elasticity of demand. Cross elasticity is dened as the percentage change in the quantity demanded of a good, given a percentage change in the price of some other good. Cross elasticities between any two goods, X and Y, can be calculated using a formula that is analogous to Equation (4.5): xy [ Qx (Qx1 Qx2) 2] [ Py (Py1 Py2) 2] (4.8) Unlike price elasticities, which are invariably positive (at least when you multiply them by 1), cross elasticities can be either positive or negativesubstitutes have positive cross elasticities whereas complements have negative cross elasticities. Whether a commodity has strong substitutes or complements depends, in part, on how nely the commodity is dened. Pepsi and Coke might have relatively large cross elasticities. The cross elasticities between colas, more broadly dened, and other soft drinks are likely to be smaller.12 Cross elasticities are useful because managers frequently want to forecast what will happen to their own sales as other companies change their prices. The PTC board is concerned about the effects that a forthcoming increase in restaurant prices would have on its ticket demand. If meals in local restaurants and theater tickets are strong complements, a substantial increase in restaurant prices would cause a serious decline in the demand for PTC tickets. In this case, the PTC board might want to offset this shift in demand by lowering ticket prices or advertising more heavily. In contrast, if meals and tickets are weak complements, the increase in meal prices would have little effect on ticket demand. In our example, a $10 increase in meal prices will result in 33 fewer ticket sales per night. Using the formula in Equation (4.8), the corresponding cross elasticity between these two points [(200, $40); (167, $50)] is 0.81: For every 1 percent 12 Next, we extend this discussion to show how cross elasticities can be used by managers to dene a rms industry. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 103 The McGrawHill Companies, 2009 4. Demand Chapter 4 Demand 119 MANAGERIAL APPLICATIONS Russian Cola Wars In 1999, Crazy Cola had a 48 percent market share in Krasnoyarsk, Russia. Crazy Cola, produced locally by OAO Pikra, is headed by a 60-year-old former Communist factory worker named Yevgeniya Kuznetsova. Ms. Kuznetsova formerly bottled Pepsi at a state-run plant. A 1.5-liter bottle of Crazy Cola sold for about 39 cents, compared to 77 cents for a two-liter bottle of Coke or Pepsi. Krasnoyarsk is a poor community, and many residents are unwilling to pay a premium for brand-name colas. To quote one 25-year-old graduate student, Viktoria Pimenova, Crazy Cola is fun, and its our local product. But it is a drink for people who dont have money. Coke and Pepsi taste better. This statement suggests that Crazy Cola is an inferior good, while Coke and Pepsi are normal goods. This implies that if the incomes of local residents increase, demand for Coke and Pepsi will increase, while the demand for Crazy Cola will decrease. Source: B. McKay (1999), Siberian Soft-Drink Queen Outmarkets Coke and Pepsi, The Wall Street Journal (August 23), B1. increase in meal prices over this range there is, on average, a 0.81 percent decline in ticket sales. This elasticity suggests that PTC tickets and restaurant meals are rather strong complements. Income Normal versus Inferior Goods Another factor that frequently affects the demand for a product is the income of potential buyers. As a persons income increases, more products are purchased, and the combined expenditures across all products rise. The demand for specic products, however, can either rise or fall as income increases. The demand for goods such as gourmet foods or jewelry would be expected to increase with income, whereas the demand for other goods like canned processed meat or cabbage might decline. Goods for which demand increases with income are called normal goods. PTC tickets are normal goods. Goods for which demand declines with income are called inferior goods. Income Elasticities The sensitivity of demand to income is measured by their income elasticity. The income elasticity is dened as the percentage change in the demand for a good, given a percentage change in income (I). Income elasticities can be calculated using the following formula. I [ Q (Q1 Q2) 2] [ I (I1 I2) 2] (4.9) The income elasticity is positive for normal goods and negative for inferior goods. The income elasticity of a rms product has important implications. Firms producing products with high income elasticities are more affected by cyclical uctuations; they tend to grow more rapidly in expanding economies but contract more sharply in depressed economies. Managers must anticipate these uctuations in managing cash ows and hiring decisions. Demands for products with small income elasticities are more stable over economic cycles. Studies indicate that goods like domestic servants, medical care, education for children, and restaurant meals tend to have relatively large income 104 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 120 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics MANAGERIAL APPLICATIONS Estimates of Income Elasticities Economists have estimated the income elasticities for various products. Below are a few of these estimates: Flour 0.36 Margarine 0.20 Milk 0.07 Meat 0.35 Dentist services 1.41 Restaurant consumption 1.48 According to these estimates, our and margarine are inferior goods. People spend less on these goods as their incomes rise. The other goods are normal goods (expenditures on the products rise with income). Dentist services and restaurant consumption are particularly sensitive to income changes. Source: E. Manseld and G. Yohe (2004), Microeconomics (W. W. Norton: New York), 135. elasticities, whereas goods such as most food products, gasoline, oil, and liquor have relatively small (in absolute value) income elasticities. Income elasticities also can inuence location decisions. For instance, PTC has a relatively high income elasticity (above 1.6). This elasticity was one of the factors that motivated the founders to locate their theater in a community with a high per capita income. They anticipated that they would have fewer customers if they located in a less afuent area. Other Variables We have concentrated on three of the more important independent variables in most demand functionsthe products own price, prices of related products, and income. Other variables, such as advertising expenditures, also can be important. In all cases, the MANAGERIAL APPLICATIONS A Pampered Dog Loses His Stylist Even wealthy consumers are affected by changes in income. Consider Betsy Illium, a marketing consultant to medical practices, who owns four Manhattan apartments (three are investment properties). In early 2008 Illium became concerned that her income would decline given forecasts of a looming economic recession. Its frightening, she said, with much of her wealth being tied up in real estate. In response she decided to replace her dog Dobbins regular groomer, who charged $130, with one from Petco, which charged only $65. She also decided to send her dirty sheets and towels to a laundry service rather than to a higher priced dry cleaner. In her case, high-priced dog groomers and dry cleaning are normal goods (her demands for these products increase with income), while laundry services and low-priced dog groomers are inferior goods (her demands for these products decreased with income). Illium became appalled when she calculated that Dobbins grooming, her own weekly hair, nail and massage appointments, gourmet groceries, restaurant meals and Starbucks coffee cost nearly $2,000 a month. She realized that she would have to tighten her alligator belt due to the threat of recession. Source: S. Rosenbloom (2008), Tightening the Alligator Belt, nytimes.com (January 27). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand The McGrawHill Companies, 2009 Chapter 4 Demand 105 121 analysis is similar. Demand responds to a change in some other variable. Sensitivity can be measured by the appropriate elasticityfor instance, an advertising elasticity. Obviously, managers do not have the time to consider all the conceivable variables that might have trivial impacts on the demand for their products. Good decision making requires that managers understand the effects of the more important factors, which usually include the products own price, the prices of close substitutes and important complements, and incomes. Industry versus Firm Demand Industry Demand Curves Although we have concentrated our analysis on rm-level demand, demand functions and demand curves can be dened for entire industries. For instance, a demand function could be specied for the entertainment industry in PTCs market area. Such a function would relate the total ticket sales for all entertainment events to factors that affect this demand. Managers often are interested in total industry demand because it provides important information on the size of their potential markets and trends that affect them. For instance, a companys executives might judge the performance of a store manager that reports at sales quite differently if market demand is shrinking and the store is increasing market share versus a case where market demand is increasing but the store is losing market share. Moreover, estimates of industry demand sometimes can be obtained at modest cost from outside analysts or business publications. Firms within an industry compete directly and their products are likely to be relatively strong substitutes. The overall industry, on the other hand, is less likely to have strong substitutes. A person wanting to go to an entertainment event might choose among several options based on price. Entertainment events more broadly dened have fewer alternatives. Thus, demands facing individual rms within an industry tend to be more price-elastic than those for the entire industry. Dening Industry and Market Area We have indicated that managers can gain important insights by analyzing industrylevel demand. One problem that managers face in conducting this type of analysis is dening the relevant industry and market area. Is PTC competing in the live theater industry or in a more broadly dened entertainment industry? Cross elasticities provide important information to answer these types of questions. The cross elasticity between PTC tickets and symphony tickets is 0.4. This relatively high value (see the box titled Estimates of Cross Elasticities presented earlier in this chapter) suggests that PTC competes against companies in a broader entertainment industry than just live theater.13 The managers at PTC also must dene the relevant geographic area of their marketplace. If PTC raises its prices, will its customers shift to theaters in other nearby cities? If so, these cities should be included in the denition of PTCs market area. 13 Cross elasticities also are used as evidence in antitrust cases. Antitrust cases generally focus on whether or not a company has signicant market power within an industry. Thus, the denition of the industry is quite important. A company might have a signicant market share (and thus apparent power) in a narrowly dened industry, but a small market share in a more broadly dened market. For instance, the government suggested that the ReaLemon Company had monopolized the reconstituted concentrated lemon juice marketsupplying over 90 percent of that market. The company responded that the appropriate market denition was broader: They faced vigorous competition from reconstituted natural-strength lemon juice, fresh lemons, frozen lemonade, lime juice, and so on. 106 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 122 Part 2 II. Managerial Economics 4. Demand The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS 9/11 Causes Massive Shifts in Demand Curves The 9/11 terrorist attacks on the World Trade Center and Pentagon caused the demand curves for numerous goods and services to shift. Consider the following examples of industries whose demand curves shifted left: Hotels lowered room rates by 30 to 40 percent. Marriott Hotels put a third of its hourly employees on part-time schedules. Disney trimmed the hours at its theme parks and reduced the hours of 40,000 part-time workers. The U.S. airline industry estimated that 9/11 caused it to lose about $7 billion and saw its revenues drop by 40 percent. Following 9/11 many planes ew at less than 50 percent capacity and airlines laid off 20 percent of their workforce. The U.S. government provided $15 billion of loan guarantees to the airline industry. Spot prices for crude oil plummeted to a two-year low of $20 per barrel. But not all rms saw the demand for their goods and services adversely affected by 9/11. The following rms saw their demand curves shift to the right: Sales of American ags skyrocketed. Annin & Co., a 675-employee ag maker and market leader, increased output of its most popular 3-by-5-foot U.S. ags from 30,000 to 100,000 a week by October 2001. Its ve plants added shifts, but dealers still had to wait 15 weeks for delivery of popular-size ags. The demand for biometric systems increased. Biometric systems identify people using digital face, iris, or ngerprint scans. The biometric trade association doubled its worldwide revenue growth forecasts following 9/11. The three largest publicly traded biometric rms stock prices rose on average 70 percent when the stock markets reopened on 9/17. InVision Technologies, which makes bomb-detection systems for airports that cost between $750,000 and $1.5 million each, saw its stock price increase 165 percent on the rst trading session after the markets reopened. The U.S. government increased the defense departments budget by $10 to $15 billion. A signicant percentage of those funds were expected to be spent on intelligence technologies, such as those produced by software companies like Narus, which track and sort data and e-mail on the Internet to identify potential terrorists. On average, publicly traded companies suffered signicant reductions in demand. For instance, the stock market suffered its worst one-week loss since the Great Depression; following 9 11 stocks lost $1.2 trillion in market value. Nonetheless, there were winners scattered among the larger population of losers. Sources: E. Brown, Heartbreak Hotel? Fortune (November 26, 2001), 161163; M. Gunther, The Wary World of Disney, Fortune (October 15, 2001), 104; K. Marron, Systems That Use Physical Traits to Control Access to Sensitive Data Are Catching On in Post-Sept. 11 Era, The Globe and Mail (March 28, 2002), B16; B. OKeefe, Securing the Air, One Bag at a Time, Fortune (October 15, 2001), 244; B. Powell, The Economy under Siege, Fortune (October 15, 2001), 87108; J. Simons, Greed Meets Terror, Fortune (October 29, 2001), 145146; S. Tully, From Bad to Worse, Fortune (October 15, 2001), 118128; D. Voreacos, As Country Wraps Itself in Flags, Company Strains to Make Them, Houston Chronicle (October 7, 2001), 10; M. Warner, Web Warriors, Fortune (October 15, 2001), 148. Network Effects For some products, demand increases with the number of users. For example, fax machines and telephones are not particularly useful unless there is a network of users. This consideration is quite important for many of todays communication and information products. For instance, consumers were reluctant to buy new products, such BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand 107 The McGrawHill Companies, 2009 Chapter 4 Demand 123 MANAGERIAL APPLICATIONS Store Layout Affects Demand Paco Underhill calls himself a retail anthropologist. His consulting rm videotapes consumers as they shop at his clients stores such as Sears, The Gap, and McDonalds. He then offers recommendations for store layout. For example, most North Americans turn right after entering a store while most British and Australian customers turn left. Consumers tend to avoid narrow aisles; they apparently dislike being jostled from behind (what he calls the butt-brush factor). Junk food should be placed on low or middle shelves so kids can reach them. After nding that women spend only half the time in the store when accompanied by a man, he recommends placing numerous chairs around stores so men can sit comfortably while the women shop. Source: K. Labich (1999), Attention Shoppers: This Man Is Watching You, Fortune (July 19), 131133. as high-denition television sets, DVD players, and new word processing programs, until they became convinced that the products had the potential for widespread use. Consumers understand that if a product does not garner sufcient demand, important complementary products, such as DVD movies, will not be produced in high volumes or at attractive prices. Also they worry that it will be difcult to acquire reliable, inexpensive service for the new product. They may learn to use a new technology only to nd that it is discontinued because of insufcient demand. For example, despite vigorous efforts by Sony to promote its Betamax technology for video recording, the technology was displaced completely by the VHS format. Now VHS is suffering a similar fate from DVDs. Products where these network concerns are important often have relatively elastic demands. When price is lowered, there are two effects. One is the standard price effect: Consumers purchase more of the product because it is being sold at a more attractive price. The second is the network effect: Demand for the product increases even more because more people are using the product. When a new consumer purchases the product, there is an externality for other users; because there is an additional user of the product, the product becomes more attractive for other current and potential users. Network effects have important implications not only on product pricing (we discuss this issue more in Chapter 7), but on product design as well. For instance, when software manufacturers are designing software upgrades, they have to decide whether to make the new product compatible with prior versions of the software and with competing products on the market. Making a new product compatible with competing products can reduce the uniqueness of the product. However, the net effect can be to increase overall demand for the product because of network effects. For example, a consumer might be more willing to buy Microsoft Word if it is compatible with WordPerfect because of the enhanced ability to interact with WordPerfect users. MANAGERIAL APPLICATIONS Demand Elasticity for Gasoline The industry-level demand for gasoline is relatively inelastic: The price of gasoline can change substantially and have little effect on the overall quantity demanded. The demand elasticities facing individual gas stations are much larger. If several gas stations are located at the same intersection, an individual station can suffer a remarkable loss of business to its local competitors by raising its price. 108 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 124 Part 2 II. Managerial Economics 4. Demand The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS eBay and Network Effects The largest online auction site, eBay, is one of the few thriving dot-coms. Recently it has been growing at 72 percent per year. It had reported prots in 2001 of $90 million on revenues of $749 million and attracted 37 million customers per quarter. Over a million sellers offer products on eBay. Many sellers work full-time selling products over eBay; for example, Angie Cash sells $10,000 a month of items that cost no more than $20 each. eBay does lots of things right such as clever e-mails, message boards, and a self-monitoring system whereby buyers and sellers rate each others performance. eBay is a trading platform that exploits network effects. The demand for eBay increases as the number of buyers and sellers increases. More sellers increase the chance that buyers will nd what they want, and more buyers increase the chance sellers will sell their products at higher prices. Source: E. Brown, How Can a Dot-Com Be This Hot? Fortune ( January 21, 2002), 7884. Product Attributes Thus far, we have taken the attributes of the product as given. Our analysis of the demand for PTC tickets is based on the existing quality and selection of plays, their starting times, the quality of seating, and so on. Given these characteristics, we examined how price and other factors affect the demand for PTC tickets. Understanding consumer demand also plays an important role in the design of the product. For instance, do local patrons prefer Shakespeare or more contemporary plays? Do they prefer mysteries or musicals? Do they value comfortable seating with additional leg room or seating that is closer to the stage? Can the anticipated decrease in demand from increased restaurant prices be offset by changing the starting time of the plays? (Delaying the starting time by thirty minutes might give people more time to eat at home before going to the play.) Answers to these types of questions are important in managerial decision making and establishing corporate strategy. Indeed, when managers speak of the importance of understanding consumer demand, they often are referring to understanding the specic product attributes that are important to customers. (We discuss corporate strategy formulation in Chapter 8.) Marketing managers are responsible for understanding the broad range of product attributes that affect demand. These include price, product design, packaging, promotion and advertising, and distribution channels.14 This broad focus on demand has played an especially important role in management innovations like total quality management programs (see Web Chapter 23, http://www.mhhe.com/ brickley5). An important problem facing most rms is how to incorporate information that may be held by many people throughout a rmfor example, about such matters as consumer demandinto the decision-making process for product design. We defer discussions of this problem until Parts 3 and 4 of this book. These sections provide insights into how to design the rms organizational architecture to help ensure that relevant information is incorporated in the decision-making process. 14 For a more formal economic analysis of the demand for product attributes, see K. Lancaster (1966), A New Approach to Consumer Theory, Journal of Political Economy 74, 132157. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand 109 The McGrawHill Companies, 2009 Chapter 4 Demand 125 MANAGERIAL APPLICATIONS Understanding What Consumers Want When Chung Mong Koo, the eldest son of the late founder of the Korean Hyundai Motor Company, became CEO in 1999, he immediately set out to make fundamental changes at the company. Consumers had viewed Hyundai as a producer of cheap, low-quality, boring cars. Days after taking over as CEO he visited one factory and began to inspect cars being produced. He saw loose wires, tangled hoses, and bolts painted different colors. He ordered that the bolts be painted black and that no car be released until all was orderly under the hood. Chung fumed, The only way we can survive is to raise our quality to Toyotas level. Chung established a quality-control unit and shufed management. He also changed the stodgy styling of Hyundais vehicles to appeal to American motorists; promoted a pair of U.S. designers, who were charged with developing a new SUV for the U.S. market; and ordered Korean executives to act on their recommendations. When Hyundais Santa Fe SUV debuted in 2000, it was a big hit. Americans were critical of the Toyota and Honda SUVs introduced several years earlier as too small and underpowered with four-cylinder engines. Hyundai gave Americans what they wanted, such as a V-6 engine (to feed Americans demand for horsepower), a large cup holder for liter soft-drink bottles, and extra power ports for cell phones. In addition, the Santa Fe is wider and longer than the competing Toyota and Honda models and comes with a longer drive-train warranty. Chung rescued Hyundais quality and design image; both sales and prots rose, especially in the United States; and its quality rating improved relative to other car companies. This case illustrates the impact of product attributesespecially qualityon the demand for the product. Source: M. Ihlwan, L. Armstrong, and K. Kerwin (2001), Hyundai Gets Hot, BusinessWeek (December 17), 8486. Product Life Cycles Our discussion of product attributes suggests that managers constantly seek to develop new and better ways to identify and respond to consumer demands. This activity leads to the introduction of new products. Managers generally recognize that market demand for a new product is unlikely to remain stable over time. Often, the industry demand curve for a new product shifts outward as the product becomes more widely known. Eventually, however, the demand is likely to shift inward as consumers shift toward other new and improved products. This pattern in the demand for new products is known as the product life cycle. As depicted in Figure 4.6 (page 126), the product-life-cycle hypothesis suggests that the demand for a product can be categorized into four main phases: introduction, growth, maturity, and decline. In the growth phase, the industry-level demand increases rapidly. In the maturity phase, the demand continues to increase and then begins to decrease. In the decline phase, the demand continues to fall. Eventually, the product is withdrawn from the market. Managers should recognize these trends in new-product planning, as well as in entry, exit, and pricing decisions for given products. The increase in demand during the growth phase encourages new rms to enter the industry. For instance, the growth in the demand for personal computers (PCs) during the 1990s prompted many rms to begin production. Given the entry of new rms, original rms typically lose market share. If industry demand grows at a faster rate than the number of rms, existing rms realize sales growth, even though their market share falls. If the number of new rms grows faster than industry demand, existing rms will experience a reduction in demand, depressing prices and rm prots. This discussion suggests that the rst rms to introduce a successful product sometimes can have rst-mover advantages. In this case, they enjoy high prots until competitive BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 126 Part 2 Figure 4.6 Cycle II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics Product Life The product-life-cycle hypothesis suggests that the industry demand for a new product goes though four main phases: introduction, growth, maturity, and decline. Q Industry quantity of output 110 Introduction Growth Maturity Decline Product life cycle T Time entry occurs. They also can develop a customer base and have a longer time period to learn how to produce the product efciently. These advantages explain why rms frequently strive to be the rst to develop and launch new products. However, in attempting to exploit an innovative product, managers must anticipate the impact that their policies are likely to have on entry decisions by potential rivals (see Chapters 8 and 9). The analysis also suggests that managers should be careful in evaluating whether to enter an industry during its growth phase. Competition during this phase can be intense; MANAGERIAL APPLICATIONS First-Mover Advantages and Financial Innovation U.S. investment banks have introduced an impressive list of innovative nancial products. Bankers estimate that developing a new nancial product requires an investment of $50,000 to $5 million. Yet the securities they create cannot be patented and SEC regulations compel innovators to disclose quite detailed information about product design. Thus rivals can copy the product at low cost and exploit the innovating banks investment in educating investors, issuers, and regulators. Bankers estimate that imitators incur costs that are only 25 to 50 percent of the costs incurred by innovators. If nancial innovation is protable, these cost disadvantages of innovation must be offset by other benetsrst-mover advantages. These advantages might include higher prices, lower costs, or larger volume. In a study of 58 nancial innovations that raised almost $280 billion over the period 19741987, there is no evidence that innovative banks charged higher prices during the brief monopoly period prior to the introduction of imitative products. In the longer run, they actually charged prices that were lower than those of their imitative rivals. This result should not be extraordinarily surprising. Investment banks typically choose one of their best customers as the issuer of an innovative product. These issuers will have special burdens placed on them in explaining the innovative product to investors, regulators, and rating agencies. If immediately following the offering, a rival rm were to issue an imitative product at a lower cost using a competing investment bank, the rms managers might be understandably annoyed. Thus, investment banks prot from product innovation in ways other than charging higher prices. They underwrite more offers of the products they innovate than do imitating rivals. Innovation also appears to lower costs by allowing banks to exploit economies of scope and learning effects. (In Chapter 5, we discuss such cost considerations.) Source: P. Tufano (1989), Financial Innovation and First-Mover Advantages, Journal of Financial Economics 25, 213240. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand The McGrawHill Companies, 2009 Chapter 4 Demand 111 127 moreover, the demand they face is expected to decline at some point in the future. To prosper in such an environment, a new rm must have some type of competitive advantage over its rivals (for example, being a low-cost producer). We discuss this issue in more detail in Chapter 8. Demand Estimation15 In our PTC example, we know the demand function. Most managers are not so lucky: They must estimate their demand functions. Sometimes it is easy to estimate demand, at least for the very near term. Other times it is quite difcult. Some companies employ statistical techniques to provide numerical estimates of demand functions. Other companies use more qualitative approaches. Demand estimation is a complex topic that is largely beyond the scope of this book. Here, we simply provide a brief discussion of three general techniques used in estimating demand: interviews, price experimentation, and statistical analysis. Our intent is to provide insights into the basic costs and benets of each approach. These insights make managers more informed consumers of demand estimates and offer guidance as to the type of demand analysis to employ in a given situation. Although each approach has its limitations, the approaches are not mutually exclusive. Because the limitations differ, many managers employ several methods and aggregate the estimates to increase their understanding of demand. Interviews Interview approaches attempt to estimate demand through customer surveys, questionnaires, and focus groups. Perhaps the most naive version of this approach is simply to ask consumers what they would purchase if faced with different prices. The answers to these questions can be remarkably unreliable. First, people have incentives to be less than completely truthful since customers would like the rm to offer lower prices. Second, even if they try to be truthful, they might have difculty forecasting what they would actually purchase given the array of alternatives available in the marketplace. More sophisticated approaches to customer interviews are possible. For example, an individual might be asked about the difference in price between two competing products. Now if you found that individuals had purchased one of the products but did not know the price of the other, you might conclude that customers were relatively insensitive to price. Sometimes companies use a simulated market where people are given play money and asked to simulate purchase decisions. These experiments can yield useful insights. Again, however, the decisions people make with play money need not mirror the decisions they would make with their own money. Consumer surveys play a particularly important role in providing information about the attributes that are valued by customers. Many businesses request that buyers ll out customer-service and complaint forms. Businesses often follow up sales or service with telephone calls to customers to ask about product and service quality and customer satisfaction. Among the most important sources of information about customer preferences are the direct contacts that salespeople and other company representatives have with their customers. 15 This section draws on W. Baumol (1977), Economic Theory and Operations Analysis (Prentice Hall: Englewood Cliffs, NJ), 234236. 112 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 128 Part 2 II. Managerial Economics 4. Demand The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Using Technology to Assess Demand The ACNielsen Corporation has been using handheld computers known as la maquinitathe little machineto collect information on the buying habits of Hispanics. As part of a pilot program in Los Angeles, 500 Latino households are taking all their household purchases and scanning their bar codes into the device. This is the rst time that comprehensive information has been collected on the buying habits of the Latino community, which now represents 12 percent of the U.S. population. Eventually ACNielsen will sell the database to consumer product rms. The likely upshot will be more spending on ads targeting the Latino community. In 1998, an estimated $1.71 billion was spent on advertising to Latinos, representing about 2 percent of total advertising dollars in the United States. This example highlights how new technologies are being employed to assess product demand. Source: R. Wartzman (1999), A Push to Probe Buying Habits in Latino Homes, The Wall Street Journal (August 5), B1. All the interview approaches, however, can produce remarkably inaccurate information if the sample is not representative of the population of the rms customers. For instance, if you are interested in estimating demand for a good with a negative income elasticity, distributing surveys at an upscale mall might be a poor way to proceed. More subtle problems with eliciting interview information also can arise. One team of researchers cautions, The curious, the exhibitionistic, and the succorant are likely to overpopulate any sample of volunteers. How secure a base can volunteers be with such groups over-represented and the shy, suspicious, and inhibited underrepresented?16 Price Experimentation A second approach is to undertake price experiments. For instance, the board might decrease PTCs ticket price to $25 and carefully track changes in ticket sales. However as part of the companys marketing strategy, PTC prints brochures that detail the seasons plays, costs, dates, and ticket prices. Thus, experimenting with their ticket prices would require reprinting their brochures. This raises the cost to PTC of this type of price experimentation. Some other types of rms incur few costs in changing prices; for instance, it is particularly easy for companies that market through the Internet to experiment with their prices. Many rms are unlike PTC in that they operate at multiple locations. If a rm has the exibility to vary prices across different geographic markets, it has the potential to gain more information than if it is limited to experimenting at a single location. But care must be exercised. Ideally, the local markets are separated geographically and have their own media outlets. Thus advertising lower prices in one market will not shift demand from the rms other locations. There are at least three limitations in the use of price experimentation. First, demand can differ, depending on whether customers anticipate that a price change will be permanent or temporary. And it can be difcult to identify customers expectations about 16 E. Webb, D. Campbell, R. Schwartz, and L. Schrest (1966), Unobtrusive Measures (Rand McNally College Publishing Company: Chicago). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand 113 The McGrawHill Companies, 2009 Chapter 4 Demand 129 ACADEMIC APPLICATIONS On Estimating Demand Curves for Common Stocks There has been a long-running debate over the demand elasticities of common stocks of individual rms. Many economists argue that these demand curves are perfectly elastic, since there are numerous stocks with similar risk-return characteristics available in the market. In this case, the demand curves for individual stocks are horizontal. Others argue that each stock is unique and has very few substitutes. Here, the individual demand curves would be downward-sloping. Managers care about the slopes of the demand curves for their common stock since these slopes affect the price at which they can sell new securities. If demand curves slope downward, price must be decreased below the current market price to sell new securities. If demand curves are horizontal, new securities can be issued at the current market price. Managers, of course, want to sell new stock at the highest possible price. The existing empirical evidence suggests that stock prices decline by about 3 percent when rms announce new issues of common stock. This nding seems to suggest that the demand curves for common stocks are downward-sloping. This nding, however, is subject to alternative interpretations. If investors think that managers tend to issue new stock when they believe it is overvalued, an announcement of a new issue will cause the entire demand curve to shift down and price will decline (since investors infer from the announcement that the rm is overvalued). The observation that price declines when new stock is issued is not sufcient to allow us to identify the price elasticity of a rms common stockthe price decrease might be due to either a shift in demand or a shift in quantity demanded. This example illustrates that it is not always easy to estimate demand curves, even when extensive data on prices and quantities are readily available. Indeed, the data on prices and volumes for publicly traded securities are among the best available in the world. Source: C. Smith (1986), Investment Banking and the Capital Acquisition Process, Journal of Financial Economics 15, 329. future prices. Second, direct-market tests are not controlled experiments; several changes might be occurring simultaneously. For instance, the board might lower PTCs ticket prices at the same time that the symphony changes its prices. The observed change in demand would reect both effects. Third, some managers worry that price experimentation is risky. They are concerned that customers lost as a result of a price increase might be difcult to regain even if subsequently the price were lowered. Alternatively, it might be difcult to raise the price once a rm had lowered it (customers might be annoyed and purchase from rivals). Statistical Analysis Often, companies use statistical techniques such as regression analysis to estimate demand functions. Computers and large databases on sales, prices, and other relevant factors have increased the usefulness of this approach materially. By using statistical techniques, the effects of specic factors often can be isolated. It is possible to analyze large samples of actual market data to obtain more reliable results. Even though statistical approaches can provide managers with important information on demand, they must recognize that there are potential problems. Just because a researcher can produce reams of computer output formatted into tables and multicolored graphs implies neither that the analysis is well done nor that the results are reliable. Below, we briey discuss three types of problems that managers encounter regularly in statistical approaches to estimating demand. 114 130 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics 2006 Income (I) Advertising (A) Price (P) Sales (S) True demand Estimated demand S S $3,000 2 10 236 120 2P 8A 140 48A 2007 2008 $4,000 3 10 284 $3,500 2.5 10 260 0.04I Table 4.1 An Example of the Omitted-Variables Problem The true demand curve of the company in this example is Sales 120 2P 8A 0.04I. The data for 2006 to 2008 are presented in the table. If the analyst omits income and uses statistical techniques to estimate a relation between advertising and sales, the analyst will obtain the following equation: Sales 140 48A. The model predicts sales perfectly (based on the data in the table). The estimated equation, however, signicantly overstates the inuence of advertising. The omitted-variables problem is present whenever important variables are left out of the analysis that are correlated with the explanatory variables that are included in the analysis. Omission of Important Variables The problem of omitted variables can be illustrated by an example. Assume that the actual demand function for a company is Sales 120 2P 8A 0.04I (4.10) where P is the price of the product A is advertising expenditures I is income Table 4.1 presents the data for 2006, 2007, and 2008. While this data is potentially available to the marketing manager, Brendis Isaccsdottir, who wants to estimate demand, does not necessarily know that both advertising and income are important determinants of demand. Suppose that Brendis ignores income and uses statistical techniques to estimate a relation between sales and advertising.17 Standard regression techniques would yield the following equation: Sales 140 48A (4.11) The model appears to predict sales perfectly (based on the data in the table). The equation, however, materially overstates the true inuence of advertising and can lead to spectacular mistakes in decision making. Based on this analysis, Brendis might budget far too much for advertising. This omitted-variables problem is present whenever important excluded variables are correlated with explanatory variables that are included in the statistical analysis.18 Including unimportant variables does not bias estimated coefcients for the other variables (however, including irrelevant variables reduces the precision of the various estimates). 17 The manager does not have to worry about controlling for price, since it was constant over the period ($10). The problem does not always result in overstated coefcients on the explanatory variables. Depending on the nature of the correlation among the explanatory variables, the coefcients can be either overstated or understated. The estimated coefcient in this example is overstated because advertising and income are positively correlated. 18 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 4 Figure 4.7 An Example of the Identication Problem Demand 131 $ Price (in dollars) An analyst has collected data on past prices and sales for her rms industry. The demand and supply curves have shifted over the three years. Connecting the three price-quantity points provides a poor estimate of the current industry demand curve (labeled D3 in the graph). The three points are equilibrium points, given all conditions that affect the demand and supply of the product at each point in time. They are not three points along the same demand (or supply) curve. 115 The McGrawHill Companies, 2009 4. Demand S1 S2 D1 P1 P2 D2 P3 S3 D3 Estimated demand Q Q1 Q 2 Q3 Quantity Multicollinearity If the factors that affect demand are highly correlated (tend to move together), it might be impossible to estimate their individual effects with much precision. For instance, two important variables in the demand function might be income and education. If in the data set to be analyzed high income is always associated with high education, it might be impossible to separate the two effects. Identication Problem Another potentially important problem that can confront Brendis is the identication problem. This problem also can be illustrated by example. Suppose the marketing manager has collected data on past prices and sales for a given industry with the aim of estimating an industry demand curve. In the past three years, the following sales and price combinations have been observed: (10, $10), (12, $8), and (14, $6). Is it valid for Brendis to connect these three points as an estimate of the demand curve? Because of the identication problem, the answer is generally no. Each data combination reects the intersection of the demand curve and supply curve for the industry for each year. If the demand curve has shifted over the three years due to changes in factors such as personal income, the points come from three different demand curves. Connecting the points does not provide an estimate of the current demand curve. In fact, if supply considerations have been stable while demand has shifted, it will trace out the industrys supply curve. Suppose in our example that both the demand and supply curves have shifted in each year. As shown in Figure 4.7, the resulting combinations of price and quantity are observed equilibrium points, given the conditions during the relevant time periods. Connecting these points provides a poor estimate of the current demand curve D3. Sometimes, Brendis will not have enough information to solve the identication problem and is better off using consumer interviews or market experiments to estimate 116 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 132 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics demand. Other times, she has enough information to identify the demand function (she needs to be able to specify factors that inuence demand, but not supply, and vice versa). One special case in which Brendis does not have to worry about the identication problem is when the demand curve is stable. Suppose the demand curve did not shift over the three years and all the different sales-price combinations were caused by changes in supply. In this case, she can obtain a reasonable estimate of the demand curve simply by connecting the observed sales-price combinations. Implications We have discussed some of the difculties that managers face in trying to estimate the demand for their product. These problems can be difcult to solve. Nonetheless, estimates of demand play a critical role in decision makingespecially the pricing decision. Successful managers address these problems the best they can, given imperfect knowledge and limited resources. ANALYZING MANAGERIAL DECISIONS: Personal Video Recorders Personal video recorders (PVRs) are digital video recorders used to record and replay television programs received from cable, satellite, or local broadcasts. But unlike VCRs, which they replace, PVRs offer many more functions, notably the ability to record up to 80 hours of programs and easy programming. A PVR consists of an internal hard disk and microprocessor. After the owner installs the hardware, the PVR downloads all upcoming TV schedules to the hardware via a phone or cable connection. Users merely enter the name of the show(s) they want recorded and the system nds the time and channel of the show and automatically records it. Users must subscribe to a cable or satellite system if they wish to record programs off these channels. Besides ease of programming and much larger recording capacity than videotape, PVRs allow the user to watch a prerecorded show while the unit is recording a new program, pause watching live programs (for example, if the phone rings) and then resume watching the rest of the live broadcast, view instant replays and slow motion of live programs, and skip commercials. In effect, PVRs, like older VCRs, allow viewers to control when they watch broadcast programs (called time shifting). However, PVRs provide much sharper pictures and are much simpler to operate than VCRs, and PVRs allow the user to download the television schedule for the next week. Two companies currently sell the hardware and provide the subscription service: TiVo and ReplayTV. Both rms started in 1997. As of 2008 TiVo had nearly four million subscribers and ReplayTV had been purchased by DirectTV. Companies are developing new technologies that make it even easier for users to snip commercials. Cable companies have begun offering a combined cable box and PVR in one unit for a small additional monthly charge. This further simplies setup and operation, and the user gets a single bill. 1. Discuss how PVRs will affect the demand from advertisers? 2. Suppose you are in charge of setting the price for commercial advertisements shown during Enemies, a top network television show. There is a 60-minute slot for the show. However, the running time for the show itself is only 30 minutes. The rest of the time can be sold to other companies to advertise their products or donated for public service announcements. Demand for advertising is given by: Qd 30 0.0002P 26V BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 4 where Qd quantity demanded for advertising on the show (minutes), P the price per minute that you charge for advertising, and V is the number of viewers expected to watch the advertisements (in millions). a. All your costs are xed and your goal is to maximize the total revenue received from selling advertising. Suppose that the expected number of viewers is one million people. What price should you charge? How many minutes of advertising will you sell? What is total revenue? b. Suppose price is held constant at the value from part (a). What will happen to the quantity demanded if due to PVRs the number of expected viewers falls to 0.5 million? Calculate the viewer elasticity based Summary 117 The McGrawHill Companies, 2009 4. Demand Demand 133 on the two points. Explain in words what this value means. 3. As more viewers begin using PVRs, what happens to the revenues of the major networks (CBS, NBC, ABC, and FOX)? 4. Discuss the long-run effects if a signicant proportion of the viewers begin adopting these advertising snipping systems. 5. What advice would you give the major commercial networks and producers of programming for these networks as more consumers adopt PVRs? SOURCE: J. Gudmundsen (2002), Video Gizmos Change the Rules, Democrat and Chronicle (August), 5E and 8E; B. Fisher (2003), TiVo and Replay TV Have Features to Satisfy Any TV Junkie, Detroit News (June 24); R. Reilly (2003), Great Invention Period, Sports Illustrated (December 22). Understanding product demand is critical for many managerial decisions such as pricing, setting production levels, undertaking capital investment, and establishing an advertising budget. This chapter provides a basic analysis of demand. A demand function is a mathematical representation of the relations among the quantity demanded of a product over a specied time period and the various factors that inuence this quantity. We focus on three independent variables in the demand function: the price of the product, the prices of related products, and customers incomes. A demand curve for a product displays how many units will be purchased over a given period at each price holding all other factors xed. Movements along a demand curve reect changes in price and are called changes in the quantity demanded. Movements of the entire demand curve are caused by other factors, such as changes in income, and are referred to as changes in demand. Demand curves generally slope downward to the right: Quantity demanded varies inversely with price. This relation often is referred to as the law of demand. Demand curves vary in their sensitivities of the quantity demanded to price. Price elasticity is dened as the percentage change in quantity demanded from a percentage change in price (expressed as a positive number). The price elasticity tends to be high when there are close substitutes for the product and when the good represents a signicant expenditure for the consumer. Demand tends to be more elastic over the long run than over the short run. How total revenue from a product changes with price depends on the price elasticity. A small price increase results in an increase in expenditures when demand is inelastic and a decrease in expenditures when demand is elastic. Total expenditures remain unchanged when the demand elasticity is unitary. An important concept in economics is marginal revenue, which is dened as the change in total revenue given a one-unit change in quantity. Marginal revenue for a linear demand curve is given by the line with the same intercept as the demand curve but with twice the negative slope. Total revenue increases with quantity when marginal revenue is positive and decreases with quantity when marginal revenue is negative. 118 134 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics The price of related products can affect the demand for a product. Goods that compete with each other are referred to as substitutes. Products that tend to be consumed together are complements. One frequently used measure of substitution between two products is the cross elasticity of demand. The cross elasticity is positive for substitutes and negative for complements. Another factor that can affect the demand for a product is the income of potential buyers. The sensitivity of demand to income is measured by the income elasticity. The income elasticity is positive for normal goods, and negative for inferior goods. Demand curves can be dened for individual rms or entire industries. The price elasticities for individual rms within an industry are generally higher than for the industry as a whole. Cross elasticities can be helpful in dening the appropriate industry. For some products, demand increases with the number of users. For example, fax machines and telephones are not very useful unless there is a network of users. Products where these network concerns are important often have relatively elastic demands. When price is lowered, there is both a standard price effect and a network effect. The standard economic analysis of demand takes the attributes of the product as given. Information about consumer demand, however, is also important in the initial design of products. Parts 3 and 4 of this book provide important insights into how to design the rms organizational architecture to help ensure that this type of information is incorporated in the decision-making process. Managers use three basic approaches to estimate demand: interviews, price experimentation, and statistical analysis. All three approaches can suffer from potentially serious problems. Managers have to do the best they can given imperfect information and limited resources. Knowledge of the potential pitfalls can make managers more intelligent producers and users of demand estimates. Appendix Demand19 In the chapter, we presented formulas for arc elasticities that estimate elasticities between two points on the demand curve. This appendix shows how to calculate elasticities at single points on the demand curve. It also derives the equation for marginal revenue for a linear demand curve and discusses a special (log-linear) demand function. Point Elasticities Elasticities measure the percentage change in quantity demanded for a percentage change in some other variable. There are several ways to express the formula for an elasticity. One way, using price elasticity as an example, follows: ( Q Q ) ( P P) ( Q P ) (P Q ) (4.12) By denition, as the change in P goes to zero, the limit of the rst term ( Q P ) is the partial derivative of Q with respect to P. At a particular point on the demand curve, the elasticity of demand for small changes in P is given by (Q 19 P) This appendix requires elementary knowledge of calculus. (P Q ) (4.13) BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Chapter 4 Demand 119 135 As an example, consider the demand function for PTC theater tickets: Q 117 6.6P 1.66PS 3.3PR 0.0066I (4.14) The point elasticity at the current price-quantity combination of $30 and 200 tickets is ( 6.6) (30 200) 1 (4.15) Recall that this is the value that we derived graphically in the text (see Figure 4.5). Other point elasticitiesfor example, point cross elasticitiescan be calculated in a similar fashion. Simply substitute the appropriate variable (for example, the price of another product) for P in Equation (4.13). Marginal Revenue for Linear Demand Curves Marginal revenue (MR) is the change in total revenue for an additional unit of quantity. As the change in quantity becomes very small, the limit of this denition is the partial derivative of total revenue with respect to Q. Linear demand curves take the following form: P Thus, total revenue, P a bQ (4.16) Q can be written: TR (a bQ ) Q aQ bQ 2 (4.17) Marginal revenue is MR TR Q a 2bQ (4.18) This formula is a line that has the same intercept as the demand curve, but with twice the negative slope. Marginal Revenue and Demand Elasticity In this section, we derive the relation between marginal revenue and demand elasticity. This relation is useful in a number of contexts. For example, it underlies a formula that we use in Chapter 7 to analyze how a rms optimal price markup over cost relates to the products demand elasticity. By denition: TR MR PQ TR Q (4.19) ( Q Q) P Multiply the quantity in parentheses by P P : MR (P Q (P Q P [1 Q ) (P P ) P Q P) P P 1] (4.20) Equation (4.20) indicates that marginal revenue is equal to price when demand is perfectly elastic. In this case, the rm can sell one more unit at the market price without having to lower the price. MR falls as the elasticity decreases and is negative when demand is inelastic ( 1). 120 136 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics Log-Linear Demand Functions The following demand function is frequently used in empirical demand estimation: Q where PI (4.21) Q is the quantity demanded P is price I is income (Other variables such as advertising and the price of other goods are commonly included as other explanatory variables.) An important property of this demand function is that the price and income elasticities are constant (they do not vary along the demand function) and are equal to and , respectively. In particular: ( Q P ) (P Q ) ( P 1I ) (P P I ) (4.22) Similarly, is the income elasticity. Taking the natural logarithm of the demand function in Equation (4.21) yields lnQ ln lnP lnI (4.23) This equation is linear in the logarithms; it thus can be estimated by standard regression analysis using data on Q, P, and I. The estimated coefcients and are estimates of the price and income elasticities. Other types of elasticitiesfor example, cross elasticities can be estimated by including other variables in the demand equation. Suggested Reading Self-Evaluation Problems G. Stigler (1987), The Theory of Price (Macmillan: New York), Chapter 3. 41. Suppose Product A has the demand function QA 10 5PA 2PB 0.01I. The initial values of the variables are QA 15, PA $4, PB $2.5 and I $2,000. a. When PA moves to $3.4, keeping other variables at their initial values, QA becomes 18. What is the corresponding own-price arc elasticity of demand? b. If income, I, increases to $2,250 per period with all other variables held at their initial values, QA becomes 17.5. What is the corresponding income arc elasticity of demand? c. If PB increases to $3 with all other variables held at their initial values, QA becomes 16. What is the corresponding cross-price arc elasticity of demand? d. Is Product A an inferior or normal good? Are Product A and Product B substitutes or complements? Explain. e. Is the rm charging the revenue maximizing price for Product A at the initial values? Explain. f. Compute the marginal revenue (MR) at the initial values. 42. Suppose your rm faces a demand curve of P 90 .30Q. Find the revenue maximizing output and price. Calculate the total revenue. Is this outcome on the elastic, inelastic, or unitary elastic part of the demand curve? Is this price the optimal price for your rm to charge? Display this choice graphically (showing the demand and marginal revenue curves). 43. The BJC Company has the following demand function: Q 300 30(price) 0.01(income) Currently, price is $5 and income is $20,000. a. Calculate the point elasticities for price and income. b. Is the product a normal or an inferior good? BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Chapter 4 Demand 121 137 44. Last year, Americans bought 5,000 Ferraris. The average retail price of a Ferrari was $100,000. Statistical studies have shown that the price elasticity of demand is 0.4. Assume the demand curve is linear. Estimate it using the above information. a. Is demand elastic or inelastic? b. What will happen to revenue if the company raises its price? Solutions to Self-Evaluation Problems 41. a. hQAA = Absolute Value P b. hQA = I - (18 - 15) 3.7 3 3.7 * = * = 1.12 (3.4 - 4) 16.5 0.6 16.5 (17.5 - 15) 2125 2.5 2125 * = * = 1.307 (2250 - 2000) 16.25 250 16.25 (16 - 15) 2.75 1 2.75 * = * = 0.354 (3 - 2.5) 15.5 0.5 15.5 d. Product A is normal because an increase in income leads to an increase in demand for the product. Products A and B are substitutes because the cross-price elasticity is positive. Consumers substitute away from B and purchase more A when the price of B increases. e. No, the rm is not charging the revenue maximizing price for Product A. Revenue maximization occurs at the price/quantity where MR 0 and the corresponding own-price elasticity is one. But here we see that hQAA 1.12 1. Therefore, the current price is higher than the P revenue maximizing price, while the current quantity is lower than the revenue maximizing quantity. f. Start with the original demand function, QA 10 5PA 2PB 0.01I and insert the initial values for PB and I. Solve for PA. This produces the demand curve for Product A: PA 7 .2Q. The MR function has the same intercept at the demand curve, but twice the negative slope: MR 7 .4Q. At the initial quantity of 15, MR $1. Producing an additional unit increases revenue by $1. c. hQBA = P 42. The demand curve is given by P 90 .30Q. The marginal revenue curve is MR 90 .60Q. Revenue maximization occurs where MR 0. Thus the revenue maximizing quantity, Q *, is: 90 .60Q * 0 Q * 150. The revenue maximizing price is found by inserting Q * into the demand function: P* 90 (.3 150) $45. Revenue maximization occurs at the midpoint of the demand curve, where the price elasticity is equal to one (unitary elasticity). The goal of the rm is to maximize prots, not revenue. Thus the rm is not at the optimal price unless the marginal cost of production $0. Graphically the analysis is pictured as follows: $/unit $90 Unitary price elasticity $45 Demand Q 15 MR 122 138 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics 43. a. The current quantity demanded is 350. Thus, the price elasticity is 30(5 350) income elasticity is .01($20,000 350) .571. b. The product is a normal good (income elasticity is positive). .429. The 44. If the demand curve is linear, it must take the form P a bQ. To estimate a and b we use the information provided. The elasticity of demand can be expressed as: abs [( 1 b) We are told that at P 100,000 and Q tuting the known values, solve for b : 0.4 b Review Questions 5,000, the elasticity of demand equals 0.4. Substi- (1 b) 50 100,000 a (P Q )] (100,000 5,000) a 50 5,000 350,000 41. What is the difference between a demand function and a demand curve? 42. How will each of the following affect the position of the demand curve for videocassette recorders (VCRs)? a. An increase in the price of VCR tapes. b. A decrease in the price of VCRs. c. An increase in per capita income. d. A decrease in the price of movie tickets. 43. If the demand for a product is inelastic, what will happen to total revenue if price is increased? Explain. 44. What signs are the cross elasticities for substitute products? Explain. 45. Distinguish between normal and inferior goods. 46. Is it true that a normal good must have an income elasticity that is more than one? Explain. 47. Suppose that the price of Product A falls from $20 to $15. In response, the quantity demanded of A increases from 100 to 120 units. The quantity demanded for Product B increases from 200 to 300. Calculate the arc cross elasticity between Product B and Product A. Is B a substitute or complement for A? Explain. Does Product A follow the law of demand? Explain. 48. How can cross elasticities be used to help dene the relevant rms in an industry? 49. Suppose the price of heating oil increases signicantly. Discuss the likely short-run and longrun effects. 410. The Alexander Machine Tool Company faces a linear demand curve. Currently, it is selling at a price and quantity where its demand elasticity is 1.5. Consultants have suggested that the company expand output because it is facing an elastic demand curve. Do you agree with this recommendation? 411. For three years in a row, income among consumers has increased. Alexander Machine Tool has had sales increases in each of these three years. Does Alexander Machine Tool produce inferior or normal goods? Forecasts predict that income will continue to rise in the future. Should Alexander Machine Tool anticipate that demand for its products will continue to rise? Explain. 412. The cross elasticity between product A and product B is 10. Do you think that product A is likely to face an elastic or inelastic demand curve? Explain. 413. Vijay Bhattacharya is interested in estimating the industry demand curve for a particular product. He has gathered data on historical prices and quantities sold in the industry. He knows that the industry supply curve has been stable over the entire period. He is considering estimating a regression between price and quantity and using the result as an estimate of the demand curve. Do you think this technique will result in a good estimate of the demand curve? Explain. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 4. Demand The McGrawHill Companies, 2009 Chapter 4 Demand 123 139 414. Maria Tejada, a civil engineer, uses data on population trends to forecast the use of a particular highway. Her forecasts indicate severe road congestion by the year 2010. She suggests building a new road. Comment on this approach. 415. Alexander Machine Tool faces the demand curve P $70 0.001Q. What price and quantity maximize total revenue? What is the price elasticity at this point? 416. Studies indicate that the income elasticity of demand for servants in the United States exceeds 1. Nevertheless, the number of servants has been decreasing during the last 75 years, while incomes have risen signicantly. How can these facts be reconciled? 417. Prior to a price increase, the price and quantity demanded for a product were $10 and 100, respectively. After the price increase, they were $12 and 90. a. Calculate the arc elasticity of demand. b. Is the demand elastic or inelastic over this region? c. What happened to total revenue? 418. Dene marginal revenue. Explain why marginal revenue is less than price when demand curves slope downward. 419. In 1991, Rochester, New York, had a serious ice storm. Electric power was out in houses for days. The demand for power generators increased dramatically. Yet the local merchants did not increase their prices, even though they could have sold the units for substantially higher prices. Why do you think the merchants adopted this policy? 420. Seven teenagers, four boys and three girls, were given $200 each to go on a shopping spree. An advertising agency, which specializes in youth markets, gave the teens the money. An account executive accompanied the teens while they were shopping. The agency wanted to learn not only what they bought, but also what they talked about to see what was on their minds. Its not so much to stay in tune with trends, because trends are elusive. Its more whats really happening with teens and whats important to them.20 a. Discuss the trade-offs between sample size (7 teens), cost, and reliability of what is learned from this experiment. b. An agent accompanied the teens while they were shopping. Why didnt the ad agency avoid this expense and just look at what the teens bought? 421. Southwest Airlines estimates the short-run price elasticity of business air travel to be 2 and the long-run elasticity to be 5. Is ticket demand more elastic in the short run or long run? Does this seem reasonable? Explain. 422. Gasoline prices increased substantially in 2004 and 2005. What adjustments did people make to minimize the long-term effects of this price increase? 423. Assume that demand for product A can be expressed as QA 500 5PA 3PB and demand for product B can be expressed as QB 300 2PB PA. Currently, market prices and quantities for these goods are PA 5, PB 2,QA 481, and QB 301. a. Suppose the price of product B increases to 3. What happens to the quantity demanded of both products? b. Calculate the arc cross-elasticity between product A and product B using prices for product B of 2 and 3. c. Are these goods substitutes or complements? 424. The Zenvox Television Company faces a demand function for its products that can be expressed as Q 4,000 P 0.5I, where Q is the number of televisions, P is the price per television, and I is average monthly income. Average monthly income is currently equal to $2,000. Answer the following questions. a. Graph the demand curve (sometimes called the inverse demand curve) faced by Zenvox at the current income level. Be sure to label this and all graphs you draw carefully. On the same graph, depict marginal revenue. At what price and quantity is Zenvoxs total revenue maximized? What is the marginal revenue at this point? Show the calculation. 20 Teens Track Retail Trends for Ad Agency, Democrat and Chronicle (September 5, 1999), 1E. 124 140 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 4. Demand Managerial Economics b. What is the price elasticity of Zenvoxs demand function at the price and quantity derived in part (a)? Explain what this value means in words. c. Why might Zenvox choose to produce at a price and quantity different than that derived in part (a)? 425. According to an article in Forbes (March 2001) teen cigarette smoking declined signicantly between 1975 and 2000. The most dramatic decline occurred in the years 19751981. Since then teen smoking has increased in some years and declined in others. Between 1975 and 1981 there was a slight decrease in the price of cigarettes. Thus the dramatic decline in smoking is not attributable to an increase in cigarette prices. One theory is that the signicant increase in gasoline prices over this period motivated many teens not to smoke. a. Discuss how a rise in gasoline prices might affect the demand for cigarettes among teens. b. Suppose there are two goods in the world, cigarettes and gasoline. Draw a gure that shows how an increase in gasoline prices can result in a decline in both gasoline and cigarette consumption. Use the standard consumer behavior graph with budget lines and indifference curves. Be sure to label your gure appropriately. c. In the late 1990s the price of cigarettes increased from $2.50 per pack to $3.25 per pack. In one community during this time period, the number of packs of cigarettes consumed by teenagers fell from 10,000 to 9,000. Assume that everything except cigarette prices remained the same. Calculate the arc price elasticity among teens between these price points. d. Calculate the total expenditures on cigarettes by the teens in part (c) both before and after the price increase. Did total revenue increase or fall? Discuss how this answer is implied by the arc elasticity that you calculated in part (c). 426. In an article appearing in the Dow Jones News Service on February 5, 2004, the agency cites Saudi Arabias concern about the overproduction of oil by the OPEC cartel. Assume the current daily demand for OPECs oil is given by the following equation: P 50 0.001Q where P is the price per barrel (ppb) and Q is the quantity of barrels sold daily (in thousands). Moreover, suppose the marginal cost of producing a barrel is constant at zero. a. Would it surprise you to learn that OPECs declared objective is to sell 25 million barrels a day for an average price of $25 per barrel? Why or why not? Explain. You may use a graph to support your argument. b. Assume that after OPECs meeting this week, the new demand for OPEC oil will be given by P 40 .001Q. Would OPECs stated objective (25 million barrels at an overall price of $25) be attainable after this change? Explain. Assume OPEC ignores the demand shift. Whats the maximum price per barrel they can charge if they decide to keep producing 25 million barrels per day? What is the prot in this case? c. Now suppose that OPEC recognizes that demand has changed (as in [b]) and wants to maximize prots. What is the daily quantity they should supply? At what price? What is the prot in this case? What is the price elasticity of demand at this price/quantity combination? Explain. 427. As a result of strikes in Canada the world price of nickel rose by 20 percent in December. Over the same period, the quantity demanded of nickel decreased from 10,000,000 to 8,500,000 pounds worldwide. The world price of nickel was 70 cents per pound before the strikes. a. Show graphically the effect of Canadian strikes on the market for nickel. b. Given the information above, whats the price elasticity of the world demand for nickel over the relevant price range? c. Did the total expenditure for nickel increase, decrease, or remain constant after the strikes? How is this consistent with your answers to parts (a) and (b)? Explain clearly and concisely. 428. Assume the demand curve for gasoline is given by the equation P 10 0.0005Q, BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 125 The McGrawHill Companies, 2009 4. Demand Chapter 4 Demand 141 where P is the price per gallon and Q is the quantity of gasoline in gallons. Assume that the only supplier of gasoline in the region is General Gasoline Co. and that the marginal cost of production is constant at zero. a. If the company is currently charging $4.00 a gallon, is it maximizing prot? If so, prove it. If not, nd out the price that maximizes its prot, and compare the prots at the two prices. b. Discuss the likely effect of the introduction of a fuel-efcient car in the region; that is, what would happen to the equilibrium quantity. Show the changes on a graph that displays (you dont need to show actual numbers) General Gasolines pricing solution and explain. 429. The accompanying chart presents data on the price of fuel oil, the quantity demanded of fuel oil, and the quantity demanded for insulation. Fuel Oil Insulation Price per Gallon Quantity Demanded (millions of gallons) Quantity Demanded (millions of tons) $3.00 $5.00 $7.00 100 90 60 30 35 40 a. Calculate the price elasticity (arc elasticity) of demand for fuel oil as its price rises from 30 cents to 50 cents; from 50 cents to 70 cents. Calculate the change in total revenue in the two cases. Explain how the changes in revenue relate to your estimated elasticities. b. Calculate the arc cross elasticity of demand for insulation as the price of fuel oil rises from 50 cents to 70 cents. Are fuel oil and insulation substitutes or complements? Explain. 430. Japan has 4,350 miles of expresswayall toll roads. In fact, the tolls are so high that many drivers avoid using expressways. A typical 3-hour expressway trip can cost $47. A new $12 billion bridge over Tokyo Bay that takes 10 minutes and costs $25 rarely is busy. One driver prefers snaking along Tokyos city streets for hours to save $32 in tolls.21 Assume that the daily demand curve for a particular stretch of expressway is: P 800 yen .16Q a. At what price-quantity point does this demand curve have a price elasticity of one? b. Assume the government wishes to maximize its revenues from the expressway, what price should it set? And how much revenue does it generate at this price? c. Suppose that trafc engineers have determined that the efcient utilization of this particular toll road is 4,000 cars per day. This trafc level represents an optimum tradeoff between congestion (with its associated reduction in speeds and increase in accidents) between expressways and surface roads. If 4,000 cars per day is the socially efcient utilization of the toll road, what price should be set on the toll road? And how much revenue is collected by the government? d. Which price, the one in part (b), or the one in part (a) would you expect the government to set? 21 J. Singer (2003), Lonesome Highways: In Japan, Big Tolls Drive Cars Away, The Wall Street Journal (September 15), A1 and A15. 126 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 5. Production and Cost The McGrawHill Companies, 2009 Production and Cost CHAPTER 5 CHAPTER OUTLINE Production Functions Returns to Scale Returns to a Factor Choice of Inputs Production Isoquants Isocost Lines Cost Minimization Changes in Input Prices Costs Cost Curves Short Run versus Long Run Minimum Efcient Scale Learning Curves Economies of Scope Prot Maximization Factor Demand Curves Cost Estimation Summary Appendix: The Factor-Balance Equation T he global demand for steel has increased in recent years due in part to major construction projects in countries, such as China and India. Construction crews have been particularly busy in China on projects related to the 2008 Beijing Olympics, the 2010 Shanghai World Exposition, and new housing. While the world production of steel also has increased, the net result has been an increase in steel prices from their historic levels. In the early 1990s, steel sold for under $100 per ton. At year end 2007, the price was $545 per ton. This more than vefold increase in the price of steel far exceeds the rate of ination, which was about 50 percent over the same period. Steel prices not only increased over this period but also displayed high volatility. In spring 2002 the price was as low as $222, but by fall 2004 it had increased to above $700. It subsequently retreated to under $500 in late 2005, but by mid-2006 approached $650. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 127 143 Automobile manufacturers are a major user of steel. High steel prices provide strong incentives to these companies to nd ways to mitigate the price increase, for example, by substituting away from steel toward relatively less expensive production materials. Consider their actions during the 1990s when steel prices began their recent assent. In 1994, domestic steel prices increased as the U.S. economy recovered from a recession.1 Indeed, the steel market was the strongest it had been in 20 years: Specically, steel prices had risen from below $90 to over $135 per ton between 1992 and 1994. After signicant price increases earlier in the year, domestic steel companies were planning to increase sheet-steel prices by another 10 percent at years end. In the tight electrogalvanized markets, price increases as high as 20 percent were expected. (In fact, in 1995, prices exceeded $142 per ton.) To counter the effects of the increase in domestic steel prices, U.S. auto companies actively pursued new overseas suppliers. For instance, in July 1994, General Motors invited bids for sheet steel from foreign companies such as Sidmar, Solldac, Thyssen, and Klockner. The increases in steel prices affected both companies pricing and output decisions. The increases in steel prices also placed pressure on U.S. automakers to use other raw materials in the production process. For example, auto companies increased their use of aluminum in engines, transmissions, body components, heating and cooling systems, and suspension systems in 1995.2 (Aluminum prices had been relatively stable; they were $.534 per pound at the beginning of 1992 and $.533 in January 1994.) Potential applications focused on replacing cast iron or steel with aluminum. In addition, auto companies increased research on new ways to use plastics, magnesium, and recyclable materials in their production process. This example raises a number of questions that are of interest to managers. First, how do rms choose among substitutable inputs in the production process? How does the optimal input mix change with changes in the input prices? How do changes in input prices affect the ultimate cost of production and the output choices of rms? In this chapter we address these and related questions. Major topics include production functions, choice of inputs, costs, prot maximization, cost estimation, and factor demand curves. In the appendix, we derive the factor-balance equation. Production Functions A production function is a descriptive relation that links inputs with output. It species the maximum feasible output that can be produced for given amounts of inputs. Production functions are determined by the available technology. Production functions can be expressed mathematically. For instance, given current technology, an automobile supplier is able to transform inputs like steel, aluminum, plastics, and labor into nished auto parts. In its most general form, the production function is expressed as Q f (x1, x2, . . . xn) (5.1) where Q is the quantity produced and x1, x2, . . . xn are the various inputs used in the production process. 1 Details of this example are from General Motors Eyes Imports to Counter Price Increases, Metal Bulletin (July 11, 1994), 21. 2 A. Wrigley (1994), Automotive Aluminum Use Climbing in 1995s Models: Automotive Applications Will Use Some 120 Million Lbs. in 1995, American Metal Market (August 9), 1. 128 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 144 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics MANAGERIAL APPLICATIONS Increasing Returns to Scale at Volkswagen In order to compete globally, VW and Volvo have long had a cooperative relationship whereby VW modies or installs VW engines in certain Volvo cars. This allows VW to produce proportionally more engines with a proportionally smaller increase in inputs, thereby increasing VWs productivity. Some of these productivity gains can be passed through to Volvo in terms of lower engine costs. This is an example of increasing returns to scale. (Note, in 1999, Ford purchased Volvos car division.) Source: B. Mitchener, A. Latour, and S. Moore (1998), VW-Volvo Talks Suggest Success Isnt Enough to Ensure Independence in Automotive Industry, The Wall Street Journal ( July 2), A17. To simplify the exposition, suppose that the auto part in this example is produced from just two inputssteel and aluminum. An example of a specic production function3 in this context is S 1 2 A1 Q 2 (5.2) where S is pounds of steel, A is pounds of aluminum, and Q is the number of auto parts produced. With this production function, 100 pounds of steel and 100 pounds of aluminum will produce 100 auto parts over the relevant time period, 400 pounds of steel and 100 pounds of aluminum will produce 200 auto parts, and so on.4 Returns to Scale The term returns to scale refers to the relation between output and the proportional variation of all inputs taken together. With constant returns to scale, a 1 percent change in all inputs results in a 1 percent change in output. For example, Equation (5.2) presents a production function with constant returns to scale. If the rm uses 100 pounds of each input, it produces 100 auto parts. If the rm increases both inputs by 1 percent to 101 pounds, it produces 101 auto parts.5 With increasing returns to scale, a 1 percent change in all inputs results in a greater than 1 percent change in output. An example of such a production function is Q SA (5.3) Here, 100 pounds of steel and 100 pounds of aluminum produce 10,000 auto parts, while 101 pounds of steel and aluminum produce 10,201 auto parts (a 2 percent increase in output). Firms often experience increasing returns to scale over at least some range of output. One major reason is that a rm operating on a larger scale can engage in more extensive specialization. For instance, if an automobile company has only three employees and three machines, each employee and each machine has to perform a myriad of 3 This production function is an example of a Cobb-Douglas production function, which takes the general form Q S A . Cobb-Douglas production functions are used frequently in empirical estimation. Not all firms, however, have production processes that are well described by this particular type of production function. 4 1001 2 1001 2 10 10 100, and 4001 2 1001 2 20 10 200. 5 [(100 1.01)1 2] [(100 1.01)1 2] 101. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 129 145 tasks for the company to produce automobiles. Given the broad array of tasks that each worker and machine has to perform, efciency is likely to be low. In contrast, a large rm employing thousands of workers and machines can engage in much greater specialization. (As noted in Chapter 3, specialization often produces efciency gains.) With decreasing returns to scale, a 1 percent change in all inputs results in a less than 1 percent change in output. An example is Q S 1 3 A1 3 (5.4) The likelihood that a rm will choose to operate where it experiences decreasing returns to scale is open to debate. Some economists argue that rms should seldom display decreasing returns to scale. If a facility of a given size can produce a given output, why cant the rm simply replicate that facility and produce twice the output with twice the inputs? Indeed, most empirical studies on the subject suggest that the typical rm initially experiences increasing returns to scale, followed by constant returns to scale over a quite broad range of output. On the other hand, several empirical studies indicate that some rms probably do experience decreasing returns to scale.6 Also, casual observation suggests that some larger rms suffer from inefciencies to a greater extent than do smaller rmsfor example, coordination and control problems become more severe as a rm becomes larger. (We focus on these organizational issues in Part 3.) In our examples, the returns to scale are the same over all ranges of output. For instance, Equation (5.2) always displays constant returns to scale, while Equation (5.4) always displays decreasing returns. Most production functions vary in returns to scale over the range of output. Most frequently, production functions have increasing returns to scale when output is relatively low, followed by constant returns to scale as output continues to increase, and possibly decreasing returns to scale when output is high. Other combinations are possible. Returns to a Factor Returns to a factor refers to the relation between output and the variation in a single input, holding other inputs xed. Returns to a factor can be expressed as total, marginal, or average quantities. The total product of an input is the schedule of output obtained as that input increases, holding other inputs xed. The marginal product of an input is the change in total output associated with a one-unit change in the input, holding other inputs xed. Finally, the average product is the total product divided by the number of units of the input employed. To illustrate these concepts, consider the production function in Equation (5.2): Q S 1 2 A1 2. Table 5.1 (page 146) presents the total, marginal, and average product of S, holding A xed at 9.7 For this production function, total product increases as S increases; marginal product, however, declines. This means that although total product increases with S, it does so at a decreasing rate. Average product also decreases over the entire range. More generally, marginal and average products do not have to decline over the entire range of output. Indeed, many production functions display increasing marginal and 6 For example, E. Berndt, A. Friedlaender, and J. Chiang (1990), Interdependent Pricing and Markup Behavior: An Empirical Analysis of GM, Ford, and Chrysler, working paper, National Bureau of Economic Research, Cambridge, MA. 7 The production function assumes that production does not have to take place in discrete units. For instance, output might be expressed in tons; clearly production in fractions of tons is possible. 130 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 146 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics Units of S Units of A Total Product of S Marginal Product of S Average Product of S 1 2 3 4 5 9 9 9 9 9 3.00 4.24 5.20 6.00 6.70 3.00 1.24 0.96 0.80 0.70 3.00 2.12 1.73 1.50 1.34 Table 5.1 Returns to a Factor This table shows the total, marginal, and average products of steel for the production function Q S1 2 A1 2. Aluminum is held xed at 9 units. The total product of S is the total output for each level of S; the marginal product of S is the incremental output from one additional unit of S; and the average product of S is output divided by the total units of S. average products over some ranges. However, most production functions reach a point after which the marginal product of an input declines. This observation often is called the law of diminishing returns (or law of diminishing marginal product), which states that the marginal product of a variable factor eventually will decline as its use is increased. To illustrate this principle, consider the classic example of farming a plot of land. Land is xed at 1 acre, and no output can be harvested without any workers. If 10 bushels of grain can be produced by one worker, the marginal product of the rst unit of labor is 10 bushels. The change in output might be even greater as the rm moves from one to two workers. For instance, two workers might be able to produce 25 bushels of grain by working together and specializing in various tasks. The marginal product of labor is 15 bushels and thus, over this range, marginal product is increasing. Eventually, as the rm continues to add more workers, while holding land xed, output will grow at a slower rate. At some point, total output might actually decline with MANAGERIAL APPLICATIONS Studying for an Examthe Law of Diminishing Returns Your performance on the CPA exam depends on both your effort and aptitude in the subject. Your aptitude is largely a xed input (e.g., it is hard to increase your basic IQ). Effort on the other hand is something you can control and vary. For example, if you are preparing for the CPA exam, you could spend many or few hours reading books and working problems. The choice is yours. If you exert no effort, you are unlikely to do well on the exam and easily could fail. If you study, your performance is likely to improve. Initially, as you begin to allocate additional hours to studying for the exam, your rate of improvement might be quite large. For instance, you might expect to increase your exam score by 20 points if you study one versus no hours. As you continue to spend more time on studying for the exam, your rate of improvement (marginal product of effort) is likely to declineyou will reach a point of diminishing returns. Indeed at some point your exam score could decline with additional effort as you become too tired to take the examthe marginal product of effort becomes negative. Basic economics says that you should study for an exam up to the point where the incremental benets of studying additional time are equal to the incremental costs. The law of diminishing returns implies that the incremental benets will eventually become smaller and possibly negative. The incremental costs of studying an additional hour depend on your opportunity cost of timefor what else could you use the time? It is unlikely to be optimal for you to spend additional time studying for the CPA exam if that causes you to neglect your other professional responsibilities. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 5 Production and Cost 147 Q Quantity of auto parts Figure 5.1 Returns to a Factor: A Common Case Total product S Q /S Quantity of auto parts per unit of steel This gure illustrates a common pattern for total product, marginal product, and returns to a factor. In the lower panel, marginal product rises, then falls, and eventually becomes negative. When marginal product is rising (between zero and S1), total product increases at an increasing rate (the curve is convex) in the lower panel. When marginal product is falling but positive (between S1 and S2), total product continues to increase but does so at a decreasing rate. Beyond S2, marginal product is negative and total product falls with additional output. Average product is rising where it is below marginal product and is falling where it is above marginal product. Average and marginal products are equal where average product is at a maximum. 131 The McGrawHill Companies, 2009 5. Production and Cost Average product Marginal product S S1 S2 Quantity of steel additional workers because of coordination or congestion problems. In this case, the marginal product is negative. Figure 5.1 illustrates returns to a factor in this common case. The upper panel displays total product, and the lower panel displays marginal and average products. As the use of input S goes from zero to S1, marginal product rises. Over this range total product is convextotal product increases at an increasing rate.8 At S1, diminishing returns set in and the marginal product begins to fall. Between S1 and S2, marginal product is positive and so total product continues to increase. However, it does so at a decreasing rate (the curve is concave). Beyond S2, marginal product is negative, hence total output falls with increases in S. Average product is rising where marginal product is above average product and is falling where marginal product is below average product. Marginal 8 Technical note: The marginal product at a point is equal to the slope of the total product curve at that point (MP TP/ S ). Thus, marginal product is decreasing when the total cost curve is concave and increasing when it is convex. 132 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 148 Part 2 II. Managerial Economics 5. Production and Cost The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Baseball Batting Averages Marginal product is above average product when average product is rising and below average product when average product is falling. This relation is a general property of marginals and averages. A useful illustration is a baseball players batting average. The batting average is dened as the number of hits divided by the number of times at bat. Suppose a player starts a game with an average of .300. If the player gets two hits with four at bats, the marginal batting average for the day is .500 and the players batting average must rise. If the player gets one hit with four at bats, the marginal is .250 and the overall average must drop. product and average product are equal where average product is at its maximum.9 This relation is a general rule.10 The accompanying box on baseball batting averages illustrates the intuition behind this relation. Choice of Inputs Production Isoquants Most production functions allow some substitution among inputs. For example, suppose that Alexi Dyachenko is chief operating ofcer, managing a rm with the production function Q S1 2 A1 2, and he wants to produce 100 auto parts. In this case, there are many different combinations of steel and aluminum that will yield 100 auto parts. For instance, 100 auto parts can be produced using 100 pounds of steel and 100 pounds of aluminum, 25 pounds of steel and 400 pounds of aluminum, or 400 pounds of steel and 25 pounds of aluminum. Figure 5.2 displays all the possible combinations of inputs that can be used to produce exactly 100 auto parts. Obviously, 100 auto parts also could be produced with more inputspoints above or to the right of a point on this isoquant but those points represent inefcient production methods. This curve is called an isoquant (iso, meaning the same, and quant from quantity). An isoquant shows all input combinations that produce the same quantity assuming efcient production. There is a different isoquant for each possible level of production. Figure 5.2 shows the isoquants for 100, 200, and 300 auto parts. Production functions vary in terms of how easily inputs can be substituted one for another. In some cases, no substitution is possible. Suppose that in order to produce 100 auto parts you must have 100 pounds of aluminum and 100 pounds of steel, to produce 200 auto parts you must have 200 pounds of aluminum and 200 pounds of steel, and so on. Having extra steel or aluminum without the other metal yields no additional output they must be used in xed proportions. As shown in Figure 5.3, isoquants from xedproportion production functions are shaped as right angles. At the other extreme are perfect substitutes: The inputs can be substituted freely one for another. Suppose that one auto part always can be produced using either 2 pounds of steel or 2 pounds of aluminum. In this case, the rm can produce 100 auto parts by using either 200 pounds of aluminum or 200 pounds of steel, or any combination in between. As shown in 9 Graphically, marginal product is the slope of a line drawn tangent to the total product curve of that level of output; average product is the slope of the line connecting a point on the total product curve with the origin. 10 Averages and marginals also are equal when the average is at a minimum. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 5 Figure 5.2 133 The McGrawHill Companies, 2009 5. Production and Cost Production and Cost 149 A Isoquants Quantity of aluminum An isoquant displays all possible ways to produce a given quantity. There is a different isoquant for each possible level of production. This gure shows the isoquants for 100, 200, and 300 auto parts for the production function Q S1 2A1 2. 300 200 100 S Quantity of steel Figure 5.3, the corresponding isoquant is a straight line. Most production technologies imply isoquants that are between these two extremes. As depicted in Figure 5.3, typical isoquants have curvature, but are not right angles. The degree of substitutability of the inputs is reected in the curvature: The closer the isoquant is to a right angle (the more convex), the lower the degree of substitutability. Generally, isoquants are convex to the origin (as pictured in the center panel in Figure 5.3the typical case). Convexity implies that the substitutability of one input Figure 5.3 Isoquants for Fixed Proportion Production Functions, Perfect Substitutes, and the Normal Case Production functions vary in terms of how easily inputs can be substituted for one another. In some cases, inputs must be used in xed proportions and no substitution is possible. Here, isoquants take the shape of right angles. At the other extreme are perfect substitutes, where the inputs can be freely substituted for one another. Here, isoquants are straight lines. Most production functions have isoquants that are between the two extremes. The isoquants in the normal case have curvature but are not right angles. A Quantity of aluminum A A S S S Quantity of steel Quantity of steel Quantity of steel Fixed proportions Normal case Perfect substitutes 134 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 150 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics MANAGERIAL APPLICATIONS Substitution of Inputs in Home Building Builders in the Pacic Northwest use large quantities of wood in the construction of residential houses. For instance, wood is used for framing, siding, oors, roofs, and so on. Home builders in the Southwest (for example, Arizona) use much more stucco and tile in home construction. An important reason for this difference is that, in contrast to the Pacic Northwest, the Southwest does not have large nearby forests. This example suggests that home builders are able to substitute among inputs in building a home. Home builders in the Southwest, however, still use wood to frame the house: The substitution of other inputs for wood is not complete. for another declines as less of the rst input is used. In our example, if the rm is using a large quantity of steel and little aluminum, it can eliminate a relatively large quantity of steel with the addition of only a small quantity of aluminum while keeping output the same (see Figure 5.2). In this case, aluminum would be much better suited than steel to construct some components of the auto part. But as the rm uses higher proportions of aluminum, its ability to substitute aluminum for steel declines: Steel is better suited for other components. Most production processes display this property. Isocost Lines Given that there are many ways to produce a given level of output, how does Alexi choose the most efcient input mix? The answer depends on the costs of the inputs. Suppose that the rm faces competitive input markets and can buy as much of each input as it wants at prevailing market prices. The price of steel is denoted Ps, whereas the price of aluminum is denoted Pa. Total cost (TC) is equal to the sum of the quantities of each input used in the production process times their respective prices. Thus, TC P sS Pa A (5.5) Isocost lines display all combinations of S and A with the same cost. Suppose Ps $.50 per pound and Pa $1 per pound, and the given cost level is $100. In this case, $100 $.50S $A (5.6) or equivalently, A 100 0.5S (5.7) MANAGERIAL APPLICATIONS General Motors Is Shanghaied In the late 1990s, General Motors participated in a $1.5 billion joint venture with a state-owned enterprise in China. The Buicks produced at the resulting state-of-the-art Shanghai plant were considered to be the highest-quality cars of that model being produced anywhere in the world. Production costs, however, were extremely high. One important reason for the high costs was government regulation. The Chinese government dictated what products could be built, as well as how many, and at what price. The government also restricted the input mix. For instance, in 1999 GM was required to use locally made components equaling 40 percent in terms of value and 60 percent in 2000. Thus although GM has shared important technology with its Chinese partner, government constraints have precluded efcient production. High costs have limited the joint ventures ability to export cars to other countries. Source: L Kraar (1999), Chinas Car Guy, Fortune (October 11), 238246. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 5 Isocost lines display all combinations of inputs that cost the same. In this example, Ps $.50 per pound and Pa $1 per pound. The gure shows isocost lines for $100 and $200 of expenditures. The slope of an isocost line is 1 times the ratio of the input pricesin this example, 0.5. Isocost lines for different expenditure levels are parallel. Production and Cost 151 A Isocost Curves Quantity of aluminum Figure 5.4 135 The McGrawHill Companies, 2009 5. Production and Cost 200 100 $100 line $200 line S 200 400 Quantity of steel Figure 5.4 graphs this isocost line. Note that the intercept, 100, indicates how many pounds of aluminum could be purchased if the entire $100 were spent on aluminum. The slope of 0.5 is 1 times the ratio of the two prices (Ps Pa): Since aluminum is twice as expensive as steel, 0.5 pounds of aluminum can be given up for 1 pound of steel and costs remain the same. Holding the prices of the inputs constant, isocost lines for different cost levels are parallel. Figure 5.4 illustrates this property using the isocost lines for $100 and $200. Note that the further away the line is from the origin, the higher the total cost. Thus, holding output constant, the rm would like to be on the lowest possible isocost line. The slope of an isocost line changes with changes in the ratio of the input prices. As depicted in Figure 5.5, if the price of steel increases to $1, the line becomes steeper (slope of 1). Here, the rm must give up 1 pound of aluminum to obtain 1 pound of steel. Alternatively, if the price of steel falls to $.25 (not depicted in the gure), the line becomes atter (slope of 0.25). In this case, the rm has to give up only 0.25 pound of aluminum for every pound of steel. Similarly, the slope of the line also changes with changes in the price of aluminum. What determines the slope of the line are the relative prices (recall the slope is Ps Pa). A This gure depicts the effect of changes in input prices on the slopes of isocost lines. The solid line shows the isocost line when the price of aluminum is $1 and the price of steel is $.50. The dotted line shows the isocost line where the prices of both inputs are $1. Total cost in each case is $100. Quantity of aluminum Figure 5.5 Isocost Lines and Changes in Input Prices 100 Ps = $1.00 per pound Ps = $.50 per pound S 100 Quantity of steel 200 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 152 Part 2 Figure 5.6 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics A Cost Minimization The input mix that minimizes the cost of producing any given output, Q*, occurs where an isocost line is tangent to the relevant isoquant. In this example, the tangency occurs at (S*, A*). The rm would prefer to be on an isocost line closer to the origin. However, the rm would not have sufcient resources to produce Q*. The rm could produce Q* using other input mixes, such as (S , A ). However, the cost of production would increase. Q* A Quantity of aluminum 136 A* S S* S Quantity of steel Cost Minimization For any given level of output, Q *, Alexi will want to choose the input mix that minimizes total costs. As shown in Figure 5.6, the cost-minimizing mix (S *, A *) occurs at the point of tangency between the isoquant for Q * with the isocost line. Alexi would like to produce the output less expensively (using an isocost line closer to the origin). However, lower-cost production is not feasible. Alexi could select other input mixes to produce Q *.11 But any other input mix would place the rm on a higher isocost line. Consider the combination (S , A ) in Figure 5.6. This combination of inputs also produces Q * units of output. Yet this output can be produced at a lower cost by using less aluminum and more steel. In the appendix to this chapter, we show that at the optimal input mix, the following condition holds: MPs Ps MPa Pa (5.8) where MPi is the marginal product of input i. (Recall that the marginal product of an input is described in Table 5.1.) Condition (5.8) has a straightforward interpretation. The ratio of the marginal product to price indicates how much additional output can be obtained by spending an extra dollar on the input. At the optimal output mix this quantity must be the same across all inputs. Otherwise, it would be possible to increase output without increasing costs by reducing the use of inputs with low ratios and increasing the use of inputs with high ratios. For instance, if the ratio is 10 units per dollar for aluminum and 20 units per dollar for steel, the rm could hold costs constant but increase output by 10 units by spending one less dollar on aluminum and one more dollar on steel. Alexi has not chosen an optimal input mix when such substitution is possible. 11 Note the similarity between this cost minimization problem and the consumers utility maximization problem introduced in Chapter 2. The mathematics are the sameboth are constrained optimization problems. The consumer maximizes utility for a given budget. Cost minimization is equivalent to maximizing output for a given budget (where the budget is that associated with the lowest-cost method of producing the output). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 137 The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 153 ACADEMIC APPLICATIONS Minimum Wage Laws The minimum wage in the United States was increased to $5.85 on July 24, 2007. Proponents of this action argued that the poor would be substantially better off as a result of this increase. The analysis in this chapter indicates why many economists and politicians are skeptical about this claim. Although it is true that the increase in minimum wage makes some workers better off by increasing their wages, other individuals would be made worse off. In particular, the increase in the wage rate is likely to motivate rms to substitute away from low-skilled workers toward more automation and additional high-skilled workers. Thus, the number of employees hired at the minimum wage is likely to decline with an increase in the wage. Estimates suggest that when the minimum wage was increased from $3.35 to $4.25, employment among teenage men fell by 7.29 percent, and employment among teenage women fell by 11.34 percent; employment among teenage blacks fell by 10 percent. Minimum wage workers who retain their jobs are better off; but individuals who want a job, yet cannot nd one, are worse off. Source: D. Deere, K. Murphy, and F. Welch (1995), Employment and the 19901991 Minimum-Wage Hike, American Economic Review 85:2, 232237. Changes in Input Prices An increase in the relative price of an input will motivate Alexi to use less of that input and more of other inputs. Figure 5.7 illustrates how the optimal input mix for producing Q * changes as the price of steel increases: Alexi chooses less steel and more aluminum to produce the output. This effect is called the substitution effect. The strength of the substitution effect depends on the curvature of the isoquant. The greater the curvature, the less Alexi will substitute between the two inputs for any given change in prices. The substitution effect helps explain the reactions of automobile companies to the 1994 increases in domestic steel prices. These companies increased their use of foreign steel. They also searched for additional ways to replace steel with other inputs such as aluminum. A Figure 5.7 Optimal Input Mix and Changes in Input Prices Q* Quantity of aluminum This gure illustrates how the optimal input mix for producing a given output, Q*, changes as the price of an input increases. In this example, the price of steel increases and the rm uses less steel and more aluminum to produce the output. This effect is called the substitution effect. The strength of the substitution effect depends on the curvature of the isoquant. The greater the curvature, the less the rm will substitute between the two inputs. A* 2 A* 1 High steel price * S2 Low steel price S * S1 Quantity of steel 138 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 154 Part 2 II. Managerial Economics 5. Production and Cost The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Choosing the Mix of People and Machines to Ticket Airline Customers You work for an airline company as a ticket operations manager at a major airport. Your unit is responsible for issuing boarding passes to 30,000 customers per day. Currently you employ 80 customer service representatives (CSRs) and 10 electronic ticketing machines to meet this demand. Each CSR is paid wages and fringe benets of $150/day. It also costs you $150/day to lease each machine (price includes installation, software support, and servicing). A computer vendor has offered to lease you additional electronic ticketing machines at this price. You estimate that by leasing 10 more machines you can meet your service requirements with 30 fewer CSRs. Should you lease the additional machines or continue to service your customers with your current input mix? 1. You conduct additional analysis and estimate that you can service the 30,000 customers with the following combinations of CSRs and machines. Calculate the total costs for each of these combinations. What combination of inputs serves the customers at the lowest possible cost? CSRs Ticketing Machines 80 50 30 20 15 12 2. Plot the input combinations in the table on a graph that contains CSRs on the vertical axis and machines on the horizontal axis. Connect the points by lines to approximate an isoquant as pictured in Figure 5.6. Add the cost minimizing isocost curve to the graph (you can derive this line from the input prices and the total cost of the low-cost input combination). How do the slopes of the isocost curve and isoquant compare at the optimal input combination? 3. Suppose that the marginal product of CSRs at the optimal input combination is 1,000. Explain in words what this means. What is the marginal product of machines at this point? Explain why. 4. Suppose that the cost of leasing a machine declines to $45/day. What is your new optimal input mix? How does this affect your graph? 5. Are there any other factors that should be considered in making this decision on the optimal mix of machines and CSRs? Discuss briey. 10 20 30 40 50 60 Costs We have analyzed how rms should choose their input mix to minimize costs of production. We now extend this analysis to focus more specically on costs of producing different levels of output. Analysis of these costs plays an important role in output and pricing decisions. Cost Curves The total cost curve depicts the relation between total costs and output. Conceptually, the total cost curve can be derived from the isoquant/isocost analysis discussed above. For each feasible level of output, there is a least-cost method of productionas depicted BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 139 The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 155 MANAGERIAL APPLICATIONS Job Seekers Use Internet Posting rsums on the Internet is growing in popularity. In 1999, there were almost ve million rsums on the Internet 200 times as many as 1994, according to Computer Economics. The number of job-related Web sites is expected to grow from about 200 in 1998 to 1,200 in 2002. This explosive growth is a response to the advantages of online job searches. Candidates can reach a larger audience with greater ease. And recruiters can reduce paperwork and travel. But this ood in job candidates sending online rsums has created unexpected headaches for employers. Some companies are getting thousands of rsums dumped into their e-mail boxes each day. Others, fed up with mass e-mailed rsums, yearn for a more personal touch. You get tons of stuff from people who arent qualied, says Michael Erbschloe of Computer Economics, a Carlsbad, California, research rm. The content of the e-mail is horrible. It bogs down your mailbox and your server. Thus in the production of job offers, the reduced cost of distributing information about themselves using the Internet has resulted in huge volume of e-mail as job seekers substitute away from more expensive alternatives. Source: S. Armour (1999), Online Rsums Bogging Down Employers, Democrat & Chronicle ( July 19), 1F. by the tangency between the isoquant and the isocost line. The total cost curve simply displays the cost of production associated with the isocost line and the corresponding output. For instance, if the least-cost method of producing 100 auto parts is $1,000, one point on the total cost curve is (100, $1000). If the least-cost method of producing 200 parts is $1,500, another point is (200, $1,500). Marginal cost is the change in total costs associated with a one-unit change in output. Average cost is total cost divided by total output. Managers sometimes refer to marginal cost as incremental cost, whereas they use the term unit cost to refer to average cost. Figure 5.8 (page 156) displays the total, marginal, and average cost curves for a hypothetical rm. (This gure illustrates a common pattern for cost curves, although not all rms have cost curves with this same shape.) The upper panel indicates that total cost increases with output. Between zero and Q1, total cost increases but at a decreasing rate (the curve is concave). As shown in the lower panel, over this range, marginal cost decreases.12 Past Q1, total cost increases at an increasing rate (the total cost curve is convex) and marginal cost increases. Average cost is declining where marginal cost is below average cost and is rising where marginal cost is above average cost. Average cost equals marginal cost where average cost is at its minimum point. As previously discussed, these relations are general rules: They apply to average costs and average products, as well as batting averages and GPAs. Production Functions and Cost Curves With constant input prices, the shapes of cost curves are determined by the underlying production function. For instance, if the production function displays increasing returns to scale over some range of output, long-run average cost must decline over that range. With increasing returns to scale, a 1 percent increase in input expenditures results in a greater than 1 percent increase in output and average cost must fall. In contrast, with decreasing returns to scale, a 1 percent increase in input expenditures results in a less than 1 percent increase in output and average cost must rise. Finally, constant 12 Technical note: The marginal cost at a point is the derivative of total cost (MC TC Q). Graphically, it is equal to the slope of the total cost curve at that point. Thus, marginal cost decreases when the total cost curve is concave and increases when it is convex. The average cost curve is the slope of the line connecting a point on the total cost curve with the origin. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 156 Part 2 Figure 5.8 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics $ Cost Curves This gure displays the total, marginal, and average cost curves of a hypothetical rm. The upper panel pictures total cost. Total cost increases with output. Between zero and Q1, total cost increases but at a decreasing rate (the curve is concave). As shown in the lower panel, over this range, marginal cost decreases. Past Q1, total cost increases at an increasing rate (the curve is convex) and marginal cost increases. Average cost declines where marginal cost is below average cost and rises where marginal cost is above average cost. This relation is a general rule. Total costs (in dollars) 140 Total cost Q $ Cost per unit of output (in dollars) Marginal cost Average cost Q1 Q2 Q Quantity of output returns to scale imply constant average cost. U-shaped curves (as pictured in Figure 5.8) normally are used to illustrate average costs. This slope suggests an initial region of increasing returns to scale, followed by decreasing returns to scale.13 There is also a direct link between the marginal cost curve and the underlying production function. Recall from Equation (5.8) that cost minimization requires the ratio of the marginal product to price to be equal across all inputs. For illustration, suppose at the optimal input mix to produce 100 auto parts, the ratio of the marginal product to price for both steel and aluminum is 2. By expending $1 more on either input, output increases by 2 units. The reciprocal of this ratio (1 2) is their marginal cost of producing one additional unit of outputif 2 units are produced with $1 of additional expenditure on inputs, the marginal cost of producing one extra unit is $.50. This example indicates that, holding input prices constant, marginal cost is determined by the marginal productivity of the inputs: The higher their marginal productivity, the lower the marginal cost. If the marginal productivities in our example were doubled, the ratio of the 13 Some economists argue that the typical long-run average cost curve is at to the right of its minimum efcient scale. Once that output is reached, additional output can be produced at a constant average cost by simply replicating the process (the production function does not experience decreasing returns to scale). But this argument presumes that organizational costs do not increase disproportionally with rm size. See P. McAfee and J. McMillan (1995), Organizational Diseconomies of Scale, Journal of Economics and Management Strategy 4:3, 399426. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 141 157 MANAGERIAL APPLICATIONS Industry Responds to Higher Metals Prices Metals prices rose substantially between 2003 and 2004; for instance, the price of hot rolled steel increased by more than 80 percent. Manufacturing rms limited the impact of these higher raw materials prices through improved productivity and switching to less expensive substitutes. For example, some stainless steel makers began to use more chromium and manganese and less nickel. Craig Yarde of Yarde Metals in Southington, CT, said that the run-up in prices had beneted his company by increasing the market value of the 40 million pounds of metals in its inventory, mostly aluminum and stainless steel. Source: B. Simmon (2004), Surge in Cost of Metal Squeezes Pricing and Prots, New York Times (February 26), C.1. marginal product to price would be 4 and the marginal cost would be $.25. The inverse relation between marginal productivity and marginal cost makes intuitive sense. If with a given increase in inputs more output can be produced, the marginal cost of producing that output is lower. Input prices also can affect the shapes of the cost curves. For instance, a declining average cost can be motivated by discounts on large volume purchases. Similarly, a machine that produces 20,000 units might not be twice as expensive as a machine that produces only 10,000 units. Alternatively, if the rm bids up the price of inputs with large purchases, average cost can rise with increased output. Thus, the long-run average cost curve can slope upward even if the rm does not experience decreasing returns to scale. Opportunity Costs Managers must be careful to use the correct set of input prices in constructing cost curves. In Chapter 2, we dened opportunity cost as the value of a resource in its next best alternative use. Current market prices for inputs more accurately reect opportunity costs than historical costs. For instance, if an auto supplier purchases 1,000 pounds of aluminum for $600 and subsequently the market price increases to $900, the opportunity cost of using the aluminum is $900. If the company uses the aluminum, its replacement cost is $900. Alternatively, the current inventory could be sold to another rm for $900. In either case, the rm forgoes $900 if it uses the aluminum in its production process. The relevant costs for managerial decision making are opportunity costs. It is important to include the opportunity costs of all inputs whether or not they have actually been purchased in the marketplace. For instance, if an owner spends time working in the rm, the opportunity cost is the value of the owners time in its next best alternative use. Short Run versus Long Run Cost curves can be depicted for both the short run and the long run. The short run is the operating period during which at least one input (typically capital) is xed in supply. For instance, in the short run it might be infeasible to change plant size or change the number of machines. In the long run, the rm has complete exibilityno inputs are xed. The denitions of short run and long run are not based on calendar time. The length of each period depends on how long it takes the rm to vary all inputs. For a cleaningservices rm operating out of rented ofce space, the short run is a relatively brief periodperhaps only a few days. For a large manufacturing rm with heavy investments in long-lived specialized plant and equipment, the short run might be a relatively long time periodit might be a matter of years. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 Figure 5.9 The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics $ Short-Run Cost Curves This gure displays the short-run cost curves of a hypothetical rm. The upper panel depicts total cost (TC) and total variable cost (TVC). Fixed costs simply shift the position of the variable cost curve. The lower panel depicts marginal and average costs. Average xed cost declines with output since the xed cost is being spread over more units. Marginal cost (MC) declines to Q1 and then increases beyond that point due to diminishing returns. Marginal cost depends only on the variable input factors and is completely independent of the xed cost. Average total cost (ATC) and AVC decline as long as marginal cost is lower than the average cost and increase beyond that point. Marginal cost is equal to both ATC and AVC at their respective minimum points. Average total cost is always larger than AVC, since ATC AFC AVC. However, this difference becomes smaller as output increases and AFC declines. Total cost Total costs (in dollars) 158 II. Managerial Economics Total variable cost Q $ Average total cost Marginal cost Cost per unit of output (in dollars) 142 Average fixed cost Q1 Q2 Q3 Average variable cost Q Quantity of output Short-run cost curves sometimes are called operating curves because they are used in making near-term production and pricing decisions. For these decisions, it often is appropriate to take the plant size and certain other factors as given (since these factors are beyond the control of the managers in the short term). Long-run cost curves frequently are referred to as planning curves, since they play a key role in longer-run planning decisions relating to plant size and equipment acquisitions. Fixed and Variable Costs In the short run, some costs are xed and do not vary with output. These xed costs are incurred even if the rm produces no output. For instance, the rm has to pay managers salaries, interest on borrowed capital, lease payments, insurance premiums, and property taxes whether or not it produces any output. Variable costs change with the level of output. These costs include items like raw material, fuel, and certain labor costs. In the long run, all costs are variable. Short-Run Cost Curves Figure 5.9 displays the short-run cost curves for the TAM Corporation. For this rm, suppose that the basic plant size is xed and that all other inputs are variable. The upper panel depicts total cost. Total cost is the sum of the xed cost (FC) and total variable cost (TVC). The shape of the total cost curve is completely determined by the shape of the BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 143 159 MANAGERIAL APPLICATIONS DeLorean Automobiles The difculties of competing with plant sizes signicantly below the minimum efcient scale are highlighted by the experience of the DeLorean Motor Company. John Z. DeLorean had been a high-ranking executive at General Motors. He left GM in 1979 to form his own automobile company, the DeLorean Motor Company. The strategy of the new company was to specialize in high-priced luxury sports cars. The companys rst (and only) car was the stainless-steel DMC12 with a list price of $29,000quite a high car price in the early 1980s. This car subsequently was featured in the Back to the Future movies. Although the minimum efcient scale is relatively large in auto production, DeLorean felt he could compete by designing higher-quality sports cars than the large auto companies. Planned production for 1980 was 3,000 cars. The company soon ran into nancial difculties. In 1982, DeLorean was accused of conspiring to buy and distribute 220 pounds of cocaine valued at $24 million. Federal ofcials asserted that DeLorean was entering the drug business to help save his ailing automobile company. Although DeLorean was later acquitted on these charges, the company still faced insurmountable nancial difculties and soon went out of business. total variable cost curve. Fixed costs simply shift up the location of the curve. Between 0 and Q1, the total cost curve is concave. Over this range, the marginal productivity of variable factors increases (assuming xed input prices). Past Q1, the total cost curve is convex and the marginal productivity of variable factors decreases. This type of pattern is expected given the law of diminishing returns. At low output levels, xed inputs are not efciently utilized. Increasing the variable inputs increases output materially. Over this range, total cost increasesbut does so at a decreasing rate. Eventually, the marginal productivity of the variable inputs declines and it becomes increasingly expensive to produce extra units of output. The lower panel depicts marginal and average costs. Average xed cost (AFC) is total xed cost divided by output. Average xed cost declines with output since the xed cost is spread over more units. Marginal cost (MC) declines up to Q1 and then increases beyond that point due to diminishing returns. Note that marginal cost depends only on the variable input factors and is completely independent of the xed cost. Average variable costs (AVC) are total variable costs divided by output. Both average total cost (ATC) and average variable cost decline as long as marginal cost is lower than average cost; they increase beyond that point. Marginal cost is equal to both average total cost and average variable cost at their respective minimum points. Average total cost is always larger than average variable cost, since ATC AFC AVC. However, this difference becomes smaller as average xed cost declines with higher output. Long-Run Cost Curves In the short run, rms are unable to adjust their plant sizes. In the long run, however, if a rm wants to produce more output, it can build a larger, more efcient plant. In the long run, the average cost (LRAC) of production is less than or equal to the short-run average cost of production. Indeed, the LRAC curve can be thought of as an envelope of the short-run average cost curves. Figure 5.10 (page 160) illustrates this concept. The gure shows four potential plant sizes. Each of the four plants provides the low-cost method of production over some range of output, assuming that only these four plant sizes are feasible. For instance, the smallest plant provides the lowest-cost method of producing any output from zero to Q1, while the next largest plant provides the low-cost method of producing outputs from Q1 to Q2, and so on. The heavy portion of each curve indicates the minimum long-run average cost for producing each level of output. 144 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 160 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics $ Figure 5.10 Long-Run Average Costs as an Envelope of Short-Run Average Cost Curves SRAC1 SRAC2 SRAC3 SRAC4 Cost per unit of output (in dollars) In the long run, the average cost (LRAC) of production is less than or equal to the short-run average cost (SRAC) of production. The LRAC curve can be thought of as an envelope of the shortrun average cost curves. The gure shows four potential plant sizes. Each of the four plants provides the low-cost method of production over some range of output. For instance, the smallest plant provides the lowest-cost method of producing any output from zero to Q1, while the next largest plant provides the low-cost method of producing output from Q1 to Q2, and so on. The heavy portion of each curve indicates the minimum long-run average cost for producing each level of output, assuming that there are only these four possible plant sizes. Q Q1 Q2 Q3 Quantity of output If we extend this analysis by assuming there are many different feasible plant sizes that vary only slightly in size, the resulting LRAC curve will be relatively smooth, as pictured in Figure 5.11. This gure also pictures the long-run marginal cost curve (LRMC). As we have discussed, the marginal cost is below average cost where average cost is falling and above average cost where it is rising. The two are equal at the minimum average cost. Minimum Efcient Scale Minimum efcient scale is dened as that plant size at which long-run average cost rst reaches its minimum point. In Figure 5.11, this minimum occurs at Q *. The minimum efcient scale affects both the optimal plant size and the level of potential competition. MANAGERIAL APPLICATIONS Public Utilities The production of electric power typically is associated with large economies of scale: The average cost of producing electricity decreases with the quantity produced. This production characteristic implies that it is generally more efcient to have one large plant that produces power for an area than several smaller plants. A problem with having one producer of electric power in an area, however, is that the rm has the potential to overcharge consumers for electricity since there are limited alternative sources of supply. Concerns about this problem provide one motivation for the formation of public utility commissions that regulate the prices that utility companies can charge consumers. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 5 Production and Cost 161 $ Figure 5.11 Long-Run Average and Marginal Cost Curves Long-run marginal cost Long-run average cost Cost per unit of output (in dollars) If there are many different plant sizes that vary only slightly in size, the resulting long-run average cost (LRAC) curve is relatively smooth, as pictured in this gure. The long-run marginal cost (LRMC) is below average cost where average cost is falling and above average cost where it is rising. The two are equal at the minimum average cost. The minimum efcient scale is dened as the plant size at which LRACs are rst minimized (Q* in this example). 145 The McGrawHill Companies, 2009 5. Production and Cost Q Q* Quantity of output Average production cost is minimized at the minimum efcient scale. As we discuss in the next chapter, competition provides incentives for rms to adopt this plant size. If rms build plants that depart materially from minimum efcient scale, they will be at a competitive disadvantage and could be forced out of business. One complicating factor is transportation costs. If transportation costs are high, cost disadvantages of smaller regional plants can be more than offset by cost savings in transporting the product to customers. In this case, when total production and distribution costs are considered, rms with plants that are smaller than the minimum efcient scale can survive in a competitive marketplace. Generally, the number of competitors will be large and competition more vigorous when the minimum efcient scale is small relative to total industry demand. For instance, suppose that Kate Polk is evaluating the possibility of entering an industry where she sees established rms reporting substantial prots. If her rm would have to produce 10 percent of the markets output to be cost-efcient, Kate should be concerned that her entry is likely to drive the price down and thus she would be less likely to enter the market than if she needed to produce only 1 percent of the markets output for efcient production. MANAGERIAL APPLICATIONS Size Doesnt Always Matter Regis Corporation has 10,000 beauty salons, buys shampoo by the train load, spends millions on advertising, and uses sophisticated technology to track performance at each salon. Nonetheless, the 300,000 independent salons in the United States still compete quite effectively with Regis chains like Supercuts. Paul Finkelstein, Regis CEO says. We dont run a big business. We run 10,000 $300,000 businesses. He says its salon location rst, and quality of stylists second that make the business go. Those are factors that an independent salon also can offer. Source: F. Bailey (2003), In Some Businesses, Size Is Irrelevant to Success, The Wall Street Journal (November 11), B13. 146 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 162 Part 2 II. Managerial Economics 5. Production and Cost The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Developing Economies of Scale for Malaysias Proton Holdings Proton Holdings Bhd is a national carmaker in Malaysia. In late 2007, the Malaysian government owned 43 percent of the company. The remaining stock of the company traded on public stock exchanges. Proton was among Malaysias worst performing companies in 2005, after competition from foreign carmakers and a lack of new models cost the rm signicant market share and prots. It has since hired a new chief executive, sold its loss-making MV Agusta motorbike rm, and pledged to nd a new technology partner. The company has been under substantial pressure, with its share of domestic sales falling to 44 percent from 75 percent over the past decade. Analysts polled in late 2007 noted that the companys new management had made several moves to revamp the company and that these efforts were bearing fruit in terms of increased sales volume and market share. New models such as the Persona, a sport edition of Savvy, and Satria Neo were relative successes. The management also implemented stringent cost controls. Nonetheless, analysts concluded that Protons long-run ability to survive depends on whether it can achieve increased production volume and economies of scale. Without sufcient scale it is unlikely that the company will survive the intense local and worldwide competition. The analysts assert that by itself, Proton would nd it hard to achieve economies of scale and to develop new technologies. Suppose that you are hired as a consultant to advise Protons management. What do the analysts mean when they say that Proton needs to achieve economies of scale to be competitive? Discuss at a general level the types of actions that the company might want to consider to achieve the necessary scale. SOURCE: Kathy Fong (2007), No Economies of Scale for Proton without Global Partner, StarBiz, thestar online (November 21). Industries where average cost declines over a broad range of output are characterized as having economies of scale. Signicant economies of scale limit the number of rms in the industry. For instance, if the minimum efcient scale is 25 percent of total industry sales, there is room for only four rms to produce at that volume. The level of competition among existing rms can vary signicantly, even if there are only a few rms in the industry. However, threat of entry is less pressing than in industries where scale economies are low. The threat of potential new competitors is often an important consideration in a rms strategic planning. In subsequent chapters, we examine how a rms market structure affects managerial decision making. Learning Curves For some rms, the long-run average cost of producing a given level of output declines as the rm gains production experience. For example, with more output, employees might gain important information on how to improve production processes. They also become more procient as they gain experience on the job.14 A learning curve displays the relation between average cost and cumulative production volume. Cumulative production is the total amount of the product produced by the rm across all previous 14 A. Alchian (1959), Costs and Outputs, in The Allocation of Economic Resources, by M. Abramovitz and others (Stanford University Press: Palo Alto, CA), 2340. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 5 Production and Cost 163 $ Figure 5.12 Learning Curve Average cost of producing Q * units Cost per unit of output (in dollars) A learning curve displays the relation between average cost for a given output period, Q*, and cumulative past production. In this example, there are signicant learning effects in the early stages of production. These effects become minimal as the rm continues to produce the product. 147 The McGrawHill Companies, 2009 5. Production and Cost Learning curve Q Cumulative quantity of output produced production periods. Figure 5.12 presents an example where there are signicant learning effects in the early stages of production. Eventually, however, these effects frequently become minimal as the rm continues to produce the product. Figure 5.13 illustrates the difference between economies of scale and learning effects. Economies of scale imply reductions in average cost as the quantity being produced within the production period increases. Learning effects imply a shift in the entire average cost curve: The average cost for producing a given quantity in a production period $ This gure shows the average cost curves for a rm when it has experienced high and low cumulative volume. In both cases, there are economies of scale (average cost declines with output). The average cost for each level of output, however, is lower where the rm has experienced high cumulative volume because of learning effects. Cost per unit of output (in dollars) Figure 5.13 Economies of Scale versus Learning Effects Learning effect Average cost with low cumulative volume Average cost with high cumulative volume Q Quantity of output 148 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 164 Part 2 II. Managerial Economics 5. Production and Cost The McGrawHill Companies, 2009 Managerial Economics ACADEMIC APPLICATIONS Economies of Scale and Learning Effects in the Chemical Processing Industry Marvin Lieberman studied economies of scale and learning effects in the chemical processing industry. He found that for each doubling in plant size, average production costs fell by about 11 percent. For each doubling of cumulative volume, the average cost of production fell by about 27 percent. Thus, there is evidence of both economies of scale and learning effects in the chemical processing industry. The size of the estimates suggests that learning effects are more important than economies of scale in explaining the observed decline in costs within the industry from the 1950s to the 1970s. Source: M. Lieberman (1984), The Learning Curve and Pricing in the Chemical Processing Industries, Rand Journal 15, 213288. decreases with cumulative volume. Learning effects sometimes can provide existing rms in an industry a competitive advantage over potential entrants; it depends on the nature of the information (Chapter 3) and the distribution of that information across employees. We discuss this issue in more detail in Chapter 8. Economies of Scope Thus far, we have focused on the production of a single product. Most rms, however, produce multiple products. Economies of scope exist when the cost of producing a set of products jointly within one rm is less than the cost of producing the products separately across independent rms. Joint production can produce cost savings for a variety of reasons. Efciencies can result from common use of production facilities, coordinated marketing programs, and sharing management systems. Also, the production of some products provides unavoidable by-products that are valuable to the rm. For instance, a sheep rancher jointly produces both mutton and wool. Economies of scope help explain why rms produce multiple products. For instance, PepsiCo is a major producer of soft drinks; yet it also produces a wide range of snack foods (for example, corn chips and cookies). These multiple products allow PepsiCo to leverage its product development, distribution, and marketing systems. These issues are discussed in greater detail in Chapter 8. Economies of scope and economies of scale are different concepts. Economies of scope involve cost savings that result from joint production, whereas economies of scale involve efciencies from producing higher volumes of a given product. It is possible to have economies of scope without having economies of scale and vice versa. MANAGERIAL APPLICATIONS Economies of Scale and Scope in Apartment Management Home Properties of New York is a real estate investment trust specializing in apartment communities in select Northeast, Midwest, and Mid-Atlantic markets. Board Chairman Norman Leenhouts noted: Since the beginning of last year, we have more than doubled the size of our owned portfolio. By concentrating our growth in our core markets, we are realizing material scale economiesespecially in advertising and personnel costs. Moreover, we exploit scope economies by identifying best practices in our different markets and exporting those practices to properties throughout our portfolio. Source: Home Properties Reports Record Second Quarter 1999 Results (August 5, 1999), PR Newswire. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 149 The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 165 MANAGERIAL APPLICATIONS Economies of Scale and Scope in DSP Production In 1999, Texas Instruments was the leading producer of DSPs (digital signal processors). Its DSPs powered roughly two of every three digital phones, most high-performance disk drives, and a third of all modems. They also are used in a myriad of other products including digital cameras, Internet audio, digital speakers, handheld information appliances, printers, electric-motor controls, and wireless networking equipment. DSPs are programmable. They are especially good at performing superfast real-time calculations, which come in handy when you want to compress, decompress, encrypt, or lter signals and images. Programmability helps keep DSPs inexpensive to produce. The major difference between a DSP in a cell phone and a DSP in a digital camera is the software that tells it what to do. Thus TI gains substantial economies of scale in DSP production even though the chips are used in a variety of products. It also enjoys economies of scope. For instance, it is able to leverage technological developments across its various products. While TI was notably slow in technology development a few years ago, it is now a world leader. For example, it makes a cell-phone chip whose circuits are just 0.18 micron (millionth of a meter) apart. This distance is similar to Intels and IBMs most advanced chips. Source: E. Schonefeld (1999), Hotter Than Intel, Fortune (October 11), 179184. Prot Maximization Thus far, we have focused on the costs of producing different levels of output. However, what output level should a manager choose to maximize rm prots? To answer this question, we return to the concept of marginal analysis that we initially introduced in Chapter 2. Marginal costs and benets are the incremental costs and benets that are associated with a particular decision. It is these incremental costs and benets that are important in economic decision making. An action should be taken whenever the incremental benets of that action exceed its incremental costs. In deciding whether or not to produce one more unit of a product, the incremental benet is marginal revenue (see Chapter 4), while the incremental cost is equal to marginal production cost (including any distribution costs)xed costs do not affect the decision. Therefore the rm should produce extra units so long as marginal revenue exceeds marginal cost; the rm should not produce extra units if marginal revenue is less than marginal cost. At the protmaximizing level of production, the following condition holds15: MR MC (5.9) As we saw in Chapter 4, marginal revenue depends on the demand curve for the product. The effective demand curve that the rm faces will be affected by the degree of competition in the product market. In Chapter 6, we examine how the output decisions of rms vary across different market settings. The changes in metal prices throughout the 1990s changed the total cost of automobile manufacturing. Typically, such changes are accompanied by changes in the marginal cost of production. For example, a reduction in steel prices would mean not only a substitution toward steel from other inputs but also an increase in output. Figure 5.14 (page 166) illustrates this effect. Note that this analysis holds other factors constant. If the 15 Technical note: Since prots equal total revenues minus total costs, Equation (5.9) is the rst-order condition for prot maximization. This condition holds at both minimum and maximum prots. At the maximum, the marginal cost curve cuts the marginal revenue curve from belowthe second-order condition. 150 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 166 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics $ Figure 5.14 Optimal Output and Changes in Marginal Cost MC1 Cost/revenue per unit (in dollars) This gure illustrates that a decrease in marginal cost (from MC0 to MC1) raises the optimal output of the rm (from Q* to Q* ). 0 1 MC0 MR Q * Q1 Q0 * Quantity of output demand for automobiles is falling at the same time (thus shifting marginal revenue downward), the net effect could be an increase in output. However, the increase in output would be less than if steel prices were constant. Factor Demand Curves In discussing the optimal input mix, we noted that the following condition must hold for efcient production: MPi Pi MPj Pj (5.10) for all inputs i and j. The ratios of marginal product to price reect the incremental output from an input associated with an additional dollar expenditure on that input. The reciprocals of these ratios reect the dollar cost for incremental output or the marginal cost: Pi MPi Pj MPj At the prot-maximizing output level, MR level the following condition must hold: Pi MPi MC (5.11) MC. Therefore, at the optimal output MR (5.12) or equivalently, Pi MR MPi (5.13) 16 Equation (5.13) is the rms demand curve for input i. It has a straightforward interpretation. The right-hand side of the equation represents the incremental revenue that 16 Technical note: The marginal product of input i can depend on the levels of other inputs used in the production process. Thus, the demand curve for an input must allow other inputs to adjust to their optimal levels as the price of input i changes. This adjustment is not important if the marginal product of input i is not affected by the levels of the other inputs. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 151 The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 167 MANAGERIAL APPLICATIONS China Becomes the Worlds Smokestack This chapter focuses on how rms choose among alternative input mixes. Another important production decision is where to locate the plant. Costs can vary dramatically across locations due to differences in the prices for labor, land, transportation and other inputs, taxes, environmental and safety regulations, threat of terrorism, and political risk. The steel industry provides a good example of how changes in costs can motivate signicant changes in the location of production. During the 1950s the Ruhr Valley in Germany had the worlds highest growth rate in steel production. In its heyday, Germany produced 10 percent of the worlds steel supply. In 2006, Germany produced only 3.8 percent of the worlds supply, ranking seventh behind China, Japan, the United States, Russia, India, and South Korea. China, which has displayed meteoric growth in steel production, supplied 34 percent. German steel production slowed in the 1960s as miners had to dig deeper for coal and taxes and labor costs continued to increase. Another important factor was new environmental regulation. The emissions from steel plants had made the Ruhr Valley one of the most polluted places in the world. The air was dark and grimy. The residents suffered from an inordinate incidence of lung and other pollution-related diseases. The white shirts that men wore to church on Sunday turned to grey by the time they came home. In an effort to green the country, the government imposed costly pollution control requirements on the steel companies. Differences in labor costs and environmental standards motivated a shift in steel production from Germany to China as smoke-spewing plants were disassembled in the Ruhr Valley and moved 5,000 miles away to China. The Phoenix steel mills in Dortmund had been among Germanys largest since before World War II. In the late 1990s they were slated for closure and were likely headed for the scrap heap. The Chinese realized that they could buy a relatively sophisticated German blast furnace for a small fraction of what a new one would cost. A Chinese company sent workers to Dortmund who labeled every part of the seven-story blast furnace, disassembled it, and packed it into wooden crates for the voyage to China. They worked day and night to accomplish this task in a much shorter time than it would have taken German workers, who were governed by strict union and government work rules. The Hebei Province is Chinas new Ruhr Valley. Its air is heavily polluted and its citizens suffer from a variety of associated health problems. Meanwhile, the Ruhr Valleys pollution level has substantially improved. This improvement, however, has come at a cost. Dortmund, which in 1960 had 30,000 residents working in the steel industry, now has less than 3,000. While Dortmund continues to have high unemployment, the decline in steel jobs has been offset to some extent by new jobs in other less-polluting industries. Source: Joseph Kahn and Mark Landler (2007), China Grabs Wests Smoke-Spewing Factories, nytimes.com (December 21). the rm obtains from employing one more unit of the input (the incremental output times the incremental revenue). We call this incremental revenue the marginal revenue product (MRPi) of input i. Figure 5.15 (page 168) illustrates the demand curve for an input.17 At the current input price of P *, the rm optimally uses Q * units of the input. i i The rm optimally employs additional units of the input up to the point where the marginal cost of the input (its price with constant input prices) is equal to the marginal revenue product of the input. Intuitively, if the marginal revenue product is greater than the input price, the rm increases its protability by using more of the input. If the marginal revenue product is less than the price of the input, the rm increases protability by reducing the use of the input. Prots are maximized when the two are equal. Our discussion of the prot-maximizing output level and the optimal use of an input might appear to suggest that these decisions are two distinct choices. The two 17 Technical note: The second-order condition for maximum prots ensures that the demand curve for the input is the downward-sloping portion of the marginal revenue product curve. Thus, Figure 5.15 displays only the downward-sloping portion of the curve. 152 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 168 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics $ The demand curve for a factor of production is the marginal revenue product curve (MRP) for the input. The marginal revenue product is dened as the marginal product of the input times the marginal revenue. It represents the additional revenue that comes from using one more unit of input. The rm maximizes prots where it purchases inputs up to the point where the price of the input equals its marginal revenue product. Cost/revenue per unit of input (in dollars) Figure 5.15 Factor Demand Curve P* i MRPi Q i* Qi Quantity of input i MANAGERIAL APPLICATIONS Hog Producers React to Increase in Corn Prices Pork is the most widely eaten meat in the world, providing about 38 percent of daily meat protein worldwide (despite the fact that it is not consumed by some people due to religious restrictions). The United States Department of Agriculture reports that in 2006 the per capita consumption of pork was 43.9 kg., 40 kg., and 29 kg. in China, Europe, and the United States, respectively. Ironically in Chinas Year of the Pig (2007), there was a crisis that increased pork prices by 50 percent. The dramatic price increase caused both social unrest and government intervention to help keep pork affordable. Feed cost is typically about 50 to 60 percent of the total cost of producing pork. In the United States, corn accounts for about 80 percent of the typical hog feed. Rising fuel prices and government policies promoting the use of corn-based ethanol as an alternative to gasoline caused U.S. corn prices nearly to double in the summer and fall of 2006. This price change signicantly increased the cost of feeding hogs. U.S. hog producers were able to lessen the effects of the increase in corn prices by switching to feed mixes that use less corn and more distilled grain and solubles (DDGS). DDGS is a by-product of the corn-based ethanol production process. Scientists found that DDGS could be substituted for corn at a 10 percent inclusion rate without having a signicant effect on the efciency, growth, or carcass traits of the hogs. Corn and DDGS, however, are not perfect substitutes in the hog production process. Feed mixes with higher DDGS inclusion rates (e.g., 20 percent and 30 percent) produced smaller pigs that offset the advantages of lower feed costs. This example highlights a general point about production costs. Increases in input prices increase production costs. The effect of the price increase often can be mitigated by shifting the input mix toward relatively less expensive inputs. The ability to reduce costs in this manner depends on the degree of substitutability among the inputs. Source: John Lawrence (2006), Impact on Hog Feed Cost of Corn and DDGS Prices, Iowa Farm Outlook (November 15, 2006), 14. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 153 169 MANAGERIAL APPLICATIONS Demand for Labor Falls Following 9/11 Terrorist Attacks Organized labor was particularly hard hit by the terrorist attacks following 9/11. Of the 760,000 job cuts in the three months following the attack, roughly 50 percent were union membersnearly four times organized labors 13.5 percent of the U.S. workforce. The reason for this higher-than-average job loss is because unions are disproportionately represented in the hard-hit travel and tourism industries. Unions representing public employees also are seeing large job cuts as state and local budgets are cut. September 11 signicantly shifted consumers demand for travel and tourism services to the left. When demand shifts to the left, so does the rms marginal revenue curve. Hence, labors marginal revenue productthe product of marginal revenue and the factors marginal product (holding constant labors marginal product)also shifts to the left. Source: A. Bernstein (2001), A Sock in the Eye for Labor, BusinessWeek (December 17), 44. decisions, however, are linked directly. Once the rm chooses the quantities of inputs, output is determined by the production function. Thus, prot-maximizing rms choose the output where marginal revenue equals marginal cost and produce that output so that the price of each input is equal to its marginal revenue product. In our auto example, an increase in steel prices would be expected to motivate simultaneous adjustments in both the number of automobiles produced and the methods used to produce them. Cost Estimation Our discussion indicates that a detailed knowledge of costs is important for managerial decision making. Short-run costs play an extremely important role in operating decisions. For instance, when the marginal revenue from increased output is above the short-run marginal cost of production, prots increase by expanding production. Alternatively, if marginal revenue is below short-run marginal cost, reducing output increases prots. Long-run costs, in turn, provide important information for decisions on optimal plant size and location. For instance, if economies of scale are important, one large plant is more likely optimal with the product transported to regional markets. Alternatively, if scale economies are small, smaller regional plants, which reduce transportation costs, are more likely optimal. If managers are to incorporate costs in their analyses in this manner, they must have accurate estimates of how short-run and long-run costs are related to various factors both within and beyond the control of the rm.18 Among the most commonly used statistical techniques for estimating cost curves is regression analysis. A regression estimates the relation between costs and output (possibly controlling for other factors, such as the product mix or the weather, which affect costs). The data for this analysis can be either time-series data on costs, output, and other variables, or cross-sectional 18 In addition, some rms estimate cost curves to obtain insights into their underlying production functions. Recall that the shapes of cost curves depend on the underlying production functions. Thus, it often is possible to infer the characteristics of a production function from the shape of the corresponding cost curves. Typically, the data for estimating cost curves is more readily available than the necessary data for estimating production functions. 154 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 170 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics data, which includes observations on variables across rms or plants at a point in time. For instance, in many applications, it is assumed that short-run total costs are approximately linear19: VC a bQ (5.14) where VC is total variable costs for the period and Q is the quantity of output produced. A detailed discussion of cost estimation is beyond the scope of this book. Sufce it to say that similar problems arise in cost estimation as arise in the case of demand estimation (for example, omitted-variables problems). Among the most common problems in cost estimation are difculties in obtaining data on relevant costs. Cost estimates often are based on accounting reports, which record historical costs. As we have indicated, these historical costs do not necessarily reect the opportunity costs of using resources. Moreover, there is the issue of choosing the appropriate functional form. Equation (5.14) presumes a linear model. However, cost curves need not be linear. For instance, it might be appropriate to use a quadratic model, which would include an additional Q 2 term. One of the more serious problems complicating cost estimation is the fact that most plants produce multiple products. Multiple products are produced in the same plant because there are economies of scope. Rather than produce two different types of cereals in two separate plants, it typically is cheaper to produce them in one plant; xed resources can be used more efciently. If a plant produces multiple products, total and average costs for each product can be calculated only by allocating xed costs across the products. This allocation often is arbitrary and complicated further by the existence of ANALYZING MANAGERIAL DECISIONS: Rich Manufacturing Gina Picaretto is production manager at the Rich Manufacturing Company. Each year her unit buys up to 100,000 machine parts from Bhagat Incorporated. The contract species that Rich will pay Bhagat its production costs plus a $5 markup (costplus pricing). Currently, Bhagats costs per part are $10 for labor and $10 for other costs. Thus the current price is $25 per part. The contract provides an option to Rich to buy up to 100,000 parts at this price. It must purchase a minimum volume of 50,000 parts. Bhagats workforce is heavily unionized. During recent contract negotiations, Bhagat agreed to a 30 percent raise for workers. In this labor contract, wages and benets are specied. However, Bhagat is free to choose the quantity of labor it employs. 19 Bhagat has announced a $3 price increase for its machine parts. This gure represents the projected $3 increase in labor costs due to its new union contract. It is Ginas responsibility to evaluate this announcement. 1. Why do many rms use cost-plus pricing for supply contracts? 2. What potential problems do you envision with cost-plus pricing? 3. Should Gina contest the price increase? Explain. 4. Is the increase more likely to be justied in the short run or the long run? Explain. 5. How will a $3 increase in the price of machine parts affect Ginas own production decisions? Variable costs are normally estimated with an intercept. Although variable costs undoubtedly are zero when output is zero, most cost curves are nonlinear. Forcing the intercept to be zero yields a less precise estimate of this slopethe change in costs associated with a change in output. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 155 171 joint costs. Cost accountants use accounting records to track costs of individual products. Fixed and variable resources used by each product are recorded. These product costs, calculated by the cost accountants, typically are used to estimate short-run and long-run average and marginal costs. Despite these estimation problems, cost curves play an important role in managerial decision making. Nonetheless, it is important that managers maintain a healthy skepticism when using these estimates. For instance, in making major decisions, it generally is instructive for managers to examine whether a proposed decision is still attractive with reasonable variation in the estimated parameters of the cost functionthat is, to conduct sensitivity analysis. Summary A production function is a descriptive relation that connects inputs with outputs. It species the maximum possible output that can be produced for given amounts of inputs. Returns to scale refers to the relation between output and a proportional variation in all inputs taken together. A production function displays constant returns to scale when a 1 percent change in all inputs results in a 1 percent change in output. With increasing returns to scale, a 1 percent change in all inputs results in a greater than 1 percent change in output. Finally, with decreasing returns to scale, a 1 percent change in all inputs results in a less than 1 percent change in output. Returns to a factor refers to the relation between output and the variation in only one input, holding other inputs xed. Returns to a factor can be expressed as total, marginal, or average quantities. The law of diminishing returns states that the marginal product of a variable factor will eventually decline as the use of the input is increased. Most production functions allow some substitution of inputs. An isoquant displays all combinations of inputs that produce the same quantity of output. The optimal input mix to produce any given output depends on the costs of the inputs. An isocost line displays all combinations of inputs that cost the same. Cost minimization for a given output occurs where the isoquant is tangent to the isocost line. Changes in input prices change the slope of the isocost line and the point of tangency. When the price of an input increases, the rm will reduce its use of this input and increase its use of other inputs (substitution effect). Cost curves can be derived from the isoquant/isocost analysis. The total cost curve depicts the relation between total costs and output. Marginal cost is the change in total cost associated with a one-unit change in output. Average cost is total cost divided by total output. Average cost falls when marginal cost is below average cost; average cost rises when marginal cost is above average cost. Average and marginal costs are equal when average cost is at a minimum. There is a direct link between the production function and cost curves. Holding input prices constant, the slopes of cost curves are determined by the underlying production technology. Opportunity cost is the value of a resource in its next best alternative use. Current market prices more closely reect the opportunity costs of inputs than historical costs. The relevant costs for managerial decision making are the opportunity costs. Cost curves can be depicted for both the short run and the long run. The short run is the operating period during which at least one input (typically capital) is xed in supply. During this period, xed costs can be incurred even if the rm produces no output. In the long run, there are no xed costsall inputs and costs are variable. Short-run cost curves are sometimes called operating curves because they are used in making near-term production and pricing decisions. Fixed costs are irrelevant for these decisions. Longrun cost curves are referred to as planning curves, since they play a key role in longer-run planning decisions relating to plant size and equipment acquisitions. 156 172 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics The minimum efcient scale is dened as that plant size at which long-run average cost is rst minimized. The minimum efcient scale affects both the optimal plant size and the level of potential competition. Industries where the average cost declines over a broad range of output are characterized as having economies of scale. A learning curve displays the relation between average cost and the cumulative volume of production. For some rms, the long-run average cost for producing a given level of output declines as the rm gains experience from producing the output (that is, there are signicant learning effects). Economies of scope exist when the cost of producing a joint set of products in one rm is less than the cost of producing the products separately across independent rms. Economies of scope help explain why rms often produce multiple products. The prot-maximizing output level occurs at the point where marginal revenue equals marginal cost. At this point, the marginal benets of increasing output are offset exactly by the marginal costs. The marginal revenue product of input i (MRPi) equals the marginal product of the input times marginal revenue. Prot-maximizing rms use an input up to the point where the MRP of the input equals the input price. At this point, the marginal benet of employing more of the input is offset exactly by its marginal cost. Managers often use estimates of cost curves in decision making. A common statistical tool for estimating these curves is regression analysis. One common problem in statistical estimation is the difculty of obtaining good information on the opportunity costs of resources. Another problem with estimating cost curves involves allocating xed costs in a multiproduct plant. Cost accountants track the costs and estimate product costs. Appendix The Factor-Balance Equation20 This appendix derives the factor-balance equationEquation (5.9) in the text: MPi Pi MPj Pj (5.15) This condition must hold if the rm is producing output in a manner that minimizes costs (assuming an interior solution). Recall that at the cost-minimizing method of production, the isoquant curve and isocost line are tangent. Thus, they must have equal slopes. The factor-balance equation is found by setting the slope of the isoquant equal to the slope of the isocost line and rearranging the expression. In the text, we showed that the slope of the isocost line is Pj Pi. We now derive the slope of an isoquant. Slope of an Isoquant The production function in the two-input case takes the following general form: Q f (xi, xj) (5.16) To nd the slope of an isoquant, we totally differentiate Equation (5.16). We set this differential equal to zero, since quantity does not change along an isoquant: dQ [Q xi dxi] [Q xj dxj] 0 (5.17) The slope of the isoquant is dened by dxi dxj. Thus, Slope of an isoquant 20 This appendix requires a basic knowledge of calculus. ( Q xj) ( Q MPj MPi xi) (5.18) (5.19) BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 157 The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 173 This expression has a straightforward interpretation. For illustration, assume that at some xed combination of xi and xj, the marginal product of i is 1 and the marginal product of j is 2. At this point, the slope of the isoquant is 2. This means that 2 units of i can be given up for 1 unit of j and output will stay the same. This is true by denition since j has twice the marginal product of i. Factor-Balance Equation When employing the cost-minimizing production method, the slope of the isoquant is the same as the slope of the isocost line: MPj MPi Pj Pi (5.20) Rearranging this expression gives us the factor-balance equation: MPi Pi MPj Pj (5.21) This expression immediately generalizes to production functions with more than two inputs. Suggested Reading Self-Evaluation Problems G. Stigler (1987), The Theory of Price (Macmillan: New York), Chapters 610. 51. The Zimmerman Company digs ditches. It faces the production function, Q L1 2K, where Q is the number of ditches dug, L is hours of labor, and K is the number of digging tools. a. Complete the following table: K L L L L 0 K 1 K 2 K 3 0 1 2 3 b. Does the production function display increasing, decreasing, or constant returns to scale? Explain. c. Are the marginal products of K and L increasing, decreasing, or constant? Explain. d. Assume constant input prices. Draw the general shapes of the following: (1) long-run average cost; (2) short-run marginal cost, assuming L is xed; (3) short-run marginal cost, assuming K is xed. 52.21 A product is produced using two inputs x1 and x 2 costing w1 $10 and w2 $5 per unit, respectively. The production function is y 5x11.5x 22 where y is the quantity of output, and x1, x2 are the quantities of the two inputs. The marginal products of inputs 1 and 2 for this production function are: MP1 7.5 x10.5 x 22 MP2 10 x11.5 x2 a. What input quantities (x 1*, x 2*) minimize the cost of producing 10,000 units of output? b. What is the total cost of producing the 10,000 units? 21 This problem requires slightly higher skills in algebra than most of the other problems in the book. Readers who cannot work the problem on their own should study the general approach used in the solution to obtain a better understanding of the conditions for cost minimization. 158 174 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics Solutions to Self-Evaluation Problems 51. a. K L L L L 0 1 2 3 0 K 0 0 0 0 1 0 1 1.41 1.73 K 2 K 0 2 2.83 3.46 3 0 3 4.24 5.19 b. The production function shows increasing returns to scale. As you increase both inputs by the same proportion, output goes up by a higher proportion (e.g., if you double both inputs output goes up by more than double). This can be seen along the diagonal of the table. c. The marginal product of K is constant. If you hold L xed and increase K, the marginal increase in output is constant as you add more and more K. You can see this along the rows of the table. The marginal product of L is decreasing. If you hold K xed and increase L, the marginal increase in output declines as you add more and more L. This can be seen along the columns of the table. d. Increasing returns to scale implies that LRAC declines as output is increased (there are economies of scale). The marginal product of K is constant. Thus SRMC is constant when K is the variable input. For example, when L 1, each additional unit of K produces one extra unit of output (see the second row of the table). In this case, SRMC (the cost of producing an additional unit of output) is simply the price of K. The marginal product of L is decreasing. Thus SRMC is increasing when L is the variable input. The general shapes of the graphs are: $/unit $/unit $/unit Q LRAC Q SRMC (L fixed) Q SRMC (K fixed) 52. a. The cost-minimizing combination of inputs is found by equating their marginal product to price ratios: MP1 P 1 MP2 P 2. This condition can be expressed as: MP1 w1 = w2 MP2 So, we get w1 7.5 * x 10.5 * x 22 MP1 10 = Q = 1.5 w2 MP2 5 10 * x 1 * x 2 Q x2 3 -1 8 x1 x2 = 2 Q = x1 4 3 Thus: x1 (3 8) x 2 (i) Use the relation in equation (i) to express the production function for y 10,000 units as a function of only x 2. You now have a solvable equation with one unknown variable, x 2: y = 5x 11.5x 22 Q 10,000 = 5 a 3x2 1.5 2 3 1.5 b (x2) = 5 a b x 23.5 8 8 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 159 The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 175 Solving for x 2 and obtaining x 1 from equation (i): x 2* x 1* 13.357 5.01 b. The total cost of producing the 10,000 units is simply the sum of the expenditures made to acquire each of the inputs: TC Review Questions w1x 1 w 2x 2 10 5.01 5 13.357 $116.87 51. Distinguish between returns to scale and returns to a factor. 52. Your company currently uses steel and aluminum in a production process. Steel costs $.50 per pound, and aluminum costs $1.00 per pound. Suppose the government imposes a tax of $.25 per pound on all metals. What affect will this have on your optimal input mix? Show using isoquants and isocost lines. 53. Your company currently uses steel and aluminum in a production process. Steel costs $.50 per pound, and aluminum costs $1.00 per pound. Suppose that ination doubles the price of both inputs. What affect will this have on your optimal input mix? Show using isoquants and isocost lines. 54. Is the long run the same calendar time for all rms? Explain. 55. You want to estimate the cost of materials used to produce a particular product. According to accounting reports, you initially paid $50 for the materials that are necessary to produce each unit. Is $50 a good estimate of your current production costs? Explain. 56. Suppose that average cost is minimized at 50 units and equals $1. What is marginal cost at this output level? 57. What is the difference between economies of scale and economies of scope? 58. What is the difference between economies of scale and learning effects? 59. Suppose that you can sell as much of a product as you want at $100 per unit. Your marginal cost is: MC 2Q. Your xed cost is $50. What is the optimal output level? What is the optimal output, if your xed cost is $60? 510. Discuss two problems that arise in estimating cost curves. 511. Suppose that the marginal product of labor is: MP 100 L, where L is the number of workers hired. You can sell the product in the marketplace for $50 per unit, and the wage rate for labor is $100. How many workers should you hire? 512. Textbook authors typically receive a simple percentage of total revenue generated from book sales. The publisher bears all the production costs and chooses the output level. Suppose the retail price of a book is xed at $50. The author receives $10 per copy, and the rm receives $40 per copy. The rm is interested in maximizing its own prots. Will the author be happy with the book companys output choice? Does the selected output maximize the joint prots (for both the author and company) from the book? 513. Suppose your company produces one product and that you are currently at an output level where your price elasticity is 0.5. Are you at the optimal output level for prot maximization? How can you tell? 514. Semiconductor chips are used to store information in electronic products, such as personal computers. One of the early leaders in the production of these chips was Texas Instruments (TI). During the early period in the development of this industry, TI made the decision to price its semiconductors substantially below its production costs. This decision increased sales, but resulted in near-term reductions in prots. Explain why TI might have made this decision. 515. The AFL-CIO has been a steadfast proponent of increasing the minimum wage. Offer at least two reasons why they might lobby for such increases. 160 176 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Managerial Economics 516. Mountain Springs Water Company produces bottled water. Internal consultants estimate the companys production function to be Q 300L2K, where Q is the number of bottles of water produced each week, L is the hours of labor per week, and K is the number of machine hours per week. Each machine can operate 100 hours a week. Labor costs $20 per hour, and each machine costs $1,000 per week. a. Suppose the rm has 20 machines and is producing its current output using an optimal K L ratio. How many people does Mountain Springs employ? Assume each person works 40 hours a week. b. Recent technological advancements have caused machine prices to drop. Mountain Springs can now lease each machine for $800 a week. How will this affect the optimal K L ratio (i.e., will the optimal K L ratio be smaller or larger)? Show why. 517. The Workerbee Company employs 100 high school graduates and 50 college graduates at respective wages of $10 and $20. The total product for high school graduates is 1,000 100Q H, whereas the total product for college graduates is 5,000 50QC. Q H the number of high school graduates, while QC the number of college graduates. Is the company hiring the optimal amount of each type of worker? If not, has it hired too many high school or too many college graduates? Explain. 518. Q 0 1 2 3 4 5 6 7 8 9 TC TFC TVC MC 500 AC AFC AVC 80 60 50 60 75 95 120 150 185 a. Complete the above table. b. Graph TC, TFC, TVC, MC, AC, AFC, and AVC against Q. 519. Suppose the Jones Manufacturing Company produces a single product. At its current input mix the marginal product of labor is 10 and the marginal product of capital is 20. The per unit price of labor and capital are $5 and $10, respectively. Is the Jones Company using an optimal mix of labor and capital to produce its current output? If not, should it use more capital or labor? Explain. 520. Suppose the production function of PowerGuns Co. is given by Q 25LK where Q is the quantity of guns produced in the month, L is the number of workers employed, and K is the number of machines used in the production. The monthly wage rate is $3,000 per worker and the monthly rental rate for a machine is $6,000. Currently PowerGuns Co. employs 25 workers and 40 machines. Assume perfect divisibility of labor and machines. a. What is the current average product of labor for PowerGuns Co.? What is the current marginal product of machines? (Assume 1 unit increase in machines.) b. Does PowerGuns production function display increasing, decreasing, or constant returns to scale? Explain. c. What is the total cost of the current production of PowerGuns in a month? What is the average cost to produce a shooting gun? Assuming the number of machines does not change, what is the marginal cost of producing one additional gun? d. What is the law of diminishing returns? Does this production display this characteristic? Explain. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 5. Production and Cost Chapter 5 Production and Cost 161 177 521. Assume Kodaks production function for digital cameras is given by Q 100(L 0.7K 0.3), where L and K are the number of workers and machines employed in a month, respectively, and Q is the monthly output. Moreover, assume the monthly wage per worker is $3,000 and the monthly rental rate per machine is $2,000. Note: Given the production function, the marginal product functions are MPL 70(L 0.3K 0.3 ) and MPK 30(L 0.7K 0.7 ). a. If Kodak needs to supply 60,000 units of cameras per month, how many workers and machines should it optimally employ? b. What are the total cost and average cost of producing the quantity given in (a)? 522. For simplicity, throughout this problem, assume labor (L), capital (K ), and quantity produced (Q) can be innitely dividedthat is, it is ne to hire 3.3 workers, rent 4.7 machines, and/or produce 134.2 units. Answer the following questions, assuming the production function for DurableTires Corp. is Q L1 3K 1 2, where Q is the quantity of tires produced, L is the number of workers employed, and K is the number of machines rented. a. What is the quantity of tires produced when the company employs 64 workers and 36 machines? b. What are the average product of labor (L) and the average product of machines (K ) when the input mix is the one given above? Clearly and concisely, please explain how you would interpret these numbers. c. Continue to assume the input mix given above: What is the marginal product of labor (L), if the number of workers is increased by 1 unit? What is the marginal product of capital (K ), if the number of machines is increased by 1 unit, instead? Clearly and concisely, please explain how you would interpret these numbers. d. Does DurableTires production function display increasing, decreasing, or constant returns to scale? Explain. Would your answer change, if the production function were Q L 1 2K 1 2? How? Explain. e. Does DurableTires production function display increasing, decreasing, or constant returns to labor? Explain. Would your answer change, if the production function were Q L 1 2K 1 2? How? Explain. 523. Answer the following questions, continuing to assume the production function for DurableTires Corp. is Q L 1 3K 1 2, where Q is the quantity of tires produced, L is the number of workers employed, and K is the number of machines rented. Moreover, assume the wage per unit of labor (WL) is $50 and the rental price per machine is $200 (WK ). a. What is the total cost of producing the quantity of tires you found in your answer to question 523(a)? And the average cost? Assuming the number of machines rented does not change, what is the marginal cost of producing one additional tire? b. Given the production function above, the marginal product of labor and the marginal product of capital are MPL 1 3(L 2 3K 1 2) and MPK 1 2(L1 3K 1 2), respectively. Given the wage and rental rate above, is DurableTires Corp. adopting an optimal input mix to produce the quantity of tires found in question 523(a)? If yes, why? If not, why not, and how could DurableTires Corp. save money producing that same quantity of tires? Explain. c. What happens to the optimal input mix you found in question 523, if the government introduces a tax that raises the cost of labor to $150 per worker? Explain. 524. Assume DurableTires Corp. faces the following demand curve, P 250 0.1Q. If DurableTires marginal cost is constant at $35, how many tires should it produce in order to maximize its prots? Whats DurableTires prot in this case? Should the elasticity of demand be greater, equal, or less than 1 at the prot-maximizing price and quantity? Explain (hint: you may use a graph to support your argument). 162 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Market Structure CHAPTER 6 CHAPTER OUTLINE Markets Competitive Markets Firm Supply Competitive Equilibrium Barriers to Entry Incumbent Reactions Incumbent Advantages Exit Costs Monopoly Monopolistic Competition Oligopoly Nash Equilibrium Output Competition Price Competition Empirical Evidence Cooperation and the Prisoners Dilemma Summary T he market for cable television has grown tremendously since Home Box Ofce (a subsidiary of Time Inc.) began broadcasting in 1975.1 Today millions of subscribers purchase multichannel packages from cable companies. Historically consumers in most local markets had but one choicepurchase cable TV from the one local provider or watch the locally broadcast free channels. Subject to regulatory constraints, local cable companies could set their prices without fear that they would be undercut by the competition. Correspondingly, annual price increases for cable TV often exceeded the rate of ination. According to the Federal Communications Commission, the average monthly price for cable TV rose by more than 90 percent between 1995 and 2005. In 1994, DirecTV began providing an alternative to cable TVsatellite TV. Initially, customers had to shell out up to $850 for installation and dish equipment, and pay 1 Some of the details for this example are from P. Grant (2002), The Cable Guy Cuts His Rates, The Wall Street Journal (September 25); and Reuters Limited (2002), FCC: Cable Prices Rose 7.5% over 12 Months (April 4); C. Wexler (2006), Ask Yourself Why . . . Cable Rates Got So High, Common Cause (October). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 163 179 ongoing fees to acquire satellite TV. Many consumers did not consider satellite TV to be a viable alternative because of the price. Fledgling satellite companies were able to attract some customers because they offered more channels and a clearer digital picture than the typical cable company; they were most successful in locations not served by cable companies. By 2002, the market had changed signicantly. Local cable companies were offering more channels and higher quality reception, while satellite companies had lowered their prices and hookup fees substantially. To many consumers, satellite TV and cable TV had become relatively close substitutes. Thus the decision between these two services began to depend more on their relative prices. One consumer who switched from paying $80/month for Mediacoms cable service to DirecTVs satellite service priced at $50/month summed it up when he said, I feel like I got everything I had with digital cable but at a lot cheaper price. In response to the increased competition, cable company managers initiated several major policy changes. First, they became more competitive in the pricing of cable TV services. Charter Communications, with 6.8 million subscribers in 40 states, is one example of a cable company that altered its pricing policies in response to competitive pressures. In St. Louis, for instance, its basic service of 99 channels for $45.85/month competed with Dish Networks service of 60 channels for $22.99/month. In explaining a price freeze in that market, Charters CEO stated in 2002, Weve got to think twice about rate increases. Second, cable companies began offering new price/channel packages to cater to various consumer groups with different price/channel sensitivities. For example, AT&T introduced a premium package of 150 channels for $50/month and an economy package of 100 channels for $40/month. Previously they only had offered a 125-channel package for $43 month. Third, increased competition also affected the companies advertising strategies. For example, companies began promoting the relative benets of cable TV (such as access to local channels, reduced rain fade, not having satellite equipment detract from the appearance of the home, and so on). The example of cable TV illustrates how policy choicessuch as pricing, product design, and advertisingare inuenced critically by the market environment. Policies that work within a protected market environment often have to be amended materially when facing a more competitive environment. It is important that managers understand the rms market environment and how this set of market circumstances affects decision making. Our purpose in this chapter is to enhance that understanding by exploring the implications of alternative market structures. Our primary focus is on output and pricing decisions within different market structures. In subsequent chapters, we examine in more detail how other policies, such as aspects of the rms strategy and organizational architecture, depend on the market environment. We begin by discussing markets and market structure in greater detail. We then provide an analysis of competitive industries. Perfect competition is at one end of a continuum based on the environment in which prices are determined within the industry. Competitive markets provide important managerial implications for rms operating within a broad class of market settings. Next, we discuss barriers to entry that can limit competition within an industry. This section is followed by an analysis of the market structure at the other end of the continuum: monopoly. In a monopolistic industry, there is but one rm. In contrast to rms in competitive industries, a monopolist has substantial discretion in setting prices. After a brief discussion of a hybrid structure monopolistic competitionwe consider the case of oligopoly, where a small number of rival rms constitute the industry. 164 180 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics Markets A market consists of all rms and individuals who are willing and able to buy or sell a particular product.2 These parties include those currently engaged in buying and selling the product, as well as potential entrants. Potential entrants are all individuals and rms that pose a sufciently credible threat of market entry to affect the pricing and output decisions of incumbent rms. Market structure refers to the basic characteristics of the market environment, including (1) the number and size of buyers, sellers, and potential entrants; (2) the degree of product differentiation; (3) the amount and cost of information about product price and quality; and (4) the conditions for entry and exit. We begin our analysis of alternative market structures by examining competitive markets. Competitive Markets Economists generally characterize competitive markets by four basic conditions: A large number of potential buyers and sellers Product homogeneity Rapid dissemination of accurate information at low cost Free entry into and exit from the market Although few markets are perfectly competitive, many markets closely approximate this description. Moreover, competition establishes a benchmark that yields useful insights into other market settings. An example of a market that comfortably satises the conditions for a competitive market is the market for soybeans. In this market, a relatively large number of farmers grow soybeans, and a large number of rms and individuals purchase soybeans. Soybeans are a relatively homogeneous commodity; the product varies little across producers. There are limited informational disparities, and entry as well as exit are essentially costless. In competitive markets, individual buyers and sellers take the market price for the product as givenno single participant has any real control over price. If a seller charges more than the market price, buyers simply will purchase the product from other suppliers. And rms always can sell their output at the market price; thus they have no reason to offer discounts to attract buyers. In this setting, rms view their demand curves as horizontala rm can sell any feasible output at the market price, P *but sells no output at a price above P *. Figure 6.1 illustrates a horizontal demand curve. With a horizontal demand curve, both marginal revenue (MR) and average revenue (AR) equal price. Firm Supply Short-Run Supply Decisions In the last chapter, we saw that a rms prot is maximized at the output where marginal revenue equals marginal cost. The intuition of this result is straightforwardit makes sense to expand output as long as incremental revenue is greater than incremental cost. 2 The specic characteristics of a product often vary across rms. Knowing which rms and individuals to group together as a market, therefore, is not always straightforward. As discussed in Chapter 4, cross elasticities are helpful in dening markets. Products with high cross elasticities can be considered in the same market because they are close substitutes. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Figure 6.1 Market Structure 165 181 Firm Demand Curve in Perfect Competition In competitive markets, rms take the market price of the product as given. The demand curve is horizontal. Both marginal revenue and average revenue are equal to the market price. $ $ Price (in dollars) S P* Di = MRi = ARi P* D Q Qi Quantity (market) Quantity (firm i ) Past this point, prots decline with additional output since incremental revenue is less than incremental cost. In a competitive market, marginal revenue is equal to price (P ). In the short run, the rm takes its plant size (and possibly other inputs) as given. The relevant cost is short-run marginal cost (SRMC). The condition for short-run prot maximization in a competitive industry is P* SRMC (6.1) This conditionone of the more important propositions in economicsindicates that at any price, a competitive rm should produce the output where price equals short-run marginal cost. The rm, however, has the additional option of producing no output at all. When the price of the product is insufcient to cover its average variable cost (AVC), the rm is better off if it ceases production. With no output, the rm loses money since it generates no revenue to cover its xed costs. However, this loss is smaller than the one it would incur if the rm produced any other level of output (since revenue from sales would be lower than its variable production costs). Hence the shutdown condition for the short run is P* AVC (6.2) A rms supply curve depicts the quantity that the rm will produce at each price. Therefore the rms short-run supply curve is that portion of its short-run marginal cost curve above average variable cost. Figure 6.2 (page 182) displays this supply curve. Long-Run Supply Decisions Firms can lose money in the short run yet still nd it optimal to stay in business. In the long run, however, a rm must be protable or it is better to exit this market. Price must equal or exceed long-run average cost (LRAC). Thus, the shutdown condition for the long run is P* LRAC (6.3) In the long run, a rm can adjust its plant size. The long-run supply decision of a rm is based on long-run marginal costs (LRMC). The long-run supply curve of a rm is 166 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 182 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Managerial Economics Figure 6.2 The Firms Short-Run Supply Curve SRMCi ATCi AVCi Costs per unit of output (in dollars) The rms short-run supply curve is the portion of the short-run marginal cost (SRMC) curve that is above average variable cost (AVC). At prices below average variable cost, the rm is better off not producing any output. $ Qi Quantity (firm i ) that portion of its long-run marginal cost curve above long-run average cost. This supply curve is depicted in Figure 6.3.3 Competitive Equilibrium In Chapter 3, we explained that the market price in a competitive market is determined by the intersection of the industry demand and supply curves. The industry demand curve depicts total quantities demanded aggregated across all buyers in the marketplace at each price. Similarly, the industry supply curve is the sum of all individual supply decisions (discussed above). Consider, as an example, the supply and demand curves, labeled S 0 and D 0 in the right panel of Figure 6.4. Here, the market price is P *. The left panel depicts the long0 run supply decision of a typical rm in the industry, rm i. At the price P *, rm i pro0 duces the quantity of output Q *0. Cost curves are dened to include a normal rate of i prot (a normal return on capital is one component of LRAC). Thus, at the price P *, 0 rm i is earning an economic prot (above normal prot). This economic prot is the LRAC) times the output Q *0 and is depicted by the shaded prot per unit (P * 0 i rectangle. The existence of economic prots will motivate other rms to enter the industry.4 This entry will shift the supply curve to the right; inventories will build above their desired levels because of the increased production; hence rms will lower price. 3 There is no inconsistency between short-run and long-run prot maximization. The LRMC at any given output is equal to the SRMC, given that the rm has the optimal plant size for the output. Hence, the rm simultaneously can choose an output where P * SRMC LRMC. 4 Prots reported by rms are based on the accounting denition: sales revenue minus the explicit costs of doing business. The calculation of accounting prots, therefore, does not include the opportunity cost of the owners entrepreneurial effort or equity capital. Economic prots include these costs. Positive economic prots attract entry because the returns are higher than the returns in the alternative activities. Positive accounting prots do not always invite entrythe returns do not always cover the opportunity costs of the owners. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 6 The Firms Long-Run Supply The long-run supply curve for rm i is the portion of the long-run marginal cost (LRMC) curve that is above long-run average cost (LRAC). If price is below LRAC, the rm should go out of business. Market Structure 183 $ LRMCi Cost per unit of output (in dollars) Figure 6.3 Curve 167 The McGrawHill Companies, 2009 6. Market Structure LRACi Qi Quantity (firm i ) Figure 6.4 Competitive Equilibrium The left panel illustrates the long-run supply decision of rm i, a representative rm in the industry. In the right panel, supply and demand curves (labeled S0 and D0) determine the market price, P * . At the price, P * , 0 0 the rm produces Q*0. At the price P * , the rm is earning an economic prot. This economic prot is the i 0 prot per unit (P * LRACi) times the total output Q*0 and is depicted by the shaded rectangle. Economic 0 i prots will motivate other rms to enter the industry. This entry will shift the supply curve to the right and lower the price. Additional entry will occur up to the point where there are no economic prots. This condition occurs at a price of P * . Here, there are no incentives for rms to enter or leave the industry 1 (incumbents are earning a normal rate of prot and inventories are stable at their desired levels), and the market is in equilibrium. In a competitive equilibrium, rms produce output at the low point on their average cost curves ( P * LRMCi LRACi). Thus, the equilibrium is associated with efcient production. 1 Price and cost per unit of output (in dollars) $ $ S0 LRMCi LRACi P0 * P0 * P1 * P1 * S1 D0 Q *1 Q *0 i i Quantity (firm i ) Qi Q * Q0 Quantity * Q1 168 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 184 Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Entry in Low-Carb Food An estimated 32 million Americans are following the Atkins Diet and spending $2.5 billion a year on low-carb foods. Atkins Nutritionals sells 120 products and has licensed its name to dozens of companies. Theres not much growth in the food industry and Atkins is getting most of it, says John McMillin of Prudential Securities. But food giants from Kraft to General Mills are beginning to offer competing products. For instance, Heinz is introducing its One Carb ketchup. Competition is inevitable, says Atkins President, Scott Kabak. Source: B. Grow (2003), The Low-Carb Food Fight Ahead, BusinessWeek (December 22), 48. Additional entry will occur up to the point where there are no longer economic prots. This condition is pictured in Figure 6.4 at a price of P *. Here, there are no incentives 1 for rms to enter or leave the industry (incumbents are earning a normal rate of prot); inventories are stable at their desired levels and the market is in equilibrium. In a competitive equilibrium, rms produce output at the minimum point on their average cost curves (P * LRMC LRAC). Thus, this equilibrium is associated with efcient production. Strategic Considerations Although few markets exactly match economists idealized conditions for perfect competition, many markets approximate this structure. In most industries, there are strong competitive forces that reduce economic prots over time. These forces imply that many strategic advantages (for example, being the rst in a new market) are likely to be shortlived. If the conditions in the market resemble the competitive model, it is important to move quickly to take advantage of transitory opportunities. In addition, potential entrants should realize that observed economic prots in an industry are likely to be bid away as time passes. This consideration can affect both long-range capital spending and entry decisions. For instance, given the increased competition, cable TV companies ACADEMIC APPLICATIONS Phantom Freight Most plywood in the United States is produced in the Pacic Northwest. Due to this dominance, plywood prices throughout the country are essentially the Northwest price plus shipping. If this condition did not hold, Northwest suppliers would curtail shipping plywood to cities with low prices and increase shipping to cities with high prices. The changes in supply would affect the prices in the cities until, in equilibrium, the prices across cities would differ only by transportation costs. In a U.S. court case, Southeast timber producers were sued for charging customers Northwests price plus shipping and then delivering locally produced plywood. It was ruled that these companies were making unjust prots because they did not actually incur the shipping costs. The jury awarded billions of dollars to the customers. Were these companies really making economic prots? The answer is probably not. The local production in the Southeast had a shipping advantage to Southeast customers. The factor that made this advantage possible was scarce timber land in the Southeast. Presumably, the price of this scarce timber land was bid up to the point where plywood producers were making only a normal prot given the prevailing price for plywood in the Southeast (which was the Northwest price plus shipping). Source: A. Alchian and W. Allen (1983), Exchange and Production: Competition, Coordination and Control (Wadsworth Publishing: Belmont, CA), 228231. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 169 The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 185 ANALYZING MANAGERIAL DECISIONS: United Airlines The WSJ recently presented data suggesting that United Airlines was not covering its costs on ights from San Francisco to Washington D.C. The article quoted analysts saying that United should discontinue this service. The costs per ight (presented in the article) included the costs of fuel, pilots, ight attendants, food, etc. used on the ight. They also included a share of the costs associated with running the hubs at the two airports, such as ticket agents, building charges, baggage handlers, gate charges, etc. Suppose that the revenue collected on the typical United ight from San Francisco to Washington does not cover these costs. Does this fact imply that United should discontinue these ights? Explain. increased the level of scrutiny they applied to internal investment proposals. In a competitive market, rms must strive for efciency and cost control; inefcient rms lose money and are forced out of the market. Superior Firms Even in relatively competitive industries, there are rms that do exceptionally well over long time periodsfor example, by being a low-cost producer or having some particular advantage, such as location, relative to competitors. The excess returns, however, often do not go to the owner of the enterprise but rather to the factor input responsible for the particular advantage. For example, land close to a highway interchange (and thus offering customers convenient access) often sells for higher prices than land farther from the interchange. Similarly, the salary of an exceptionally talented manager will be bid up by other rms. In many cases, the rms employing these superior factors of production earn only a normal rate of return (P * LRAC). (These issues are discussed in more detail in Chapter 8.) Barriers to Entry5 Although the competitive model is a reasonable approximation in many markets, there are other industries where rms have notable market poweroutput decisions of individual rms have a noticeable impact on prices. A necessary condition for market power to exist is that there are effective barriers to entry into the industry. To understand what constitutes an effective entry barrier, it is useful to consider the decisions of individual rms to enter an industry. Firms consider entering a new market when they observe extant rms reporting large prots. For instance, if Wen Ho observes a rm such as a cable TV company reporting large prots, his rm (like a number of other rms) is likely to consider entering the industry. Entry decisions depend on three important factors: First, Wen will be concerned about whether his entry will affect product prices. This depends, at least in part, on how existing rms are likely to respond to a new entrant. For example, are they likely to cut prices? Second, Wen will be concerned about incumbent advantages. Do existing rms have advantages that an entering rm will have difculty duplicatingones that make it unlikely that the new rm will enjoy similar prots? Third, Wen will be concerned about costs of exit. How much will it cost to leave the industry if this incursion fails? We discuss each of these factors in turn. 5 This section provides a brief summary of the literature in economics on barriers to entry; it draws on S. M. Oster (1994), Modern Competitive Analysis (Oxford Press: New York). 170 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 186 Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics Incumbent Reactions Specic Assets Specic assets are assets that have more value in their current use than in their next best alternative use. Consider the case of the Alaskan Pipeline. It has a high value in its current use. Yet it is completely specialized for transporting oil from the North Slope to Prudoe Bayit has virtually no other use. Moreover, it could be moved only at enormous expense. If existing rms in an industry have invested heavily in assets quite specialized to that market, they are likely to ght harder to maintain their positions than if their assets are less specic and can be shifted at low cost to alternative activities. Scale Economies Industries with signicant economies of scale have minimum efcient scales that occur at high output levels (see Chapter 5). In such industries, a new entrant must produce at high volume to be cost-effective. Large-scale production is more likely to have a material effect on price. For example, if the minimum efcient scale is 30 percent of total market demand, price certainly will decline if a new entrant tries to capture such a large share of the marketits entry undoubtedly would trigger vigorous price competition from incumbents. Note that the absolute size of the minimum efcient scale is not as important as is this scale relative to the size of the total market. Minimum efcient scale varies enormously across industries. In one study, estimates of minimum efcient scale, as a percentage of industry capacity, ranged from 0.5 percent (fruit/vegetable canning) to 33 percent (gypsum products).6 Globalization of markets increases effective market size, thereby reducing this entry barrierfor example, consider the size of American versus global automobile markets. Reputation Effects Potential entrants can be inuenced by the reputations of existing rms in the industry for reactions to new entrants. In certain circumstances, it can pay for an existing rm to react more aggressively than would be implied by considering only its immediate interests. For example, facing a new rival, the rm might engage in extensive price cutting to establish a reputation as a formidable competitor. Note, however, that threats by rms to cut prices if entry occurs sometimes lack credibility. If new rms actually enter, existing rms might not follow through with their threats because they would be harmed by their own price cuts. Thus, it can be reasonable for a potential entrant to ignore threatsif the entrant believes that incumbents are blufng. We examine these considerations in greater detail in Chapter 9. HISTORICAL APPLICATIONS Excess Capacity at Alcoa In 1940, Alcoa Aluminum lost an important antitrust case involving its production strategy of maintaining excess capacity. The judge ruled that he could think of no better effective deterrent to entry. 6 K. Lancaster and R. Dulaney (1979), Modern Economics: Principles and Policy (Rand McNally: New York), 211. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 171 The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 187 MANAGERIAL APPLICATIONS Entry in Consumer Electronics Since its founding two decades ago, Xoceco Inc. has evolved from producing low-cost color TVs for the Chinese market to producing at-screen TVs. They plan to market in the United States by supplying companies like Dell or Hewlett-Packard. Rather than spending lavishly in chips and software to power their products, they buy the components, assemble the gadgets, and undercut the industry leaders prices. Thus, consumer electronics leaders like Sony and Matsushita are threatened the way IBM was by the rise of the PC clone. Source: E. Ramstad and P. Dvorak (2003), Off-the-Shelf Parts Create New Order in TVs, Electronics, The Wall Street Journal (December 16), A1. Excess Capacity If rms with excess capacity cut production, they can be confronted with much higher average costs (depending on the slopes of their average cost curves). Also, rms with excess capacity are better able to satisfy the demands of new customers should they lower price and force a rival out of business. Potential competitors, therefore, may be less likely to enter when there is excess capacity in the industry because they anticipate more aggressive reactions on the part of incumbents.7 Excess capacity frequently exists for completely innocuous reasons. For example, a rm facing cyclical production or anticipating growth has excess capacity over some time spans because it has invested in additional capacity to satisfy peak demands better. In other cases, excess capacity may be chosen specically to deter entry. Incumbent Advantages Precommitment Contracts Existing rms often have long-term contracts for raw materials, distribution outlets, shelf space, and delivery of the nal product. These contracts can serve as a deterrent to entry, since they limit the opportunities for customers and suppliers to switch from incumbent rms to new entrants. Licenses and Patents Sometimes, entry is limited through government restrictions such as licensing requirements and patents. For instance, the number of doctors is limited effectively by state medical licensing requirements. This restriction allows doctors to charge higher prices than if entry were unrestricted. Regulators and licensed physicians justify such restrictions with arguments based on consumer protection. Yet, whether or not consumers benet from stringent licensing is debatablegiven that they pay higher prices. Normal patent life is 17 years. Over this period, other rms are prohibited from copying the innovation; thus a patent provides a rm with potential market power. Patents also provide important incentives to innovate. From a practical standpoint, the effectiveness of a patent in blocking entry varies dramatically (some patents can be circumvented by clever design, for example). 7 Excess capacity can occur because of signicant declines in industry demand. In this case, prots are likely to be low and entry will not be attractive. Our current discussion focuses on cases where incumbents are making economic prots and have excess capacity. These economic prots might not induce entry because of the fear of price cutting by incumbents. 172 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 188 Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Government Restrictions on Exit Some regulators want to restrict companies from closing plants. These regulators appear motivated by concerns over people who lose their jobs when a company closes a plant. Restrictions on plant closings, however, are likely to reduce the desirability of entry into an industryrms will be reluctant to enter an industry if they cannot exit easily when they are losing money. Thus, potential effects of government restrictions on exit are less vigorous competition in the affected industries, higher consumer prices, and lower levels of employment. Learning-Curve Effects In Chapter 5, we discussed how average costs are reduced in some industries through production experience. As production experience accumulates, the rm learns how to lower unit costs. Learning-curve effects can result in new rivals having a cost disadvantage relative to existing rms. Whether these effects are important depends on whether the new entrants simply can copy the techniques learned by existing rms through their experience. Pioneering Brand Advantages Sometimes, a rm benets from being rst in an industry. In some industriesover-thecounter drugs, for examplea satised customer might be reluctant to switch brands even if the price of a competing product is substantially lower. This tendency is likely to be strongest in experience goods, which have to be tried by the customer to ascertain quality. For instance, customers might hesitate to try a new pain reliever because they fear that it might not be as effective as their regular brand. Where quality can be judged by inspection prior to purchase, this advantage of incumbents is lower. Sometimes the incumbents advantage with an experience good can be overcome by a new entrant through free samples, endorsements, or government certication. Each of these methods entails additional coststhese costs of overcoming incumbent advantages deter entry. Exit Costs Another important entry consideration centers on the costs of exit. In some industries, it is possible to hit and run. For instance, forming a new company to seal asphalt driveways requires little investment in specialized equipment or training. A new rm can enter quickly when the prot potential is high and exit at low cost if prots decline. In other industries, especially those with specic assets, exit costs can be high. In such industries rms bear signicant costs, such as moving employees to new locations and liquidating plants and other assets when they decide to exit. High exit costs deter initial entry. Monopoly Effective barriers to entry limit the threat of competition and give incumbent rms market power. Although competitive markets are at one end of the spectrum, at the other end is monopoly where there is but a single rm in the industry. Here, industry and rm demand curves are one and the same. Prot Maximization Suppose that a monopolist charges the same price to all customers. (As discussed shortly, such a pricing policy might be motivated by either government regulation or the inability BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 6 Monopoly This gure illustrates the price and output decisions of a monopolist. In the example, demand is P 200 Q. Marginal costs are $10. The protmaximizing output occurs at 95 units, where MR MC. To sell this output, the rm charges a price of $105. The rm makes $95 per unit prot ($105 $10) for a total prot of $9,025 ($ 95 95), as indicated by the shaded rectangle abcd. Some consumers are willing to pay more than the marginal cost of production, yet do not receive the product. The associated loss in potential gains from trade is pictured by the shaded triangle cde. The rm does not lower the price to sell to these consumers because it does not want to lower the price for other customers. Market Structure 189 $ Price and cost per unit of output Figure 6.5 173 The McGrawHill Companies, 2009 6. Market Structure P* 105 b 10 c d a Q* e 95 MC = AC Q MR Quantity to prevent resale among customers; in Chapter 7, we relax this restriction.) The rms objective is to choose the price-quantity combination along the demand curve that maximizes prots. This combination occurs where marginal revenue equals marginal cost. For purposes of illustration, consider the following linear demand curve: P 200 Q (6.4) (Assume that marginal cost is constant at $10.) Recall from Chapter 4 that the marginal revenue curve for a linear demand curve is a line with the same intercept and twice the negative slope. Figure 6.5 displays the demand curve, marginal revenue curve, and marginal cost curve in this example. Optimal output occurs at 95 unitswhere MR MC. To sell this output, the rm charges a price of $105. The rm makes $95 prot per unit ($105 $10) for a total prot of $9,025 ($95 95); this is indicated by the shaded rectangle abcd. As opposed to pure competition, monopolistic suppliers charge customers more than the marginal and average costs of production and distribution; the rm thus earns an economic prot. Monopolies restrict output compared to competitive industries. In our example, if the industry were competitive, the market price would be $10 (marginal cost) and total quantity sold would be 190 units. (Pricing by rms with market power is discussed in additional detail in Chapter 7.) Unexploited Gains from Trade Given the monopolists output and pricing choices, some consumers are willing to pay more than the marginal cost of production and distribution, yet do not purchase the product. Thus, not all gains from trade are exhausted. The associated loss in potential gains from trade is pictured by the shaded triangle cde in Figure 6.5. Consumers along this segment of the demand curve value the product at more than $10 but 174 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 190 Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Pricing and Investment Decisions You work for a drug manufacturing company that holds a patent on Hair Grow, the worlds most effective drug for restoring hair. Your job is to analyze the pricing and investment decisions facing the rm. Your marketing group estimates that Hair Grow has the following demand curve: P 101 .00002Q 1. Your marginal cost for producing a Hair Grow pill is $1. What is the prot-maximizing price and quantity? What is your prot? 2. Suppose that your production facility can only produce 1,000,000 pills. What is your optimal price and quantity given the production constraint? What are your prots? 3. Suppose that you could increase the capacity of your plant to 3,000,000 pills within a two-year period for a cost of $30,000,000. Should you undertake the investment (for simplicity, assume you can borrow the funds for the expansion at a 0 percent interest rate)? less than $105. The rm does not lower the price to sell to these consumers because it does not want to lower the price for other customers (recall that in this chapter we presume the rm charges the same price to all customers). From the rms standpoint, any gain from selling to additional customers would be more than offset by the loss from lowering its price to all its customers.8 Monopolistic Competition As the name implies, monopolistic competition is a market structure that is a hybrid between competition and monopoly. In this market structure, there are multiple rms that produce similar products. There is free exit from and entry into the industry. Yet competition does not eliminate market power because the rms sell differentiated products. Examples of such industries include toothpaste, golf balls, skis, tennis rackets, shampoo, and deodorant. For instance, although Colgate and Crest compete directly, many customers do not view these brands of toothpaste as perfect substitutes. These companies thus have some market power. New toothpaste rms are likely to enter the industry if the existing rms report large protsthere are no signicant barriers to entry. Monopolistic competition is similar to monopoly in that rms under both market structures face downward-sloping demand curves: A toothpaste company can raise its price without losing all sales. Given that the rms face downward-sloping demand curves, each strives to select the price-quantity combination that maximizes its prots. The output decision is based on the same analysis as for the pure monopolistchoose that output where MC MR. 8 Economists frequently refer to these lost gains from trade as a dead-weight loss. This inefciency (or social cost) is one reason why governments might pass regulations like antitrust laws to restrict the formation of monopolies. But these regulations also can be motivated by concerns about the higher prices that consumers pay when they face monopolistic suppliers. Although government regulation has the potential to reduce inefciencies and wealth transfers from consumers to rms, it is important to keep in mind that government regulation is not costless. There are salaries for regulators and court costs, for instance. From a societal viewpoint, the costs of government regulation should be weighed against the benets. These issues are discussed in greater detail in Chapter 21. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 175 191 MANAGERIAL APPLICATIONS Monopolistic Competition in Golf Balls There are many brands of golf balls. Some golfers view the balls as perfect substitutes and simply purchase the lowest-priced brand. Other golfers prefer one brand to another. For instance, they might believe one brand of ball ies farther or provides greater control than competing brands. These golfers are willing to pay a higher price for their favorite ball than for competing balls. However, they often will substitute if the price difference is more than a few dollars a dozen. Also, if a company develops some popular feature, like a larger number of dimples on the ball, the feature is typically copied by other companies within a short time period. Since a golf equipment company has a monopoly in producing its own brand, it has some market power. However, this power is limited given the competition in the industry. The difference between monopoly and monopolistic competition is that in monopolistic competition, economic prots invite entry. If a toothpaste with a new whitening formula is a hot seller, other companies will imitate the product. This entry will shift the original rms demand curve to the left and reduce prots. Zero economic prots exist when the demand curve is shifted to the point where average cost equals price. Figure 6.6 (page 192) shows this condition. Entry will tend to force prots to zero. Yet some brands continue to be more distinctive than others. Also, costs can vary because of differences in production techniques and inputs. It is possible for some rms to earn economic prots in monopolistic competition.9 Oligopoly Within oligopolistic markets, only a few rms produce most of the output. Examples of oligopolistic industries include automobiles and steel during the 1950s. These industries had important scale economies and other substantial entry barriers. In 1995, the top four cereal makers in the United States produced about 90 percent of industry output, while the top eight accounted for virtually all production. Products may or may not be differentiated. Firms can earn substantial prots. These prots are not reduced through new entry because of effective entry barriers. Yet as we shall see, economic prots sometimes can be eliminated in oligopolistic industries through competition among the existing rms. In our analysis of other market structures, we assume that rms take the prices of their competitors as given. A rm was not expected to respond to announcements of changes in prices by rival rms. This assumption certainly is reasonable in the case of competitive markets with many small rms, as well as in the case of monopoly with only one large rm. But this assumption rarely is valid within oligopolistic industries. For instance, when American Airlines considers lowering its prices on particular routes, it obviously must be concerned about whether United Airlines and its other competitors will follow suit. In fact, rms in oligopolistic industries ordinarily will be quite concerned about how their rivals will react to most major policy decisions, be they advertising campaigns or product design decisions. Decision making within these industries requires strategic thinking. Decision makers must realize that competitors are rational parties operating in 9 Monopolistic competition does not exhaust all gains from trade for two reasons. First, as in monopoly, the rms do not sell to all consumers who value the product at above marginal cost. Second, rms do not operate at the bottom of their average cost curves (see Figure 6.6). Lower average cost would be obtained with fewer rms, each producing more output. Nonetheless, regulation to address these inefciencies is unlikely to be effective. Consumers value product differentiation and are arguably better off with more variety at slightly higher average cost than with lower variety produced at lower average cost. Second, with few entry barriers, the market power of rms is unlikely to be great. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 192 Part 2 Figure 6.6 II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Managerial Economics Monopolistic Competition In monopolistic competition, rms sell differentiated products. This gure shows the demand curve for rm i in such an industry. The curve is downward-sloping. Similar to monopoly pricing, the rm selects the output where marginal revenue equals marginal cost. Monopolistic competition differs from monopoly in that abnormal prots will invite entry. Entry shifts the demand curve for the rm to the left (as some of the customers buy from the new rms). The rm makes no economic prots when price is equal to average cost. This condition occurs at price P * and quantity Q*. i i $ Price and cost per unit of output (in dollars) 176 LRACi LRMCi P i* Di Qi Q i* MRi Quantity (firm i ) their own self-interest. Thus, it is important for decision makers to place themselves in their rivals positions and consider how they might react. (This basic principle, which we now examine briey, is developed more completely in Chapter 9.10) Nash Equilibrium To analyze oligopolies, we need an underlying principle to dene an equilibrium when rival rms make decisions that explicitly take each others behavior into account. Previously, we used the concept that a market is in equilibrium when rms are doing the best they can given their circumstances and have no reason to change price or output. For example, in a competitive equilibrium, there is no reason for entry or exit (existing rms are making normal prots). No existing rm has any reason to change its output level (all are producing where MC MR P * and inventories are stable at their desired levels). We can apply this same basic idea to oligopolistic markets with minor modication. In the following analysis, a rm does the best it can, given what its rivals are doing. In doing so, the rm anticipates that other rms will respond to any action it takes by doing the best they can as well. Actions are noncooperative in that each rm makes decisions that maximize its prots, given the actions of the other rms. The rms do not collude to maximize joint prots. An equilibrium exists when each rm is doing the best it can, given the actions of its rivals. Economists call this a Nash equilibrium for Nobel laureate John Nash who rst developed these general concepts. To illustrate this approach, assume a simple setting: There are two rms in an industry a duopoly. Each independently chooses a price for an identical product. The rms choose 10 This chapter presents a basic introduction to game theory. The material provides sufcient background for the game theory applications found in subsequent parts of the book (these are in the appendices of several chapters). Chapter 9 extends this introduction of game theory and discusses in more detail how managers might use this theory as a tool in decision making. Readers interested in a more detailed treatment of game theory should read Chapter 9. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure $20 $40 WonCoLow price In this example, there are two rms in an industryWonCo and TuInc. Each independently chooses a price for an identical product. The rms choose either a high price or a low price. The payoffs are given in the table (the upper-left entry in a cell displays the prots for WonCo, the lower right shows the prots for TuInc). The equilibrium is for WonCo to charge a high price and TuInc to charge a low pricethe shaded cell. Any other combination is unstable: That is, given the action of one of the rms, the other rm has the incentive to deviate. This equilibrium is called a Nash equilibrium. TuIncLow price $200 $250 Market Structure 193 $40 $0 TuIncHigh price WonCoHigh price Nash Equilibrium WonCoHigh price Figure 6.7 WonCoLow price Chapter 6 177 TuIncLow price $400 $200 TuIncHigh price either a high price or a low price. The payoffs are given in Figure 6.7. (The entry on the upper left in each cell is for WonCo, while the entry on the lower right is for TuInc.) For example, if both rms charge a high price, WonCos prots are $400 and TuIncs prots are $200.11 The equilibrium is for WonCo to charge a high price and TuInc to charge a low price. Any other combination is unstable: Given the action of one of the rms, the other rm has the incentive to change its price. For instance, if both rms charge a high price, it is in the interests of TuInc to lower its priceits prots go from $200 to $250. The other combinations of WonCo charging a low price and TuInc a high price and both rms charging a low price are similarly unstable: Each rm has an incentive to alter its price given the other rms choice. A Nash equilibrium is self-enforcing. If WonCo charges a high price, it is optimal for TuInc to charge a low price. Similarly, if TuInc charges a low price, it is optimal for WonCo to charge a high price. Given the choice of one rm, there is no reason for the other to alter its strategy. In this example, the Nash equilibrium is not the outcome that maximizes the joint prots of the two companies. Combined prots would be higher if both rms charged a high price. Conceptually, the combined prots under this pricing policy could be split in a manner that would make both rms better off than in the Nash equilibrium. For instance, the combined prots of $600 could be split, with each rm receiving $300. As this example illustrates, noncooperative equilibria do not necessarily maximize the joint value of the rms. (Since the potential gains from trade are not exhausted, it is often the case that one or more rms can be made better off, without making other rms worse off, by changing the joint decisions.) Output Competition The rst major analysis of oligopoly was published by Augustine Cournot in 1838. To illustrate his model, suppose again that there are only two rms in the industry and that 11 The prots differ due to differences in the underlying production costs. 178 194 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Managerial Economics they produce identical products. In the Cournot model, each rm treats the output level of its competitor as xed and then decides how much to produce. In equilibrium, neither rm has an incentive to change its output level, given the other rms choice. (Thus, this is a Nash equilibrium.) Suppose the duopolists face the following total industry demand: P 100 Q (6.5) where Q Q A QB. For simplicity, assume that both rms have marginal costs of zero: MCA MCB 0. Each rm takes the other rms output as xed. Thus, the anticipated demand curve for rm i (i A or B) is Pi where Q j (100 Q j) Qi expected output of the other rm. The marginal revenue for rm i is MRi (100 Q j) 2Q i (6.6) 12 (6.7) Firm i s prots are maximized by setting marginal revenue equal to marginal cost (in this case, zero). Doing so, and rearranging the expression, yields the following reaction curve: Qi 50 0.5Q j (6.8) The reaction curve indicates rm i s optimal output given the output choice of rm j. Both rms have the same reaction curve in this example, except that the subscripts are reversed. Equilibrium occurs where the two curves cross. At these output levels, each rm is maximizing prot given the other rms output choice. Neither rm has an incentive to alter its output. The equilibrium is pictured in Figure 6.8. In equilibrium, each rm produces 33 units for a total output of 66 units; the price is $33.34. This output level is lower than in a competitive market. With competition, total output would be 100 units and the price would be zero (where P * MC). In the Cournot equilibrium, rms make economic prots: Price is $33.34; average costs are zero. Each rm thus reports prots of $1,110.89. This prot is lower than the two rms could obtain if they directly colluded and jointly produced the monopolistic output of 50 units (for example, 25 units per rm). With effective collusion, joint prots would be $2,500 rather than $2,221.78. Figure 6.9 displays the three price-quantity outcomes using the original industry demand curve. (This model can be extended readily to more than two rms. The same general results hold: As the number of rms grows, outcomes approach those of a competitive market.) Price Competition In the Cournot model, rms focus on choosing output levels. An alternative possibility is that rms might focus on choosing product price.13 Here, the Nash equilibrium is for both rms to choose a price equal to marginal costthe competitive outcome. To see 12 Recall that marginal revenue for a linear demand curve is a line with the same intercept, but twice the negative slope. 13 This situation often is referred to as the Bertrand model. Bertrand was a French economist who wrote a short note almost 50 years after Cournots work was published arguing that in some markets, producers set prices rather than quantities. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 6 Figure 6.8 Cournot Equilibrium 100 195 Firm As reaction curve a 50 b 33.33 25 a = Competitive equilibrium b = Cournot equilibrium c = Collusive (monopoly) equilibrium c Firm Bs reaction curve QA 25 33.33 50 100 Quantity of output by Firm A Figure 6.9 Comparison of Prices and Outputs among Collusive, Cournot, and Competitive Equilibria $ 100 Price (in dollars) In this example, the total industry demand curve is P 100 Q. Marginal cost is zero. The gure shows the price-quantity outputs for the industry under collusive, Cournot, and competitive equilibria. The output is smallest and the price is highest for the collusive equilibrium. The output is largest and the price is smallest for the competitive equilibrium. Market Structure QB Quantity of output by Firm B The duopolists in this example face the total industry demand curve, P 100 Q, where Q is the sum of the two outputs. Both rms face a marginal cost of zero. The gure shows the reaction curves for each rm. The reaction curve indicates rm is optimal output given the output choice of rm j (i, j A or B). The Cournot equilibrium occurs where the two reaction curves cross. Each rm produces 33.33 units. The market price is $33.34. The output for the rms is lower and the prots are greater than in the competitive equilibrium. The output for the rms is greater and the prots are lower than in the collusive (monopoly) equilibrium. 179 The McGrawHill Companies, 2009 6. Market Structure Collusion 50 Cournot 33.34 Competition 0 MC = 0 Q 50 66.66 MR Quantity 100 180 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 196 Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Price Wars Firms in oligopolistic industries sometimes engage in intense price competition to their mutual detriment. Examples include the airline and wireless industries. The major airlines, including American, United, Delta, and Northwest, periodically enter fare wars that lower the price of air travel for consumers and lower the combined prots of the airline industry. For instance, in 2005 Delta lowered its fares in the hope of gaining new passengers. This price reduction was matched by other airlines within hours. In 2002 wireless carriers such as Sprint PCS, AT&T Wireless, and Verizon stole each others customers offering look-alike calling plans and rock-bottom prices. As soon as one company offered a new calling plan, the rest followed suit. Outside analysts generally agree that the rms lose prots through these price cuts. Financial analysts following the stocks of wireless service companies recommended to their clients that they reduce the amount invested in this sector because intense price competition was lowering prots. Sources: S. Mehta (2001), First That Old Sinking Feeling, Fortune (December 10), 3435; and M. Rollins (2002), Wireless Services Second Quarter Preview: Less Than a Zero Sum Game, Solomon Smith Barney Industry Note ( July 8); J. Horwich (2005), Northwest Matches Deltas Fare War . . . Reluctantly, Minnesota Public Radio (January 5). why, suppose one of the rms chooses a price, P MC. In this case, it is optimal for the rival rm to charge a price just below P to capture all industry sales. (Since we assume the rms products are identical, customers buy the product from the rm that offers the lower price.) Given that the second rm charges a price just below P , it is now optimal for the rst rm to charge a slightly lower price. This process continues; only when price equals marginal cost does neither rm have an incentive to lower price. (Lowering price further would result in selling below cost, thus generating a loss.) Of course, both rms would like to devise a way of avoiding competition and capturing higher prots. Yet as we discuss below, fostering cooperation can be difcultand in certain cases, illegal. ANALYZING MANAGERIAL DECISIONS: Entry Decision In the simple Cournot model, rms make their output choices simultaneously. In practice, rms sometimes make these kinds of decisions sequentially. Suppose that you manage one of the rms discussed in the Output Competition example in the text. The industry demand in this example is P 100 Q and the MC of each rm is zero. 1. Suppose that each rm must make an upfront investment of $1,000 to enter the market and that your competition has already paid this investment and chosen to produce 50 units. This investment is nonrecoverable (sunk). Should you make the $1,000 investment and enter the market? If so, how much should you produce and what are your prots? Continue to assume that your rm will survive for only one production period. 2. How do your prots and those of your competitor compare to the case of simultaneous decisions discussed in the text? Would you say that this example of output competition has a rst mover advantage or disadvantage? BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 181 197 Empirical Evidence There are various economic models of oligopoly. We have presented but two of them simply to illustrate that economic theory does not make unambiguous predictions about what to expect within such industries. Some models yield outcomes close to pure competitionrms sell at marginal cost and make no economic prots. Other models yield outcomes closer to pure monopoly. What actually occurs in specic oligopolistic markets is an empirical issue. Available evidence suggests that oligopolies typically result in less output than competitive markets and that rms earn economic protsat least in some industries.14 Firms sometimes compete on price, to each others detriment, and normally earn less in aggregate than a monopolist could. Cooperation and the Prisoners Dilemma As we have discussed, in oligopolistic industries it is in the private interests of rms to nd ways to cooperate and capture more prots than through competition. In principle, rms are most protable if they effectively collude and act as a monopolist in jointly setting price and output for the industry. Collusion maximizes joint prots, which then can be divided among the rms in the industry. Many governments understand these incentives and have passed a variety of antitrust laws to limit rms ability to engage in xing prices. These laws are designed to lower the prices consumers pay for products. Some of the more restrictive of these laws have been adopted in the United States. Internationally, rms tend to have more latitude in forming cooperative agreements in attempting to increase protsfor example, consider the OPEC cartel.15 Prisoners Dilemma Even when free to cooperate, rms nd that cooperation is not always easy to achieve. Individual rms have incentives to cheat and not adhere to output and price agreements. This incentive can be illustrated by the well-known prisoners dilemma. In the original prisoners dilemma, there are two suspects; hence, suppose the SEC has been investigating an insider trading scheme and their investigation is focused on two securities brokers, Avi Wasserman and Bea Haefner, who are arrested and charged with a crime. Police have insufcient evidence to convict them for insider trading violations unless one of them confesses. The police place Avi and Bea in separate rooms and try to get them to confess. If neither confesses, they are convicted of less serious crimes associated with their trading activities and are sentenced to only 2 months in jail. If both confess, they spend 12 months in jail. However, if one confesses but the other does not, the confessor is released under a plea bargain in return for testifying while the other is sentenced to 18 months in jail12 for the crime and 6 for obstructing justice. The payoffs in terms of jail time faced by each individual are displayed in Figure 6.10 (page 198). The Nash equilibrium is for both suspects to confess. Given these payoffs, it is always in the individual interests of each suspect to confess, taking the action of the other party as given. If Avi does not confess, Bea is set free by confessing. Alternatively, if Avi 14 D. Carlton and J. Perloff (1990), Modern Industrial Organization (HarperCollins: New York), Chapter 10, discusses some of the relevant empirical literature. 15 In smaller countries, much of the local production of key products is exported. In this case, it can be in the countries interests to allow the formation of cartels. Ultimately, consumers pay higher prices and there are inefciencies. However, many of these costs are imposed on people in other countries. 182 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 198 Part 2 Figure 6.10 II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Managerial Economics Prisoners Dilemma 2 months 2 months AviNo confession AviNo confession In the prisoners dilemma, there are two suspects: Suppose the SEC has been investigating an insider trading scheme and their investigation is focused on two securities brokers, Avi Wasserman and Bea Haefner, who are arrested and charged with a crime. The police do not have sufcient evidence to convict them for insider trading unless one of them confesses. The police place the suspects in separate rooms and ask them to confess. If neither confesses, they are convicted of minor violations uncovered by the investigation of their trading activities and are sentenced to 2 months. If both confess, they spend 12 months in jail. However, if one confesses and the other does not, the confessor is released under a plea bargain in return for testifying but the other is sentenced to 18 months in jail12 for the crime and 6 for obstructing justice. The payoffs in terms of jail time faced by each individual are displayed. Each entry in the table lists the jail sentences for Avi and Bea, respectively. The Nash equilibrium is for both suspects to confessthe shaded cell. Given the payoffs, it is always in the individual interests of each suspect to confess (taking the action of the other party as given). 0 months 18 months BeaNo confession 0 months BeaConfession AviConfession AviConfession BeaNo confession 18 months 12 months 12 months BeaConfession confesses, Bea reduces her jail sentence from 18 to 12 months by also confessing. Either way, it is in Beas interests to confessconfessing is a dominant strategy. Since the payoffs are symmetrical, it also is optimal for Avi to confess. Although it is in the individual interests of each party to confess, it is clearly in their joint interests not to confess. By not confessing, each only serves 2 months in jail, compared to 12 months if both confess. The prisoners dilemma suggests that any agreement for neither to confess is likely to break down when they make their individual choices unless there is some mechanism to enforce their joint commitment not to confess. (For particular crimes one such mechanism might be the Mob: Both suspects have incentives not to confess if they expect to be executed for providing evidence to the police.) Cartels Cartels consist of formal agreements to cooperate in setting price and output levels. (These activities are generally illegal in the United States.) Firms trying to maintain cartels can face a problem like the prisoners dilemmawe might call it the cartels dilemma. Members can agree to restrict output to increase joint prots. However, individual rms have incentives to cheat. If all other rms restrict output, prices will not be affected BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure $500 BeaCoLow output $600 $150 BeaCoLow output AVIncLow output $500 AVIncHigh output Two rms, AVInc and BeaCo, attempt to form a cartel. If both rms restrict output, prices are high and each rms prot is $500. If both cheat on the cartel and increase output, price will be low and each rms prot is $200. If one rm expands output while the other restricts output, the market price will be at an intermediate level; the rm with the high output will make $600 (because of the increased sales), but the other rm will only make $150 (because of the lower price). These payoffs are displayed. The Nash equilibrium is for both rms to increase outputthe shaded cell. Given the payoffs, it is always in the interest of each rm to increase output (taking the output of the other rm as given). AVIncHigh output Figure 6.11 Cartels Dilemma AVIncLow output Chapter 6 Market Structure 183 199 $150 $600 BeaCoHigh output $200 $200 BeaCoHigh output signicantly by the extra output of one rm. However, that rms prots will increase from selling additional output at the cartel-maintained high price. But if all rms react to these incentives by increasing output, the cartel breaks down. Actual cartels often unravel because of such incentives. This outcome is pictured in Figure 6.11, which displays the payoffs for two rms attempting to form a cartel. It is in their joint interests to restrict output. Yet, as in the prisoners dilemma, both rms have individual incentives to renege and increase output. The Nash equilibrium is for each to increase output. A cartel can persist if it can impose sufcient penalties on cheaters. But for these penalties to be effective, cartel members must be able to observe (or reliably infer) that a rm has cheated. To the extent that cartel members expect to interact on a repeated basis, there are greater incentives to cooperate. Repeated interaction also increases incentives to invest in developing effective enforcement mechanisms to limit cheating. Potentially, these incentives can be strong enough to resolve the cartels dilemma. In general, cooperation is easier to enforce if the number of rms in the industry is small: It is easier to identify and punish cheaters. Even when rms are not permitted to form cartels, there may be ways of cooperating to increase prots. For example, over time, a rm might become known as a price leader. Such a rm changes prices in the face of new demand or cost conditions in a way that approximates what a cartel would do. Other rms follow the price changes, thus acting like members of a cartel. Individually, rms still can have short-run incentives to cheat (for example, reducing price to get more sales). However, rms might resist this shortrun temptation to foster cooperation in the long run and hence obtain higher long-run prots.16 Another potential mechanism to foster cooperation is the structure of contracts with buyers employed by rms in the industry. Most-favored-nation clauses provide buyers 16 Indeed, economists have shown that within any long-term relationship, with no specied ending date, cooperation is a possible equilibriumthe parties need not succumb to the cartels dilemma. We discuss this issue in more detail in Chapter 9 and in the appendix to Chapter 10. 184 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 200 Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Collusion in the Lysine Industry Mark Whitacre was a high-ranking executive at Archer Daniels Midland Corporation (ADM). He has accused his former employer of engaging in price xing. Lysine is an amino acid derived from corn used in swine and poultry feed to promote growth. ADM entered the lysine market in 1991. Prior to that time, the market had been dominated by two Japanese companies. ADM quickly gained market share. However, with the competition, the price of lysine fell from about $1.30 per pound to $.60 per pound. According to Whitacre, ADM executives began discussions in 1992 with their Japanese competitors about how it would be in their mutual self-interest to collude and x prices. Collectively, the competitors were forgoing millions of dollars of prot per month because of the competition among the three companies. Whitacre indicates that a favorite saying at ADM is: The competitor is our friend, and the customer is our enemy. In 1996, ADM pleaded guilty to price xing and paid a $100 million ne. In 1999 Whitacre was sentenced to two-anda-half years in prison. Source: M. Whitacre, as told to R. Henkoff (1995), My Life as a Corporate Mole for the FBI, Fortune Magazine (September 4), 5262; K. Eichenwald (1999), Three Sentenced in Archer Daniels Midland Case, New York Times ( July 10), C1. with guarantees that the seller will not sell to another buyer at a lower price. These clauses reduce incentives of sellers to lower the price for one buyer because that same price concession would have to be offered to other buyers as well. Meeting-the-competition clauses guarantee that a seller will meet the price of a competitor. Such a clause makes it difcult for rms to cheat on an agreement not to lower price since price concessions are more likely to be brought to each others attention by customers. Our discussion of these strategic interactions among rivals in output markets is meant only to provide a basic introduction to these issues. A more extensive analysis is provided in Chapter 9. Summary A market consists of all rms and individuals who are willing and able to buy or sell a particular product. These parties include those currently engaged in buying and selling the product, as well as potential entrants. Market structure refers to the basic characteristics of the market environment, including (1) the number and size of buyers, sellers, and potential entrants; (2) the degree of product differentiation; (3) the amount and cost of information about product price and quality; and (4) the conditions for entry and exit. Competitive markets are characterized by four basic conditions: A large number of potential buyers and sellers; product homogeneity; rapid, low-cost dissemination of information; and free entry into and exit from the market. In competitive markets, individual buyers and sellers take the market price of the product as given: They have no control over price. Firms thus view their demand curves as horizontal. The rms shortrun supply curve is that portion of its short-run marginal cost curve above short-run average variable cost. The long-run supply curve is that portion of its long-run marginal cost curve above long-run average cost. In a competitive equilibrium, rms make no economic prots. Production is efcient in that rms produce at their minimum longrun average cost. Firms in competitive industries must move rapidly to take advantage of transitory opportunities. They also must strive for efcient production in order to BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 185 201 survive. Some rms in the industry can employ resources that give them a competitive advantage (for example, an extremely talented manager). Yet in such cases, any excess returns often go to the factor of production responsible for the particular advantage, rather than to the rms owners. Although the competitive model provides a useful description of the interaction between buyers and sellers for many industries, there are others where rms have substantial market powerprices are affected materially by the output decisions of individual rms. Market power can exist when there are substantial barriers to entry into the industry. Expectations about incumbent reactions, incumbent advantages, and exit costs all can serve as entry barriers. The extreme case of a rm with market power is monopoly, where the industry consists of only one rm. Here, industry and rm demand curves are one and the same. In contrast to competitive markets, consumers pay more than marginal cost and the rm earns economic prots. Output is restricted from competitive levels. With a monopoly, not all the potential gains from trade are exhausted. Monopolistic competition is a hybrid between competition and monopoly. It is like monopoly in that rms under both market structures face downward-sloping demand curves. Market power comes from differentiated products. Examples include the toothpaste, golf ball, tennis racket, and shampoo markets. The analyses of output and pricing policies are similar in the two cases. The difference between monopoly and monopolistic competition is that in monopolistic competition, economic prots invite entry that limits prots. In oligopolistic markets, only a few firms account for most production. Products may or may not be differentiated. Firms can earn substantial profits. However, these profits can be eliminated through competition among existing firms in the industry. To analyze output and pricing decisions in oligopolistic industries, we use the concept of a Nash equilibrium: A Nash equilibrium exists when each firm is doing the best it can given the actions of its rivals. In the Cournot model, each firm treats the output level of its competitor as fixed and then decides how much to produce. In equilibrium, firms make economic profits. However, these profits are not as large as would be made if the firms effectively colluded and posted the monopoly price. Other models of oligopoly yield different equilibria. For instance, one model based on price competition yields the competitive solution: Price equals marginal cost with no economic profits. Economic theory makes no clear-cut prediction about the behavior of firms in oligopolistic industries. Available evidence suggests that in some oligopolistic industries, firms restrict output from competitive levels and hence capture some economic profits. It is in the economic interests of rms in oligopolistic industries to nd ways to cooperate, thereby capturing higher prots. Even when rms are free to cooperate, effective cooperation is not always easy to achieve. Individual rms have incentives to deviate from agreed-on outputs and prices. This incentive is illustrated by the prisoners dilemma. This model highlights incentives that can cause cartels to be unstable. However, rms sometimes can cooperate successfully when they can impose penalties on noncooperative rms. Cooperation also can be sustained through the incentives provided by long-run, repeated relationships. Suggested Readings A. Dixit and B. Nalebuff (1991), Thinking Strategically (Norton: New York). G. Stigler (1987), The Theory of Price (Macmillan: New York), Chapter 3. R. Pindyck and D. Rubinfeld (1992), Microeconomics (Macmillan: New York), Chapters 813. 186 202 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 6. Market Structure Managerial Economics Self-Evaluation Problems 61. The total and marginal cost of producing Product A are: TC MC $1,000 4Q 2Q 2 The $1,000 is a xed cost in the short run, but can be avoided in the long run by shutting down (going out of business). There is only one possible plant size for this operation; thus SRMC LRMC 4Q in this problem. Derive and graph the rms short-run and long-run supply curves (on separate graphs). 62. Suppose your rm faces a demand curve of P 90 .30Q and the marginal cost of production is $10/unit. Find the prot-maximizing output and price. Display this choice graphically (showing the demand, marginal revenue, and marginal cost curves). Is this outcome on the elastic, inelastic, or unitary elastic part of the demand curve? What are your prots? 63. Genesee and Natural Light are the two sole competitors in the ultra low-end beer market in the Rochester metro area. Both rms have marginal costs of 0 and xed costs of 200. The industry demand curve is P 100 0.1Q, where Q Q 1 Q 2 a. Assume the rms compete on quantity (Cournot competition). What is the equilibrium price and total production in the market? How much prot will each rm make? b. If rms compete on price (rather than quantity), is the answer in Part A a Nash equilibrium? Explain. c. Now assume that Genesee acquires Natural Lights Rochester operations and has a total regional monopoly on ultra low-end beer. What price will they charge? How much prot will the combined rm make? Is this more or less than in Part A? Explain briey why you found what you did. d. Now suppose that neither rm has entered the Rochester market and that it costs $10,000 to build a plant. Suppose that Genesee spends $10,000 to build a plant and starts producing 400 units. Will Natural Light want to spend $10,000 to enter the market? Show why. Assume that there is only one production period to consider. Solutions to Self-Evaluation Problems 61. The rms short-run supply curve is the portion of its short-run marginal cost curve that lies above average variable cost (AVC). AVC is VC Q 2Q 2 Q 2Q. SRMC, which is 4Q, is everywhere above 2Q. Therefore the rms short-run supply curve is P 4Q : P SR supply $4 Q 1 The rms long-run supply curve is the portion of its long-run marginal cost curve that lies above long-run average total cost (LRAC). LRAC in this problem is: LRAC TC Q $1,000 Q 2Q Recall that LRAC falls when LRMC is below it and rises when LRMC is above it. LRAC is minimized when it equals marginal cost. This is the point where positive supply begins BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 187 The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 203 (the rm stays in business), since its LRMC is above LRAC at every quantity beyond this point. The point can be found by setting LRAC LRMC and solving for Q and LRMC: $1,000 Q 2Q $500 Q 4Q Q2 22.36 units At this quantity LRMC LRAC $89.44. At this price the rm breaks even (TR TC). At lower prices the rm loses money and in the long run it is optimal to shut down. Thus the rms long-run supply is 0 up to a price of $89.44 and is P 4Q at higher prices: P LR supply LRAC $89.4 Q 22.36 62. Prot maximization occurs where MR MC. The MR function has the same intercept as the linear demand curve but twice the negative slope. Therefore the prot-maximizing condition is: 90 .60Q 10. Solving this equation produces the prot-maximizing quantity, Q * 133.33 units. The optimal price (from the demand curve) is $50. The rm makes a prot of $40/unit for a total prot of $5,333.20. The outcome is on the elastic portion of the demand curve (above the midpoint on the linear curve). It is never optimal for a rm to produce on the inelastic portion of its demand curve since revenue is increased and costs reduced by lowering output. The graph for this problem is: $/unit Demand $90 $50 $10 MR 133.33 MC Q 63. a. In Cournot competition, each rm maximizes its own prots taking the other rms output as given. In the equilibrium, both of the rms expectations about the other rms output are realized. The prot-maximizing conditions for the two rms are: MR1 MR2 (100 (100 .1Q 2 ) * .1Q 1 ) * .2Q 1 .2Q 2 0 (optimal condition rm 1) 0 (optimal condition rm 2) The equilibrium occurs at quantities that jointly satisfy these two equations. In this problem, both rms are identical. Thus in equilibrium Q 1 * Q 2 . Substitute this relation into either * 188 204 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 6. Market Structure The McGrawHill Companies, 2009 Managerial Economics of the two equations to produce a single equation with one unknown variable. Using the rst equation: (100 Q1 * .1Q 1 ) .2Q 1 * *0 333.33 units Q 2 * * Q 2 ) 666.66 units. The equilibrium price from the industry * Total production (Q 1 demand curve is $33.34. Each rm makes prots of ($33.34 333.33) $200 $10,913.22. b. No the solution in Part A is not a Nash equilibrium if the rms compete on price. In a Nash equilibrium, each rm is choosing optimally given the choice of the other rm. If one rm chooses a price of $33.34, the other rm would do better by undercutting the price and capturing the entire market. For example, if rm 1 cuts its price by $1 to $32.34 it will sell 676.6 units (assuming the other rm does not match the price decrease) and make a prot of $21,674.48 (greater than the $10,913.22 in Part A). The Nash equilibrium with price competition and homogenous products is for both rms to price at MC ($0 in this problem). c. After the merger, Genesee is a monopolist. Its demand curve is the industry demand curve. It will maximize prots by setting its MR, 100 0.2Q 0. Q * 500 units and P * from the demand curve is $50. The rm makes a prot of ($50 500) $200 $24,800. This prot is higher than the combined prot in Part A ($21,826.44). As a monopoly the rm can charge the price (and produce the corresponding quantity) that maximizes total prots for the industry. Firms make economic prots in the Cournot equilibrium but their competition (noncollusion) produces an outcome with lower prots than monopoly. d. No, Natural Light will not enter. If Genesee is producing 400 units, Natural Lights demand curve is P (100 .1 400) .1Q 2. Its optimal production (setting its MR 0) is 300 units and the resulting price will be $30 (given 700 total units are being produced). Natural Lights prots are (300 $30) 200 $8,800, which are less than the $10,000 entry fee. Review Questions 61. What four basic conditions characterize a competitive market? 62. The short-run marginal cost of the Ohio Bag Company is 2Q. Price is $100. The company operates in a competitive industry. Currently, the company is producing 40 units per period. What is the optimal short-run output? Calculate the prots that Ohio Bag is losing through suboptimal output. 63. Should a company ever produce an output if the managers know it will lose money over the period? Explain. 64. What are economic prots? Does a rm in a competitive industry earn long-run economic prots? Explain. 65. The Johnson Oil Company has just hired the best manager in the industry. Should the owners of the company anticipate economic prots? Explain. 66. A Michigan court ruled in the 1990s that General Motors did not have the right to close a particular Michigan plant and lay people off. Do you think this ruling beneted the people of Michigan? Explain. 67. The Suji Corporation has a monopoly in a particular chemical market. The industry demand curve is P 1,000 5Q. Marginal cost is 3Q. What is Sujis prot-maximizing output and price? Calculate the corresponding prots. 68. Assume the industry demand for a product is P 1,000 20Q. Assume that the marginal cost of product is $10 per unit. a. What price and output will occur under pure competition? What price and output will occur under pure monopoly (assume one price is charged to all customers)? b. Draw a graph that shows the lost gains from trade that result from having a monopoly. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 189 The McGrawHill Companies, 2009 6. Market Structure Chapter 6 Market Structure 205 69. In 1981, the United States negotiated an agreement with the Japanese. The agreement called for Japanese auto rms to limit exports to the United States. The Japanese government was charged with helping make sure the agreement was met by Japanese rms. Were the Japanese rms necessarily hurt by this limited ability to export? Explain. 610. Compare the industry output and price in a Cournot versus a competitive equilibrium. Do rms earn economic prots in the Cournot model? Does economic theory predict that rms always earn economic prots in oligopolistic industries? Explain. What does the empirical evidence indicate? 611. What is a Nash equilibrium? Explain why a joint confession is the Nash equilibrium in the prisoners dilemma. 612. Candak Corporation produces professional quality digital cameras. The market for professional digital cameras is monopolistically competitive. Assume that the inverse demand curve faced by Candak (given its competitors prices) can be expressed as P 5,000 0.2Q and Candaks total costs can be expressed as TC 20,000,000 0.05Q 2. Answer the following questions. a. What price and quantity will Candak choose? b. Is this likely to be a long-run equilibrium for Candak Corporation? Why or why not? If not, what is likely to happen in the market for professional digital cameras, and how will it affect Candak? 613. Will a monopolist ever choose to produce on the inelastic portion of its demand curve? Explain. 190 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Pricing with Market Power CHAPTER 7 CHAPTER OUTLINE Pricing Objective Benchmark Case: Single Price per Unit Prot Maximization Estimating the Prot-Maximizing Price Potential for Higher Prots Homogeneous Consumer Demands Block Pricing Two-Part Tariffs Price DiscriminationHeterogeneous Consumer Demands Exploiting Information about Individual Demands Using Information about the Distribution of Demands Bundling Other Concerns Multiperiod Considerations Strategic Interaction Legal Issues Implementing a Pricing Strategy Summary I ntuit began as a small software company in 1983 with its new program Quicken a personal nance program that addressed the common household problem of balancing the familys checkbook. By 2008, Intuit had grown to a company with over $2.6 billion in annual revenue and publicly traded stock on the Nasdaq Stock Market. Its agship products in 2008 included Quicken, QuickBooks, and TurboTax. A visit to Intuits cyberstore in 2008 would have revealed many interesting pricing decisions. Consider the following examples: (1) Intuit offered four versions of Quicken including its Starter, Deluxe, Premier, and Home & Business editions. The prices for these products ranged from the Starter Edition selling at $29.99 to the Home & Business Edition selling at $99.99. (2) The three higher priced editions could read computer les produced by older versions of Quicken, but the low-priced Starter Edition could not. Thus existing users who wanted to update their software, but use existing data les, had to purchase one of the higher priced alternatives. (3) Existing Quicken users were offered BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 191 207 $20 off the retail price and a $10 mail-in rebate (good until July 31, 2008) if they purchased Quicken Deluxe 2008. (4) Customers who purchased Quicken were given the option to purchase TurboTax (Intuits popular tax preparation software) and Quicken as a bundle at a $20 savings from the normal online prices. (5) Online customers could enter a Special Offer Code that gave discounts to customers with the code. (6) Intuit offered free trials on some of their products, such as Quicken Bill Pay. (7) Packages of 250 standard checks were priced at 26 cents per check, while packages of 2,000 checks were priced at only 9 cents per check. Pricing is a key managerial decision. These examples illustrate some of the complexities associated with product pricing. For example, how should managers set their basic prices? Why do rms use coupons and rebates? Why are some customers charged higher prices for the same product than others? Why do rms bundle products? Why do rms offer a line of similar products at different price points? Why would a rm ever give its products away? Why do some rms offer volume discounts? In this chapter we present a basic analysis of pricing with market power and provide answers to these and related questions. We organize the remainder of the chapter as follows: First, we discuss the underlying objective of pricing decisions. Next, we analyze the benchmark case where the rm charges the same price to all customers. Subsequently, we consider more complex pricing policies. The chapter ends with a brief discussion of several other issues, including multiperiod considerations, strategic interactions, legal and implementation issues. Pricing Objective A rm has market power when it faces a downward-sloping demand curve. Firms with market power can raise price without losing all customers to competitors. The ultimate objective is to choose a pricing policy that maximizes the rms value. We continue with the standard economic analysis in which managers seek to maximize prots over a single period. Although managers actually seek to maximize the present value of all future prots, if the business setting is expected to be stationary, these problems are equivalent. Later in this chapter, we discuss how concerns about future prots can affect the current pricing decision. Nonetheless, our single-period analysis provides useful insights into pricing decisions. Figure 7.1 (page 208) depicts a rms demand curve for and its marginal cost of producing the product. The demand curve reects what consumers are willing to pay for the product. Only in quite special cases is it in the interests of the rm to sell the product at below marginal cost: It can do better by not producing the product. (Later in this chapter, we examine multiperiod considerations that might prompt rms to set current price below marginal cost.) Thus, the maximum potential gains from trade are given by the shaded triangle. If the rm were to sell the product at marginal cost, all the gains would go to consumers in the form of consumer surplus. Consumer surplus is dened as the difference between what the consumer is willing to pay for a product and what the consumer actually pays when buying it. Prot-maximizing managers try to devise pricing policies that capture as much of the available gains from trade as possible: The managerial ideal would be to capture all the potential consumer surplus as company prot. We begin by reviewing the benchmark case where the rm charges all customers the same price. In this case (which was introduced in Chapters 4 and 6), the rm captures some, but not all, of the potential gains from trade. Subsequently, we consider more complex pricing policies. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 208 Part 2 Figure 7.1 Power II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Managerial Economics Pricing with Market The demand curve reects what consumers are willing to pay for the product. It typically is not in the rms interest to sell the product below marginal cost, since it can do better by not producing the product. Thus, the maximum potential gains from trade are given by the shaded triangle. The rms objective is to select a pricing policy that maximizes its share of the gains from trade and thus rm prot. $ Price (in dollars) 192 Demand MC Q Quantity Benchmark Case: Single Price per Unit Prot Maximization Suppose that Intuit sells a software product called Checkware. It purchases the product at $10 from the manufacturer and selects a retail price to post on the Internet. All customers buy at this price regardless of the quantity purchased. Intuit has no other incremental costs and faces the following demand curve: P 85 0.5Q (7.1) where P is price and Q is quantity (in thousands of units). What price should Intuit select to maximize prots? Chapter 6 shows that prots are maximized by selecting a price-output combination in which marginal revenue equals marginal cost. Marginal revenue is 85 Q, and marginal cost is $10 in this example.1 Thus the optimal quantity and price are 75 and $47.50, respectively. Prots are $2,812,500. Figure 7.2 illustrates the solution graphically. We focus on this example to provide a number of basic insights. It is important to note, however, that this analysis simplies the pricing problem in at least four important ways. First, all consumers are charged the same unit price, regardless of the quantity purchased. Thus more complicated pricing strategies are not considered. Second, the rm sells only one product; thus interactions among products are not considered. Third, the demand curve is for a single period. The analysis focuses on maximizing single-period prots and abstracts from longer-term considerations (for example, how pricing this period might affect either demand or costs in future periods). Fourth, the demand curve assumes that the prices of competing products are constant no matter what price 1 For expositional simplicity, we assume marginal cost is constant; in the more general case, one also requires information about the marginal cost function. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 7 Single Price per The marginal cost of the software is $10. Demand is given by P 85 0.5Q. If the rm charges a single unit price to all customers, prots are maximized by setting marginal revenue (MR 85 Q) equal to marginal cost. The resulting optimal price and quantity are $47.50 and 75, respectively. Pricing with Market Power 209 $ 85.00 Price (in dollars) Figure 7.2 Unit 193 The McGrawHill Companies, 2009 7. Pricing with Market Power P* 47.50 Demand MC 10.00 MR Q* 75 Q 170 Quantity of Checkware Intuit charges. In some markets, there is likely to be interaction in the pricing decisions of rms within the industry. Later in this chapter, we consider the implications of relaxing these assumptions. Relevant Costs Managers maximize prots by setting marginal cost equal to marginal revenue. As emphasized throughout this book, sunk costs are irrelevant for the pricing-output decisiononly incremental costs matter. Suppose that Intuit previously had spent $100,000 for developing its cyberstore and establishing and promoting its Web site. While Intuit hopes to generate enough prots to offset this initial investment, this expenditure is sunk and hence is irrelevant for its current pricing decisions. Also as discussed earlier, it is important for managers to focus on opportunity costs not accounting costs. For instance, suppose that Intuit has some Checkware packages in inventory that it had previously purchased from the manufacturer at $18. This historical cost is not relevant for the current pricing decision. Rather what is important is the current cost for replacing the inventory$10.2 Price Sensitivity Additional insights into a monopolists pricing decision can be developed using the concept of price elasticity introduced in Chapter 4. Recall that the price elasticity is a measure of price sensitivity. The higher the price elasticity, the more sensitive is the quantity demanded to price changes. With some algebra, it is easy to show that the monopolists optimal pricing policy of setting marginal revenue equal to marginal cost can be rewritten as P* 2 MC* [1 1 *] (7.2) Suppose that a package was sold for $16 that originally cost $18. The company reports an accounting loss of $2. However, if the company has to pay $10 to replace the unit in inventory, it actually has an economic prot of $6. (When the wholesale price fell from $18 to $10, Intuit lost $8 for each unit in inventory.) 194 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 210 II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Part 2 Managerial Economics Figure 7.3 Price Sensitivity and Optimal Markup The optimal price markup above marginal cost depends on the elasticity of demand. The optimal markup decreases with the elasticity of demand. The demand curve for Checkware on the left is less elastic than the demand curve for Illustrator on the right. Correspondingly, the optimal markup is higher for the demand curve on the left ($37.50 above cost versus $16.25 above cost). At the optimal price and quantities, the elasticity for the demand curve on the left is 1.27 versus 1.62 for the demand curve on the right. Price (in dollars) $ $ 85.00 85.00 P* 47.50 42.50 Demand 10.00 MR Q* 75 P* 26.25 10.00 MC 170 Demand MR Q Q* 65 Quantity of Checkware 170 Q Quantity of Illustrator Less elastic demand MC More elastic demand where P * is the prot-maximizing price, MC* marginal cost, and * is the elasticity of demand, both at the optimal output level. Recall that the elasticity of demand ranges between 0 (totally inelastic) and q (innitely elastic). Market power decreases as demand becomes more elastic. No rm should operate on the inelastic portion of its demand curve ( 1). To understand why, assume this conclusion is falsesuppose the prot-maximizing price is on the inelastic portion of the demand curve. Now consider a price above this protmaximizing price. With inelastic demand, total revenue increases with an increase in price. As price is increased, fewer units are sold and production costs fall. With an increase in revenue and a reduction in costs, prots must increase with price: Hence the maximum prot cannot lie on the inelastic portion of the demand curve. Thus, we are assured that * 1, and the optimal price is greater than marginal cost. If a rm has substantial market power, its demand will be less elastic and the markup over marginal cost will be high. In contrast, if the rm has limited market power (for example, there are many good substitutes), elasticity will be high and the markup low. In our Checkware example, the markup is $37.50 ($47.50 $10). The elasticity at the optimal price-quantity combination is 1.27 (using the technique from the appendix of Chapter 4). Figure 7.3 compares this case with that of a more elastic demand curve for another software productIllustrator :3 P 42.50 0.25Q (7.3) With this demand curve, the optimal output and price are 65 and $26.25, respectively. The elasticity at this combination is 1.62 compared to 1.27 in the rst example. Correspondingly, the markup is lower ($16.25 versus $37.50). 3 For both demand curves, the quantity sold is 170 when price is zero; however, quantity declines more rapidly with price increases in the second case. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 195 211 Estimating the Prot-Maximizing Price Economic theory suggests that managers should price so that marginal revenue equals marginal cost. One practical problem in applying this principle is that managers often do not have precise information about their demand curves and thus their marginal revenue (Chapter 4 discusses methods of estimating demand). Thus, depending on the circumstances, policies like cost-plus or markup pricing (discussed below) can serve as useful approximations or rules of thumb in the absence of better information. Linear Approximation One technique that can be employed with limited information is the linear approximation method. This method requires that the pricing manager have estimates of the current price, the current quantity sold, the quantity sold if the price is changed (say by 10 percent), and the marginal cost of production. Suppose in our example that Intuits pricing manager, Sally McGraw, currently is pricing the product at $70 and is selling 30 units per period. Sally estimates that if she lowers her price by $5 to $65, she will sell 40 units. This estimate might be based on knowledge about what competitors are charging for the same product, results from past experiences in altering prices, and so on. This information gives her two points on her demand curve. If she assumes the demand curve is linear, she can solve for it. The estimated slope of the line is (65 70) (40 30) 0.5. To solve for the intercept, she would substitute the current price and quantity into the equation for a line (P a bQ ) with slope equal to 0.5: $70 a 0.5(30) Solving for the intercept yields a 85 Hence, the estimated demand curve is P 85 0.5Q (7.4) Given this estimated demand curve and a marginal cost of $10, Sally has enough information to solve for the optimal price. Whether or not this price is close to the true optimal price depends on the accuracy of her cost and sensitivity estimates and on whether the demand curve is close to being linear. Even if the demand curve is not linear, linearity may not be a bad approximation, especially if she begins at a price that is not too far from the optimal price. In this example, Sally will lower the price from $70 to $47.50. The corresponding change in sales will provide her with additional information, which will be useful for future pricing decisions. For example, was the increase in sales close to what she predicted? If she worries that a $23.50 price reduction is too drastic given her lack of condence in her estimate of the demand curve, she might start with a smaller price reduction, followed by more reductions if additional experience warrants them. Alternatively, she might offer a $23.50 rebate on the purchase price. If the price cut turns out to be a bad idea, she can simply eliminate the rebate (rebates are discussed in greater detail below). Cost-Plus Pricing One of the more common pricing methods used by rms is cost-plus pricing. Firms that use this technique calculate average total cost and then mark up the price to yield a target rate of return. 196 212 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Managerial Economics For example, suppose that Sally expects to sell 75 (000) units and targets a 20 percent return on sales. Total costs include the $10 variable cost per unit and the $100,000 of up-front investment. To solve for the implied price, Sally begins by calculating a unit cost: Unit cost variable cost (xed costs unit sales) (7.5) so in this case, Unit cost $10 ($100 75) $11.33 The price that will yield the target rate of return is found using the following formula: Price unit cost (1 target rate of return) (7.6) In our example, this formula implies a price of $14.16 ($11.33 0.8); the return on sales is ($14.16 $11.33) $14.16 0.20. We have stressed how prot-maximizing pricing considers only incremental costs and depends on the price sensitivity of customers. Cost-plus pricing appears to ignore both of these considerations. The cost-plus price marks up average total cost and seems to ignore the demand for the productjust because Sally targets a 20 percent return on sales does not mean that customers will necessarily buy the product in the required quantities at the implied price. Firms that consistently use bad pricing policies will nd themselves earning lower prots than they could with better pricing techniques and even may go out of business. But if cost-plus pricing is so unsound, why is it so widely used in the marketplace? One explanation is that managers implicitly consider market demand in choosing the target return and the target sales volume. If Sally knows that the product faces little competition, more units can be sold and a higher return will be chosen than when facing greater competition. Conceptually, there is always some target volume and return on total cost that produces the prot-maximizing price. In our rst example, the optimal price and quantity for Checkware are $47.50 and 75,000, respectively. The corresponding unit cost is $11.33. Using a target return of 76.15 percent in Equation (7.6) will produce the prot-maximizing price of $47.50. If Sally realizes that she has substantial market power with Checkware, she should select a high target return (in this case76.15 percent). Correspondingly, for products where she has less market power, she should choose lower target returns. Thus, if Illustrator (see Fig. 7.3) has the same unit costs as Checkware, then Sally could choose a target return of 56.84 percent and the implied price would be $26.75. The idea that managers choose target returns and volumes based on market power when they do cost-plus pricing is supported empirically. Consider the price markups by a typical grocery store employing cost-plus pricing. Stable products such as bread, hamburger, milk, and soup are relatively price-sensitive and carry low margins (markups of under 10 percent above cost). Products with high margins tend to be those with relatively inelastic demand, such as spices, seasonal fresh fruit, and nonprescription drugs (markups as high as 50 percent or more). One experienced grocery store manager noted that price sensitivity is the primary consideration in setting margins.4 Cost-plus pricing is more useful when the rate of return that yields the prot-maximizing price on the product is relatively stable over the relevant range of cost variation for a given product and varies little across a related set of products. 4 J. Pappas and M. Hirschey (1990), Managerial Economics, 6th edition (Chicago: Dryden Press). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 197 The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 213 Markup Pricing Many managerial economics textbooks suggest a pricing rule-of-thumb based on Equation (7.2): P * MC* [1 1 *]. This methodtypically referred to as markup pricing consists of substituting an estimate of the current price elasticity and marginal cost into Equation (7.2) and solving for the optimal price. Equation (7.2) is a condition that holds at the optimal price-quantity combination. But price elasticity generally varies along the demand curve. If Sally currently is at a suboptimal price-quantity combination, the current elasticity will differ from that at the optimal point. Using the current price elasticity in Equation (7.2) will yield a good estimate of the optimal price only when it is close to the elasticity at the optimal price and quantity. Although price elasticity varies along a linear demand curve, it can be constant across a range of values for some nonlinear demand curves. In fact, some demand curves are isoelasticthe elasticity is constant along the entire demand curve. (An example is P a Q ; see the appendix to Chapter 4.) On the Importance of Assumptions In our Checkware example, the underlying demand curve is linear. Unsurprisingly, the linear approximation works well. In contrast, the markup pricing rule does not. If Sally begins at a price of $70 and a quantity of 30, then the demand curve (P 85 0.5Q ) implies that the elasticity is 4.66 (this can be derived using the technique from the appendix of Chapter 4). Plugging 4.66 into Equation (7.2) along with the marginal cost of $10 yields an optimal price of $12.73. But this is far from the optimal price of $47.50. This illustration highlights the importance of considering the underlying assumptions in choosing among alternative rules of thumb. It is easy to imagine other circumstances where the markup pricing rule yields a better estimate than the linear approximation. MANAGERIAL APPLICATIONS Parker Hannin Increases Prots by Adopting an Economically Sound Pricing Policy Parker Hannin Corporation is a large industrial parts maker. Throughout much of its 90-year history, the company used a simple cost-plus pricing policy for its 800,000 parts. It calculated how much it cost to make and deliver each product and added a at percentage on top, usually 35 percent. This policy was inconsistent with basic economic theory, which indicates that prices should not be based on just costs. The customers willingness to pay (price sensitivity) also should be considered. Donald Washkewicz became Parkers CEO in early 2001. He quickly decided that Parker should stop thinking like a widget maker and start thinking like a retailer, determining prices by what a customer is willing to pay rather than what a product costs to make. He reasoned that sports teams raise ticket prices if theyre playing a well-known opponent. Why shouldnt Parker do the same? In October 2001, Washkewicz unveiled a new plan, which involved creating a new senior position for pricing and bringing in outside consultants. In implementing this plan, Parkers analysts divided its many products into four basic categories (A, B, C, and D) based on their estimated demand sensitivities. They determined that about a third of Parkers products fell into niches where there was little or no competition or where Parker offered some other unique value. Based on this analysis, Parker increased it prices for many of it products, while lowering the prices of other products where they faced stiff competition. Parkers management believes that this new pricing strategy was a major reason why the companys operating income increased by $200 million between 2002 and 2006. From the end of 2002 to April 2007, Parkers stock price increased nearly 88 percent, compared to a 25 percent jump in the S&P 500. Parkers experience highlights how pricing decisions can have a major effect on prots and value. Source: Timothy Aeppel (2007), Seeking Perfect Prices, CEO Tears Up the Rules The Wall Street Journal (March 27). 198 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 214 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Managerial Economics Figure 7.4 Potential for Higher Prots b Price (in dollars) If the rm charges a single unit price, prots are maximized at a price of P* and a corresponding quantity of Q*. The corresponding prots are shown by the shaded rectangle. Consumers receive surplus equal to triangle abc, and triangle def represents unrealized gains from trade (consumers who value the product at more than cost but who do not buy the product because of the high price). The rm can earn higher prots if it can devise a more complex pricing policy that allows it to capture some of the potential prots represented by these two triangles. $ P* a c e Demand d f MR MC Q Q* Quantity of Checkware Potential for Higher Prots In the benchmark case of a single unit price, the rm captures some, but not all, of the potential gains from trade. Figure 7.4 provides an illustration. Firm prots are displayed by the shaded rectangle, and the associated consumer surplus is the triangle labeled abc. The triangle, labeled def , shows additional potential gains from trade that would accrue from selling to customers who value the product above its marginal cost, but who do not buy the product at the higher unit price P *. Below we discuss how a rm might increase prots through more complicated pricing strategies that allow it to capture some of the gains from trade displayed by these two triangles. MANAGERIAL APPLICATIONS Microsofts Market Power and Pricing In 2001 Microsoft managed to boost sales of its Windows products by 16 percent while worldwide shipments of PCs fell 4 percent. Windows and Microsoft Ofce software continue to throw off amazing amounts of cash. Analysts agree that most customers really dont have a choice of desktop software. While analysts rarely describe Microsoft as a monopoly, they frequently refer to Microsofts unique market position or huge installed base. Microsoft has found several ways to raise prices. Microsoft stopped requiring companies to buy individual copies of software for each employee. It now offers a multiyear site license where one copy of the software can be downloaded to individual PCs. With multiyear licenses, organizations pay an annual fee to Microsoft to use its software rather than buying individual copies. Alvin Parks, an analyst with the research rm Gartner Inc., argued that this new pricing policy means a price increase for everybody, eventually. Walter Casey, an analyst at Banc One, says, Theyre trying to force people to upgrade, [though] maybe theyd say encourage. Source: R. Buckman (2002), Microsoft Sprints On as Tech Sector Plods, The Wall Street Journal (August 8), C1C3. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 199 215 ANALYZING MANAGERIAL DECISIONS: Prot Potential for a Microbrewery You operate a small microbrewery in Germany. The demand curve for cases of your beer is P 50 .002Q. Your marginal cost for producing beer is 10 euros per case. Currently you are charging all customers the same price for a case of your beer. 2. Calculate the surplus that goes to consumers. 3. Describe how you might capture additional prots using a more sophisticated pricing policy that does not involve capturing more of the consumer surplus. Explain. 1. What are the optimal price, quantity, and prots under this pricing policy? Homogeneous Consumer Demands Potential customers for a product might have quite similar individual demand curves, or they might vary widely in their demands for the product. For instance, in the Checkware example, all potential customers might be willing to pay the same maximum price for the product. Alternatively, some customers might be willing to pay a high price for the product, whereas others might be willing to buy the product only at a quite low price. We begin our discussion of more complicated pricing strategies by examining the case of homogeneous demandsin the limit they are identical. These same techniques are appropriate where the rm is dealing with a given individual customer and wants to extract maximum prot. Subsequently, we examine the case of heterogeneous demands. Block Pricing The overall demand curve for a product is the sum of individual customer demands at each price. For many products, individual customers are likely to consider purchasing multiple units (consider fruit, clothes, and so on). In the benchmark case, individuals pay the same price per unit independent of the number of units purchased. More generally, the company might charge a given customer different prices per unit, depending on the number of units purchased. Prots sometimes can be increased (even if all customers have similar demand curves) by basing the price per unit on the number of units purchased.5 Individuals have downward-sloping demand curves for products. This implies that the marginal value that a customer places on each additional unit declines as the quantity purchased increases. Conceptually, the rm could capture all the potential gains from trade with a customer by charging a price equal to the marginal value of each unit. Thus it might charge $100 for the rst unit, $99 for the second unit, $98 for the third unit, and so on (where the prices represent the customers marginal value of each additional unit). Practical considerations often will preclude such a pricing policy. However, the policy might be approximated by charging a high price for the rst purchase block (for example, up to 5 units) and declining prices for subsequent blocks. For example, the 5 Note that in the following discussion we assume that a customer cannot buy a large quantity and then resell it to other customers (undercutting the prices that the company charges to other customers for small quantities). The importance of this assumption is discussed in more detail below. 200 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 216 Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Block Pricing at Hickey-Freeman Hickey-Freeman is a prominent producer of high-quality business suits and mens clothing. Each winter and fall it holds a sale at its factory stores in New York, Indiana, and Maine. During a recent spring sale, Hickey-Freeman priced its high-end suits in the following manner. One suit could be purchased at about $700. A second suit could be purchased for around $600. Greater discounts were given if the customer purchased more than two suits. Many customers require at least one good business suit for work. Thus they are likely to be willing to pay a high price for the rst suit. While many customers might prefer to own several suits, their willingness to pay is likely to decline signicantly with the number of suits purchased. Hickey-Freemans pricing policy generates more prots than if it sold all suits at 1 unit price. For instance, if it priced all suits at $700 it would fail to sell additional suits to customers who value the extra suits more than the marginal cost of production (but less than $700). Alternatively, if the company priced all suits at $600, it would miss out on the opportunity to charge higher prices for some of the suits. company might charge the customer $98 per unit up to the rst ve, $95 per unit for the next ve, and so on. As discussed below, block pricing not only is used to extract additional prots from a homogeneous customer base, but also is used in certain circumstances to increase prots when customer demands are more heterogeneous.6 Another type of block pricing can occur in the choice of package size. Suppose a typical customer values a rst T-shirt at $16 and a second at $10. If vendors sell the shirts individually, they have to charge $10 or less to get a customer to buy two. If they offer them in packs of two at a price of $26, customers will be willing to purchase the package and pay an effective price of $13 per shirt. Two-Part Tariffs With a two-part tariff, the customer pays an up-front fee for the right to buy the product and then pays additional fees for each unit of the product consumed. A classic example is an amusement park, where a customer pays a fee to get in and then so much per ride (Disneyland used to price in this manner).7 Golf and tennis clubs, computer information services, and telephone service providers are examples of companies that frequently use two-part tariffs. The benchmark case of charging one price to all customers is a special case of a twopart tariff. The entry fee is zero and the additional units can be purchased at the quoted price. Making more general use of two-part pricing (charging a positive entry fee and a subsequent usage price) can sometimes increase prots substantially relative to the benchmark case. As we discuss below, this is most likely to be true when potential consumers have relatively homogeneous demands for the product. For illustration, consider an example where all consumers have identical demands for the product, P 10 Q. Figure 7.5 displays a demand curve for a representative consumer. The marginal cost of producing the product is $1. The potential gains from trade 6 For expositional simplicity, we have assumed that marginal cost is constant. If production, packaging, or distribution costs decline with volume, price discounts would be offered, even without market power. With market power, the price reductions will exceed the cost reduction as volume increases. 7 W. Oi (1971), A Disneyland Dilemma: Two-Part Tariffs for a Mickey Mouse Monopoly, Quarterly Journal of Economics (February), 7796. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 7 Two-Part Tariff In this example, all potential customers had identical demands. The gure displays a demand curve for a representative consumer, P 10 Q. The managerial cost of producing the product is $1. The potential consumer surplus that the rm could capture is $40.50, as shown by the shaded triangle. Maximum prots can be extracted by charging an up-front fee equal to all the consumer surplus (or slightly less) and then charging a price equal to marginal cost, $1. Under such a scheme, the consumer purchases 9 units. Pricing with Market Power 217 $ 10 Price (in dollars) Figure 7.5 201 The McGrawHill Companies, 2009 7. Pricing with Market Power Demand P* 1 MC MR Q* 9 Q Quantity that the rm could capture are shown by the shaded triangle and are equal to $40.50 (0.5 9 9). Maximum prots can be extracted by charging an up-front fee equal to all the gains from trade (or slightly less) and then charging a price equal to marginal cost, $1. With this pricing strategy, the consumer purchases 9 units. In contrast, if the rm charged a single unit price, the best it could do would be a prot of $20.25 (by setting marginal revenue equal to marginal cost). In this case, the price is $5.50 per unit and the consumer purchases 4.5 units. Two-part tariffs also can be used protably when customers demands are not identical, but this pricing strategy tends to be less effective the more consumers vary in their demands for the product. From a practical or legal standpoint, a company might have to offer the same two-part tariff to all potential customers. Charging a high entry fee allows the rm to extract more surplus from customers who have high demands for the product, but potential customers with lower demands will choose not to purchase. When customers demands vary widely, it often is best to charge a low entry fee (possibly zero) and then charge a price above marginal cost for use. In this case, given the costs of implementing a two-part tariff (devising the pricing strategy, collecting the fees, and so on), the rm frequently is better off just charging a single price. However, in the next section, we discuss other pricing policies that can increase prots facing more heterogeneous demand. Price DiscriminationHeterogeneous Consumer Demands Potential customers often vary materially in their willingness to pay for a product. In our benchmark case, the rm charges the same price to all potential customers. With a heterogeneous customer base, the company can make higher prots if it is able to charge higher prices to those customers who are willing to pay more for the product. Price discrimination occurs whenever a rm charges differential prices across customers that are not related to differences in production and distribution costs. With price discrimination, the markup or prot margin realized varies across customers. Two conditions are necessary for protable price discrimination. First, different price elasticities of demand must exist among potential customers for the product (demand must be 202 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 218 Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Two-Part Pricing for Capital Goods Frequently a consumer buys a capital good from a rm and then purchases another good to obtain the services of the capital good. For example, Gillette sells razors and razor blades, Hidden Fence installs pet containment systems that require special batteries sold only by Hidden Fence. This situation is like a two-part tariff. With homogeneous consumers, prots are maximized by setting the price of the capital good to extract all consumer surplus and pricing the consumable at marginal cost. With heterogeneous consumers it is typically optimal to charge a lower price for the capital good and a price above marginal cost for the consumable. heterogeneous). Otherwise, there is no point in segmenting the market. When different price elasticities do exist, it is generally more protable to charge higher prices to those customers who are less sensitive to price. Second, the rm must be able to identify submarkets and restrict transfers among consumers across different submarkets. Otherwise, any attempt to charge differential prices to customers will be undercut by resale across the submarkets. One group of consumers can buy at the low price, then resell to the other groups at a price below the rms prices to these groups. Sometimes managers have quite good information about individuals product demands (which specic customers are willing to pay more for the product). For instance, if Andrew Leone has sold automobiles to the same customers on repeated occasions, he is likely to have relatively good information about each customers price sensitivity. In cases, other managers have poor information about individual product demands. For example, early in his career Andy had less experience and accumulated information to differentiate among customers who came to the dealership. But he still might be able to engage in certain kinds of price discrimination with information only about the range or distribution of customer demands. We begin our examination of price discrimination by considering the case where the manager has good information about individual demands. We then consider the case where the manager has information only about the distribution of demands. MANAGERIAL APPLICATIONS As Cigarette Prices Soar, a Gray Market Booms To price-discriminate successfully, a rm must not only be able to identify submarkets with different price elasticities, but also be able to restrict trade among these submarkets. Cigarette companies have historically charged much higher prices in the United States than in foreign markets for their premium brands. Due to the fact that United States cigarette prices increased by over 50 percent in the late 1990s, madefor-export cigarettes have appeared for sale in United States stores and on the Internet. These cigarettes were returned for sale in the United States through a complex and shadowy distribution network. For example, in 1999, Cigarettes Cheaper, a California-based discounter, sold 45,000 cartons of made-for-export cigarettes produced by R.J. Reynolds. A carton of diverted Winstons sold for $24.89 compared to $30.39 for regular Winstons. In addition, the Internet contains Web sites offering cheap export cigarettes. In Brasschaat, Belgium, merchant Tehoe Rooijakkers indicated that hits have increased to 150 a day on his Web site advertising low-cost Marlboros, Winstons, and Camels. These activities reduce the ability of the tobacco companies to charge premium prices for their brand-name cigarettes in the United States. Source: S. Hwang (1999), As Cigarette Prices Soar, a Gray Market Booms, The Wall Street Journal ( January 28), B1. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 203 The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 219 MANAGERIAL APPLICATIONS Tuition Pricing Firms engaging in personalized pricing strive to extract the maximum willingness to pay from each customer. While colleges and universities do not engage in perfect personalized pricing, they effectively charge different prices to students for tuition through the use of nancial aid packages. Stated tuition is the maximum price that any student is charged. Low-income students, who are likely to be relatively price-sensitive, typically are offered more nancial aid than high-income students. In addition, top students who are likely to have numerous scholarship offers, and thus more options, are offered signicant discounts. An average student from a high-income family typically pays much higher effective tuition than other students. Exploiting Information about Individual Demands Personalized Pricing8 Suppose there are many potential customers. Each customer places a value on the product that signies the maximum that that individual would pay for the producttheir reservation price. For simplicity, suppose each customer purchases at most 1 unit. Personalized pricing ( rst-degree price discrimination) extracts the maximum amount each customer is willing to pay for the product. Each consumer is charged a price that makes that customer indifferent between purchasing and not purchasing the product. In this case, the rm extracts all the potential gains from trade. This extreme form of price discrimination is rare and typically is possible only when the number of customers is extremely small and resale is impossible. With personalized pricing, the rm sells to all customers who are willing to pay more than the marginal cost of production. All gains from trade are exhausted, and the outcome is efcient. All the gains from trade, however, go to the rm. While perfect personalized pricing is rare, new technologies are making it easier for companies to customize quoted prices. For example, companies selling over the Internet can vary quoted prices based on past buying histories, demographic information obtained through electronic registration, clickstreams, and so on. Similarly, companies that sell through catalogs canand often doinclude personalized inserts, where the quoted prices vary depending on the customers buying history and personal characteristics (for example, zip code). This type of personalized pricing was more difcult under older printing technologies and before the existence of computerized databases that store customer information.9 Group Pricing Managers sometimes can gauge an individuals price sensitivity by observing a characteristic of the individual such as age, income, or dress. In these cases, the manager can have a fairly good idea of a specic individuals demand for the product, even if the manager never has interacted with the customer in the past. 8 Economists often categorize price discrimination as rst-, second-, or third-degree. These terms were originated by A.C. Pigou (1950), The Economics of Welfare (Macmillan: London). Unfortunately, they are not very descriptive. Following C. Shapiro and H. Varian (1999), Information Rules: A Strategic Guide to the Network Economy (Harvard Business School Press: Boston), we use more descriptive terms like personalized pricing, menu pricing, and group pricing. 9 Note, however, that the Internet also lowers information costs and that this makes market segmentation (a necessary aspect of effective price discrimination) more difcult (recall the earlier Managerial Application on made-for-export cigarettes). 204 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 220 Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Virtual Vineyards Virtual Vineyards offers premium wines and specialty foodssuch as El Serpis Anchovy Stuffed Olives and Foxs Fine Foods Killer Corn Relishover the Internet. The company also offers advice, monthly wine programs, and a variety of other services at its Web site. Virtual Vineyards tracks the clickstream of each user and instantaneously makes special offers based on the behavior. In a similar vein, Amazon.com tracks the purchases of each consumer and recommends additional related books the next time the user logs on. The Internet has made possible many marketing opportunities not available through other media. Source: C. Shapiro and H. Varian (1999), Information Rules: A Strategic Guide to the Network Economy (Harvard Business School Press: Boston). Group pricing (third-degree price discrimination) results when a rm separates its customers into several groups and sets a different price for each group. For example, utility companies charge different rates to individual versus commercial users, computer companies give educational discounts, and airlines charge different rates based on the amount of notice given for the reservation. Beyond.com charged government agencies and large companies lower prices than other customers. As illustrated in the following example, a rm that can segment its market maximizes prots by setting marginal revenue equal to marginal cost for each market segment. Firms use a variety of characteristics to divide customers into groups. Three prominent examples are age, time of purchase, and income. For instance, movie theaters frequently give discounts to senior citizens and students, price lower for matinees than for evening performances, and vary prices across locations depending on the average income in the area. The objective is to charge a higher price to the groups who are less price-sensitive. MANAGERIAL APPLICATIONS Pricing of Books Firms divide customers into groups based on various characteristics. Different prices are charged to each group, depending on their elasticity of demand. One characteristic used by book publishers to segment the market is time. When a new book comes to market, it usually is offered only in a hardcover edition for a relatively high price. Subsequently, it is offered in paperback at a substantially reduced price. Individuals who have a high demand for the book (and thus a low price elasticity) do not want to wait for the paperback edition and thus pay a high price for the book. Those with lower demands wait for the cheaper edition. After the paperback edition comes to market, the publisher generally will continue to offer the hardcover edition. Hardcover books are likely to make better gifts than paperbacks. Also customers are likely to prefer hardcovers for their libraries. Thus there continues to be a market for both types of books. Continuing to offer multiple versions is an example of menu pricing (discussed below). In late 2005, Amazon.com offered several versions of Harry Potter and the Order of the Phoenix by J. K. RowlingBook 5 in the popular series. The hardcover version sold for $19.20, whereas the paperback sold for $9.99. In addition, there was a hardcover, large-print edition for $29.95, an audio CD for $47.25, and an audiocassette for $29.70. The hardcover and paperback editions also could be purchased as part of specially boxed ve book sets. Book 6, Harry Potter and the HalfBlood Prince, is available only in hardcover. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 205 221 Consider the Snowsh Ski Resort, which can separate its demand into local skiers and out-of-town skiers. The marginal cost of servicing a skier of either type is $10. Suppose the resort faces the following demand curves: Out of town: Local: Q0 Q1 500 500 10P 20P (7.7) (7.8) Total demand at any one price is the sum of the demands for the two types of consumers:10 Q 1,000 30P (7.9) If the company sells all tickets at one price, prot maximization will occur at:11 P* $21.66; Q * 350; Q* 0 283; Q* 1 67; Prot $4,081 The company can make higher prots by charging different prices to the two sets of skiers. The optimal prices are found by setting the marginal revenue equal to the marginal cost in each of the two market segments. Under this pricing policy, the following prices, quantities, and prots are observed: P* 0 $30; Q * 0 200; P * 1 $17.50; Q* 1 150; Prot $5,125 where P 0 and P 1 prices charged to out-of-town and local skiers, respectively. The re* * sort charges higher prices to the out-of-town skiers, who are less sensitive to ticket prices than local skiers. Figure 7.6 (page 222) displays the optimal pricing policy for each market segment. Snowsh treats the two markets as separate and charges the optimal monopolistic price to each segment. Consistent with Equation (7.2), the optimal markup is lower in the more elastic local market. Using the point-elasticity formula developed in the appendix to Chapter 4, it can be shown that at the optimal prices, the elasticities for the local and out-of-town skier markets are 2.33 and 1.5, respectively. The respective markups above marginal cost are $7.50 and $20 (given the prices of $17.50 and $30). There are a number of methods that Snowsh might use to charge the two groups different prices. Discount coupons might be sold at supermarkets away from major resort hotels. Presumably, most of the sales at these supermarkets will be to local customers. Alternatively, discount books of tickets (nontransferable) could be sold locally prior to the start of ski season. Ski resorts use both techniques. These policies are more protable the more difcult it is for out-of-town skiers to buy the tickets at prices less than $30. Using Information about the Distribution of Demands In some settings, the manager does not have enough information to divide customers into meaningful groups. For instance, even if a retailer knows that low-income individuals are more price-sensitive than high-income individuals, the retailer may not be able 10 This demand curve assumes that price is lower than or equal to $25. At higher prices, the local skiers purchase no tickets and the total demand curve is simply the demand curve for out-of-town skiers (Q 500 10P ). 11 The reader should know by now that the solution to this problem is found by setting marginal revenue equal to marginal cost and solving for Q *. Price can then be found from the equation for the demand curve. For instance, the total demand curve can be obtained by rearranging Equation (7.9): P 33.33 0.033Q. When the tickets are sold at one price to all consumers, the marginal revenue is MR 33.33 0.067Q. Since marginal cost is $10, the optimal quantity is 350; price is $21.66. The optimal prices and quantities for the individual market segments are found by completing similar calculations using Equations (7.7) and (7.8). 206 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 222 Part 2 Figure 7.6 II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Managerial Economics Optimal Pricing at Snowsh Ski Resort Snowsh can segment its customers into two market segments, out-of-town skiers and local skiers. The marginal cost of serving either type of skier is $10. The optimal pricing policy is to set the monopoly price in each market segment. The markup is higher for out-of-town skiers because they have less elastic demands than local skiers. At the optimal prices ($30 and $17.50), the demand elasticities are 1.50 for out-of-town skiers and 2.33 for local skiers. Price (in dollars) $ $ 50.00 50.00 * = 1.50 P* 30.00 * = 2.33 25.00 P* MC 10.00 MR Q* 200 Quantity of passes for out-of-town skiers 17.50 10.00 MC MR Q Q* Q 150 Quantity of passes for local skiers to gauge the incomes of customers when they come to the store.12 Nonetheless, the manager might have enough information about the range or distribution of individual demands to engage in protable price discrimination. In this section we discuss two prominent methods that can be used in this setting. Both rely on the principle of self-selection. Consumers are provided with options. They then reveal information about their individual price sensitivities by their choices. Menu Pricing With menu pricing (second-degree price discrimination), all potential customers are offered the same menu of purchase options. The classic example involves block pricing, where the price per unit depends on the quantity purchased.13 For instance, cellular phone companies typically give customers a choice among several rate plans, where the price per minute varies with the minimum number of minutes per month specied across each plan. Customers use their private information about likely usage to select the best rate plan for themselves. By carefully constructing the menu of options, the company makes more prots than if it simply offered the product at one price to all potential customers (for example, offering phone service at $.10 per minute, independent of volume). If such quantity discounts are based solely on costs, then there is no price discrimination. However, large-quantity users have incentives to search and thus are likely to be more price-sensitive than low-quantity users, and thus block pricing allows different rates to be charged to the two groups even if per-unit costs are similar. Public utilities frequently price in this manner. 12 Also it may be neither practical nor legal to charge customers different prices for the same products even if the retailer knows each customers income. 13 Recall that block pricing can be used to increase prot either by extracting more prots from a given homogeneous customer population or by increasing prots through charging different prices to high- versus lowvolume customers. In this section, we discuss the second use. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 207 The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 223 Sophisticated User Starter edition Premier edition Unsophisticated User $ 45 $100 $30 $30 Table 7.1 Example of Menu Pricing In this example, there are two versions of TurboTax and two types of users. The table shows the maximum price that customers in each group are willing to pay for the product (reservation prices). The marginal cost of producing both versions is zero. With menu pricing, the rm will price the Starter version at $30 and the Premier version at $85 (or just below). Consumers acting to maximize consumer surplus will self-select. Sophisticated users will buy the Premier version (surplus of $15) and the unsophisticated users will buy the Starter version (surplus of zero). Unlike personalized pricing, the rm does not obtain all the gains from trade. The rm would like to charge more for the Premier version. However, if it does, the sophisticated user will purchase the Starter version. A related pricing strategy is to offer potential customers a menu of price-quality combinations. For instance, TurboTax markets both a Premier and Starter version of its software. The Premier version contains additional features that are likely to appeal to sophisticated users. The marginal costs of producing and distributing both versions are virtually equal. However, the company marks up the Premier version more because the typical customer choosing this version is likely to be less price-sensitive than the typical customer choosing the Starter version. Table 7.1 presents a numerical example of menu pricing. Here the market for TurboTax software is divided into two types of users, sophisticated and unsophisticated. The reservation prices (maximum willingness to pay) for the Premier and Starter versions are given for both types of users. The marginal cost of producing the software is assumed to be zero. Ideally, the company would like to identify the two types of users prior to purchase. Assuming it could prevent reselling the software among consumers, the company would maximize prots by charging $100 to the sophisticated users, who would purchase the Premier version, and $30 to the unsophisticated users, who would purchase the Starter version. This pricing strategy is equivalent to personalized pricing. Since the company cannot identify the type of user prior to purchase, it prices the Starter version at $30 and the Premier version at just below $85 (for example, $84.99). The customers, who know their own type, choose the quality-price combination that maximizes their individual consumer surplus. Unsophisticated buyers purchase the Starter version (which yields no consumer surplus). Sophisticated users buy the Premier version and gain a surplus of just over $15 (the surplus they would enjoy if they purchased the Starter version). The potential for the sophisticated buyer to purchase the lower-quality product limits the price that can be charged for the Premier version. If the Premier version were priced above $85, sophisticated users would obtain greater consumer surplus by buying the Starter version at $30, and no Premier versions would be sold. The end result is that the sophisticated user gains some consumer surplus and the company makes less prot than it would if it could engage in personalized pricing. Nonetheless, it does better than if it offered only one version at a single price. For instance, the company could offer just the Premier version at a price of $100 and sell only 208 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 224 Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Cell Phone Pricing You are a pricing manager for a cell phone company. You have two types of customers with different demand curves for your service. The demand curves for an individual customer from each group for hours of talk time per month are: Type A customer: P Type B customer: P 10 10 2Q Q Your marginal cost for providing hours of phone service is zero (all your costs are xed). There are 1,000 customers of each type. You know the demand curves for the two types of customers. However, it is impossible for you to identify when a person purchases a plan whether the customer is from one group or the other. much as possible from the Type B customers given that they have the option to purchase your rst plan (it might help to graph the problem). Each of the plans on the menu must offer a maximum number of hours of talk time per month for a xed monthly fee. (Hint: You can extract all the consumer surplus from the Type A customers using a two-part tariff. Such a plan can be expressed in terms of offering a xed number of minutes at a xed price.) 2. What are the total prots from offering the two plans? 3. What happens if you increase the monthly fee for the plan designed for the Type B customers? Explain. 1. Design a menu plan that extracts all of the consumer surplus from the Type A customers and as to sophisticated users or offer the product at $20 and sell to both types of users. But it generally is more protable to offer both quality-price combinations.14 Coupons and Rebates Firms frequently use coupons in product pricing. For instance, most Sunday papers contain numerous coupons offering discounts to customers who use the coupons before their expiration dates. Coupons also are distributed through direct mailings, product packages, and magazines. Firms also frequently offer rebates, where the customer using the rebate is refunded some portion of the purchase price. For instance, automobile manufacturers often offer signicant rebates (for example, $1,000) to customers who purchase cars during the rebate period. Software manufacturers and retailers (for example, Microsoft and Intuit) frequently offer rebates for their products. Rebate offers often are attached to products sold at grocery stores. Coupons and rebates give price discounts to customers. Price discounts in turn might be given to attract new customers (new rst-time users or brand switchers) or to increase sales among current customers. Price discrimination is one primary reason why some rms use coupons and rebates to make price discounts rather than simply lowering the price. Coupons have to be clipped and brought to the store, whereas rebates often require customers to complete and mail rebate forms to the manufacturer. Many customers do not use available coupons or rebates to purchase products because the money saved 14 If the number of sophisticated users is large relative to the number of unsophisticated users, it might be better to price the deluxe version at $100 and sell only to sophisticated users. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 209 225 through using the coupon or rebate is less than the value of the time it takes to search for and redeem the offer. The typical coupon/rebate user is likely to have a relatively low opportunity cost of time (for example, the user might not work outside the home or have a low income); such consumers are likely to be more price-sensitive. Thus, issuing coupons and rebates is one way of lowering effective prices to potential customers with more elastic demands.15 In 2008 Intuit offered a $10 rebate to existing customers for the Windows version of Quicken Deluxe (along with a $20 price reduction). Customers who applied for the rebate paid a net price of $29.99, whereas a customer failing to use the rebate paid a price of $39.99. Customers failing to redeem a $10 rebate presumably place a high value on their time and are not likely to be very price-sensitive.16 Consistent with the principles of price discrimination, they pay a higher price for the product than the more pricesensitive rebate user. Coupon/rebate programs can be expensive. For instance, there are the costs of designing the promotion, printing and circulating the offer, and handling the redemptions. These costs can be signicant and should be compared with the benets in choosing whether to offer the program. We have indicated that one of the potential benets of these programs is increased prot through price discrimination. Another potential benet is the increase in future prots from new customers who start using the product due to the sales promotionif they like the product, they might purchase it in the future even if a coupon is not offered.17 Coupons (for example, in newspapers and magazines) sometimes can be viewed as a form of advertising that lowers information costs about the product. Interestingly, some retail companies have offered rebates for the Windows version of a program while offering no rebate for the Mac version. Mac users typically have fewer software products to choose among than Windows users. Also, local software retailers are likely to carry a smaller inventory of Mac products than Windows products. The lack of available substitutes implies that the typical Mac customer is likely to be less pricesensitive than the typical Windows customer. Beyond.com thus offered coupons to Windows users with their higher price sensitivity, and no coupons to the less price-sensitive Mac users. Finally, if Macs are less likely to exist in the future than Windows-based machines, the benets of attracting new customers for future sales are lower for Macs. Bundling To this point, we have considered the case where the rm sells a single product. This section extends the analysis by providing an introduction to the case of multiple products.18 15 Price discrimination is unlikely to explain the use of large rebates or coupons. For instance, if an automobile company offers customers a $2,000 rebate, it reasonably can forecast that all customers will take advantage of this offer. Thus rebates are essentially equivalent to giving all customers a price reduction, for example, by having a sale. Rebates, however, often are offered by the manufacturer who does not have direct control over the retail price. For a more extensive discussion of coupons, rebates, and other sales promotions see R. Blattberg and S. Neslin (1990), Sales Promotion: Concepts, Methods, and Strategies (Prentice Hall: Englewood Cliffs, NJ). 16 Alternatively, customers might be purchasing the product on the behalf of their companies. In this case, since the company is paying the bill, the customer might be less price-sensitive. 17 Sometimes this benet can be obtained simply by lowering the purchase price (for example, by having a sale). 18 The topic of pricing multiple products can be relatively complex. Our intent here is to provide a brief introduction to the idea of product bundling. For a more detailed analysis, see J. Long (1984), Comments on Gaussian Demand and Commodity Bundling, Journal of Business 57, S235S246. 210 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 226 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Managerial Economics Microsoft Streets Consumer Professional Microsoft Trip Planner Bundle $ 70 $110 $80 $60 $150 $170 Table 7.2 Product Bundling In this example, there are two types of customers, professionals and ordinary consumers. The table shows the reservation prices for Microsoft Streets and Microsoft Trip Planner. The reservation price for the bundle for any consumer is the sum of the reservation prices for the individual products. If the rm wants to sell both products to both types of customers, it must price the products at $70 and $60, respectively (the minimum reservation prices for each product). It collects $130 from each customer. It can do better by selling the products in a bundle for a price of $150 (the minimum reservation price for the bundle). Bundling increases prots in this example because the two groups have opposite relative valuations of the two products. Professionals are willing to pay more for Streets than consumers, while the opposite holds for Trip Planner. Thus the minimum reservation price for the bundle is greater than the sum of the minimum reservation prices for the separate products. Companies frequently bundle products for sale. For example, in 2008 Microsoft offered Microsoft Streets and Trips as a bundle for $149.95. This bundle combined leading street-mapping and trip-planning technologies into a single package. Other examples include retailers bundling free parking with a purchase at their store, newspapers selling advertising in both morning and afternoon editions at one price, season tickets for sporting events, and restaurants offering xed-price complete dinners. One reason for bundling products is to extract additional prots from a customer base with heterogeneous product demands.19 For example, Table 7.2 presents two types of potential customers: consumers and professional users. The gure displays the maximum prices (reservation prices) that the individual customers within each group are willing to pay for Microsoft Streets and Microsoft Trip Planner. The marginal cost of producing either product is assumed to be zero (to simplify the discussion). Potential customers purchase a product only if their surplus is nonnegative. Why would the company bundle the products, forcing customers to purchase both in order to obtain either? If the rm did not bundle, the most it could charge for Streets would be $70 and for Planner $60 if it wanted to sell to both groups of customers. Using this policy, it would collect $130 from each customer. The company, however, MANAGERIAL APPLICATIONS Bundling Videogames Microsoft is in a battle with Sony Corp. over videogames. Microsoft is bundling a limited edition Xbox packaged with Microsofts Halo (a shooting game) for $169. Halo is one of the most popular titles exclusive to the Xbox, and Microsoft expects that this offer will spur sales of the game machine itself. Source: R. Guth (2004), Game Gambit: Microsoft to Cut Xbox Price, The Wall Street Journal (March 19), B1. 19 Products also might be bundled to reduce packaging costs or to reduce the costs to customers who want to buy the products together. For example, to the extent that most users want to utilize features of both street mapping and trip planning, the technologies might be combined to lower the costs to the customer of obtaining and integrating the two products. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 211 The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 227 MANAGERIAL APPLICATIONS Harry Potter: An Example of Price Discrimination Sears.com offers a set of Harry Potter DVDs for around $55 plus shippingapproximately $10. A customer could obtain a rebate form online and mail it in for a rebate of the shipping costs. What was going on? Why not simply sell the DVDs for $55? Wouldnt costs be avoided? For example, it is expensive to issue and mail checks and print posters and coupons. To illustrate the potential benets of the policy, suppose that if the DVDs were sold at the same price to all customers, Sears.com would have priced them at $65. With this policy, Sears.com would lose the potential prot from selling to consumers who are willing to pay a price above Sears production cost but less than $65. Sears apparent objective was to nd a way to sell to these consumers at a lower price without having to lower the price to other consumers. One way to accomplish this objective was through rebates. Presumably, those willing to purchase at $65 had higher opportunity costs for their time (on average). This made them less likely to ll out and mail in the coupons. Customers not using the rebates paid a price of $65, whereas the customers using the rebate coupons paid $55. There are a variety of issues to consider in deciding on such a rebate program. First, there is a trade-off between the costs of administering the program and the benets of additional sales. Also, there is the loss of $65 sales to customers who would have purchased at $65 but now use the rebate coupons. can bundle the products and sell to all customers at a price of $150 (the minimum reservation price for the bundle). Although both groups in this example value Streets higher than Planner, the two groups have opposite relative valuations of the two products. Professional users are willing to pay more for Streets than consumers, whereas consumers are willing to pay more for Planner than professional users. This feature implies that the minimum bundle value is greater than the sum of the minimum reservation prices for each product. Thus the rm makes more money selling the bundle than selling the products separately. If one group valued both products higher than the other group, there would be no gains from bundling (the minimum bundle value would be the same as the sum of the minimum individual reservation prices). Typically, products are not just sold in bundles. Often rms use a tactic of optional (mixed ) bundling, where the products can be purchased separately or in a bundle at a price below the sum of the individual prices. Optional bundling can be more protable than pure bundling when some customers value one product highly, but value the other product below the marginal cost of production. For these customers, the extra revenue the rm would earn from selling the bundle would be lower than the extra cost of producing it. Other Concerns Multiperiod Considerations Future Demand In 2008 Intuit offered a free edition of TurboTax. If the objective is to maximize singleperiod prots for a single product (as in the benchmark case), it is never optimal to set a price of zero.20 Managers, however, are concerned with sales not only in the current period but in future periods as well. Giving the product away can attract new customers; 20 Firms that sell multiple products can have incentives to price selected individual products below marginal cost. For example, McDonalds sometimes has promotions that sell hamburgers at below cost. The intent is to attract customers who buy other products, such as fries, at prices substantially above cost. 212 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 228 Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics it lowers the full cost of consuming the good below that of competing products. If they like the product, they will purchase the product in the future at positive prices. This is a major reason why rms provide free samples of their products. Offering a product at a price below marginal cost is a more effective pricing strategy if information costs are higher. If a customer takes the time to learn to use a new software product (such as TurboTax), the customer often will want to continue to use the program in the future, rather than take the time to learn to use a new product. Thus at similar prices, the customer is more likely to buy upgrades and future editions of the current software rather than switch to a competing product. This lock-in effect is a reason why some software rms give away their products (or charge low prices) in the early stages of development. Firms sometimes charge lower prices than the optimal single-period price because they value maintaining customer goodwill. For example, a severe ice storm in March 1991 produced a major power outage in upstate New York. Electric generators were in high demand and could have been sold to customers at extremely high prices. But local stores did not raise the price of generators substantially. One concern apparently was that they would be seen as taking advantage of customersthat such a tactic would undermine the rms reputation and reduce future demand. Drug companies face similar concerns when they price new drugs. These companies have an incentive to charge the prot-maximizing price for their products (to help reimburse them for their development and other costs). However, in setting prices these companies have to consider the reactions of public-interest groups and government regulators. Future Costs As discussed in Chapter 5, for some rms the long-run average cost of producing a given level of output declines as the rm gains experience in producing the product. For instance, employees can gain important information on how to improve the production process as they gain more experience. When these learning effects are important, it can be optimal for the rm to produce a high volume of the product initially to gain experience and thus a cost advantage over competing rms in subsequent periods. This high output is correspondingly sold at lower prices than if the rm produced the lower volume associated with optimizing single-period prots. Many cost advantages, however, are short-lived because competing rms often can copy innovations. Thus managers should consider carefully whether it is appropriate to adopt such a high-volume, low-price strategy. Chapter 8 discusses this issue in more detail. HISTORICAL APPLICATIONS Early Use of the Free Sample In 1870, after over 10 years of experimentation, Robert Chesebrough opened a factory in Brooklyn to produce petroleum jelly. But he initially sold not one bottle. He tried giving samples to doctors and druggists, but they failed to reorder. Finally, he decided he had to create a market for his salve, so he loaded a wagon full of 1-ounce bottles and drove around the state giving away samples. When people ran out, they went to their druggist, who then began to place orders. Chesebrough soon had a dozen wagons canvassing the countryside. As demand continued to expand, Vasoline made the persistent Chesebrough an extremely wealthy man. Source: I. Flatow (1992), They All Laughed (Harper Perennial, New York). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 213 229 MANAGERIAL APPLICATIONS Apple Apologizes for Its Pricing of iPhones Apple Computer provides a good example of how concerns about customer relations and long-term reputation can affect pricing decisions. In June 2007, Apple and AT&T launched the iPhone at a price of $599. Many iPod acionados waited in line for hours to obtain the new, much-hyped phone. Within three months, Apple reduced the price by $200 to $399. Customers who had purchased the phone at $599 were outraged. Many of these customers had been long-term users of Macs and iPods. They felt betrayed when newer, less loyal customers were able to buy the iPhone at a much lower price by waiting only a few months. Some of the iPhone purchasers said the timing of the price cut would discourage them from buying Apple products in the future. This is like a slap in the face to early adopters, said John Keck, an executive at an advertising agency in Detroit. Apple received a ood of negative e-mails. The customers strong reactions were featured prominently in media reports, including the major television networks and newspapers. In response to this criticism, Apples CEO Steve Jobs issued a rare public apology and offered a $100 credit to all customers who had paid the original price. In a letter posted on Apples Web site Jobs wrote, Our early customers trusted us, and we must live up to that trust with our actions in moments like this. If Apple expected to go out of business in the near future, there would be little economic reason to give the credit to people who had purchased products in the past. Jobs, however, presumably expects Apple to be around for the long term and that the company must maintain a reputation for being fair and honorable. He is concerned not only about this years prots, but about future years prots as well. Source: Nick Wingeld (2007), Steve Jobs Offers Rare Apology, Credit for iPhone, The Wall Street Journal Online (September 7), A1. Storable Products In a number of our examples, we have discussed products that can be stored. This means that sales of the product and consumption of the product are not the same. When you lower the price of a product, the quantity demanded rises for two reasons: rst, as demand curves slope downward and as prices drop, there is more consumption; second, if the price reduction is temporary, the customer will purchase additional units to be consumed at future dates. Thus, for storable goods, the variation over time in sales is greater than the variation in consumption. Strategic Interaction The analysis in this chapter starts with a demand curve for the product. Managers choose a pricing strategy that maximizes prots given demand and its cost structure. This approach holds the price of substitute products constant (recall the denition of a demand MANAGERIAL APPLICATIONS Market Segmentation The European Commission is trying to forge a single market for drugs across the European union. Bayer, the German drug maker, was found guilty of striking agreements with French and Spanish wholesalers to dissuade them from exporting Adalat, a heart treatment, from Spain and France where the price is low, to Britain, where it is 40 percent higher. The European Court of Justice, Europes highest court, ruled that the commission was wrong to ne Bayer for their actions. This ruling provides Bayer with more exibility to segregate markets and offer the same product at different prices. Source: P. Meller (2004), Europe Effort to Control Pricing Is Set Back, New York Times ( January 7), W1. 214 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 230 Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics curve). Thus the reactions of competitors are not considered explicitly in this analysis. Although this approach serves as a useful starting point in many practical applications, it is not as useful in markets with only a few major competitors. In these markets, it often is foolish to ignore rival reactions in setting prices. For example, if United Airlines offers a major price reduction to customers, it is reasonably safe to assume that the other major airlines (for example, American Airlines and Delta) will not hold their prices constant. In these situations, it is important for managers to consider explicitly the reactions of rival rms when they price their products. Chapter 9 provides a set of tools (derived from game theory) that is useful for managers within these interactive settings. Legal Issues There are laws that restrict the ability of rms to charge different prices to different customers. For example, the Robinson-Patman Act limits the ability of rms in the United States to charge retailers different prices unless they can justify the price by showing differences in their costs. (Companies have had considerable difculty sustaining a cost justication defense.) Laws also can limit price levels. For instance, in many countries public utility companies are regulated in terms of the prices that they are allowed to charge customers. Managers contemplating the pricing policies discussed in this chapter should check the legality of a proposed pricing policy given the laws they face within a particular jurisdiction. Legal constraints can drive the rms choice of pricing policy in a variety of ways. For example, prompted by complaints from local merchants in a college town, the city council passed a law to restrict street vendors from selling handcrafted items. But the law exempted ower vendors: There was a long-standing tradition of giving owers to dates before ball games, and local ladies had long supplied the cut owers. Undeterred, the craft vendors bought bouquets of owers, placed a ower on each item, and sold the ower with the craft item bundled in for free. Some manufacturers in the 1970s began offering rebates rather than cutting product price because when price controls were implemented by the U.S. government, rebates and price cuts were treated differentially. Xerox originally leased copying machines with a requirement that the lessee buy Xerox paper; it set the paper price above marginal cost. IBM used a similar pricing plan for tabulation machines and IBM cards. The intent was to extract higher prots from higher-volume users. The government successfully charged both companies with MANAGERIAL APPLICATIONS Apple Settles Antitrust Case by Lowering iTune Prices in Britain Many forms of price discrimination are legal in countries around the world. Companies, however, can face regulatory constraints in pricing. On January 9, 2008, Apple Computer settled an antitrust suit by agreeing to cut prices on iTunes digital music in Britain to align them with those in continental Europe. The European Commission had accused Apple in spring 2007 of unfairly charging British consumers more than their counterparts in the euro zone for tracks from iTunes. British consumers had been paying about $1.55 per song, while other European consumers were paying about $1.46 per song. Under the settlement, all the European customers will be charged the same price. Source: Eric Pfanner (2008), Apple to Cut the Prices of iTunes in Britain. nytimes.com (January 10). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 215 The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 231 employing illegal tying arrangements. These companies could have achieved the same result with less legal exposure. Rather than sell paper and punch cards at prices above cost, they could have offered these items at marginal cost and rented the machines with a two-part pricea xed amount per month plus so much per copy or card read. Implementing a Pricing Strategy In this chapter we have provided an introduction to product pricing. We have discussed the objective of pricing decisions, presented a basic economic analysis of product pricing, and provided a rationale for many observed pricing policies. We also have identied key features in the business environment (such as the nature of the customer base and the type of information that is held by managers) that can affect pricing decisions. Given our intent, we have kept the analysis relatively simple. The actual implementation of a pricing strategy is complicated by a number of important factors: First, the pricing policies presented in this chapter are not mutually exclusive. Many companies can and do use a combination of pricing policies (consider Beyond.com). Second, our analysis has focused primarily on pricing a single product, but most firms sell a variety of products. In many cases, it is important to consider the interactions of demands and costs of multiple products in developing a pricing strategy. Third, optimal pricing policies can change across time. Fourth, firms have to give more detailed consideration to the legal and strategic issues in formulating pricing strategies. Thus, to manage product pricing effectively, it is important to supplement the basic material presented in this chapter with industry experience and additional training in the economics of pricing.21 ANALYZING MANAGERIAL DECISIONS: iTunes Music Pricing22 Consumers have been able to purchase digital music and audiobooks over the Internet through Apple Computers iTunes Music Store (Music Store) since April 2003. The Music Store is integrated with Apples iTunes Software, which allows users to manage their digital music libraries and to interface with their iPodsApples highly successful MP3 player. Apples stock price increased from $7/share in April 2003 to over $46/share in August 2005 21 (adjusted for stock splits). This 6.5-fold price increase compares to Microsofts stock, which increased from about $21.50/share to $27/share over the same period. Apples strong performance was fueled by the growth in its digital music business (increased PowerBook sales also contributed). In August 2005, Apples market shares for downloaded music and MP3 players in the United States were approximately 75 percent and 80 percent respectively. Apples net sales by product from 2002 through 2004 can be The high salaries paid to successful pricing managers highlight the complexity of pricing decisions in some industries. For instance, senior pricing managers are among the highest-paid professionals in major airline companies. 22 This case is based on a number of articles and SEC lings in the public domain. Among the most important of these sources are J. Leeds (2005), Apple, Digital Musics Angel, Earns Record Industrys Scorn, New York Times (nytimes.com, August 27); and J. Leeds (2008), Free Song Promotion Is Expected from Amazon, nytimes.com (January 14). 216 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 232 Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics found on page 28 of Apples 10-K statement for scal year 2004. (The 10-K can be obtained online from the SECs Edgar Archives Web site at http://www.sec.gov/Archives/edgar/data/320193/ 000104746904035975/a2147337z10-k.htm.) Since the inception of the Music Store, Apple priced all downloaded music at $.99 per song. About $.70 per song is paid to the major record companies that have the rights to the songs. The record companies were initially happy with this arrangement since it provided a way to collect at least some revenue from downloaded music. Prior to the development of iTunes, many consumers downloaded music through services such as Napster, with no royalties paid to the music companies or artists. By August 2005, Music Store sales had become big business and two of the four major record companies expressed dissatisfaction with the $.99 price. Sony BMG Music Entertainment and Warner Music favored a more complex pricing scheme that would price songs by popularity. A popular new single, for example, might sell for $1.49, while a golden oldie might sell for substantially less than $.99. Executives from these two music companies argued that their revenue stream could be enhanced by exible pricing. They complained that Apple had an incentive to sell downloaded music at too low a price to promote the sale of iPods. To quote one music company executive, Mr. Jobs has got two revenue streams: one from our music and one from the sale of his iPods. Ive got one revenue stream that it would require a medical professional to locate. Its not pretty. Not all of the major record companies shared the same view. For example, as of August 2005, the Universal Music Group (a unit of Vivendi Universalthe industrys biggest company) supported Apples desire to maintain the price of $.99 a track. The difference in opinion among the four record companies reected varying views on whether the rapidly expanding digital market was stable enough to bear a mix of pricesparticularly a higher top-end price. Millions of consumers were still trading music free on unauthorized le-swapping networks and an increase in price would increase the incentives to engage in this practice. One music executive noted, I dont think its time yet. We need to convert a lot more people to the habit of buying music online. I dont think a way to convert more people is to raise the price. As of January 2008, Apple had substantial power in negotiating with the record companies. It was unlikely that any one of the music companies would try to force Apple to change its pricing policies by withholding their music. Analysts, however, forecasted that Apples leverage over the music companies could fall in the future due to increased competition, for example, from major wireless companies who were likely to begin offering downloaded music services to cell phone customers. In January 2008 all four major music companies agreed to allow Amazon.com to sell their entire music catalogs in the MP3 format, without digital locks that restrict users from making copies of the songs. Each of the companies, except for the EMI Group, continued to require Apple to sell their music wrapped in digital rights management software. All four of the music companies also agreed to participate with Pepsi-Cola in a free song promotion program through Amazon that would be featured during the 2008 Super Bowl. The companies participated in a similar deal between Apple and Pepsi in 2004. 1. Provide an argument for why a variable pricing policy might increase the sales revenue from Apples Music Store (compared to the at pricing policy). 2. Discuss other potential pricing policies that might increase the revenue from Music Store sales. 3. What are the risks and potential costs of implementing more sophisticated pricing schemes for the downloaded music? 4. Is Apples pricing objective to maximize the revenue it receives from the sales of downloaded music? Is this the objective of the major record companies? Explain. (Hint: review the revenue/product data from Apples 10-K.) 5. Do you think that Apples ability to control the pricing of downloaded music is likely to change in the future? Explain. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Summary Pricing with Market Power 217 233 A rm has market power when it faces a downward-sloping demand curve. Firms with market power can raise price without losing all customers to competitors. The ultimate objective is to choose a pricing policy that maximizes the value of the rm. Consumer surplus is dened as the difference between what the consumer is willing to pay for a product and what the consumer actually pays when buying it. Managers, in maximizing prots, try to devise a pricing policy that captures as much of the gains from trade as possible. Thus, they try to capture potential consumer surplus as company prot. In the benchmark case, the rm chooses a single per-unit price for all customers. Prots are maximized at the price and output level where marginal revenue equals marginal cost. Fixed and sunk costs are irrelevant; only incremental costs matter in the pricing decision. The optimal price markup over marginal cost depends on the elasticity of demand. The optimal markup decreases as demand becomes more elastic: It is optimal to charge high prices when customers are not very price-sensitive. Economic theory suggests that managers should price so that marginal revenue equals marginal cost. One practical problem in applying this principle is that managers often do not have precise information about their demand curves and thus their marginal revenue (Chapter 4 discusses methods of estimating demand). The linear approximation technique can be used when the demand curve is roughly linear and the manager has basic information about current price-quantity, price sensitivity, and marginal cost. Markup pricing is a technique that managers can use when they have limited information and reason to believe that price elasticity varies little across the demand curve. One of the most common pricing methods used by rms is cost-plus pricing. Managers using this technique calculate average total cost and mark up the price to yield a desired rate of return. Cost-plus pricing appears inconsistent with prot maximization since it includes xed and sunk costs and does not consider consumer demand explicitly. Managers, however, can consider consumer demand implicitly by choosing appropriate target returns (lower target returns are chosen when demand is more elastic). The widespread use of this pricing policy suggests that it can be a useful rule of thumb in some settings. The benchmark policy charges the same price to all customers independent of the quantity purchased. Sometimes a rm can do better with more complicated pricing policies. With block pricing a high price is charged for the rst block and declining prices for subsequent blocks. Block pricing either can be used to extract additional prots from a set of customers with similar demands or can be used to price-discriminate. With a two-part tariff, the customer pays an up-front fee for the right to buy the product and then pays additional fees for each unit of the product consumed. Two-part tariffs tend to work best when customer demand is relatively homogeneous. Price discrimination occurs whenever a rm charges differential prices across customers that are not related to differences in production and distribution costs. With price discrimination, the markup or prot margin realized varies across customers. Two conditions are necessary for protable price discrimination. First, different price elasticities of demand must exist in various submarkets for the product (customers must be heterogeneous). Second, the rm must be able to identify submarkets and restrict transfers among consumers across different submarkets. Personalized pricing extracts the maximum amount each customer is willing to pay for the product. Each consumer is charged a price that makes him or her indifferent between purchasing and not purchasing the product. Group pricing results when a rm separates its customers into several groups and sets a different price for each group. A rm that can segment its market maximizes prots by setting marginal revenue equal to marginal cost in each market segment (higher prices are charged to the less price-sensitive 218 234 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 7. Pricing with Market Power The McGrawHill Companies, 2009 Managerial Economics groups). Both personalized and group pricing require relatively good information about individual customer demands. Even if the manager does not have detailed information about individual demands, price discrimination still is possible with sufcient information about the distribution of individual demands. With menu pricing all potential customers are given the same menu of options. The classic example involves block pricing, where the price per unit depends on the quantity purchased. Customers use their private information to select the best option for them. By carefully constructing the menu of options, the rm makes more prots than if it simply offered the product at one price to all potential customers. Coupons and rebates offer price discounts to customers. Price discrimination is one reason why rms use coupons and rebates to make price discounts rather than simply lowering the price. Price-sensitive customers are more likely to use coupons and rebatesand thus are charged lower effective pricesthan customers who are less price-sensitive. Similar to menu pricing, customers self-select, depending on private information about their personal characteristics. Coupon and rebate programs are expensive to administer. These costs have to be compared to the benets in deciding whether to adopt such a program. Firms frequently bundle products for sale. One reason for bundling products is to extract additional prots from a customer base with heterogeneous product demands. Bundling can be more protable than selling the products separately when the relative values that the customers place on the individual products vary. This chapter focuses on a single-period pricing problem, in which managers face a xed demand curve and cost structure. The prices of competing products are held constant. In some situations, concerns about future demand and costs, as well as the reactions of competitors, can motivate managers to choose pricing policies that would not be appropriate in the simple single-period analysis. Chapter 9 addresses issues of strategic interaction in greater detail. Suggested Readings T. Nagel 1994, The Strategy and Tactics of Pricing: A Guide to Protable Decision Making, 2nd edition (Prentice Hall: Englewood Cliffs, NJ). C. Shapiro and H.Varian (1999), Information Rules: A Strategic Guide to the Network Economy (Harvard Business School Press: Boston). Self-Evaluation Problems 71. ABC Software Solutions allows customers to access its software application remotely. Let P be the price charged to the customer each time the customer accesses the software. Customers are homogenous in their demands. The demand curve for each customer is P 20 0.1Q, and ABCs total cost for each customer is TC 10 2Q. a. If ABC charges a single price per access (and no up-front fee), what are the optimal P and Q per customer? Compute the prots per customer for ABC for this scenario. b. Suppose ABC charges a menu of 2 prices where each customer is charged one price per access for up to 90 uses and a lower price per access for additional uses. What are the optimal prices and what quantity would be purchased per customer? Compute the prots per customer for this scenario. c. If ABC decides to levy a two-part tariff, what are the optimal up-front fee, P, and Q per customer? What is the prot per customer for this scenario? 72. Consider a rm with MC AC 1 trading with two buyers, whose demand functions are Q 1 5 2P1 and Q 2 7 3P2. The rm can distinguish the buyers and is able to price discriminate. Suppose the rm charges each buyer a single price per unit for the product. What is the optimal price-discrimination strategy? Calculate the rms prots. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 219 235 73. The Key Club is a bar that has two types of potential customers: legal and underage drinkers. It is illegal to allow entry to underage drinkers, but there is no way to perfectly identify underage drinkers (fake IDs, etc.). Assume that Key Clubs marginal cost is $3.00 per drink. The drink demand for a representative customer in each of the two groups is given by: PL PU 5 QL 3.5 Q U (legal drinkers) (underage drinkers) Design a pricing policy that will extract all of the prot from the legal drinkers without appealing to underage drinkers at all. 74. Company XYZ supplies two products, DVD discs and DVD storage cases, to two different segments of customers (1 and 2). The following table summarizes the value that the typical customer in each segment assigns to the products offered by XYZ: Customer Type Discs $6 $7 1 2 Cases $8 $5 Assume that there are 10 customers of each type, that XYZ has no xed costs, and that the marginal costs of producing discs and cases are both constant at zero. a. What is the pricing strategy for XYZ if it prices the products individually? What is the corresponding prot? b. If XYZ decides to offer discs and cases in a bundle, what price should it charge for the bundle? Is bundling a better strategy in this case? Why? Explain. Solutions to Self-Evaluation Problems 71. Homogenous Consumer Demands a. Prot maximization occurs when MR MC: 20 .2Q 2. Thus, Q * 90 and P * $11. Prots/customer (TR TC) $800. b. ABC should set the same price as in Part A for the rst 90 units, $11 (the demand curve implies that $11 is the highest price the rm can charge for 90 uses). This leaves a residual demand curve of P 11 .1Q (at a price of $11, no additional units are purchased; additional units are purchased as the rm lowers it price below $11 for the additional uses). Setting MR MC for this curve: 11 .2Q 2 yields and optimal price and quantity of Q ** 45 and P ** $6.50. Each customer ends up purchasing 135 units, (the rst 90 at the $11 price and 45 additional units at the $6.50 price). Total prots/customer for the two-price scheme: (90 $11) (45 6.50) (10 270) $1,002.50. Graphically the solution looks as follows: P $2 Residual demand curve $1 $6.50 90 135 200 Q 220 236 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Managerial Economics c. Under the profit-maximizing two-part tariff, ABC would charge MC $2 for each access to its software. Each customer would access the software 180 times at this price (from the demand curve). Charging marginal cost maximizes the gains from trade (a customer accesses the software whenever he values it more than the cost to the company). ABC should charge an up-front fee that extracts all of these gains from trade as profit. Graphically the up-front fee is represented by the area of Triangle A in the following diagram. The area of a triangle is .5(base height). Therefore the optimal up-front fee is .5(18 180) $1,620. Total revenue from both the price/unit and the fee is $1,980. Total costs are ($10 (2 180)) $370. Thus the profit/customer from this scheme is $1,610. ABC obtains the maximum possible profits using the two-part tariff. It does worse with the quantity discount scheme and even worse with the single price scheme. P Optimal up-front fee is equal to the area of Triangle A: .5(18 180) = $1,620. $2 Optimal price for each access of the software A $2 180 200 Q 72. Group Pricing Begin by solving the demand functions to obtain the two demand curves: P1 2.5 P2 (7 3) .5Q 1 1 3 Q2 Set marginal revenue equal to marginal cost in both segments (the profit-maximizing conditions): 2.5 73 Q1 1 2 3 Q2 1 Next, solve for the equations to obtain the optimal quantities. Use these quantities and the demand curves to nd the optimal prices: Q 1 1.5 and P 1 $1.75, and Q 2 2 and * * * P 2 $1.67. Prots are equal to the sum of total revenue total costs for the two segments: * ($1.125 $1.34) $2.46. 73. Using Information about Distribution of Demands A two-part tariff can be used to extract all the profit from the legal drinkers. Charge them $3.00 per drink (MC) and have a cover charge equal to their consumer surplus (1 2 (.5 .3) 2) $2.00. The underage drinkers will not pay the cover charge to enter since their consumer surplus given a price of $3.00 per drink is less than the cover charge of $2.00. Their consumer surplus (before the cover charge) is (1 2 ($3.50 3.00) .5) $.125. This pricing scheme motivates potential customers, who know their own types, to make choices that are consistent with the Clubs objective of maximizing profits from legal drinkers without serving underage drinkers. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 221 The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 237 P P $5 CS = $2 CS = $.125 $3.50 $3 $3 2 Legal drinkers Q .5 Underage drinkers Q 74. Product Bundling a. Because marginal cost is equal to zero, maximizing revenue is the same as maximizing prot. By charging the low reservation price for each product (discs and cases), XYZ will sell both products to both types of customers. This is the prot-maximizing strategy for the separate pricing case in this example. Therefore, XYZ should charge: $6 for discs $5 for cases The corresponding prot is 6 20 5 20 $220. Prot is lower if either or both products are priced higher and sold to only one customer type. (Note that if the number of customers of each type changed, other strategies might be optimal. For example, if there were 100 Type 1 customers and one Type 2 customer, XYZ would be better off to price cases at $8 and sell them only to the Type 1 customers. The optimal solution could also be different with different assumptions about the marginal costs of producing the two products.) b. A Type 1 customer has a reservation price of $14 (8 6) for the bundle, whereas a Type 2 customer has a reservation price of $12 (7 5). Given that there are an equal number of customers of each type, prots are maximized by charging $12 for the bundle. This is the highest price that ensures XYZ can sell the bundle to both types of customers. The corresponding prot is $12 20 $240. Bundling results in higher prots than the individual product pricing strategy because the reservation prices for the two customer types are negatively correlated. Type 1 customers value cases more than Type 2 customers (8 5 1), whereas Type 2 customers value discs more than Type 1 customers (7 6 1). This negative correlation implies that the minimum reservation price for the bundle ( $12 for the Type 2 customers) is greater than the sum of the minimum reservation prices for the individual products ($11 $5 $6). Thus more revenue per customer is collected by selling the products in a bundle for $12, rather than pricing them separately at $5 and $6. Review Questions 71. Macrosoft is a new producer of word processing software. Recently, it announced that it is giving away its product to the rst 100,000 customers. Using the concepts from this chapter, explain why this might be an optimal policy. 72. The local space museum has hired you to assist them in setting admission prices. The museums managers recognize that there are two distinct demand curves for admission. One demand curve applies to people ages 12 to 64, whereas the other is for children and senior citizens. The two demand curves are PA PCS 9.6 0.08QA 4 0.05QCS 222 238 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Managerial Economics where PA is the adult price, PCS is the child/senior citizen price, QA is the adult quantity, and QCS is the child/senior citizen quantity. Crowding is not a problem at the museum, and so managers consider marginal cost to be zero. a. What price should they charge to each group to maximize prots? b. How many adults will visit the museum? How many children and senior citizens? c. What are the museums prots? 73. Textbook publishers have traditionally produced both United States and international editions of most leading textbooks. The United States version typically sells at a higher price than the international edition. (a) Discuss why publishers use this pricing plan. (b) Discuss how the Internet might affect the ability of companies to implement this type of policy. 74. Suppose in Table 7.2 (Product Bundling) that the professional user values Expedia Streets at $15 rather than $30. Keep all other valuations the same. Discuss how this change affects the optimal pricing strategy. 75. Explain why perfect personalized pricing is typically more protable than menu pricing. Why then do companies use menu pricing? 76. In the example in this chapter, the linear approximation method produced the prot-maximizing price, whereas the markup pricing rule did not. Does this imply that the linear rule is always better than the markup rule? Explain. 77. Why do companies grant discounts to senior citizens and students? 78. You own a theater with 200 seats. The demand for seats is Q 300 100P. You are charging $1.25 per ticket and selling tickets to 175 people. Your costs are xed and do not depend on the number of people attending. Should you cut your price to ll the theater? Explain. What other pricing policies might you use to increase your prots? 79. The Snow City Ski Resort caters to both out-of-town skiers and local skiers. The demand for ski tickets for each market segment is independent of the other market segments. The marginal cost of servicing a skier of either type is $10. Suppose the demand curves for the two market segments are: Out of town: Q0 600 10P Local: Q1 600 20P a. If the resort charges one price to all skiers, what is the prot-maximizing price? Calculate how many lift tickets will be sold to each group. What is the total prot? b. Which market segment has the highest price elasticity at this outcome? c. If the company sells tickets at different prices to the two market segments, what is the optimal price and quantity for each segment? What are the total prots for the resort? d. What techniques might the resort use to implement such a pricing policy? What must the resort guard against, if the pricing policy is to work effectively? 710. All consumers have identical demand for a product. Each persons demand curve is P 30 2Q. The marginal cost of production is $2. Devise a two-part tariff that will exhaust all consumer surplus. 711. Xerox sells both copiers and a toner for their copiers. While customers are not required to buy Xerox toner, most do because specied machines use toner only for that machine. The Xerox toner and machines are closely designed and non-Xerox toner in Xerox machines produces inferior copies. Evaluate the statement: Xerox makes 75 percent of its prots selling toner and 25 percent of its prots selling machines. 712. Some tennis clubs charge an up-front fee to join and a per-hour charge for court time. Others do not charge a membership fee but charge a higher per-hour fee for court time. Consider clubs in two different locations. One is located in a suburban area where the residents tend to be of similar age, income, and occupation. The other is in the city with a more diverse population. Which of the locations is more likely to charge a membership fee? Explain. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 7. Pricing with Market Power Chapter 7 Pricing with Market Power 223 239 713. Consider three rms: a shoe store at the mall, an automobile dealership, and a house painting rm. a. Which rm would you expect to engage in the most price discrimination? Why? b. How has the Internet changed the pricing policies of these businesses? 714. Cellwave is a cellular phone company. Answer the following questions relating to its pricing policies: a. When Cellwave started, it sold to a group of homogeneous retail customers. Each persons monthly demand for cell phone minutes was given by P $2 0.02Q , where P the price per minute and Q the quantity of minutes purchased each month. Cellwaves marginal cost is 10 cents per minute. Suppose that Cellwave charges a single per minute price to all customers (independent of the number of minutes they use each month). What is the prot-maximizing price. Depict this choice on a graph. On a per customer basis, what are the companys prot, consumer surplus, and the deadweight loss? b. Suppose that Cellwave chooses to charge a two-part tariff (with a monthly xed charge and a per minute rate) rather than a single per minute price. What two-part tariff extracts the entire consumer surplus? What are the companys prots (on a per customer basis)? How many minutes does each customer use per month? What is the deadweight loss? c. After several years of operation, Cellwave developed a new group of business customers (in addition to its old customer base). The business customers had homogeneous demands. Each of these customers monthly demand for cell phone minutes was given by P $2 0.004Q. Graph the two demand curves for the two customer groups on the same gure along with the marginal cost. Suppose that Cellwave wants to menu price by offering two plans with different monthly xed charges. Each plan would allow free calls up to some maximum limit of minutes per month. No calls are allowed beyond these maximums. Assume that Cellwave designs a plan that extracts all consumer surplus from the retail customers. Shade the area of the graph that shows how much consumer surplus must be given to each business customer to make the plan work. Explain why. 715. The Hewl-Pact Company produces a popular printer than prints over 100 pages per minute. It recently announced that it was introducing a lower priced model of the printer that can print 30 pages per minute. While not revealed to the public, it turns out that it costs the company more to produce the lower priced product. The two models are identical except for a $20 internal part for the low-priced model that slows the printer from 100 to 30 pages per minute. Provide an economic explanation for why the company decided to produce a new lower priced, but more costly, model of the printer. 224 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value CHAPTER 8 CHAPTER OUTLINE Strategy Value Creation Production and Producer Transaction Costs Consumer Transaction Costs Other Ways to Increase Demand New Products and Services Cooperating to Increase Value Converting Organizational Knowledge into Value Opportunities to Create Value Capturing Value Market Power Superior Factors of Production A Partial Explanation for Wal-Marts Success All Good Things Must End Economics of Diversication Benets of Diversication Costs of Diversication Management Implications Strategy Formulation Understanding Resources and Capabilities Understanding the Environment Combining Environmental and Internal Analyses Strategy and Organizational Architecture Can All Firms Capture Value? Summary O ver the past four decades, Wal-Mart Stores has become the largest, most protable retailer in the world.1 Its discount-variety stores, Sams Clubs, and supercenters (combined discount retail and bulk grocery stores) are located throughout the United States, as well as in many other countries such as 1 Details for this example are from S. Foley (1994), Wal-Mart Stores, Inc., Harvard Business School Case 9-794-024; the nancial press; company reports; and C. Loomis (2000), Sam Would Be Proud, Fortune (April 17), 131144. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 225 241 Mexico, Canada, Puerto Rico, Germany, Brazil, Argentina, China, Korea, and Indonesia. In 2008, it reported $374.5 billion in sales and net income of over $12.9 billion; its 2 million employees worked at 7,266 stores. Since Sam Walton opened his rst store in 1962, the overall nancial performance of the company has been nothing less than phenomenal. To illustrate, suppose you had purchased $1,000 of the stock at the initial offering in 1970 and held it through 2008. This investment would be worth more than $5 million (that is a 20 percent compound annual growth rate); in addition, you would have received cash dividends paid by the company over the period. This performance made Wal-Mart one of the hottest stocks in the market. While Wal-Marts performance has been spectacular, it has not been awless. Following a large price drop in the second quarter of 1993 (over 20 percent during a period when the overall market was essentially at), Wal-Marts stock price continued to decline modestly until 1997, even though the overall stock market was rising. In addition, sales growth at Wal-Mart, which for a long period had outpaced the retail industry, fell to unremarkable levels. In an attempt to increase performance during the 1990s, Wal-Mart took a variety of actions; it opened new stores internationally, opened new supercenters, and logged on to the world of electronic commerce. In 1998, Wal-Mart entered the traditional grocery store business in Arkansas where it opened three experimental 40,000-square-foot grocery stores (about the same size as traditional supermarkets). During the period from 1997 to 2005, Wal-Mart stock again performed well relative to the general stock market and retail industry. All managers would like to outperform the general market as well as their specic industry over a sustained period. Examples like that of Wal-Mart suggest that such performance is possible but raise at least ve important questions: What accounts for the success of these rms? Should all properly managed rms expect sustained superior performance? What actions can managers take to generate superior performance? Can managers enhance nancial returns through diversication (as Wal-Mart was attempting to do by opening grocery stores)? Do all high-performing rms ultimately fall back with the rest of the pack as Wal-Mart did in the mid-1990s? This chapter applies the basic economic concepts developed in this book to address these and related questions. Strategy Strategy refers to the general policies that managers adopt to generate prots. For example, in what industries does the rm operate? What products and services does it offer and to which customers? In what basic ways does it compete or cooperate with other rms within its business environment? Rather than focusing on operational detail, a rms strategy addresses broad, long-term issues facing the rm.2 Typically, strategies do not remain constant but evolve through time. For example, Wal-Marts strategy focuses on discount retailing and the related grocery industry. It owns and operates four basic 2 In the strategy literature, these questions are frequently divided: Corporate strategy refers to the choice of industries, whereas business strategy refers to the choice of how to compete within the chosen industries (for example, whether to focus on cost or quality, or on some combination of the two). 226 242 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics types of stores: discount-variety stores, Sams Clubs, supercenters, and grocery stores. It offers a wide product assortment, every-day low prices (supported by a low cost structure), limited advertising, and friendly, well-informed sales associates. Originally, its strategy focused on placing stores in the rural Southeast. Wal-Mart now operates stores throughout the world. Ultimately, along with the organizational architecture of the rm (discussed in Part 3 of this book), strategy is a key determinant of the success or failure of the enterprise. The ultimate objective of strategic decision making is to realize sustained prots.3 To achieve this objective, managers must devise ways both to create and to capture value. Earlier chapters focused on how value might be created and captured through input, output, and pricing policies. Those chapters also analyzed how competition constrains the ability of managers to capture value. But they (like most discussions in the traditional managerial economics literature) focused on a single product, taking the industry and product characteristics as given. This chapter takes a broader look at how managers create and capture value. The next section presents an analysis of value creation within a given industry. Subsequent sections examine capturing value and the choice of industries. The nal section offers a general framework for implementing the concepts in this chapter. Firms often compete against a few identiable rivals. For example, Boeing competes largely with Airbus in the production of commercial jet airplanes. It is particularly important for managers in such rms to consider likely responses of rivals when making strategic decisions about pricing, new investment, advertising, and so on. For example, it would be foolish if Boeing failed to consider likely responses by Airbus in setting the prices of its wide-body jets. In this chapter, we concentrate on the broader issues of how rms create and capture value; thus we abstract from how reactions by rivals might be incorporated explicitly in the analysis. In Chapter 9 we use game theory to provide an explicit analysis of reactions by rivals in the strategic decision-making process. Value Creation Figure 8.1 displays supply and demand curves for an industry. It differs from our previous supply and demand gures in one important respect: It explicitly displays both consumer-borne and producer-borne transaction costs. Consumer transaction costs include such things as the costs of searching for the product, learning product characteristics and quality, negotiating terms of sale with a supplier, and enforcing agreements. If these costs were lower, demanders would be willing to pay more for the product. For example, automatic teller machines increase the demand for banking services by reducing the amount of time customers spend in line and by providing basic banking services around the clock. The dotted demand curve indicates potential demandwhat demand would be if consumer-borne transaction costs could be eliminated. The solid demand curve displays the effective demand given a per-unit consumer transaction cost of a. Similarly, producers bear costs in transacting with consumers and suppliers (for example, negotiating terms with customers and paying attorneys to draft a supply agreement). The dotted 3 If stock market participants anticipate that a company is going to earn sustained abnormal prots, its stock price will be bid up so that in equilibrium investors will earn only a normal rate of return by buying the stock. The fact that Wal-Mart has generated both high prots and high stock returns over a long time period suggests that the company has repeatedly surprised the stock market with its earnings performance. In this book, we concentrate on the underlying prots of the company, not the stock market valuation of these prots. Stock market valuation is covered in nance classes and textbooks. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 8 Figure 8.1 Value Ways to Create Economics of Strategy: Creating and Capturing Value 243 $ a Co ns um er- bo Price (in dollars) This gure displays the supply and demand curves for an industry. It differs from previous diagrams by including producer and consumer transaction costs. [Per-unit consumer-borne (producer-borne) transaction cost is a(b).] Consumers would demand more and producers would supply more if these costs were lower. Value consists of the sum of consumer and producer surplus. Managers can increase value by reducing transaction or production costs or by increasing the demand for the productfor example, by increasing the perceived quality of the product. 227 The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value rne tra Effective supply given transaction costs ns ac tio nc Potential supply if no transaction costs os ts Consumer surplus P* ts Producer surplus ne bor ruce ct nsa ion cos Potential demand if no transaction costs tra d b Pro Effective demand after transaction costs Q Q* Quantity supply curve indicates potential supplythe willingness of producers to supply the product if producer transaction costs were eliminated. The solid line shows the effective supply curve given a per-unit transaction cost of b.4 The area under the dotted demand curve (demand before transaction costs) and above the dotted supply curve (supply before transaction costs) up to Q * is divided into four areas. First, there are the two parallelograms representing consumer and producer transaction costs. Next, there are the two triangles representing consumer and producer surpluses. Total value created by the industry is the sum of producer surplus and consumer surplus. An important rst step in making prots is discovering ways to create value. The second step, discussed below, devises ways to capture this value. Figure 8.1 suggests at least four general ways that managers within the industry might increase value. They can take actions to lower production costs or producer transaction costs, thus shifting the effective supply curve to the right. Managers can implement policies to reduce consumer transaction costs, thus shifting the effective demand curve to the right. They can take actions other than reducing consumer transaction costs to increase demand, shifting both the potential and effective demand curves to the right. Managers can devise new products or servicesin essence creating a new gure. We discuss each of these strategies in turn.5 4 For expositional convenience, we assume that supply costs are separable into production and transaction costs. Admittedly, there are cases where they are joint or where the allocation is arbitrary. 5 We depict demand and supply at the industry level because it facilitates our subsequent discussion on capturing value (which involves competition within the industry). The same four value-creating factors are important if we frame the problem using a rms demand and costs. 228 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 244 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Creating Value: Reducing Consumer Waiting Time Many rms undervalue their customers waiting time. In 2000 the Oregon Public Utilities Commission ordered Qwest Communications to pay consumers $270 million to compensate them for installation delays. Cable companies require a four-hour window for appointments. Some doctors keep four busy patients waiting for every one that the doctor is seeing. If U.S. citizens value their time at $20 per hour, taxpayers waste $26 billion a year ling tax returns. Some companies recognize the value of their customers time and build successful corporate strategies around reducing waiting time. Suppose a customers after tax earnings is $100,000 and she works 2,000 hours; then her time is worth $50 per hour. If she decides to buy the latest $20 John Grisham novel at a bookstore, she might spend an hour driving to and from the store, parking, locating the book, and purchasing it. The total cost to purchase the book is $70. Compare that to spending 6 minutes buying the book online for $20 plus $4.50 shipping. The total cost of the online purchase is $29.50 (including the $5 for her time) for a saving of $40.50. In fact, most e-tailers business strategy is to economize on customer time. The following companies have strategies aimed at conserving customer time. Wendys serves up Big Bacon Classic within two-and-a-half minutes. Enterprise offers at-home pickup and return of rental cars. Virgin Atlantic Airlines picks up and checks in upperclass passengers from their chauffeured cars. J.C. Penny offers three-day delivery or in-store pickup of online orders. Source: K. Barron (2000), Hurry Up and Wait, Forbes (October 16), 158164. Production and Producer Transaction Costs Chapter 5 discussed how managers should choose inputs to minimize production costs. Over time, managers can discover new technological opportunities to reduce these costs and increase value. They also can devise ways to lower the costs of transacting with customers and suppliers. For example, when personal computers rst were developed, they were relatively expensive to produce. Over time, companies learned to reduce these production costs. As a result, the quantity of personal computers sold in the market has MANAGERIAL APPLICATIONS Dell Computers: Reducing Producer and Consumer Transaction Costs Dell has developed a system to collect information from customers and assemble a customized product quickly and inexpensively. This allows Dell to bypass a dealer network and their required markups, thereby offering computers at lower prices. (Note, however, that by adopting this strategy of competing with more emphasis on price than service, their product is more likely to appeal to knowledgeable customers than rst-time buyers.) This strategy has allowed Dell to grow revenues to over $30 billion, to capture about 25 percent of the PC market, and to make Michael Dell the richest man under 40 in the world. Extending their original strategy, Dell now offers customized services to large accounts. For example, Dell has loaded specialized software along with associated peripherals and network servers at Dells factory for First Union Capital Markets Group, which uses over 2,500 PCs. The package is tagged and shipped directly to the specic First Union trader for whom the unit had been designed. This process reportedly has saved First Union over $500,000 annually. These are just a few examples of how companies like Dell have reduced transaction cost, for both themselves and their customers. Sources: A. Serwer (1998), Michael Dell Rocks, Fortune (May 11), 5870; A. Serwer (2002), Dell Does Domination, Fortune ( January 21), 7075; and G. McWilliams (2003), Dell Pins Hopes on Services to Boost Prot, The Wall Street Journal (November 11), A1. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 229 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 245 increased substantiallyas has the total value (consumer plus producer surplus) created within this industry. Computer manufacturers also have devised ways to reduce their costs of transacting with suppliers and customers. For instance, large computer manufacturers have developed electronic connections with major software producers to lower the cost of ordering software programs. They also have developed computer links that reduce their costs of transacting with customers. Wal-Mart has shifted its effective supply curve to the right by developing new lowcost methods for producing and distributing retail services and products. For example, its extremely efcient hub-and-spoke distribution system has lowered the costs of stocking its stores. Wal-Mart has reduced transaction costs through direct computer links with major suppliers, such as Procter and Gamble. These links have cut the costs of restocking products and essentially have eliminated writing paper checks to these vendorsthey are paid through an electronic payment system. Consumer Transaction Costs Reducing consumer transaction costs also can increase value. For example, early WalMarts were established in small rural towns. One way that these stores added value was through reducing travel time for local residents, who previously had to drive to urban centers to do a larger part of their shopping. Wal-Mart also reduces consumerborne transactions costs by the layout of their stores. For example, Wal-Mart captures market-basket data from customer receipts at all its stores. By analyzing this data, Wal-Mart can tell which products are likely to be purchased together. Wal-Mart uses this knowledge to reduce customers costs in navigating their stores by placing commonly purchased bundles of products together. Examples of such pairings include bananas with cereal, snack cakes with coffee, bug spray with hunting gear, tissues with cold medicine, measuring spoons with baking supplies, and ashlights with Halloween costumes.6 MANAGERIAL APPLICATIONS Reducing Consumer Transactions Costs: Kraft Lunchables Kraft Foods plant in Avon, New York, had once been on the verge of being closed. But during the late 1990s it tripled employment thanks to booming demand for Lunchables, a line of nger food for kids. Kraft prepackages kids meals to provide a convenience for parents preparing lunch for children to take to school. Included in the package are meat, cheese, crackers, and a drink box. Other Lunchables include pizzas, tacos, hot dogs, and hamburgers. With ashy, colorful packaging and contests to win trips to amusement parks, Lunchables are aimed dead center at the kid consumer. And since the product line was introduced in 1988, they have become huge business for Kraft. The basic idea for Lunchables was to provide convenience for parents while allowing kids to have a hand in selecting their lunches, said Jeff Meyer, brand manager for Lunchables. Grocery Manufacturers of America spokeswoman Lisa McCue added, People dont have to think about shopping for ingredients, slicing and dicing, cooking, etc. And nowadays, people arent spending a lot of time in the kitchen. The whole convenience trend just is going to continue. Source: M. Daneman (1999), Cashing In on Kids, Democrat and Chronicle ( July 19), 1F. 6 E. Nelson (1998), Why Wal-Mart Sings, Yes, We Have Bananas! The Wall Street Journal (October 6), B1. 230 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 246 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Terrorist Attacks Affect Value Creation in the Airline Industry A rm does not have complete control over its demand or costs. Thus a rms ability to create value can be affected by factors that are beyond its control. Consider how the terrorist attacks of September 11, 2001, affected value created by the airline industry. In the customers eyes, the perceived value of ying fell, causing a dramatic drop in the number of airline seats sold, which resulted in an immediate decline in the potential for value creation. Moreover, because of the increased costs to the airlines of additional screening and tightened security procedures, as well as the additional transaction costs to passengers in the form of longer lines, longer airport waits, and limits on carry-on luggage, there also was a decline in the value created on the seats that were being sold. Most airlines reported losses in the third quarter of 2001their average costs exceeded the price they charged for tickets. Customers would y only if the value of their trip exceeded its total cost. Thus, the net value created by the airline industry in that quarter was considerably reduced and was arguably even negative (if the losses by the airlines exceeded the net value captured by their customers). The entire industry that involves marketing over the Internet is another example of reducing consumers transaction costs. For instance, when prospective customers use search engines to explore the Internet, they often are presented with a list of related books, which can be ordered electronically at a discount through companies such as Amazon.com. This service reduces consumer transaction/search costs by identifying books of potential interest and making it easier to place an order. Other Ways to Increase Demand The demand curve holds variables other than the price of the product constant. We already have discussed how managers can increase demand by reducing consumer transaction costs. They also can increase the effective demand for their productsand thus total value created by their transactionsby affecting variables such as expected product quality, prices of complements, or prices of substitutes. Product Quality Actions that enhance perceived quality increase demand; total value also is raised unless these actions entail larger increases in production costs. For instance, the innovation of titanium golf clubs increased the demand for golf equipment, while the invention of parabolic skis increased the demand for skis and skiing. The resulting increases in demand have been greater than the associated increases in production coststhus, total value created within these industries has increased. HISTORICAL APPLICATIONS Giving Away Razors to Increase Demand for Blades King Gillette gave away free razors after he invented his famous double-edge disposable blade in 1885United Cigar Store gave razors to customers who bought boxes of Cuban cigarsbanks bought razors for pennies and gave them away as part of shave-and-save campaigns. These promotions were designed to get razors in the hands of consumers who would then buy a stream of future blades for the razors. In 2005 Gillette was acquired by P&G for $57 billion. Source: No Charge, Attach (September 1999), 1416; and A. Coolidge (2005), Gillette: P&G Not Our First Choice, The Cincinnati Post (March 16). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Chapter 8 Economics of Strategy: Creating and Capturing Value 231 247 Price of Complements Managers sometimes can act to reduce the price of complements, thus increasing the demand for their products. To illustrate, consider CompuInc, which produces personal computers, and PrintCo, which produces a complementary printer.7 For simplicity, suppose that customers purchase either both products or neither product (they are quite strong complements), the price of CompuIncs personal computers is Pc , the price of PrintCos printers is Pp , and the demand for each product is Q 12 (Pc Pp) when Pc Pp is 12 or less, 0, otherwise (8.1) Q is the number of computer-printer combinations sold, and the marginal cost of producing both products is 0. If the two companies do not cooperate in setting prices, each will seek to maximize its individual prots, given its expectation of the other rms price.8 In this case, CompuInc will view its demand curve as (12 Pp) Q (8.2) (12 Pc Pc) Q (8.3) while the PrintCo demand curve is Pp where Pp and Pc represent the expectations about the other rms price.9 Each rms prot is maximized by setting MR MC (0 in this example): CompuInc (12 Pp) 2Q 0 (8.4) PrintCo Pc) 2Q 0 (8.5) (12 Substituting for Q from Equation (8.1) and rearranging the terms yields the following two reaction curves (see Chapter 6): Pc 6 0.5Pp (8.6) Pp 6 0.5Pc (8.7) The equilibrium consists of prices P * and P c that simultaneously solve these two equa* p tions.10 Figure 8.2 (page 248) displays the solution graphically. In equilibrium, both rms choose a price of 4; thus, a total of 4 units of each product is sold. The prot for each rm is 16; combined prots are 32. Now consider what happens if the rms were to coordinate the prices, for example, through a joint venture. To maximize the combined prots, the companies jointly set marginal revenue equal to marginal cost:11 12 2Q 0 (8.8) They sell 6 units at a combined price of 6 (for example, Pc Pp 3). In this case, they make a combined prot of 36. This combined prot by pricing the products jointly is 7 See A. Brandenburger (1996), Cheap Complements? (mimeographed text, Harvard Business School). We solve for these prices using basic algebra and the fact from Chapter 4 that marginal revenue (MR) for a linear demand curve is a line with the same intercept as the demand curve but with twice the negative slope. (The problem could be solved in fewer steps using calculus.) 9 There is no reason to place a subscript on Q. By assumption it is the same for both rms. 10 Recall from Chapter 6 that a Nash equilibrium is where each rm is doing the best it can, given the actions of its rivals. 8 11 Note that the combined demand curve is (Pc Pp) 12 Q ; MR 12 2Q. 232 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 248 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Managerial Economics Pp Figure 8.2 Noncooperative Pricing for CompuInc and PrintCo 12 Price of PrintCo printers If the two companies price noncooperatively, each will try to maximize its individual prots given its expectation of the other rms price. The equilibrium consists of prices P* and P* , which p c simultaneously solves these maximization problems. This gure displays the solution graphically. The gure displays the reaction curves for each rm. A given rms reaction curve indicates its optimal price given the pricing decision of the other rm. In equilibrium (where the reaction curves cross), both rms choose a price of 4. A total of 4 units of each product is sold. The prots for each rm are 16. Combined prots are 32. (Note that if they cooperate and set each price at 3, combined prots will be 36.) CompuIncs reaction curve in choosing Pc 6 Pp = 4 * PrintCos reaction curve in choosing Pp Pc Pc = 4 * 6 12 Price of personal computers higher than the prot they would receive if they do not cooperate in setting prices. When the rms price independently, they do not consider the negative effect that their higher prices have on the others prot. Pricing cooperatively, they take this interaction into account. Note that in this case, consumers also would be better off because the prices of both products are lower. Prices of Substitutes Low-priced substitutes reduce the demand for a product. Sometimes managers can affect the price of substitutes. For example, movie theaters frequently prohibit patrons from bringing food into the theater. These restrictions on lower-priced substitutes increase the demand for snacks offered by the theater. New Products and Services To this point, our discussion has focused on creating value by increasing demand or reducing the costs of producing existing products. Inventing new products and services also creates value. For example, consider the consumer electronics industry. Many of todays products that create signicant value are relatively new developments: for instance, MP3 technology, digital cameras, and digital video displays (DVDs).12 12 Obviously, the classication of whether a product development constitutes a product improvement or a new product is somewhat arbitrary. For example, are miniature cell phones separate products or improvements over previously existing larger cell phones? This classication problem is not central to our focus. The basic point is that value often is created through the development/enhancement of products. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 233 249 MANAGERIAL APPLICATIONS Airlines Restrict Cell Phone Use American Airlines warns passengers that cell phones may interfere with the aircrafts communication and navigation systems. What American does not tell passengers is that there is no scientic evidence to support these claims. A 1996 study commissioned by the Federal Aviation Administration looked at records from thousands of ights; it failed to nd even one instance where equipment was affected by a cell phone. Also plane makers Boeing and Airbus have bombarded their aircraft with cell-phone frequencies and found no interference with communication, navigation, or other systems. A likely reason for the lack of impactcell phones do not operate on any of the frequencies used by airplane systems. Why then are airlines so interested in restricting cell phone use? Airlines have an economic incentive to restrict itthey receive about 15 percent of the revenues from the telephones installed on board. The two major providers of air-phone service, GTE and AT&T, charge about $6 for a 1-minute call (more than 20 times the typical cell-phone rates). Source: J. Auerbach, (1999), Connecting Flights: Cell-Phone Use Aloft May Not Be the Danger That Airlines Claim, The Wall Street Journal (October 5), A1. Cooperating to Increase Value Our example of computers and printers shows that rms sometimes can increase value through cooperating with each other, rather than competing. In this case, the companies were producers of complementary products. Opportunities to increase value through cooperation also can arise with customers, suppliers, and even competitors. For instance, cooperating with suppliers and customers in developing computer and information links can reduce supply costs and lead to the production of more valuable products ones more tailor-made for the customer (recall the example of Dell Computers). One way competitors cooperate is in development projects to reduce joint costs. For example, major automobile manufacturers have participated jointly in the research and development of batteries for electric cars. Longer-lived batteries are essential for these companies to market electric cars on a wide-scale basis. If each company acts independently, development costs are expected to be higher. As another example, offshore drilling is quite expensive. Prior to soliciting bids for offshore sites, the U.S. government allows the oil rms to conduct a survey of the area jointly; they share the data and divide the costs. American antitrust laws generally make it illegal for rival rms to cooperate for the purpose of monopoly pricing. Nonetheless, many forms of cooperation are legal and increase the welfare of both producers and consumers (again consider our PC-printer example). MANAGERIAL APPLICATIONS Technology and Value During the latter part of the twentieth century there has been a massive change in information, communication, and production technologies. This technological change has provided important opportunities for increasing value. For example, the business process reengineering movement in the 1990s used computer and information technology to lower costs (for instance, by streamlining systems used to process orders, shipments, payables, and receivables). Flexible production technologies have allowed rms to custom-design certain products to t specic customer demands better. Computer and information technologies have been used to reduce the costs of transacting with suppliers. Using technology to increase value is likely to remain a signicant focus well into the twenty-rst century. 234 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 250 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Advanced Photo System Firms can increase value by cooperating with other rms. Consider the Advanced Photo System developed in the 1990s. This system has several advantages over traditional photo systems. The lm comes in a convenient cartridge that is loaded easily into the bottom of the camera. There are special encoding features that allow enhanced developing. The customer is able to choose from several formats when taking pictures (for example, classic 35 mm or panoramic). The customer receives a print of all photos with index numbers for subsequent reprints. To maximize consumer appeal for the new system, it was important for the lm and camera companies to adopt a set of standard system parameters. Demand for the products would be lower if the market consisted of several incompatible brands of lms, cameras, and nishing equipment. The mutual benets from cooperating on the development of this photo system motivated traditional rivals such as Kodak and Fuji in lm and Canon, Minolta, and Nikon in cameras to cooperate in a joint venture to develop the new system. Now that the system is developed and a set of standard parameters adopted, these companies again compete in their respective lm and camera markets. Converting Organizational Knowledge into Value To create value, employees must convert their existing knowledge about production processes, transaction costs, customer demand, and so on, into ideas that can be implemented by the rm. In Chapter 3, we discussed the knowledge conversion process. In this section, we consider the strategic implications of this analysis. Recall that the resources within a rm can be divided into three general categories. First are its tangible assets, which include property, plant, and equipment. Second are its intangible assets, such as patents, trademarks, and brand-name recognition. These assets typically are not shown on the rms balance sheet but can be signicant in creating and capturing valuethe rms methods of doing business, its formulas and recipes, are a particularly important type of intangible asset. Third and perhaps most important are its human resources. Firms in Silicon Valley frequently refer to these three types of resources as hardware, software, and wetware. Hardware consists of physical assets. Software is broadly used to describe the rms soft assets, such as its formulas and recipes for creating value. Wetware refers to employee brainpower, that is, wet computers. A rm owns and can capture value from its hardware and software, but it only rents its wetware. Wetware is the private property of individual employees, who can take it with them to another rm if they so choose. To create and capture value, managers must nd ways to convert the knowledge contained in employee wetwareeven knowledge the employees may not realize they haveinto software. The evolution of the McDonalds Corporation provides a good example of how this process occurs. The rst McDonalds unit was established in 1956. The companys hardware consisted of property and equipment. However, its most important asset was its software. McDonalds major source of value was its new formula for selling hamburgers, fries, and drinks to customers. This formula translated into a business approach that McDonalds was able to duplicate in locations around the world. If McDonalds had stopped innovating in 1956, new companies that copied and improved on the original formula, such as Burger King, Wendys, and Kentucky Fried Chicken, eventually would have forced McDonalds out of business. But McDonalds continually improved its formula for creating and capturing value by converting wetware into new software. Today BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 235 251 McDonalds has a much wider product offering, more effective store designs, and better production processes than it had in 1956. The development of the Filet-O-Fish Sandwich at McDonalds illustrates the conversion from wetware to software. Originally, the only food products McDonalds sold were hamburgers and fries. The product line intentionally was limited so that it could be produced efciently and quickly, according to McDonalds formula for value creation. A franchisee in a Catholic neighborhood, however, was unable to sell many hamburgers on Fridays because Catholics were not permitted to eat meat on Fridays. The franchisee worked hard to develop a tasty sh sandwich. At rst, McDonalds would not let him sell the new sandwich because it was not consistent with the image as a hamburger company. It also would lead to inconsistency across unitssomething that a franchise company generally wants to avoid. Eventually, however, McDonalds saw the value that could be created and captured by offering a sh sandwich. Specialists at the corporate level devised ways to improve the sandwich and to lower production costs (for example, by using a different type of sh that could be precut into a standard size). Ultimately, the Filet-O-Fish sandwich was introduced across all McDonalds units and has been a menu staple ever since. The idea of a sh sandwich initially was contained in the wetware of the franchisee, while the ways to improve the product were in the wetware of specialists at the corporate level. At this stage, the ideas and knowledge were not creating value. But the wetware was eventually converted into software and is now part of the McDonalds formula for creating and capturing value. A similar story lies behind the Big Mac, which was the brainchild of a franchisee in Pittsburgh who wanted a heftier sandwich to sell to steelworkers. Software is different from hardware in that it is not a scarce resource. While a given machine can be used only at one location, software can be replicated to create value at locations throughout the world. For example, McDonalds currently has over 10,000 units operating under the same business format. (Of course, as we will discuss, software also can be copied by competing rms, thus reducing its prot potential.) MANAGERIAL APPLICATIONS Sonic Drive-Ins Convert Wetware to Software Sonic Corp. is the fth largest fast-food chain and the largest drive-in restaurant brand in America with 2,400 drive-ins and 2002 sales of $2 billion. Pattye Moore, CEO of Sonic, spends half her time visiting stores and eats at Sonic with her daughters three or four times a week. We really get a lot of ideas that way. . . . We just go out and ask them what theyre xing for themselves. . . . We encourage our employees to play with their food. Sonic employees at one store were bored with buns and started eating their burgers and chicken on thick, grilled toast. Sonic introduced toaster sandwiches that became popular with customers. One customer suggested sausage on a stick wrapped in a buttermilk pancake and deep fried. It is now a breakfast staple. One store manager created the grilled chicken wrap. All of these products were launched throughout the entire Sonic chain. Notice how Sonic encourages people to experiment with new food dishes (enhancing their wetware) and then converts this wetware into software. Moore and her staff then choose which of these ideas to pursue. The recipes are rened and codied. Sonic illustrates how one company uses its organizational architecture to convert wetware at one drive-in into literally new recipes (software) that are leveraged throughout the rms other 2,400 locations. Moreover, Sonic has a corporate culture that promotes experimentation, identies potentially valuable wetware, and converts that wetware into software that is then leveraged across the rm. Source: Required Eating Makes This Job Fun, Democrat and Chronicle ( June 3, 2002), 7D. 236 252 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics Whether or not a rm is likely to be successful in converting wetware into software depends critically on how decision rights are assigned within the rm and on how employees are evaluated and rewarded (the rms organizational architecture). Part 3 of the book provides a systematic analysis of these considerations. Opportunities to Create Value The discovery of better ways to use existing resources drives much of the value that rms create. Todays personal computers are far more powerful than the 1980s mainframes, yet they take signicantly fewer resources to produce. The value created by improved computer technology has not come from the discovery of new raw materials or resources, but by using existing resources more efciently. To quote growth theorist Paul Romer,13 So it is not the raw materials or the mass of things on earth that really lies behind economic success and high standard of living, it is the process of rearrangement. And what underlies this process of rearrangement are instructions, formulas, recipes, and methods of doing things [software]. The possibilities for new value creation are immense. Consider Romers simple example of a production process that involves just 20 steps. The order of the steps can be varied in an extremely large number of ways (approximately 2.4 followed by 17 zeros) and thereby affect the value created. In most production processes, of course, there is a natural ordering that precludes certain combinations of stepsit doesnt make sense to weld the body together after the car is painted. But given all the ways to rearrange the many resources on earth, the possibilities for continued value creation are enormous. This discussion suggests that rms face an essentially unlimited set of opportunities to create better instructions, formulas, recipes, and methods for making improved products at lower cost. New opportunities are likely to emerge as technology continues to evolve. For instance, the latter part of the twentieth century saw a massive change in information, communication, and production technologies. This technological change has provided substantial opportunities for increasing value. The business process reengineering movement in the 1990s used computer and information technology to lower costs (for instance, by streamlining the systems used to process orders, shipments, payables, and receivables). Flexible production technologies have allowed rms to tailor the design of their products to t specic customer demands. Computer and information technologies reduce the costs of transacting with suppliers. Using technology to increase value is likely to remain a signicant focus well into the future. Capturing Value Creating value is an essential rst step in generating prots. But then it is necessary to capture this value. It does the rm little good to reduce transaction/production costs or to increase consumer demand (for instance, by creating new value-enhancing software) if rivals can copy these changes quickly and enter the marketsuch competition will eliminate the prots. 13 Paul Romer (1998), Bank of America Roundtable on the Soft Revolution: Achieving Growth by Managing Intangibles, Journal of Applied Corporate Finance 11(2), 827. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 8 Figure 8.3 237 The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Economics of Strategy: Creating and Capturing Value 253 Firm with Market Power versus a Firm in a Competitive Industry The rm on the left has a downward-sloping demand curve and thus has market power. It can choose price-quantity combinations. Chapters 6 and 7 demonstrated that rms with market power often can capture economic prots. The other rm is in a competitive industry and thus faces a horizontal demand curve. Chapter 6 indicated that earning economic prots in this market setting is difcult. Firms along the dotted section on the supply curve (below the market clearing price of P *) still would produce the product if demand and price fell. The shaded triangle represents producer surplus. This chapter discusses how a rm sometimes (but not always) can capture a portion of this producer surplus as an economic prot. $ $ Price (in dollars) $ Producer surplus S Dj P* P* Di D Qj Qi Quantity: Firmi MARKET POWER Quantity: Firmj Q* Q Quantity: Industry COMPETITIVE INDUSTRY Figure 8.3 compares a rm in a competitive market to a rm with market power. In competitive markets, rms face horizontal demand curves and are price takers. With market power, rms choose price-quantity combinations. Chapters 6 and 7 indicated that rms often can capture value if they exploit their market power. Sometimes rms can capture value, even without market power, if they employ superior factors of production that allow them to be more productive than their rivals. In a competitive market, there typically are rms that would be willing to continue to produce the product even if demand and the market price fell. This production corresponds to the dashed section of the supply curve (below the market price P *) in Figure 8.3. Also shown is the corresponding producer surplus. Sometimes (but not always) a rm can capture a portion of this producer surplus as economic prot. Next, we discuss the conditions under which market power and superior resources lead to economic prots.14 Market Power Entry Barriers If there are no barriers to entry, competition from new rms tends to erode prots within the industry.15 Entry barriers exist when it is difcult or uneconomic for a would-be entrant to replicate the position of industry incumbents. Chapter 6 discussed a variety of 14 A note on vocabulary: Economists say a rm is earning rents when it sells its product for a price higher than average total cost (which includes a normal rate of return). Positive rents imply that the rm is earning more than is necessary to motivate it to continue to produce the product over the long run (it is earning an economic prot). It is useful to distinguish between two types of rents: Monopoly rents, which correspond to our discussion of market power, and Ricardian rents, which correspond to our discussion of superior factors of production. 15 As discussed in Chapter 6, a rm can have market power when there are limited entry barriers (in the case of differentiated productsmonopolistic competition). Entry, however, often will eliminate or at least substantially limit economic prots. 238 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 254 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics HISTORICAL APPLICATIONS Creating but Not Capturing Value: Eli Whitney A great problem in harvesting cotton during the eighteenth century was separating cotton from its seed. This laborious task was done by hand. It was so difcult that it took a worker a whole day to clean 1 pound of staple cotton. Eli Whitney, while visiting Georgia, was intrigued by this problem. Within a few weeks, he produced a machine he called a cotton gin (a shortened form of cotton engine). It greatly increased the amount of cotton that could be cleaned in a day and soon led to cotton becoming the chief crop in the South. Clearly, the invention of the cotton gin created signicant value. Yet, before Whitneys invention was completed and patented, his rst model had been widely copied. Virtually all his prots went into lawsuits to protect and enforce his rights. He did, however, make prots in the nancial markets speculating on the price of cotton. entry barriers. These barriers include factors that make price cuts likely if a new rm enters (such as economies of scale), incumbent advantages (such as patents and brand names), and high costs of exit. Numerous researchers have tested the theoretical link between entry barriers and prot potential.16 Consistent with the theory, these studies generally conrm that price-cost margins tend to be lower in competitive industries. Researchers also have found a positive correlation between prot and specic entry barriers (such as economies of scale or advertising) as well as a positive correlation between prot and concentration (combined market share of the top few rms) across geographic markets within a given industry.17 While theory and evidence suggest that prot potential increases with entry barriers, the existence of such barriers is no guarantee of economic prots. There are at least four additional factors that are important. These include the degree of rivalry within the industry, threat of substitutes, buyer power, and supplier power.18 Degree of Rivalry When the degree of rivalry is high, prots will be low even if there are entry barriers (see Chapter 6). Two factors that help determine the level of rivalry are the number and relative sizes of competitors. The fewer the number of competitors (the more concentrated the industry), the more likely that rms will recognize their mutual dependence and refrain from cutthroat competition. The presence of a large, dominant rm (as opposed to a comparable number of similarly sized rms) also can reduce rivalry because a dominant rm frequently takes the lead in setting prices and takes actions to sanction others that do not follow its lead (within the limits of antitrust constraints). The level of rivalry can change over time as basic conditions within the industry change. For example, if existing rms within an industry have excess capacity, they often will engage in price competition in an attempt to increase their individual outputs. In general, excess 16 For a survey of some of the relevant studies, see R. Schmalensee (1989), Studies of Structure and Performance, in R. Schmalensee and R. Willig (Eds.), The Handbook of Industrial Organization (North Holland: Amsterdam). 17 In apparent contrast to the theory, researchers have failed to document a strong positive correlation between prots and concentration across industries. This is due in part to the difculty of comparing accounting data across industries. Cross-industry differences in the treatment of depreciation and other accounting choices make interindustry comparisons problematic. Another difculty is that industries can be concentrated for at least two quite different reasons. First, there can be actual barriers to entry. Second, there can be few rms in an industry because it lacks prot potential. If the sample contains both types of industries, it is not surprising that researchers nd mixed results. 18 M. Porter (1980), Competitive Strategy (Free Press: New York), labels these factors, along with the threat of entry, as the ve factors. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 239 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 255 MANAGERIAL APPLICATIONS Italian Textiles and Chinese Competition For more than six centuries wool has been cleaned, dyed, and woven in Biella, Italy. Chinese manufacturers once competed primarily on price, offering low-quality textiles. But now the quality of Chinese textiles has improved markedly. This has exerted substantial pressure on prices and has led to layoffs and closings in the Italian cloth trade. Source: C. Rhoades (2003), Threat from China Starts to Unravel Italys Cloth Trade, The Wall Street Journal (December 17), A1. capacity, high xed costs, lack of product differentiation, and slow growth all increase the degree of rivalry within the industry. Threat of Substitutes Even if entry is limited, rms within an industry are not immune to outside competition. There is the threat of substitutes. For instance, the large scale of investment potentially limits entry into the overnight delivery market (populated by rms such as FedEx and UPS). However, in recent years effective substitutes, such as e-mail and fax machines, have reduced the demand for delivery services (relative to what would have been without these technological developments). Similarly, banks face competition from money market accounts that offer customers the ability to write negotiable orders of withdrawal (usually restricted to amounts in excess of $100). Buyer and Supplier Power The nal two factors that help determine a rms market power are buyer and supplier power. If the industry has only a few large customers, prot is likely to be low since these customers will use their buying power to extract lower prices. Similarly, if the key suppliers to an industry are large, concentrated, or well organized (for example, unionized), they will attempt to extract industry prots through high input prices. During the 1990s Intel and Microsoft were essentially sole suppliers of certain key inputs for the production of personal computers; hence they have been quite protable. Market Power and Strategy Firms sometimes can increase economic prot by taking actions that promote entry barriers, reduce intraindustry rivalry, limit the availability of substitutes, or reduce buyer/supplier power. Examples of each type of activity are easy to nd. For instance, American manufacturers frequently lobby for taxes or restrictions on imports to reduce MANAGERIAL APPLICATIONS Competition and the Number of Competitors The degree of rivalry generally increases with the number of competitors. Depending on the nature of the industry, however, competition can be quite erce even with only a few competitors. Coca-Cola has had about 50 percent of global soft drink sales, while PepsiCo has had about 13 percent. Yet the battle between Coke and Pepsi has been intense for decades, with costly promotional contests and intense competition for distribution. Similarly, competition has been strong in the U.S. tobacco industry, although the three largest companies have had about 90 percent of the market. In detergents, two large companies, Procter and Gamble and Unilever, have conducted a global soap war. To varying degrees, each of these industries is characterized by excess capacity, high xed costs, lack of product differentiation, and slow growthall factors that lead to greater rivalry. 240 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 256 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Investing in a New Restaurant Concept For the last 10 years you have been working in the health food business. You have talked to many customers who have suggested a new restaurant concept. The restaurant would feature a variety of low calorie meals (under 500) made from healthy ingredients (e.g., organic fruits and vegetables and steroid/ hormone free meat). The restaurant would include a bar with an extensive organic wine list and trendy decor. There would be an emphasis on high quality, friendly service, and colorful meal presentation. Your customers suggest that they would be willing to pay around $50 to eat a meal and have a glass of wine at this type of restaurant. They lament the fact that their community has no upscale restaurants that offer this type of fare. You have conducted a detailed nancial analysis of this potential business opportunity. You believe that you have good information on the costs of starting and operating the restaurant. You project that with a $50 meal price and anticipated demand you would earn a high prot and an excellent rate of return on your investment. You have the equity capital to start the business. Several friends with MBAs argue that you would be crazy to start this business. They claim that there are few entry barriers to the restaurant industry and that every person with business training knows that you cant make prots in a competitive industry. Should you drop the idea of opening the new business based on this argument? Explain. entry by foreign rms. Firms in certain industries form cartels or other collusive agreements to reduce competition.19 As we have discussed, organizations like the AMA act to limit the availability of substitute products or services (in their case new physicians, nurses, physician assistants, and midwives). Firms reduce buyer power by opposing buyer-cooperatives. Finally, rms thwart supplier power by opposing labor unions and expanding capacity in nonunionized locationsespecially internationally. Historically, many managers focused on capturing value through anticompetitive activities, such as erecting entry barriers. They have found that often it is difcult to limit competitionespecially in a global marketplace. Also, as discussed in Chapter 21, regulators extract much of the producer surplus created through limits on competition, for example, through campaign contributions to encourage the adoption and enforcement of entry restrictions. Due to these considerations, much of the contemporary focus in strategy is on superior resourcesthe subject of the next section.20 Superior Factors of Production Both human as well as physical assets vary in productivity. For example, Sammy Sosa is a great baseball player, Michael Dell of Dell Computers is a superb CEO, land in Californias Imperial Valley is incredibly fertile, land adjacent to an expressway interchange offers customers quite convenient access, and so on. If an asset allows the rm to make a prot because of its superior productivity, other rms will compete for this resource and bid up its price. Thus, with well-functioning markets, the gains from superior productivity accrue to the responsible asset. 19 Managers must be careful undertaking these types of actions in the United States since many are illegal under American antitrust law. 20 An example of the contemporary focus in the strategy literature is D. Collis and C. Montgomery (1995), Competing on Resources: Strategy in the 1990s, Harvard Business Review, 118128. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 241 257 MANAGERIAL APPLICATIONS Sugar Prices The 2008 wholesale price for sugar in world markets was about $.16 per pound. In contrast, the price in United States markets was over $.26 per pound. The dramatic difference between the two prices is due largely to American tariffs on sugar imports. Foreign producers would be willing to sell to Americans at lower prices. Yet, they are not permitted to do so. This restriction benets domestic sugar producers, who are able to sell their sugar at the higher price. Without entry restrictions, the United States price would be driven down to about $.16 per pound. Such a domestic price decline would benet American consumers but hurt American sugar producers. Producer Surplus Captured by Superior Assets Figure 8.1 divides value into consumer surplus and producer surplus. Consumers receive the consumer surplus. Producer surplus goes to the owners of superior assets. As an example, consider Pete Irving, general manager of Speedy Modems. Last year, Speedy was just breaking even in producing modems at a price of $100. Demand for the particular type of modem increased, and the price rose to $150; Speedy began making a prot (producer surplus) of $50 per unit. This prot motivated other rms to produce the same type of modems. They were less productive than Speedy solely because they lacked a manager with Petes talents. The rival rms began making job offers to Pete. To dissuade him from going to a rival rm, Speedy had to increase Petes salary. Through this process, Petes salary increased to the point where Speedy was making only a normal rate of protthe same as the competing modem companies. The gains from his special talents went to Pete, not to the rm. This same process works for physical assets. For instance, if a rm rents a prime piece of land or a unique piece of equipment, rental rates will be bid up to reect the assets superior productivity. And if the rm owns the asset, its value (and thus the opportunity costs for continuing to use it) will be bid upin effect, the rm pays the rent to itself for its continued use of the asset. Figure 8.4 (page 258) displays a graphical analysis of this general phenomenon. The illustration on the right depicts supply and demand in the marketplace. The initial price is P *. A typical rm in the industry is depicted on the left and is making no economic 0 protsprice equals long-run average cost (LRAC0). The demand for the product increases and the market price goes up to P 1. The rm appears to have the potential to * make an economic prot, since the price is above its initial long-run average cost. Competitors and new entrants, however, will bid for the special resources (such as a talented manager or a productive piece of land) that would allow the rm to produce the product at an average cost below price. In equilibrium, the rms costs would increase to the point where its long-run average cost (LRAC1) again equals price.21 The additional producer surplus created by the increase in price (shaded area) is reected in the price or opportunity cost of the scarce assets, which made lower-cost production possible at the initial price and output level. If the rm owns the scarce asset (for example, a piece of land), the wealth of the rms owners increases as the price of the asset is bid up. It is important, however, to realize that a rm does not have to use a superior asset to realize this value. In fact, the rm might be better off selling the assets to another rm. Managers sometimes overlook this alternative. 21 This analysis assumes that the asset is equally valuable to all rms. If this assumption does not hold, the rms costs will not necessarily be bid up to the point where it makes no economic prots. Price will not be bid above the value to the next highest valued user. We discuss this issue next. 242 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 258 Part 2 Figure 8.4 II. Managerial Economics The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Managerial Economics Producer Surplus Is Captured by Superior Assets The illustration below depicts supply and demand in the marketplace. The initial price is P * . The typical rm 0 in the industry is depicted on the left and is making no economic protsprice equals long-run average cost (LRAC0). The demand for the product increases and the market price goes up to P * . The rm appears 1 to have the potential to make an economic prot, since the price is above its initial long-run average cost. Competitors and new entrants, however, will bid for the special resources (such as a talented manager or a piece of land) that would allow the rm to produce the product at an average cost below price. In equilibrium, the rms cost would have increased to the point where its long-run average cost (LRAC1) equals price. The additional producer surplus created by the increase in price (shaded area) goes to the scarce assets, which made lower-cost production possible at the initial price and output level. LRAC1 $ $ Cost per unit (in dollars) LRMC S LRAC0 P1 * P0 * D0 Qi Q 0 Q1 ** Quantity: Firm i D1 Q Q0 * Q1 * Quantity: Market As an example, it has been argued that Arco has a competitive advantage in producing retail gasoline because it owns productive oil elds in Alaska that it bought long ago at low prices.22 Potential competitors are disadvantaged, it is argued, because they have to buy oil on the open market at a higher price than Arcos current cost of extracting its oil. This argument is awed because it fails to apply the concept of opportunity costs appropriately (see Chapter 5). If Arco can sell its oil on the open market, the cost of using it internally is the revenue forgone from not selling itand that is the current spot market price of oil, not just Arcos production cost. If other companies were more efcient in rening crude oil and producing gasoline for retail customers, Arco would be more protable if it sold the oil on the open market and did not compete in the production of gasoline. The wealth of Arco shareholders is certainly greater because the company owns valuable oil rights in Alaska. But the critical question is how do managers best exploit this valuable asset. It might be best either to sell the land to another company, sell the oil from the land to another company, or rene the oil internally to produce and sell retail gasoline (or some other petroleum product). Its best course of action depends on the relative efciency of Arco versus its competitors in the extraction and distribution of oil or the production and distribution of gasoline. Arco should not extract oil or produce retail gasoline simply because it owns its oil reserves. 22 This example is taken from a strategy textbook. Similar examples are easy to nd throughout the management literature. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 8 243 The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Economics of Strategy: Creating and Capturing Value 259 Accounting Prots Selling price Sales Extraction costs Production costs Prots $4.00/gallon $4,000,000 $400,000 $2,000,000 $1,600,000 $3.60/gallon $3,600,000 $400,000 $2,000,000 $1,200,000 Economic Prots Selling price Sales Opportunity costs of oil Production costs Prots $4.00/gallon $4,000,000 $2,000,000 $2,000,000 $0 $3.60/gallon $3,600,000 $2,000,000 $2,000,000 ($400,000) Table 8.1 Arco and Opportunity Costs In this hypothetical example, Arcos cost of extracting enough oil to make a gallon of gasoline is $.40. This oil can be sold on the open market for $2.00. It costs $2.00 for rening the oil into gasoline. Arco produces 1 million gallons of gasoline a month. There are no relevant costs or revenues. The top panel shows Arcos income statement using standard accounting techniques assuming gasoline prices of either $4.00 or $3.60 per gallon. Here the cost of the oil is the $.40 per gallon production cost. Arco reports positive prots at both prices, $1,600,000 and $1,200,000, respectively. The bottom panel reproduces the income statements using the true opportunity cost of the oil (its current spot market price). In the latter case, it is better for Arco to get out of the production of gasoline and to sell the oil (or possibly the land) on the open market. It nets $1,600,000 by selling the oil compared to $1,200,000 producing gasoline. Arco is indifferent between the two alternatives when the price is $4.00 (in either case, Arco nets $1,600,000). This example emphasizes the basic point made in Chapter 5: Managers should make business decisions using opportunity costs, not accounting costs. Table 8.1 presents an illustrative example. The following assumptions are used: (1) Arcos cost of extracting enough oil to make a gallon of gasoline is $.40, (2) this oil can be sold on the open market for $2.00, (3) it costs $2.00 to rene the oil into gasoline, (4) Arco produces 1 million gallons of gasoline a month, and (5) there are no other relevant costs or revenues. The top panel shows Arcos income statement using standard accounting techniques assuming gasoline prices of either $4.00 or $3.60. Here the production cost of the oil is the $.40 per gallon. Arco reports positive prots at both prices$1,600,000 and $1,200,000, respectively. Based on these accounting prots, it might appear that it is profitable for Arco to produce retail gasoline. The bottom panel reproduces the income statements using the opportunity cost of the oil: its market price. Using opportunity costs, Arco breaks even when the price of gasoline is a dollar but loses money when the price is $3.60. In the latter case, it is more protable for Arco to get out of the production of gasoline and enter the open market to sell the oil (or possibly the landthis choice depends on Arcos efciency in extracting the oil). Arco is indifferent between the two alternatives when the price is $4.00; either way, it makes $1,600,000 over its production costs.23 This 23 Whether it produces the gasoline or sells the oil directly, it captures the difference between the $2.00 market price for oil and the $.40 production cost. At the $4.00 price, the difference between the market price of gasoline and the market price of oil is just sufcient to cover Arcos rening costs of $2.00. At the $3.60 price, the difference between the price of gasoline and oil is not sufcient to cover the rening costs. Arco makes only $1,200,000 producing gasoline, as opposed to $1,600,000 by simply selling the oil. 244 260 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics example illustrates a basic point made in Chapter 5: Managers should make business decisions using opportunity costsnot accounting costs.24 Second-Price Auctions Competition tends to take a differentiated asset to its highest-valued use, but at a price that reects its second-highest valued use. In a competitive auction, no one has an incentive to bid more than the value they place on the asset. For example, if Lena Otis is the second-highest valued user of a piece of land, she will bid only up to her value, while Jos Ricardo, the highest valued user, can obtain the asset by bidding just slightly more. This principle implies that the producer surplus captured by the winning bidder is limited to the difference between the assets rst and second-highest valued uses; the rest of the value will be reected in the price of the superior asset. Team Production Firms consist of collections of assets. For example, at Microsoft their collection consists of Bill Gates, all his senior managers and employees, the companys physical assets, brand name and other intangible assets, as well as the installed base of existing Microsoft software. Because of the interdependencies among workers and assets, the value of the inputs as a team25 sometimes can be greater than the simple sum of the values if each worker and asset were employed at its next best use across other rms. Thus, it is possible that the overall rm will be more valuable than the sum of its parts. We characterize such a rm as having team production capabilities. A rm can capture value by maintaining team production advantages only if competing rms cannot assemble comparably productive teams. If rivals can, then in a competitive marketplace the price of the product will be driven down so that the rm makes only a normal prot. How do rms create team production capabilities that are difcult to duplicate? Part of the answer is a natural consequence of differences in the past histories of rms and the now-sunk choices by managers about goods produced, markets entered, individuals hired, and capital acquired. Due to these differences, rms acquire unique combinations of assets, organizational processes, communication channels, collective learning, and so on. As environments evolve, some rms nd themselveseither through luck or foresightto be in the enviable position of having developed team capabilities that are especially productive in these new circumstances, whereas others do not. For example, Wal-Mart developed an efcient hub-and-spoke distribution network to serve its chain of successful rural stores. This system, along with its brand name and procurement advantages, also helped it compete successfully in larger urban areas. Other rms can observe the team production capabilities of successful rms. Yet, duplicating these capabilities can be difcult because, from the outside, it is frequently difcult to pinpoint the exact source of synergies within the team: Many assets are combined in the typical rm. It also can be expensive to acquire the necessary assets (contracting/transaction costs), provide training to workers, and so on. In addition, implementing signicant changes within an existing organization presents daunting challenges. That team capabilities often are expensive to duplicate is consistent with the observation that many rms enter a new business by buying an existing rm in the industry, rather than by developing the new business within the rm. 24 In Chapter 17, we shall discuss how accounting costs can play an important role in helping to control incentive conicts in organizations. 25 The word team is used in a variety of contexts in the management literature (for example, to refer specically to small work groups of people). Here, we use the term in a broader sense to refer to a combination of assets (including human, physical, and organizational assets) used to produce a particular product or service. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 245 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 261 MANAGERIAL APPLICATIONS Team Capabilities at Sharp Corporation During the 1970s, Sharp began marketing electronic calculators with liquid crystal displays. As the company developed expertise with LCD technology, Sharp began to apply it to other products, such as television sets. Through these actions, Sharp developed a set of resources and capabilities that other companies did not have. With the large growth in consumer electronics and computers during the 1980s and 1990s, potential applications for LCDs expanded rapidly. Sharp proted from this growth. Other companies could not immediately overcome Sharps competitive advantage because of time constraints and their lack of Sharps accumulated assets and experience. Companies such as Matsushita, NEC, and Canon entered the industry. Yet, Sharp was able to maintain its dominant position because of its special team capabilities. It continues as a leader in LCD technology; in 2007, Sharp negotiated a major deal with Toshiba to exchange its LCD sets for Toshiba TV chips. This advantage was not completely captured by its individual assets in their market values; individual assets would not have allowed other rms to copy Sharps advantage. Thus part of this advantage went to Sharps shareholders and was reected in its stock price. Source: D. Collis and C. Montgomery (1997), Corporate Strategy (Irwin: Chicago); K. Hall (2007), Sharp and Toshiba: Big TV Tieup, BusinessWeek.com (December 21). Team Capabilities and Organizational Architecture Part 3 of this book analyzes organizational architecture. A rms architecture consists of its assignment of decision rights and its systems for evaluating and rewarding performance. Organizational architecture is important in contributing to team capabilities. For example, a decentralized rm can respond more quickly and effectively to a new opportunity that requires rapid front-line decision making (for instance, in competitive pricing to customers) than can a rm with a more centralized decision-making structure. Alternatively, other new opportunities that require detailed coordination of activities ACADEMIC APPLICATIONS Flexible Manufacturing and Team Capabilities Economists argue that often it is appropriate to view managerial policies as a system of complementseach policy is more valuable when it is adopted along with other complementary policies. For example, a fall in the costs of exible manufacturing equipment can favor the following simultaneous changes in a rms strategy and organizational architecture: investment in more exible manufacturing equipment, increased output, more frequent product innovations, more continuous product improvements, higher levels of training, additional investment in more efcient production design procedures (computer-aided design), greater autonomy of workers and hence better use of local information, more cross-training, greater use of teams, additional screening to identify prospective employees with greater potential, and increased horizontal communication. Changing only a subset of these policies might be less productive than changing them all at once because of the complementarities: More value is created when there are reinforcing changes across this collection of policies. For example, employee training is likely to add less value if employees are not given increased autonomy to employ their new skills. Yet, making rapid changes in many elements of a rms strategy and organizational architecture is expensive. Firms in which more of these policies are already in place are more likely to take advantage of a reduction in the cost of exible manufacturing than rms without these policies in place. Because of their team capabilities, these rms are likely to generate economic prots that will not be competed away in the near term. In this example, the unexpected change in the environment is a fall in the cost of exible manufacturing technology. This change benets some rms but not others. If a different environmental change occurred, a different group of rms might benet. Source: P. Milgrom and J. Roberts (1995), Complementarities and Fit: Strategy, Structure, and Organizational Change in Manufacturing, Journal of Accounting and Economics 19, 179208. 246 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 262 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Leaving New York City for the Farmlands of Illinois You have worked as an investment banker in New York and Hong Kong for the last ve years. You are tired of living in populated cities and yearn for a more peaceful environment. The Wall Street Journal contained a story about the federal governments actions to increase corn-based ethanol as a substitute for gasoline in the United States. According to reports, new laws and regulations will result in the construction of hundreds of new ethanol factories over the next decade. Demand for corn to fuel the plants is expected to soar. Based on a Web search, you found 80 acres of prime farmland for sale in Illinois at a price of $10,000/acre. You have saved enough from your past bonuses to purchase the land. You are tempted to quit your job and purchase the land to farm corn. Several colleagues have told you that it sounds like a good idea. They say that not only would it allow you to move to a less populated area, but it is a great business opportunity with little risk. Surely the price of farmland will increase dramatically over the next few years as the ethanol plants begin operation and the demand for corn skyrockets. Good farmland is limited, and you might reasonably expect its price to double or even triple over the next few years. You are smart and should be able to learn the farm business very quickly. In addition, you will have a cost advantage over farmers who wait to buy land at much higher prices. If you decide that you do not want to be a farmer, you can always sell the land at a huge prot. Do you think your colleagues are giving you good advice? Explain. across several business units might favor the rm with the more centralized structure. As we shall discuss, it often is hard to copy another rms architecture because multiple systems have to be changed in a coordinated manner. Implementing this type of major change within an organization can be surprisingly difcult. A Partial Explanation for Wal-Marts Success Sam Waltons initial strategy in the early 1960s was to establish stores in small towns (populations under 25,000) in rural Arkansas, Missouri, and Oklahoma. His intent was to place good-sized stores in little one-horse towns, which everybody else was ignoring. As he increased the number of stores, he designed an extremely efcient distribution systema hub-and-spoke system with regional distribution centers. He also cultivated key vendor relationships, a distinctive human-resource management system, and a nonunionized workforce. As his network of stores increased, he opened additional stores in nearby states, eventually expanding throughout the country and internationally. As Wal-Mart continued to expand, its success became evident. Would-be competitors envied this success; yet, they did not have the capability to mimic Wal-Marts strategy. First, it did not make economic sense for companies to construct competing stores in the small rural communities where Wal-Mart already operated. Surely, the resulting competition with Wal-Mart would drive prices down in these one-horse towns. Wal-Marts rst-mover advantage created an entry barrier and market power. Without a sufcient number of existing stores, it also did not make economic sense to copy Wal-Marts distribution system in the rural Southeast. As Wal-Mart gained experience, it developed organizational processes and resources (team capabilities) that provided potential advantages over would-be competitors as it expanded into new geographic areas, such as the Pacic Northwest. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 247 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 263 MANAGERIAL APPLICATIONS Economic Prots without Market PowerA Summary of the Key Concepts Chapter 6 characterized competitive output markets in terms of four basic conditions: (1) many buyers and sellers, (2) product homogeneity, (3) rapid dissemination of accurate information at low cost, and (4) free entry into and exit from the product market. In this setting, rms have no market power and essentially are price takers in the output market: Firm demand curves are horizontal at the market price. The marginal rm sells the product at average cost and makes no abnormal prot. In this chapter, we discuss how it is possible for some rms in a competitive market to make economic prots (produce at an average cost below the market price). Economists refer to these rms as being inframarginal. To be inframarginal, two conditions must be met: Rivals cannot imitate the inframarginal rm and assemble teams of assets that produce the product at the same low costat least in the near term. If competitors can erode cost advantages through imitation, competition will drive the price down and eliminate above-normal prots. The full value of the rms superior productivity cannot be captured by selling its assets to other rmsunless they are sold together. Otherwise the opportunity cost of production will be the same as for their rivals (even if they cant assemble equally productive teams). The fact that rms have different histories and paths can help explain the existence of team capabilities and inframarginal rms. Nonetheless, staying inframarginal is not easy. Potential competitors have strong incentives to discover ways to imitate successful rms or otherwise counteract their advantages. Also, to the extent that an advantage is based on an identiable asset, such as a talented manager or a prime location, the opportunity cost of the resource is likely to rise from competitive pressures in factor markets. Thus, competitive pressures in both the product and factor markets make sustaining economic prots difcult. Wal-Mart could have sold some of its resources, such as individual stores, yet potential buyers would have limited incentives to bid up the price of these assets unless they could have purchased them all together. For example, the value of an individual store is lower when it is not coupled with Wal-Marts distribution system.26 Thus, although Wal-Mart faced an opportunity cost from not selling its individual assets piecemeal to other rms, the value of keeping these assets together within the same rm was likely to be far higher than this opportunity cost. If Wal-Mart were to construct an income statement based on its opportunity costs (as we did for Arco in the bottom panel of Table 8.1), it would show a positive economic prot.27 Even though Wal-Mart has had a competitive advantage in rural America, its advantage has been smaller in urban centers. Urban areas can support more than one discount storethus promoting entry by Target, Kmart, and others. There also is more competition from nondiscount stores. Competing rms have copied many of Wal-Marts innovations in distribution, vendor relationships, and so on. Thus, in urban areas Wal-Mart has had both less market power and a smaller productivity advantage. Wal-Mart also has had problems at some of its international locations, such as Brazil. Internationally, 26 A reader might wonder why Wal-Mart couldnt sell the store to a buyer at a high price and contract with the buyer to allow it to use Wal-Marts distribution system. This issue is addressed in Chapter 19. That analysis suggests that given the nature of the assets, Wal-Mart is better off owning them jointly. 27 Recall that the opportunity costs of the individual resources are their next highest valued use outside of their current use. 248 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 264 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Nomura Securities Company: It Is Not Easy to Remake a Business Japans Nomura Securities Companywhose roots go back to the nineteenth century rice exchanges of Osakawas one of the worlds most protable securities rms in the 1980s. In 1997 following a serious scandal involving payoffs to racketeers, the company almost collapsed. The company also was hurt because it strongly recommended a set of stocks to customers that then fell signicantly in value. The new president, Junichi Ujiie, vowed to remake the company. He wanted to deemphasize the selling of individual stocks to customers and focus more on asset gathering and portfolio advice. His objective was to become more like Merrill Lynch or Morgan Stanley Dean Witter & Company. By late 1999, Ujiie had failed to turn the company around. Indeed Nomura continued to lose out to smaller competitors in obtaining new business. Like many large companies, Nomura had trouble remaking itself to regain lost success and protability. Source: B. Spindle (1999), Nomura Restructuring Falters; Can Mr. Ujiie Still Remake the Firm? The Wall Street Journal (September 3), A1. Wal-Marts advantages from superior distribution and vendor relations are arguably smaller; also its brand name is less well known than in the United States. All Good Things Must End The business environment is constantly evolving with new technological innovations, changes in consumer tastes, new business concepts, new rms, and other developments. Given these changes, it is unlikely that any competitive advantage will last forever, unless the rm can nd a succession of new value-increasing strategies. Just as the elements MANAGERIAL APPLICATIONS Wal-Marts Strategy Proves Timely during the 2007 Holiday Season Wal-Marts experience in the fall of 2007 illustrates how differences in historic paths can give rms a competitive advantage of disadvantage as the economic environment changes. The 2007 holiday shopping season was the weakest in ve years. Consumer spending fell as consumers worried about oil prices that approached $100/barrel and the prospect of economic recession. Relative to the previous year, same-store sales fell during the holiday season at Target (down 5 percent), Abercrombie & Fitch (2 percent), Gap (6 percent), Macys (7.4 percent), Nordstrom (4 percent), JC Penney (7.5 percent), and Kohls (11.4 percent). Wal-Mart was one of the few retailers with strong performance, experiencing a 2.6 percent increase in same-store sales during the holiday season. Throughout November and December of 2007, Wal-Mart opened stores very early in the morning on multiple occasions enticing shoppers with big discounts on products such as high-denition DVRs. Thousands of consumers ooded Wal-Mart stores in search of bargains. Its emphasis on low prices attracted consumers throughout the holiday season. Other retail chains, such as Target, scrambled to adapt and offered similar promotions to attract price-focused customers. These efforts, however, were not as successful as Wal-Marts. Wal-Mart was known among consumers for its historic focus on everyday low prices. The historic business strategies of most of the other retailers focused on other dimensions. For example, Target had developed a reputation for good design, while Nordstrom had developed a reputation for quality products and service. Historic strategies and reputations made it difcult for many of the chains to convince consumers that they were now focusing on low prices. Costco, which has historically focused on bulk merchandise and low prices, was one of the few other success stories with a same-store sales increase of 7 percent (4 percent excluding gasoline sales). Source: Michael Barbaro (2008), Wal-Marts Strategy Proved a Timely One, nytimes.com (January 11). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 8 249 The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Economics of Strategy: Creating and Capturing Value 265 Changing Fortunes The fortunes of rms change over time. Due to competitive pressures, the top rms in one period are often not the top rms in subsequent periods. Below is a listing of the top 10 rms in terms of market value in 1970 and 2008. The lists are different. Eastman Kodak, IBM Sears, and Xerox fell off the list between 1970 and 2008. New entrants to the list by 2008 include Citigroup, Conoco-Phillips, Ford, and Wal-Mart. (Standard Oil of NJ changed its name to Exxon in 1972 and merged with Mobil in 1999; Gulf Oil merged with Standard Oil of California in 1984 and became Chevron. It then merged with Texaco in 2001.) Rank 1970 2008 1 2 3 4 5 6 7 8 9 10 IBM AT&T General Motors Standard Oil of NJ Eastman Kodak Sears Roebuck Texaco General Electric Xerox Gulf Oil Wal-Mart Stores Exxon-Mobil Chevron General Motors Conoco-Phillips General Electric Ford Motor Citigroup Bank of America AT&T erode a mountain, persistent competition erodes rm prots over time. While entry may be limited, outside rms have strong incentives to devise methods of capturing prots from successful rms. Ultimately, one of these methods is likely to work. If one compares todays top rms with those of, say, 50 years ago, the lists are quite different. For example, todays top rms like Wal-Mart, Microsoft, and Intel did not exist 50 years ago, whereas many of the top rms from yesteryear have become less successful or gone out of existence. Consistent with this view, Wal-Marts growth slowed signicantly during the rst part of the 1990s. Although its performance improved during the late 1990s, economic theory and experience suggested that continued superior performance should not be expected over the long run. Polaroids Success and Ultimate Failure to Capture Value From 1948, when Polaroid introduced its rst instant camera, until 1972, when it unveiled its SX-70, the company continually improved its almost magical product. The rst Polaroid camera produced sepia-toned prints, but over the years the company developed the ability to produce color photos that materialized right before ones eyes. In 1972, Polaroids stock price was 90 times earnings, propelling the company into the Nifty Fifty (the top 50 companies in the Fortune 500). But the SX-70 was followed by a series of ops, including Polavision, a moving version of instant photography that was inferior to video. Moreover, the development of digital imaging provided an alternative method of viewing pictures immediately and precipitated Polaroids Chapter 11 bankruptcy ling in October 2001. It stopped making cameras in 2007 and will stop making lm after 2009. The history of Polaroid illustrates two important points. First, successful rms develop new wetware that leads to innovative products, which satisfy customers demands. Second, competitionoften arising from new technologieserodes the protability of companies that, on a continuing basis, fail to innovate successfully. Source: D. Whitford (2001), Polaroid, R.I.P., Fortune (November 12), 44; Wikipedia (2008), Polaroid Corporation. 250 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 266 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics Economics of Diversication Although some rms, such as Wrigleys gum, concentrate on a single major business, most large rms engage in numerous businesses: They are diversied. This raises the question: Is corporate diversication a productive strategy to create and capture value? Economies of scope are the primary reason that diversication might enhance value. In addition, combining businesses within the same rm can promote complementary products. But diversication also entails costs. Value maximization requires that rms consider both the benets and the costs of diversication. In this section, we discuss these costs and benets and analyze the circumstances under which diversication is most likely to create value. We also examine who is most likely to capture this value. Benets of Diversication Economies of Scope As discussed in Chapter 6, economies of scope exist when one rm could produce multiple products at lower cost than separate rms could produce the products. Economies of scope might occur anywhere along the vertical chain of production.28 Consider the following examples: A diversied rm buys inputs at a discount reecting its higher volume from using the same inputs in producing different products; a rm economizes on transportation costs because it delivers different products to the same customer; a rm uses salespeople more efciently because they are able to offer customers an array of products; and a rm economizes on otation costs because it raises capital to fund several businesses at once. Sometimes, producing one product unavoidably also produces others because of a jointness in the production process. For instance, if a rm produces beef by slaughtering cows, it also frequently makes economic sense to produce other products with the bones and hides. Combining activities inside the same rm is not the only way to achieve these types of economies. There are alternative ways to organize that also accomplish this objective. For example, an independent wholesaler might offer retailers an array of products produced by separate manufacturers, centralize the billing of retailers, and provide a more efcient inventory and distribution system. Which method of organization is best depends on contracting costs. For example, negotiating and enforcing contracts among independent rms can be expensive, as can organizing activities within the same rm (see below). Sometimes integrating activities within the same rm is less expensive than MANAGERIAL APPLICATIONS Wal-Mart Diversies into the Traditional Grocery Store Business In 1998, Wal-Mart opened three experimental 40,000-square-foot grocery stores in Arkansas. The potential benets of this diversication derive from economies of scope. Wal-Mart has a quite efcient distribution system that also can be used to stock the grocery stores. It has the potential to leverage its relationships with key vendors, which currently serve its discount stores and supercenters. Wal-Marts management has signicant experience in managing stores in a related business. This experience is likely to be helpful in choosing store locations, establishing and running management systems, and so on. 28 The vertical chain of production consists of the various steps in taking raw materials and transforming them into consumer products. These steps include research and development, purchasing, production, outbound logistics, marketing, sales, service, support activities, and so on. See Chapter 19 for a more detailed discussion of this chain of activities. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 251 267 MANAGERIAL APPLICATIONS McFocus at McDonalds McDonalds decided to shed at least some of its sideline ventures into pizza, fast-casual Mexican food, and takeout dining and concentrate on revitalizing its core hamburger business. McDonalds had gotten into Chipotle Mexican Grill, Donatos Pizza, and Boston Market to offer its best franchisees expansion opportunities and to nd new growth vehicles. But neither goal has been achieved. We cannot allow anything to distract our owner-operators, suppliers, and company people from maximizing the full potential of Brand McDonalds, said CEO Jim Cantalupo. Source: R. Gibson (2003), McDonalds May Exit from Ventures, The Wall Street Journal (November 10). contracting among rms. In this case, diversication creates value. We discuss the tradeoffs in choosing among organizational forms in greater detail in Part 3 of this book. Promoting Complements Another potential reason for diversication is to promote the supply of complementary products. For example, Ford and General Motors entered the consumer credit business in 1919 and 1959, respectively.29 One potential benet of auto manufacturers entering the nancing business would be to increase the demand for automobiles by offering low-interestrate loans to customers (recall Figure 8.1).30 Contracting/transaction costs also were reduced because the consumer could ll out the loan application while purchasing the automobile. Today, communication systems, computers, and sophisticated credit-scoring systems make it easy to apply for a car loan from a lender over the phone or the Internet. This was true in neither 1919 nor 1959. Again there may have been other ways of organizing to achieve these economies (for example, the automobile companies could have contracted to process loan applications for a bank). However, the costs of these organizational alternatives evidently were higher than combining the businesses within one company. As another illustration, recall our example of PCs and printers. The companies beneted by coordinating the pricing of their products; another way of achieving this cooperation would be to merge the two rms. Costs of Diversication Although diversication has potential benets, it also has potential costs. As rms grow, they often become bureaucratic and more expensive to manage. As we shall discuss in more detail in Part 3 of this book, it is difcult to devise compensation plans that motivate managers in large companies to behave like owners of smaller companies. Owners have a natural incentive to work hard and increase the value of their rms because they keep the prots of the rm, whereas salaried managers do not. If diversication occurs through the merger of two rms (as is often the case), it also can be quite expensive to develop common personnel, communication, information, and operating systems. 29 See G. Stigler (1966), The Theory of Price, 3rd edition (Macmillan Press: London). Offering low-interest-rate loans beneted Ford and GM only if their cost of providing credit to customers was lower than the prevailing market price for credit (for example, the loan market was noncompetitive or the auto company valued repossessed cars from bad loan customers more than other lenders). Otherwise, the companies could have obtained the same increase in sales by simply reducing the price of automobiles by the amount of the credit subsidy. (However, note that the credit subsidy might make it easier to price-discriminate between cash and credit customers. Low-interest rates might be used to hide price concessions to credit customers.) See S. Mian and C. Smith (1992), Accounts Receivable Management Policy: Theory and Evidence, Journal of Finance 47, 169200. 30 252 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 268 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Proposed Megamerger Collapses in the Drug Industry The planned $35 billion merger between American Home Products and Monsanto Company collapsed in late 1998. Monsantos primary focus had been on biotechnology and drugs. For example, in the 1990s its strategy focused on being the rst to cash in on patents for crops that make drugs and repel bugs. American Home Products brought a well-developed sales force to the deal that could have moved Monsantos crop creations into drugstores and supermarkets. This merger of related companies had the potential to generate economies of scope. The merger collapsed because of the costs of integrating the two companies. The two sides disagreed on many items ranging from which company should be assigned the corporate headquarters to how much compensation should be given to top executives. Especially problematic was merging the two management styles of the CEOs who planned to share the title of co-CEO. One of the CEOs was noted for tightsted cost cutting, whereas the other was known for more liberal spending and investment. Upon the announcement of the merger collapse, Monsantos stock closed at $37 per share, down $13.38 per share, or 27 percent. American Home Products stock fell $5, or 10 percent. Apparently, the stock market thought that the merger would have created value and was disappointed when it collapsed. Source: T. Burton and E. Tanouye (1998), Another Drug Industry Megamerger Goes Bust, The Wall Street Journal (October 14), B1. Management Implications A Faulty Reason to Diversify Some managers diversify to reduce earnings volatility. For example, the former CEO of Goodyear Tire justied diversication into the oil business because it was countercyclical to the tire industry. He reasoned that when gas prices were low, the company would do well in tires, since people would be driving more. When gas prices were high, the company might do poorly in tires but would do well in the oil business. Overall, earnings would be less volatile. It is true that diversication can reduce earnings volatility. The problem with using this as a justication for diversication is that this reduction in volatility need not increase a rms value. Shareholders (the owners of public companies) can diversify within their own investment portfolios at low cost. For example, it is easy for investors to purchase shares of both a tire company and an oil company. Through this diversication investors reduce return volatility on their overall portfolios. There is no reason for investors to pay a premium for a company simply to reduce return volatility, since they can achieve this same objective by purchasing stock in the two separate companies.31 At the same time the costs of integrating diverse businesses within the same rm can be signicant. When Does Diversication Create Value? Related diversication occurs when the businesses serve common markets or use related technologies. For example, Disney Corporation operates theme parks, hotels, retail shops, and television stations. In each of these industries, Disney concentrates on employing its strong brand name to market family-oriented products. Their brand name 31 Although managers generally should be skeptical of the proposition that reductions in earnings volatility increase value, there are cases where the proposition is true, for example, if reducing earnings volatility reduces expected taxes or the costs associated with nancial distress. See C. Smith and R. Stulz (1985), The Determinants of Firms Hedging Policies, Journal of Financial and Quantitative Analysis 20, 391405. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 253 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 269 MANAGERIAL APPLICATIONS Philips Electronics A common view since at least the 1990s is that unrelated diversication can destroy value. In turn, many companies have streamlined their business portfolios to focus on core businesses. The decision to reduce the amount of diversication often has been motivated by pressures from stockholders and threats of corporate takeovers. If managers insist on remaining overly diversied, they can lose their jobs. Philips Electronics NV is based in the Netherlands and is the worlds biggest consumer-electronics company outside of Asia. In contrast to many other companies, Philips has remained stubbornly diverse. Philips produces everything from toasters to televisions, software to computer chips, electric toothbrushes to trafc management systemsit even produces 13 different models of rice cookers. During the late 1990s, the company performed quite poorly; consequently, shareholders of the company suffered. For example, income from continuing operations fell by 50 percent in 1999, and return on equity was far below the industry average. Some analysts argued that Philips should sell off many of its businesses. Yet, they have failed to do so, in part, because of the lack of an active market for corporate control in Europe. Moreover, Dutch managers receive little incentive compensation tied to stock-price performance. Robert Lyon, president of Institutional Capital, which owns about 1.4 percent of Philips stock, has stated, If Philips were an American company, it would have been acquired or busted up long ago. Source: J. Flynn and G. Zachary (1999), Philips, an Innovator in Electronics, Proves Resistant to Change, The Wall Street Journal (September 10), A1. reduces transaction costs to consumers in their search for quality products appropriate for children. Net benets are likely to be greater in related rather than unrelated diversication. If there is little interaction on either the production or demand side of the businesses, it is hard to envision where economies of scope might arise and by denition the two businesses would not produce complementary products.32 An example of unrelated diversication is the decision by KinderCare Learning Centers, a day-care management rm, to enter the savings and loan business; it subsequently entered bankruptcy. Value is increased only when the benets of diversication are larger than the costs. There is a signicant body of research on the value consequences of diversication.33 For the most part, diversied rms have not performed well. Indeed, many of the large conglomerates during the 1980s, such as ITT and Tenneco, increased their values by divesting units (through sales and spinoffs) and refocusing on a more narrow line of businesses. Research documents that 55.7 percent of exchange-listed rms had a single business segment in 1988, compared to 38.1 percent in 1979. More importantly, the research shows that this increase in focus was associated with higher stock returns.34 Diversication has been most effective in the case of related diversication where there are potential economies of scope and opportunities to promote complements.35 32 Some managers assert that their management skills are so good that they can create value in any business. Although this may be true, it is obvious that a given management team has limited capacity in the number of businesses the team can manage. Again, one might expect value to be maximized if the businesses are related. 33 See J. Barney (1997), Gaining and Sustaining a Competitive Advantage (Addison-Wesley: Reading, MA), 388389, for a summary of this research. There are, however, examples of a few rms that appear to have created value through unrelated diversication. 34 R. Comment and G. Jarrell (1995), Corporate Focus and Stock Returns, Journal of Financial Economics 37, 6787. This issue of the Journal of Financial Economics contains several other interesting articles on corporate focus. 35 There are some exceptions to this general nding. For example, General Electric is quite diversied and runs a variety of rather unrelated businesses. Nonetheless, it has performed extremely well. 254 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 270 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Diversication Problems at Xerox During the 1980s Xerox diversied into nancial services. The company had gained experience in nance through the leasing of its copy machines, and management believed that by acquiring rms in the nancial sector, it could leverage this experience. In late 1982, Xerox purchased a property/casualty insurance company, Crum and Foster, for $1.6 billion. This acquisition was followed by other acquisitions in life insurance, real estate, and investment banking. With minor exceptions, this diversication turned out to be a nancial drain on Xerox. The economies of scope were not as great as had been hoped and problems in the nancial businesses distracted top management from Xeroxs main businesscopiers. In the 1990s Xerox repositioned itself as the document company and sold its nancial businesses to other companies. Who Captures the Gains from Diversication? Even if diversication creates value, the owners of the diversied rm do not always capture this value. Again, it depends on whether the rm brings some special resource or team capability to the transaction. If it does not, the value is likely to be captured by another party due to competition. Consider the potential benets of merging a television cable company with a long-distance telephone company. Suppose that there is one cable company with the rights to operate in a specic area, whereas there are several longdistance telephone companies. Competitive bidding among the telephone companies would imply that the phone company that valued it the most would get the cable company at a price equal to its value to the second-place phone company. This process would give most of the value gains to the owners of the cable company: They own the unique resource. Consistent with this example, substantial research documents the fact that target rms obtain most of the gains in corporate takeovers.36 The average return to target rm shareholders around successful acquisitions over the past two decades has been in the neighborhood of 30 percent, whereas the average gain to the shareholders of bidding rms has been virtually zero. Strategy Formulation Developing and implementing strategies that increase a rms value require an understanding of both the internal resources and capabilities of the rm and its external business environment.37 Figure 8.5 displays these two factors. Understanding Resources and Capabilities An important rst step in strategy development is understanding the rms resources and capabilities. By resources and capabilities, we mean the rms physical, human, and organizational capital. This includes the rms hardware, software, and wetware; specically, those activities which the rm can do better than other rmsits team production capabilities. It also is important to understand the opportunity costs of using these assets within the rm. For example, how much would they be worth if they were sold to other rms? 36 See G. Jarrell, J. Brickley, and J. Netter (1988), The Market for Corporate Control: The Empirical Evidence since 1980, Journal of Economic Perspectives 2, 4968, for a summary of this research. 37 A variety of approaches to strategic management are discussed in the strategy literature. The framework outlined in this section is most consistent with the strengths, weaknesses, opportunities, and threats approach. See J. Barney (1997), Gaining and Sustaining a Competitive Advantage (Addison-Wesley: Reading, MA). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 8 Figure 8.5 255 The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Economics of Strategy: Creating and Capturing Value 271 Framework for Strategic Planning Members of the management team should have a good understanding of their rms internal resources and capabilities before they adopt their strategies. They also should understand their external business environment. The objective is to use this knowledge to discover ways of creating and capturing value. INTERNAL RESOURCES AND CAPABILITIES BUSINESS ENVIRONMENT Markets Input Output Physical Capital Human Capital Organizational Capital Technology Production Information Communication Government Regulation STRATEGY Understanding the Environment As depicted in Figure 8.5, important factors in the business environment include the rms markets (both input and output), technology (production, information, and communications), and government regulation. Effective managers monitor the business environment to keep abreast of new developments in each of these areas. Referring to Figure 8.1, managers must understand the business environment to identify opportunities for value creation. For example, what technological opportunities exist to reduce costs? It would be a poor decision to invest in an expensive technology to lower costs, if forecasted technological innovations are expected to make the investment MANAGERIAL APPLICATIONS Faulty Analysis at Kodak During the 1990s, Kodak executives decided to invest heavily to produce writable CD-ROMs. They knew they had the internal capability to produce this product. They, however, failed to consider the threats in the external business environment imposed by other potential competitors. Just because Kodak was a large producer of CD-ROMs did not necessarily mean that it had team capabilities to produce them less expensively than competitors. To quote CEO, George Fisher: [With regard to] the loss in writable CD-ROMs, I think we screwed up. A highly protable business for us, No. 1 market share in the world, and I think we were wishing that it would stay that way. We let it get in the way of making objective decisions. We should have known that prices would fall as manufacturers worldwide ramped up production. Source: Rochester Democrat and Chronicle (November 2, 1997). 256 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 272 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics obsolete soon. What opportunities exist to reduce consumer or producer transaction costs? Managers must have a detailed understanding of how transactions take place in the marketplace to discover ways of reducing transaction costs. What opportunities exist for improving consumer products? Are there ways of promoting complementary products? Are there potential industries in which the rm could leverage its resources and capabilities? Are there opportunities to create value through cooperating with other rms? An understanding of the rms environment also is important for identifying opportunities for, as well as threats to, capturing value. For example, what opportunities are there to block entry or the development of substitutes? What threats loom from potential substitutes, buyer power, supplier power, and industry rivalry? Combining Environmental and Internal Analyses Most resources and capabilities are nitechoices must be made. For example, rms have limited production capacity and human resources. Thus, they face trade-offs in deciding how to use these resources and in deciding whether it is worth the investment to supplement them. Also managers must decide whether it is best to use their marketable assets within the rm or whether to sell them to other rms. These strategic choices require managers to combine environmental and internal analyses: They must understand both their internal strengths and weaknesses, as well as the threats from and opportunities within their business environment. Typically, strategies do not remain constant, but evolve through time. It usually is important for managers to consider how other economic agents in the business environment will react to their strategic decisions (for example, rival rms, suppliers, and buyers). In Chapter 9 we examine this issue in detail. In deciding how best to use special resources and capabilities, it is important to be forward-looking. Sometimes it is more protable to invest in complementary resources and capabilities that allow the rm to exploit its unique position better in the future. For example, it might have been protable for Sharp to invest in acquiring complementary knowledge about computers and other technologies to supplement the rms skills in LCDs. Strategy and Organizational Architecture To be successful, a company must have not only a good strategy, but also an appropriate architecture. Ultimately, both strategy and organizational architecture are key determinants of value. Part 3 of the book presumes that the strategy of a rm has been MANAGERIAL APPLICATIONS Strategy at Microsoft Microsofts market share has given it an advantage over competitors in the market for operating software for PCs. Consumers are reluctant to purchase competing operating systems because they want to be compatible with other users. Also, the many users of Microsoft products would have to learn new systems if they switched. Microsoft realizes this strength, but also realizes that the opportunities are limited in selling to rst-time PC buyers. Microsofts strategy, therefore, involves producing new upgrades of system software (for example, Windows 2000). These upgrades encourage existing users to buy new software. Microsoft also promotes hardware upgrades (to faster PCs) by developing new software that runs better on machines with faster processors. New hardware sales typically are associated with additional software sales. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 257 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 273 MANAGERIAL APPLICATIONS A Retail Success Story and Luck IKEA, one of few retailers that has ourished on foreign soil, had 233 stores in 34 countries selling $14.8 billion a year worth of furniture in 2005. Its strategy is that well-designed furniture can be inexpensive without being ugly. IKEA entered international commerce by chance. Started in Sweden in the 1960s, local Swedish retailers pressured Swedish manufacturers to cut off supplies to the upstart IKEA. In response, IKEA turned to Poland for supplies and was surprised to nd that it could buy well-crafted furniture more cheaply. This transformed IKEA into thinking about foreign suppliers and retail markets. The rm now has spread to Asia and North America. IKEA is a very patient rm. While trying to gure out how to please U.S. shoppers, IKEA absorbed losses for years. They had to learn that Americans expected jumbo beds and didnt think in centimeters. Source: J. Hagerty (1999), How to Assemble a Retail Success Story, The Wall Street Journal (September 9), A24; K. Capell (2006), IKEAs New Plan for Japan, BusinessWeek (April 26). determined and presents a detailed analysis of how a rms environment and strategy inuence its organizational architecture. This approach is reasonable because the organizational design often is based on what the rm wants to do strategically. For example, if a rm adopts a strategy to react quickly to customer demands, it will probably have to design a decentralized decision system with accompanying performance evaluation and reward systems to motivate productive decisions. Figure 8.5, however, emphasizes that the effects are not all in one direction. A rms architecture also can inuence its strategyit is an important part of the rms internal resources. For example, if the rm has a well-functioning, decentralized decision-making system, it is more likely to enter markets for which this type of organizational design is well suited. ACADEMIC APPLICATIONS Contemporary Approach to Strategy A common approach to strategy in the 1990s has been to focus on core competencies. This popular terms meaning is closely related to the economic concept of team capabilities. Both terms are used to describe the ability of a rm to outperform other rms in some activity because of the rms uniqueness as a team of productive resources. Core competencies are typically taken to mean team capabilities that can be extended or leveraged across different products and markets. A common approach begins with identifying what, if any, are the rms core competencies. The next step is to identify opportunities to leverage these competencies (for example, by entering a new industry). Sometimes it is important to invest in complementary assets and training to exploit the rms core competencies effectively in the future. An example is the Sony Corporation. Through its experience in producing solid-state AM radios, the company acquired a capability (competence) in electronic miniaturization. Arguably team Sony could create more value in producing miniaturized products than competing teams. Upon identifying this competency, Sony leveraged it by producing a wide range of electronic productstelevisions, tape recorders, stereo equipment, and so on. In developing these new products, Sony acquired additional assets to supplement its team capabilities in electronic miniaturization. Some consultants argue that most rms have core competencies. They simply need to identify them and decide how best to leverage them. A contrasting view presented in this chapter is that some rms have team capabilities at a point in time, whereas others do not. Trying to leverage core competencies when they do not really exist is a waste of resources. Source: G. Hamel and C. Prahalad (1994), Competing for the Future (HBS Press: Boston). 258 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 274 Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Walmart.com Conventional wisdom holds that to succeed in electronic commerce, you have to get in early. But in late 1999, Wal-Mart decided to challenge that most sacred of Web rules. After several years of tinkering with its Web site, watching while others broke new Internet ground, the retailing giant was ready to ex some cyber muscle. Up to that point, Walmart.com had realized modest success online, ranking 43rd among Internet shopping sites. It trailed Web pioneers like eBay and Buy.com. In May 1999 Amazon.com greeted almost 10 million online visitors; Walmart.com saw only 801,000. For 1999, analysts expected Wal-Marts e-commerce activities to produce sales of less than $50 million out of the companys total sales of $157 billion. Wal-Mart faced increasing direct competition from Amazon.com. In July 1999, Amazon announced its expansion from books, music, and videos into toys and consumer electronics. Wal-Mart already was a powerhouse in these product categories through its traditional stores. It announced that it would offer products from all 25 categories carried in a typical Wal-Mart discount store. Moreover, it expected to offer a broader array of higher-priced items than its traditional stores for instance, DVD players and digital cameras. It also enabled customers to return products ordered online to any of Wal-Marts 2,451 U.S. discount stores. Like Amazon, it planned to provide tailored online specials to match the shopping habits of its repeat customers. The company announced plans to expand its online store offerings before the end of 1999 to match more closely the breadth of its traditional outlets. To facilitate this expansion, Wal-Mart penned deals with Fingerhut Business Services and Books-a-Million. Both had expertise in distributing individual orders directly to customers homesquite a different set of skills from bulk shipments, which had been Wal-Marts forte. Demographic shifts occurring in cyberspace offered the potential to help Wal-Mart. Back in 1999 Jupiter Communication projected e-retailing to grow from approximately $12 billion in 1999 to an estimated $41 billion in 2002. Much of this expansion would be concentrated in Wal-Marts existing lower- and middle-class customer base. Jupiter analyst Kenneth R. Gasser noted, Internet users are increasingly coming to resemble the population at large. By 2005 Walmart.com had logged $1 billion of Internet sales. However, the worlds largest retailer only ranked about 13th in Internet sales while Amazon.com had sales of over $10 billion. About 500 million total visitors clicked on Walmart.com in 2005, and the company predicted this to increase to 700 million in 2006. But, its online sales still only accounted for about 1 percent of Wal-Marts annual sales. In January 2007, Walmart.com launched Soundcheck, an original series of musical performances that feature punk pop and rock bands to increase its digital music offerings. In March 2007, Walmart.com announced Site to Store where Walmart.com shoppers can purchase online and have orders delivered to their local store for free. By July Site to Store sales more than doubled since its March rollout, and about 90 percent of participating stores had at least one Site to Store order within the rst 48 hours of service activation. In January 2008, only 11 months after initiating its movie download service, Walmart.com quietly dropped this service because Hewlett-Packard Co. stopped providing the application that allowed shoppers to purchase and download videos such as movies and TV shows. Placing yourself back in 1999, answer the following questions in a well-developed discussion: 1. What is the impact of Walmart.com on customer-borne transaction costs? 2. Do you think that Walmart.com is likely to create additional value? 3. Is it likely that Wal-Mart will capture any value created by Walmart.com? 4. Should Wal-Mart have pursued e-commerce more aggressively sooner? 5. What do you think the potential impact of Walmart.com will be on the companys efforts to expand internationally? BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 8 6. In November 2007, Walmart.com announced that it wants to be the most visited, most valued online retail site. Suppose you were hired by an outside consulting rm to evaluate Walmart.coms potential to achieve this goal. Write a short report listing those factors that will enhance and those factors that will impede Walmart.coms ability to achieve 259 The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Economics of Strategy: Creating and Capturing Value 275 the goal of the most visited, most valued online retail site. Sources: W. Zellner (1999), When Wal-Mart Flexes Its Cybermuscles, BusinessWeek (July 26), 82; S. Murphy (2007), The Walmart.com Way, Chain Store Age (July), 80; www.internetretailer.com; and http://today.msnbc.msn.com/id/21940867. Can All Firms Capture Value? Consultants often suggest that any rm can develop a strategy that produces economic prots, even if it does not begin with unique resources or team capabilities. What is necessary is that the manager be a visionary and make sound investments in developing the skills and capabilities to compete successfully in the future. (Of course, for a fee these gurus would be happy to help the manager accomplish this objective.) Basic economics suggests that these consultants are wrong. Even if a manager is exceptional at predicting the future and seeing what resources and capabilities are important, abnormal prots will not be earned on a systematic basis so long as there are other managers who follow the same strategy.38 Competition will bid up prices of the required resources in the factor markets and bid down prices in the output markets. As in any competitive market in equilibrium, the expected outcome is a normal rate of return not sustained abnormal performance. Although some managers may be better at systematically predicting the future, luck plays a key role in the acquisition of valuable resources. Under this view, many rms invest in projects that before the fact are expected to yield normal returns. These rms take somewhat different paths in their investment strategies, development of internal processes, hiring decisions, and so on. As time passes and environments change, some rms nd themselves with superior resources and team capabilities, whereas others do not. It takes managerial talent to identify whether or not the rm has valuable resources and capabilities and to decide how best to use them to maximize returns. Managers easily might think they have unique resources and capabilities when in fact they do not.39 Firms without special resources or capabilities at best can expect to earn a normal rate of return on their investment. Firms whose resources and capabilities do not t the changed environment should not expect to earn a normal rate of return and might prepare to go out of business.40 In a competitive environment, it can take signicant managerial effort and talent (devoted to decisions like choosing products as well as producing and marketing them efciently) just to earn a normal return. 38 And even if no other manager were to have this skill, it still would not be clear that the rm would capture the gains: It is the manager who owns this unique resource. 39 See M. Ryall (1998), When Competencies Are Not Core: Self-Conrming Theories and the Destruction of Firm Value, working paper, University of Rochester. 40 H. Kim and J. Schatzberg (1987), Voluntary Corporate Liquidations, Journal of Financial Economics 19, 311328. This article documents that, on average, shareholders incur signicant nancial gains around the announcements of voluntary liquidations. Apparently, the stock market is unsure that managers will make the tough decision to liquidate poorly performing rms and is happily surprised when liquidation announcements are made. 260 276 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 Summary II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics Strategy refers to the general policies that managers employ to generate value. Rather than focus on operational detail, a rms strategy addresses broad, long-term issues facing the rm. Ultimately, along with its organizational architecture (to be discussed in Part 3), strategy is a key determinant of the success or failure of the enterprise. The ultimate objective of strategic decision making is to realize sustained prots. To achieve this objective, managers must devise ways to create and capture value. An essential rst step in generating prots is discovering ways to create value. There are at least four general ways that managers can increase value: (1) They can take actions to lower production costs or producer transaction costs. (2) Managers can implement policies to reduce consumer transaction costs. (3) They can adopt strategies to increase demand. Demand might be increased by taking actions to increase perceived quality, lower the price of complements, or increase the price of substitutes. (4) They can devise new products or services. Sometimes more value is created by cooperating with rms than by competing against them. Successful rms nd ways to convert ideas and knowledge in their employees wetware into softwareformulas and recipes for creating value. The discovery of better ways to use existing resources drives much of the value that rms create. For example, the value created by improved computer technology has not come from the discovery of new raw materials or resources, but by using existing resources more efciently. Firms face an essentially unlimited set of opportunities to create better instructions, formulas, recipes, and methods for making improved products at lower cost. New opportunities are likely to emerge as technology continues to evolve. Creating value is a necessary rst step in making prots. It is also necessary to capture value. A rm may have reduced its transaction/production costs or increased its consumer demand, but if other rms copy these changes quickly and enter the market, the competition will eliminate the prots. Chapter 6 indicates that the potential for rms to capture value increases with market power. Sometimes it also is possible to capture value without market power if the rm has superior factors of production that allow it to be more productive than competitors. The existence of effective entry barriers is required for market power. Entry barriers exist when it is difcult or uneconomic for a would-be entrant to replicate the position of industry incumbents. The existence of barriers, however, is no guarantee of market power or economic prots. At least four other factors are important: The degree of rivalry within the industry, threat of substitutes, buyer power, and supplier power. Both human as well as physical assets vary in productivity. If an asset allows the rm to make a prot because of its superior productivity, other rms will compete for this resource and bid up its price. Thus, in a well-functioning market, gains from superior productivity go to the responsible asset. For example, if a rm owns a unique piece of equipment, its price will be bid up to reect its superior productivity. The rm itself does not always have to use such assets to realize its values. It might earn more prots by selling or leasing the assets to other rms. Competition tends to take a differentiated asset to its highest valued use, but at a market price that reects its second-highest valued use. Because of the interdependencies among employees and assets, the value of the inputs as a team sometimes can be greater than the sum of their values if each were employed at their next best uses across other rms. Thus, it is possible that the overall rm will be more valuable than the sum of its parts. We characterize such a rm as having team production capabilities. A rm can maintain team production advantages only when competing rms cannot assemble BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value Chapter 8 The McGrawHill Companies, 2009 Economics of Strategy: Creating and Capturing Value 261 277 teams that are equally productive. Firms develop different team capabilities because they have different histories and development paths. The business environment is constantly evolving with new technological developments, changes in consumer tastes, new business concepts, new rms, and so on. Given these changes, it is unlikely that any competitive advantage will last forever. Some rms concentrate on a single major business, but many large rms engage in multiple businesses: They are at least partially diversied. Economies of scope provide the primary reason why diversication might enhance value. In addition, combining businesses in the same rm can promote complementary products. Although diversication has potential benets, it also has potential costs. As rms grow, they often become bureaucratic and more costly to manage. If diversication occurs through the merger of two rms (as often is the case) it also can be quite expensive to develop common personnel, communication, information, and operating systems. Some managers diversify to reduce earnings volatility. This often is a poor reason to diversify because shareholders can diversify on their own account simply by owning the shares of multiple companies. Related diversification occurs when the businesses use related technologies or serve common markets. The net benefits are likely to be greater in related rather than in unrelated diversification. For the most part, highly diversified firms have not performed well. Diversification has been most effective in the case of related diversification where there are potential economies of scope and opportunities to promote complements. Even if diversification creates value, the owners of the diversified firm do not always capture this value. Again, it depends on whether the firm brings some special resource or team capability to the transaction. Historically, most of the gains in corporate acquisitions go to the shareholders of target firms. Developing and implementing strategies that increase a rms value require an understanding of both the internal resources and capabilities of the rm and the external business environment. A rms resources and capabilities include its physical, human, and organizational capital. Important factors in the business environment include the rms markets (input and output), technology (production, information, and communications), and government regulation. Managers monitor the external environment to identify threats and opportunities for creating and capturing value. Most resources and capabilities are nitechoices have to be made. To make optimal choices, managers must consider the rms resources and capabilities jointly, as well as threats and opportunities within the external business environment. Ultimately, the rms strategy, along with its organizational architecture, is a key determinant of a rms value. Strategy consultants often suggest that all rms can develop strategies that deliver systematic economic prots even if they do not begin with unique resources or team capabilities. Basic economics suggests that this claim is false. Even if a manager were good at predicting the future and seeing what resources and capabilities were important, abnormal prots would not be earned on a systematic basis so long as a sufcient number of other managers adopt the same strategies. If multiple managers adopt the same strategy, there will be competition in the output markets as well as in input markets to obtain the necessary resources and capabilities. As in any competitive market, the expected equilibrium outcome is a normal rate of returnnot sustained abnormal performance. Firms whose resources and capabilities do not t the changed environment will not expect to earn a normal rate of return; they should prepare to go out of business. 262 278 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Managerial Economics Suggested Readings D. Besanko, D. Dranove, and M. Shanley (2000), Economics of Strategy (Wiley: New York). A. Brandenburger and B. Nalebuff (1996), Co-opetition (Doubleday: New York). D. Collis and C. Montgomery (1995), Competing on Resources: Strategy in the 1990s, Harvard Business Review, 118128. M. Porter (1980), Competitive Strategy (Free Press: New York). P. Romer (1998), Bank of America Roundtable on the Software Revolution: Achieving Growth by Managing Intangibles, Journal of Applied Corporate Finance 11(2): 827. Self-Evaluation Problems 81. Racket Sports Inc. has invented a new long-lasting tennis ball, but only if the players use a new type of string in their rackets. The racket string is produced by the Strings and Things Company. The respective prices for the balls and strings are PB and PS. The marginal cost of producing either product is $5.00. Demand for each product is Q 20 (PB PS) when (PB PS) less than 20 or less, 0, otherwise How will the two companies price the products if they do not cooperate? What are the resulting quantities and prots? What are the prices, quantities, and prots if the two companies price cooperatively? Explain why there is a difference. 82. Opportunity Cost and Competitive Advantage The average price of jet fuel in March 2008 was $3.10/gallon. In 2000 the price was only $.87/gallon. During this time period, Southwest Airlines hedged the price of fuel in nancial markets. The hedge contracts helped lock Southwest Airlines into lower fuel pricesas the price of fuel increased, so did the value of their nancial contracts. Roughly speaking, if the price of fuel increased at $1 per gallon, the nancial contracts paid Southwest Airlines $1 per gallon, which effectively allowed them to purchase fuel at the price before the increase. Leading nancial analysts argued that these hedges gave Southwest Airlines a competitive advantage over airline companies that had not hedged fuel prices. To quote one, Southwest Airlines has a competitive advantage in the Texas market because of its fuel hedges. Its fuel costs only a $1 per gallon, while other airlines are paying $3. We do not see how these other airlines can continue to compete with Southwest on most of its routes. Do you agree with the analysts statement? Explain. Solutions to Self-Evaluation Problems 81. If the two companies do not cooperate in setting prices, each will seek to maximize its individual prots, given its expectation about the other rms price. In this case, Racket Sports will view its demand curve as: PB (20 PS *) Q while Strings and Things will view its demand curve as: PS (20 PB *) Q * * where PB and PS represent the expectations about the other rms price. Each rm will maximize its prots by setting MR MC ($5 in this problem). Racket Sports: Substitute Q 20 (20 PS *) 2Q 5 Strings and Things: (20 *) PB 2Q 5 (PB PS) and rearranging the terms yields the following two equations: (20 PS *) 2[20 (PB PS)] 5 (20 PB *) 2[20 (PB PS)] 5 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 8 263 The McGrawHill Companies, 2009 8. Economics of Strategy: Creating and Capturing Value Economics of Strategy: Creating and Capturing Value 279 The equilibrium consists of the two prices that simultaneously solve these two equations. Note that in this problem the two rms face identical problems (they face the same expected demand curve and marginal cost). Thus in equilibrium, they will choose the same prices, PS * PB * P *. Substituting and solving one of the equations: (20 P *) 2 [20 (P * P *)] 5 3P * 25 P* $8.33 Each rm sells a total of Q 20 (8.33 8.33) 3.33 units. The prot margin on each unit is $8.33 $5 $3.33. The total prots per rm are $11.11 (9 3 9 3) and combined prots of $22.22. If the two rms cooperate they will price the two products to maximize combined prots. To maximize combined prots, the companies will jointly set marginal revenue equal to marginal cost using the combined demand curve PC 20 Q and marginal cost $10, where PC the combined price of (PB PS): 20 2Q $10 They sell 5 units at a combined price of $15 (for example, PS * PB * $7.50) for a combined prot of $25 [ 5 ($15 $10)]. When the rms price independently, they do not consider the negative effect that their higher prices have on the others prot. Pricing cooperatively, they take this interaction into account. 82. The analysts comment is not consistent with sound economic analysis. Southwest makes money on its nancial contracts when fuel prices increase. This revenue helps to increase its prots relative to the prots of airline companies that do not have similar hedging contracts. However, Southwest still faces the same opportunity cost for fuel as other airlinesthe market price. In deciding whether or not to y a particular route, Southwest should consider the market price of fuel in its prot forecasts. It makes money on the nancial contracts whether it ies the route or not. The marginal cost for fuel is the current market price. Either Southwest has to purchase additional fuel in the market to y the route or use fuel that it has in inventory that could be sold to other airlines at the market price. Either way the opportunity cost of fuel is the current market price. Review Questions 81. Choose a company that markets computer products over the Internet (for example, through a Web search). In what ways does the company create value? Is it likely to capture much of this value? Explain. 82. Airbus and Boeing are two major producers of jumbo jets. Are these rms guaranteed to make high prots since there are only two large rms in the industry? Explain. 83. The Watts Brewing Company owns valuable water rights that allow it to produce better beer than competitors. The company sells its beer at a premium and reports a large prot each year. Is this rm necessarily making economic prots? Explain. 84. What are team capabilities? Give examples of rms that appear to have them. 85. Sun Resorts has a hotel on a Caribbean Island. It recently spent money to lobby the government to build a better airport and expand air service. Why did they do this? Do you think that Sun Resorts cares about how many airlines will serve the island? Explain. 86. Evaluate the following statement: Business is war. Never consort with the enemy. 87. The Long-Drive Golf Company manufactures a new line of golf clubs. The Cushion Bag Company makes a special golf bag that protects the delicate shafts on these clubs. The respective prices are Pc and Pb for the clubs and bags. The marginal cost for producing either product is 100. Demand for each product is Q 1,000 (Pc Pb) when Pc Pb is 1,000 or less, 0, otherwise 264 280 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 8. Economics of Strategy: Creating and Capturing Value The McGrawHill Companies, 2009 Managerial Economics How will the two companies price the products if they do not cooperate? What are the resulting quantities and prots? What are the prices, quantities, and prots if the two companies price cooperatively? Explain why there is a difference. 88. One CEO justied the merger of his soft-drink company with a machine tool company in the following manner: This is a great merger. First the products are unrelated. Thus our companys earnings volatility is likely to decrease. Second, our management team has proved that we are better managers than the former management team of the tool company, and thus we are likely to discover new ways to create and capture value within the tool company. Evaluate this rationale. 89. Pepsi produces Fritos and Lays potato chips in addition to its basic soft-drink products. Discuss potential ways that this business combination might increase value. 810. The Strippling Drug Company has just obtained an important patent for a new drug that increases male virility and cures male pattern baldness at the same time. Does this imply that Strippling has a competitive advantage in producing the drug? Explain. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 265 Economics of Strategy: Game Theory CHAPTER 9 CHAPTER OUTLINE Game Theory Simultaneous-Move, Nonrepeated Interaction Analyzing the Payoffs Dominant Strategies Nash Equilibrium Revisited Competition versus Coordination Mixed Strategies Managerial Implications Sequential Interactions First-Mover Advantage Strategic Moves Managerial Implications Repeated Strategic Interaction Strategic Interaction and Organizational Architecture Summary Appendix: Repeated Interaction and the Teammates Dilemma T he Boeing Company and Airbus Industrie (a consortium of four European rms) manufacture most of the worlds large commercial jetliners.1 In 2003 Airbus delivered more jetliners than Boeing for the rst time in its 34-year history. With just two major companies in the market, one might expect that the rms would cooperate regularly, avoiding competition in order to increase prices and prots. In actuality, these companies frequently compete quite aggressively. The choice between cooperation and aggressive competition is determined in part by the nature of the environment. For example, cooperation is more likely in circumstances where each rm is near capacity. But in 2001, Airbus orders had fallen by 28 percent and Boeing by 45 percent. These reductions mirrored the growing crisis in the airline industry. Airline losses were soaring in the face of declining demand for air travel in response to the combination of the slowdown in the global economy and the September 11, 2001, 1 Details for this example are from A. Bryant (1997), The $1 Trillion Dogght, New York Times (March 23), 3.1; M. Freudenheim and J. Tagliabue (1998), Regulators Investigating Price Increases by Boeing and Airbus, New York Times (November 27), C.1; and D. Michaels (2004), Airbus Zooms Past Boeing, but Strong Euro Poses a Threat, The Wall Street Journal ( January 16), A9. 266 282 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics terrorist attacks. This was recognized in Morgan Stanleys August 12, 2002, research report on Boeing. It suggests an already tough pricing environment gets tougher if Airbus needs to ll slots available. Many of the worlds markets are like the commercial aircraft industry in that there are a few large rms who are the major players. Examples include soft drinks (Coke and Pepsi), U.S. domestic airlines (American, United, and Delta), copy machines (Canon and Xerox), candy (Hershey and Mars), and lm (Kodak and Fuji). Even small rms often compete with an identiable set of rivals and can act as the major players within a given market area. Consider two competing gasoline stations at an expressway interchange. In this type of market, it generally is important for managers to consider rivals responses when making major decisions. For example, part of Airbuss marketing plan in the late 1990s was to be overtly negative toward Boeing. One Airbus brochure depicted the Boeing 777 as an aging stretch limousine painted like a yellow taxicab, while the Airbus craft was pictured as a Mercedes luxury sedan. In another ad, Airbus pointed out that the Boeing 737 entered service in 1968, the same year Richard Nixon rst was elected president of the United States. Perhaps not surprisingly, Boeing produced negative advertisements of its own. In choosing its advertising policy, Airbus presumably had to consider whether a negative campaign would be value-maximizing given Boeings likely response. Similarly, Boeing had to consider Airbuss response in choosing its actions. Game theory provides a useful set of tools for managers to use when considering rival responses in decision making. We briey introduced these tools in Chapter 6. This chapter extends this introduction by presenting a more detailed discussion of the basic theory and by showing how managers might use these tools in strategic decision making. This analysis also provides managers with a richer understanding of competition within different market settings. For example, it provides insights into why there is erce competition in some concentrated industries (such as commercial aircraft), whereas in others the competition is more benign. Although we focus primarily on interactions among rival rms in product markets, these concepts also are useful to managers when dealing with other parties, such as suppliers, employees, or government ofcials. Game Theory Game theory is concerned with the general analysis of strategic interaction.2 It focuses on optimal decision making when all decision agents are presumed to be rational, with each attempting to anticipate the likely actions and reactions of its rivals.3 Although it 2 The terms strategy and strategic are used in at least three related yet slightly different ways in the economics literature. In all three contexts, the decision maker is assumed to interact with others in the environment. Thus the optimal decision is affected by the actions of others. The focus of the analysis, however, varies among the three uses. First, as in Chapter 8, the terms are employed to refer to the general policies that managers use to generate prots. The traditional strategy literature tends to use the terms within this context (for example, what businesses should the rm be in and how should they compete?). Even though rival responses are in the background in this literature, they are not explicitly treated in the analysis. Second, the terms are used to refer more specically to decision-making contexts where it is essential for managers to consider the behavior of identiable rivals in choosing their policies (for example, in setting prices in an industry with a few large rms). This is the primary use in this chapter. Third, they also are used in a more formal sense in game theory to describe the players actions in a given game (for example, to describe a players strategy in playing a game). We introduce this usage and related concepts in this chapter. 3 Game theory is divided into two branches. In cooperative games, players can negotiate binding contracts that allow them to plan and implement joint strategies. In this chapter, we focus on noncooperative games, where negotiation and enforcement of binding contracts is not possible. This type of game theory generally is more useful for analyzing managerial decision making. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 9 267 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Economics of Strategy: Game Theory 283 ACADEMIC APPLICATIONS Game Theorists Win the Nobel Prize The Nobel Prize for Economics (or, more precisely, The Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel) is the worlds most prestigious award for contributions to the eld of economics. It is awarded annually by the Royal Swedish Academy of Sciences. The prize consists of a gold medal, a diploma bearing a citation, and a sum of money U.S. $1.3 million in recent years. It represents the ultimate recognition by ones peers. In 1994, John Harsanyi, John Nash, and Reinhard Selten won the prize for their pioneering analysis of the theory of non-cooperative games. And again in 2005, game theorists Robert Aumann and Thomas Schelling received the prize. Their contributions have had an important impact not only on theoretical economics, but on business practice as well. Firms hire game theorists as consultants in formulating strategic decisions (for example, in government auctions, corporate takeovers, pricing, labor negotiations, and competitive positioning). Several best-selling business books have illustrated how game theory can be used in everyday business decisions. Game theory now is being taught within most major business schools. And for those who do not want to take a course or read a book, the movie A Beautiful Mindbased on Sylvia Nassers award-winning biography of John Nashwon three Oscars in 2001, including Best Picture. (Note, however, that the movies illustration of a Nash equilibrium is fundamentally awed.) began as a set of methods to analyze parlor games, this theory has evolved to the point where it is applied to study a wide variety of strategic interactions ranging from politics to competitive strategy. In this chapter, we introduce its basic elements and apply them in the context of managerial decision making. Sound managerial decision making often requires putting yourself behind your rivals desk. Assuming rivals are rational and acting in their self-interest, what decisions are they likely to make and how are they likely to respond to your actions? A complicating factor is that rivals optimal choices typically will depend on their expectations of what you will do; their expectations, in turn, depend on their assessments of your expectations about them. This type of circularity or recursive thinking might appear to make the overall problem completely intractable. Yet, this situation is precisely where game theory is most useful. Strategic interactions can take a variety of forms and involve many players who choose among a variety of potential actions. In this chapter, we limit ourselves to simple two-player, two-action cases. This limitation allows us to depict strategic problems using convenient diagrams. The fundamental intuition and concepts developed in these basic cases readily extend to multiple players and multiple actions. Some problems involve simultaneous decisions by rivals, whereas others involve sequential choices. We begin by examining strategic problems involving simultaneous choices. Later, we consider sequential problems. In some cases, interaction is expected to be repeated; in others, it is not. Throughout most of this chapter, we focus on problems where the interaction is not expected to be repeated. Toward the end of the chapterand in the appendixwe consider implications of repeated interaction. Simultaneous-Move, Nonrepeated Interaction In strategic problems that involve simultaneous moves, rivals must make decisions without knowledge of the decisions made by their competitors. Nonrepeated means that the interaction is presumed to occur only once. As an example, suppose that Boeing and Airbus are asked to submit sealed bids on the price of 10 jet airliners to a foreign national airline. Both companies doubt that they will compete in similar ways in the future. Both BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Strategic Form In this case, Boeing and Airbus individually choose and simultaneously submit a bid price for 10 planes. They can submit either a high price or a low price. If one company bids high and the other bids low, the order goes to the low bidder; if both companies submit the same bid, they split the order. This diagram presents the possible prots from the interaction in strategic form (normal form). Each cell presents the payoffs (in millions of dollars) of a pair of decisions. The entry on the lower right of each cell is the payoff for Airbus, whereas the entry on the upper left is for Boeing. Both rms have a dominant strategy (shaded cell)choose a low price. $500 $500 BoeingLow pr ice Figure 9.1 Managerial Economics AirbusLow pr ice $0 $1000 AirbusLow pr ice $1000 $0 AirbusHigh pr ice BoeingHigh pr ice Part 2 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory BoeingLow pr ice 284 II. Managerial Economics BoeingHigh pr ice 268 $750 $750 AirbusHigh pr ice companies can select either a high price or a low price. If one company bids high and the other bids low, the order goes to the low bidder; if both companies submit the same bid, they split the order. Each rm has the capacity to build all 10 airplanes. Analyzing the Payoffs Both companies privately choose their bids at the same time. The resulting payoffs (prots expressed in millions of dollars) depend on both rms choices; they are displayed in Figure 9.1. This type of diagram was rst introduced in Chapter 6 and presents the potential payoffs from the interaction in what is called strategic form (also called normal form). Each cell presents the payoffs in millions of dollars for a pair of decisions. The entry on the lower right of each cell is the prot for Airbus, and the entry on the upper left is for Boeing.4 Dominant Strategies A dominant strategy exists when it is optimal for a rm to choose that strategy no matter what its rival does. In Figure 9.1, both rms have a dominant strategychoose the low price. To illustrate, consider Boeings position. If Airbus chooses a high price (the right column), Boeing captures the entire order by submitting a low price. The resulting payoff of $1 billion is higher than the payoff of $750 million if both rms price high and split the order. If Airbus chooses a low price (the left column), Boeing is clearly better off to price low 4 This diagram has two primary columns and two rows. Boeing is the row rm in the sense that its strategies vary between the two rows. Airbus is the column rm in that its strategies vary across columns. We place the payoff to the row rm at the upper left of the cell and the payoff to the column rm in the lower right. An alternative way to picture the payoffs is a simpler 2 2 table with each row representing one of Boeings potential actions and each column representing one of Airbuss potential actions. The entries in each cell of the table would list the payoff to Boeing and to Airbus, respectively (for example, 100, 0). Although this alternative is more common (possibly because it is easier to type), students seem to nd diagrams like 9.1 easier to read. Both diagrams display the same information. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory Chapter 9 269 The McGrawHill Companies, 2009 Economics of Strategy: Game Theory 285 MANAGERIAL APPLICATIONS StockliftingA Dominant Strategy? In the spring of 1998, Midwest Quality Gloves, Inc., paid about $700,000 to Lowes Home Improvement Stores (the number-two home-center chain behind Home Depot) for 225,000 pairs of gloves produced by a competing rm, Wells Lamont. Midwest then sold the gloves for about $280,000 to a rm that specializes in liquidating closed-out merchandise. The motivation for this transaction was to clear shelf space, allowing Lowes to restock the shelves with Midwests gloves. This practice, known as stocklifting, is increasing in popularity. Makers of everything from bicycle chains to party napkins are lifting truckloads of competitors products from large retail chains such as Kmart and Revco drugstores. Other stocklifting examples include cellular phones, power adapters, leather cases, pet toys, humidiers, ashlights, faucets, and glue. The stocklifted merchandise is subsequently dumped for resale by faraway (sometimes foreign) retailers. The purpose in each case is to clear shelf space for the stocklifters products. To quote one executive involved in this practice, buybacks are a necessary evil in gaining market share. From the standpoint of the manufacturer this costly practice is a potentially dominant strategy. To quote one product manager, It costs a ton of money; however, if you want to land a major retail account, youre going to have to do it. When competing manufacturers jointly engage in this practice, they can nd themselves in a rivals dilemma such as in our Boeing/Airbus example: They would each be better off if no one stocklifted. Given the private incentives of each rm, however, stocklifting is difcult to avoid. For instance, after purchasing stocklifted products from a company, one major liquidator makes courtesy calls to the victims, encouraging them to return the favor by working with his company. Source: Y. Ono (1998), Where Are the Gloves? They Were Stocklifted by a Rival Producer, The Wall Street Journal (May 15), A1. and split the order. Its alternative is to price high and sell no planes. The same logic holds for Airbus. Given these strong incentives, the likely outcome is for both rms to submit a low price. Note that the rms would be better off if they jointly were to submit high prices. But this outcome is unlikely without repeated interactions. This problem has the same structure as the prisoners dilemma introduced in Chapter 6we might call it the rivals dilemma. As we shall see, a cooperative outcome where both rms submit high prices without explicit collusion is more likely if their interaction is expected to be repeated. This logic helps explain why Boeing and Airbus frequently compete aggressively on price. Airline orders frequently are measured in billions of dollars, so each sale is important. Further, customers (normally commercial airline companies) have economic incentives to deal with a single major supplier of aircraft since this policy reduces the required inventory of spare parts, technician training (maintenance personnel work only on one make of plane), and so on.5 Thus if Boeing or Airbus lose an order, it implies the loss of potential future orders as well. Just as in Figure 9.1 the companies have strong economic incentives to submit low bids to capture orders. Boeing and Airbus have particularly strong incentives to compete on price because each rm has the capacity to sell higher output. Boeing has stated that it wants to win two-thirds of all new orders (substantially above its current level), whereas Airbus wants to win about 50 percent (approximately equal to its current level). These conicting goals promote price competition as each rm tries to take orders from the other. If the rivals faced tighter capacity constraints, it is likely that there would be less price competition. For instance in our simplied example, a rm clearly would not want to lower the price to capture the full order if it lacked the capacity to produce all 10 planes. 5 Given these considerations, our assumption that if the bids are the same, the order is split between Boeing and Airbus should be questioned. Yet, if the buyer ips a coin and gives all the order to the lucky bidder whenever the bids are equal, then the expected payoffs are the same as in Figure 9.1. 270 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 286 II. Managerial Economics Part 2 Managerial Economics Figure 9.2 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Nash Equilibrium $500 #1: $500 BoeingLow price BoeingLow price Boeing and Airbus individually submit prices for 10 planes. They can choose either a high price or a low price. If one company bids high and the other bids low, the order goes to the low bidder; if both companies make the same bid, they split the order. If Boeing loses the bid, it sells four planes at the high price through a side deal. The circle technique provides a simple way to identify a Nash equilibrium. Start by assuming that Boeing will submit a low price (the top row). Circle the maximum payoff (in millions of dollars) that Airbus can achieve. This payoff (labeled #1) is in the cell where Airbus submits a low price (if the same payoff occurs regardless of whether Airbus submits a high or low price, circle both cells). Second, move to the bottom row and assume that Boeing submits a high price; circle Airbuss highest payoff (#2). Third, assume Airbus will charge a low price (left column) and circle Boeings highest payoff (#3). Fourth, move to the right column and assume Airbus submits a high price; circle Boeings highest payoff (#4). If a Nash equilibrium exists, the payoffs in both halves of the cell will be circled. In this game, the Nash equilibrium is where Boeing submits a high price and Airbus submits a low price shaded cell. (Note: Airbus has a dominant strategysubmit a low pricethe low-price strategy is circled in both rows.) #3: $600 #2: $1000 $0 AirbusHigh price BoeingHigh price BoeingHigh price AirbusLow price #4: $1000 AirbusLow price $750 $750 AirbusHigh price Nash Equilibrium Revisited Firms do not always have dominant strategies. For instance, suppose in our example that the United States government places pressure on the foreign country to have its national airline purchase planes from Boeing (governments sometimes do this for their domestic producers). The airline still splits the order when the bids are the same and awards Boeing the entire order if Boeing is the low bidder. But due to this political pressure, if Boeing bids high and loses the bid, the airline will buy four planes from Boeing at the high price on a side deal after purchasing the 10 planes from Airbus at the low price. Figure 9.2 presents this new payoff structure. Choosing a low price is still a dominant strategy for Airbus. Boeing, however, does not have a dominant strategy. If Airbus prices high, it is optimal for Boeing to price low to capture the entire order, whereas if Airbus prices low, it is better for Boeing to price high and make the side deal. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory Chapter 9 271 The McGrawHill Companies, 2009 Economics of Strategy: Game Theory 287 ACADEMIC APPLICATIONS Are Nash Equilibria Likely? Researchers have conducted many laboratory experiments on how people act in strategic situations. One particular question of interest is, Do they choose Nash equilibria? The evidence suggests that the concept works relatively well in predicting behavior in simple single-move situationsespecially if the individuals have prior experience interacting in similar ways with different partners in the past. It appears to work less well in more complex situations (for example, situations that involve sophisticated mathematical calculations) and in repeated situations (we discuss implications of repetition later in this chapter). Also it often fails where coordination is required and there are multiple equilibria, unless there is a natural focal point. Source: D. Davis and C. Holt (1993), Experimental Economics (Princeton University Press: Princeton, NJ). Nash Equilibrium When dominant strategies do not exist, the concept of a Nash equilibrium is useful in predicting the outcome. As dened in Chapter 6, a Nash equilibrium is a set of strategies (or actions) in which each rm is doing the best it can, given the actions of its rival. In Figure 9.2, the combination of a low Airbus price and a high Boeing price is a Nash equilibrium. Neither rm would want to change its price given the price submitted by the other rm. A particular problem might have multiple Nash equilibria. Nash equilibria are not necessarily the outcomes that maximize the joint payoff of the players. For instance, in Figure 9.1 the outcome where both rms submit low prices is a Nash equilibrium. Yet both rms would be better off if they jointly submitted high prices. Identifying a Nash Equilibrium The circle technique provides a simple way to identify a Nash equilibrium in such a problem. Consider the case in Figure 9.2. Start by assuming that Boeing will charge a low price (this is in the top row). Circle the maximum payoff that Airbus can achieve. This payoff (labeled #1) is in the cell where Airbus submits a low price (if the same payoff were to occur regardless of whether Airbus submits a high or low price, circle both cells). Second, move to the bottom row and assume that Boeing submits a high price; again circle Airbuss higher payoff (#2). Third, assume Airbus submits a low price (left column) and circle Boeings higher payoff (#3). Finally, move to the right column and assume Airbus submits a high price; circle Boeings higher payoff (#4). If a Nash equilibrium exists, the payoffs in both halves of the cell will be circled. In this problem, the Nash equilibrium is the shaded cell where Boeing submits a high price and Airbus submits a low price. Dominant strategies also can be identied by this technique. If the rm has a dominant strategy, the strategy will be circled for all actions by the rival rm. In this example, submitting a low price is a dominant strategy for Airbus: For each of Boeings possible actions submitting a low price is circled. Management Implications The power of a Nash equilibrium to predict the outcome in strategic situations stems from the fact that Nash equilibria are self-enforcing: They are stable outcomes. For instance, if Boeing can forecast Airbuss choice (perhaps because it understands that Airbus has a 272 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 288 Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics ANALYZING MANAGERIAL DECISIONS: Favoring a Government Ban on Advertising Cigarette manufacturers were among the rst companies to advertise extensively on television. Older people today can still hum the heavily televised jingle, Winston tastes good like a cigarette should, and recall the Old Gold Dancing Cigarette Packan oversized dancing cigarette package with shapely female legs. They can also remember beautiful women with black eyes as they testied that Tareyton smokers would rather ght than switch. John Wayne and other movie stars appeared in commercials to endorse cigarettes. Cigarette companies were the major sponsors for many popular television shows. The companies also paid to have cigarettes included in the actual programs. In 1962 the cartoon characters Fred and Wilma Flintstone were shown smoking Winston cigarettes. The U.S. federal Government banned television advertising of cigarettes starting in 1971. Interestingly the ban was supported by the leading cigarette manufacturing companies even though they had advertised heavily on television since the 1940s. 1. Design a simple two-company game that illustrates why it might have been in the economic interests of the cigarette companies to support the ban. In designing the game, assume that there is no regulation and that the two rms simultaneously choose between advertising and not advertising. Display your hypothetical payoffs in strategic form (see Figure 9.1) and highlight the Nash equilibrium. Explain the intuition for why the rms in your example would favor regulation to ban advertising. 2. Can you conclude from your example that all rms in all industries will favor bans on television advertising? Explain. Can you ever envision a situation where one rm might favor the ban and a competing rm might be against it? Discuss. dominant strategy), it is optimal for Boeing to choose its equilibrium action, a high price. And Airbus has no incentive to avoid its equilibrium choice, a low price. Thus, even if both rms can forecast the outcome, neither rm has an incentive to choose any other action. Although the idea of a Nash equilibrium is quite useful, it is not as powerful in predicting the outcomes of strategic interactions as is the concept of a dominant strategy. When dominant strategies exist, there are strong private incentives to choose them, regardless of what the other player does. Thus, it is quite predictable that rivals will choose dominant strategies. With a Nash equilibrium, your best choice generally is contingent on what you expect your rival to do. In many cases it is reasonable to expect that a Nash equilibrium will occur. This is more likely to be true when the rivals have more experience in similar strategic problems, have better information about each other, or when the Nash equilibrium is what is called a natural focal point.6 For example, consider the problem in Figure 9.2. If 6 A focal point exists when there is a reasonable and obvious way to behave. To illustrate how a Nash equilibrium could be a focal point, consider a game where two strangers are asked to raise a hand. Both players receive a large payoff if they stick up the same hand (both left hands or both right hands) and nothing if they raise opposite hands. Two Nash equilibria exist: (1) both raising left hands, (2) both raising right hands. Nonequilibrium outcomes occur when the players raise opposite hands. The Nash equilibrium of both raising right hands is potentially a focal point. Most people are right-handed. If one player expects that the other is likely to raise his right hand because he is right-handed, she will have the incentive to raise her right hand as well. Reciprocally, it is reasonable for the other player to follow the same logic. Thus this outcome is potentially much more likely than either the nonequilibrium outcomes or the Nash equilibrium of two left hands. II. Managerial Economics The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Boeing and Airbus make simultaneous choices of new communication systems for their planes. Two technologies exist: Alpha and Beta (payoffs in millions of dollars). Both rms benet if they choose the same technology. Applying the circle technique, we can see that there are two Nash equilibria: Alpha/Alpha and Beta/Beta (shaded cells). Economics of Strategy: Game Theory BoeingAlpha Coordination Game $100 $50 AirbusAlpha 289 $50 AirbusBeta AirbusAlpha $50 273 $50 $100 BoeingBeta Figure 9.3 BoeingAlpha Chapter 9 BoeingBeta BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition $100 $100 AirbusBeta Boeing has reasonable information about potential payoffs and Airbuss lack of political power within the specic country (it understands that there is a close working relationship between the local and U.S. governments), it will realize that Airbus has a dominant strategy to submit a low price. Boeing correspondingly will choose a high pricethe Nash equilibrium. When rivals know little about the game or each other and when there is not a natural focal point, outcomes other than Nash equilibria (nonequilibrium outcomes) are more likely to occur. Competition versus Coordination To this point, we have analyzed competitive interaction. In each case, at least one of the rms has an incentive to bid low to take sales from the other and garner additional profits. This potential gain comes at the expense of the other rm. Many strategic situations, however, involve coordination rather than competition. Consider the problem in Figure 9.3, in which Boeing and Airbus make simultaneous choices of new communication systems for their planes.7 Two technologies exist: Alpha and Beta. Both rms benet if they choose the same technology. A common technology standard allows producers to exploit scale economies and increases the likelihood that other companies will invest to develop new enhancements for the system because they can sell to both companies. Similarly, companies are more likely to develop service capabilities and stock larger inventories of spare parts. The overall demand for planes also might be higher because airlines would have lower costs in learning a single system. 7 For example, suppose each company is working to introduce a new intermediate-capacity plane at next years Paris Air Show and each views its choice of communication technology as a critical selling point for its model. Each will have to commit to a technology at the design phase, and each might be reluctant to discuss such features with its rival before the new model is unveiled. 274 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 290 Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Coordination Problems with HDTV High-denition televisions are digital and capable of displaying images at very high resolution. When combined with a good sound system, HDTV can provide a theater-like experience. The introduction of HDTV into the marketplace was slowed by coordination problems between television networks and manufacturers. Both groups thought that they would benet by the development of this product. However, neither wanted to be the rst to commit. Network executives were quoted as saying that they were reluctant to move forward with plans for new programming until the television manufacturers committed themselves to producing enough affordable sets to receive it. Yet manufacturers did not want to commit until the networks indicated that there would be enough digital programming for consumers to want to buy the sets. The situation was summed up in 1997 by a senior executive at CBS, The networks are waiting to see what the TV makers are going to do, and the TV makers are waiting to see what the networks are going to do. In this situation, there are two Nash equilibria: Manufacturers and broadcasters both invest or neither invests. Due to coordination problems, rms may get stuck in the second equilibrium, even though both groups prefer the rst. Although development of HDTV was slowed by coordination problems, it became a commercial reality by 1999. Manufacturers were selling the sets and broadcasting companies were providing more and more HDTV programming (the rst equilibrium). Nonetheless, for HDTV to become a widely adopted product, available programming had to increase and set prices had to decline. Their initial prices for the sets ranged from about $5,000 to $10,000. The coordination problems illustrated in this example arise frequently with new technologies. The overall value of a technology is usually higher when there are many users (there are network effects) and there is a common standard (for example, consider how much less valuable DVDs would be if there were several different incompatible formats and only a small number of users of any one type). Adopting a uniform standard can be difcult when there are numerous players with somewhat conicting interests. Sometimes governments, joint ventures among rms, and industry trade groups play a constructive role in promoting common standards, thereby helping to realize a preferred equilibrium. For example, in December 1996 the Federal Communications Commission set a timetable for stations and manufacturers converting to digital technology and, in October 2005, the Senate Commerce Committee set April 7, 2009, as the last date at which U.S. television stations can broadcast analog signals. Source: J. Brinkley (1997), Networks and Set Makers in Standoff over HDTV, New York Times (August 29), 5; and J. Kerr (2005), Senate Looks to Spend $3B on Digital TV, WashingtonPost.com (October 20). Applying the circle technique, we can see that there are two Nash equilibria: Alpha/ Alpha and Beta/Beta. If precommitment communication is possible, it is quite likely that one of these equilibria will be reached. For instance, if both rms announce that they will choose Alpha, there is no private incentive to deviate from this choice. Coordination can prove more difcult in cases where precommitment communication is costly and/or there are many players. Some strategic interactions involve elements of both competition and coordination. In the interactions problem in Figure 9.4, Boeing and Airbus benet from choosing the same technology; again, there are two equilibria. But the companies are not indifferent between the two. Boeing prefers the Alpha technology, whereas Airbus prefers the Beta technology (the technologies are better matches for their particular aircraft design). Coordination in this setting can be more difcult than in pure coordination problems, since the rms want different outcomes. Nonetheless, if one of the rms is convinced of the choice the other rm is going to make, it has an economic incentive to follow suit and choose the same technology. Below we discuss ways that rms might be able to make credible statements regarding their upcoming choices. Economics of Strategy: Game Theory $100 $50 $25 $25 AirbusAlpha 291 $40 $40 AirbusAlpha BoeingBeta Boeing and Airbus must make simultaneous choices of new communication systems for their planes. Two technologies exist: Alpha and Beta (payoffs in millions of dollars). Both rms benet if they choose the same technology. Applying the circle technique, we can see that there are two Nash equilibria (shaded cells): Alpha/Alpha and Beta/Beta. Boeing prefers the Alpha/Alpha equilibrium, whereas Airbus prefers the Beta/Beta equilibrium. BoeingAlpha Chapter 9 Figure 9.4 Coordination/ Competition Game 275 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory BoeingAlpha II. Managerial Economics AirbusBeta BoeingBeta BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition $50 $100 AirbusBeta Mixed Strategies In the strategic problems we have examined thus far, rms have chosen one specic action (for example, a high or low price). This type of choice is known as a pure strategy. Sometimes it can pay to randomize; for example, choose a high price with probability p and a low price with probability 1 p. The benet of this so-called mixed strategy derives from the element of surprise. For example, a football team does not want its opposing team to forecast its plays with 100 percent accuracy because it will eld defenses specically designed to stop the predicted plays. The team wants to mix up its plays and surprise the other team. This same logic can hold in business. MANAGERIAL APPLICATIONS Ice Cream Wars The Indian frozen dessert market of $200 million a year is expected to grow 20 percent annually as Indians inclination to snack is increasing. Two companies compete for the majority of Indias ice cream market: Hindustan Lever, Ltd. (better known as HLL) and Amul. Between them they sell about 75 percent of the ice cream in most major markets. HLL is 51 percent owned by the British-Dutch giant Unilever, and Amul is owned by an Indian farmers cooperative of 2 million farmers. Amul sells a no frills ice cream under the slogan A taste of India and spends less than 1 percent of their revenues on advertising; whereas HLL spends 10 to 15 percent. Amul sells its ice cream at about 12 percent less than HLL. Being a full-line dairy concern, Amul markets related products, thereby reducing its distribution costs. HLL claims they produce a higher quality product, but Amul counters that they have lower costs. HLLs executive director complains, They rebuff attempts to form an ice cream trade body. How can we take them seriously? But HLL is losing ground; their ice cream revenue was down in the rst quarter of 2002, and they lost 2 percentage points of market share. It appears as if Amul has a dominant strategy built around being the low-cost producer. Not only is Amul capturing market share from HLL, but Amul has rejected HLLs attempt at a coordination game in the form of an ice cream trade body. Source: S. Rai (2002), Battling to Satisfy Indias Taste for Ice Cream, New York Times (August 20), W1. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Mixed Strategy Boeing and Airbus simultaneously must commit to an advertising campaign. The advertising can focus on either the negatives of the other companys planes or the positive aspects of the companys own planes (payoffs in millions of dollars). Boeing is the market leader and benets when both rms choose the same strategy. Airbus does better when it can differentiate itself by choosing a different strategy. The circle technique indicates that there is no equilibrium in pure strategies. There is a unique equilibrium in mixed strategies. In this equilibrium, each rm randomizes between the two campaigns (choosing each with a probability of .5). $10 $10 Airbus Negative campaign $10 $10 Airbus Negative campaign Boeing Negative campaign Figure 9.5 Managerial Economics Boeing Positive campaign Part 2 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Boeing Negative campaign 292 II. Managerial Economics Boeing Positive campaign 276 $10 $10 Airbus Positive campaign $10 $10 Airbus Positive campaign Consider the strategic problem in Figure 9.5 where Boeing and Airbus simultaneously must commit to an advertising campaign. The advertising might focus on either the negatives of the other companys planes or the positive aspects of the companys own planes. Boeing is the market leader and benets more when both rms choose the same strategy. Airbus does better when it can differentiate itself by choosing a different strategy. The circle technique indicates that there is no equilibrium in pure strategies. If the campaigns match, Airbus wants to change. If the campaigns do not match, Boeing wants to change. The problem has an equilibrium employing mixed strategies: Both rms randomize between the two actions with a probability of .5they might base their actions on a coin ip. To understand why this is an equilibrium, consider the following logic.8 This problem is what is called a zero-sum game: If Boeing gains $10 (million), Airbus loses $10 and vice versa. Thus if one company has a positive expected value in the game, the other must have a negative expected value. If Boeing selects one of the campaigns with a probability greater than .5, Airbus can choose the other campaign with a probability of 1 and gain more than half the time. For instance, suppose Boeing chooses a positive campaign with a probability of .6, Airbus would gain 60 percent of the time by always choosing a negative campaign. Its expected payoff would be .6(10) .4( 10) 2 0. Boeings expected payoff would be .6( 10) .4(10) 2 0. The only way that Boeing can ensure an expected payoff of zero is by randomizing between the two actions with equal probabilities.9 Symmetric logic holds for Airbus. The outcome of both rms 8 All strategic interactions with a nite number of both players and actions have at least one Nash equilibrium in either pure or mixed strategies. In this chapter, we give examples of mixed strategies and discuss the general intuition behind the associated equilibria. More detailed discussion on how to solve for these equilibria can be found in Dixit and Nalebuff (1993), Chapter 7. 9 If Boeing chooses each action with a probability of .5, the expected payoff is zero independent of Airbuss strategy. Suppose Airbus chooses the positive campaign with probability p. The probability of Boeing matching the positive campaign is .5p (the probability of two independent events is the product of the individual probabilities of the events). Similarly the probability of matching a negative campaign is .5(1 p). Thus the likelihood of Boeing winning is .5p .5(1 p) .5. It correspondingly loses .5 of the time and has an expected payoff of zero. This is true independent of the p Airbus actually chooses. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory Chapter 9 277 The McGrawHill Companies, 2009 Economics of Strategy: Game Theory 293 ACADEMIC APPLICATIONS A Mixed Strategy at Wimbledon Tennis players can serve to either an opponents backhand or forehand. The receiver can choose to move left or right in anticipation of the serve. Both the opponent and receiver want to mix their choices to catch the other off guard. For example, if the server always serves to the backhand, the receiver will know that it is best to favor that side. The server would then want to mix it up and serve to the opponents forehand. In a mixed-strategy equilibrium, the expected payoff to a given player is the same regardless of which action he actually ends up taking. Thus, if the server is mixing between serving to the receivers backhand and forehand, the likelihood of winning the point should be the same. To test whether the concept of a mixed-strategy equilibrium actually explains the behavior of top tennis players, researchers recorded the results of serves in 10 matches at Wimbledon. There, data supported the theorythe success rates with serves to either the forehand or backhand were the same. Source: M. Walker and J. Wooders (2001), Minimax Play at Wimbledon, American Economic Review 91, 15211538. randomizing with a probability of .5 is a Nash equilibrium. Neither rm has an incentive to change its strategy given the strategy of the other rm. In equilibrium, a rm receives the same expected payoff regardless of which of the campaigns it actually chooses. Thus if Boeing ips a coin that indicates it should choose a positive campaign, there is no reason to deviate and choose a negative campaign. In either case it gets the same expected payoff (zero). By denition, it would not be an equilibrium if either rm had an incentive to alter its choice. Managerial Implications Managers often make decisions in circumstances where the decision is not expected to be repeated and the payoff depends on the simultaneous decisions of other parties, be they rival rms, customers, suppliers, or employees. Potential examples include making a large investment decision to enter a new market or industry, pricing a new product, or making an acquisition bid for a rm. To summarize the managerial implications of our analysis, consider Valerie Black, who must submit the bid for Boeing in our example. Val should perform the following calculations: Estimate the payoffs given each of her potential actions as well as those of her rival, Airbus.10 Examine whether she has a dominant strategyif so, she should employ it. Without a dominant strategy, she should make her best estimate of what Airbus will do and identify her corresponding best action (which might be a mixed strategy). Check whether the resulting outcome is a potential Nash equilibrium: Does the forecasted action of Airbus appear optimal from their viewpoint, given her proposed action? If so, her proposed action appears reasonable; if not, Val should reexamine the underlying assumptions of her initial forecast. 10 If she has many potential actions, she may wish to simplify the problem by focusing on a few key possibilities. Sometimes it is not feasible to quantify the payoffs. In these cases, managers can go through the following steps on a more qualitative basis. 278 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 294 Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Failure to Consider Strategic InteractionsThe Paper Industry Bennett Stewart of the Stern Stewart Company made the following point in a roundtable discussion with the CFO of International Paper: Im kind of fascinated by your industry because it does go through these cycles, and some part of that cyclical market behavior seems to be self-inicted. . . . When demand does rise, all the companies tend to throw piles of money into new capacity in attempts to maintain market share. . . . Then you all suffer together from excess capacity. The CFO responded: If youre thinking about building a new paper facility, youre going to base your decision on some assumptions about economic growthand, as the result of globalization, the relevant growth measure today is worldwide rather than just domestic or U.S. growth. What we never seem to factor in, however, is the response of our competitors. Who else is going to build a plant or machine at the same time? Managers generally will make better decisions if they incorporate the responses of competitors into their analysis. Source: Stern Stewart EVA Roundtable, Journal of Applied Corporate Finance (Summer 1994). Her forecast probably should be revised if it is based on the implicit assumption that Airbus is either dumb or irrational (for example, Airbus is unable to forecast a likely action on her part or does not know what is in its own best interests). In essence, Val should place herself behind her rivals desk and ask what she would do if she worked for Airbus. She should assume that Airbus undoubtedly is trying to forecast her actions, as well. Is there some reasonable set of beliefs that Airbus might have about her actions that would motivate the rm to choose the action in the initial forecastcan it be rationalized? If not, Val probably should revise the forecast. Since Val almost surely has less than perfect information about the factors affecting Airbuss choice, she may misforecast what Airbus actually does. Nonetheless, her goal is simply to do the best she can, given her imperfect information. Sometimes it will pay to collect additional information about a rival to make a better-informed choice (when the expected incremental benets of the new information are larger than its incremental costs). If Val is reasonably condent that Airbus will choose what appears to be a nonequilibrium strategy (for example, Airbus has a track record of choosing a particular strategy in similar situations), she should reevaluate both her and her rivals available actions and their associated payoffs carefully. It is frequently the case that what appears to be a nonequilibrium move reects an incomplete understanding of the alternatives. Ultimately, if she is quite condent that Airbus will choose a particular action, she should choose the action that maximizes her payoff, even if the outcome is not a Nash equilibrium. If Val is extremely risk-averse and has little experience either dealing with this rival or managing in similar situations, one option is to choose a secure strategya strategy that guarantees the highest payoff given the worst possible case.11 In other words, she forecasts the worst payoff that could arise for each of her potential actions and chooses the action that offers the highest payoff among the worst payoffs. But following such a strategy generally will not maximize the value of the rm (for example, if the primary benet is to guard against an unlikely outcome), and often she could do better with additional analysis and thought. 11 Another name for this type of strategy is maximin. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 279 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Chapter 9 Economics of Strategy: Game Theory 295 Sequential Interactions Thus far, we have limited our attention to simultaneous-move strategic problems. Yet in many business situations managers make decisions sequentially. For example, consider the problem in Figure 9.4 where Boeing and Airbus simultaneously chose new communication technologies. Decisions of this type often are made sequentially. For example, Boeing might choose the technology in one year, while Airbus chooses it in the next. As we shall see, the equilibrium outcome of strategic interactions can vary, depending on whether they occur simultaneously or sequentially. For instance, in this technology choice problem, there are two possible equilibria given simultaneous choices, but only one given sequential choices. Extensive Form Figure 9.6 displays this strategic interaction in extensive form assuming that Boeing moves rst (for example, Boeing chooses the technology for the new model it introduces at this years Paris Air Show, while Airbus will introduce its new model next year). The extensive form depicts the sequence of actions and the corresponding outcomes. A node indicates a point at which a rm must choose an action, while the branches leading from a node display possible choices at the node. At the rst node (node 1) Boeing must choose between the Alpha and Beta technologies. Airbus makes the next move. Whether it is at node 2 or node 3 depends on Boeings initial choice. At either node Airbus must choose whether to adopt the Alpha or Beta technology. The numbers at the end of the nal branches indicate the ultimate payoffs to Boeing and Airbus, given the sequence of choices made by the two rms. (These payoffs are the same as in Figure 9.4.) If both rms choose the Alpha technology, the payoffs are 100 for Boeing and 50 for Airbus (as in the upper-left cell of Figure 9.4), while if they choose the Beta technology, the payoffs Figure 9.6 Extensive Form In this sequential game, Boeing chooses the technology rst and then Airbus makes a choice. This diagram shows the game in extensive form. A node indicates a point at which a rm must choose an action, whereas the branches leading from a node indicate the possible choices. The numbers at the end indicate the payoffs (in millions of dollars) for Boeing and Airbus, respectively. The game is solved by backward induction. Given the payoffs, Airbus will choose Alpha at node 2 and Beta at node 3. If Boeing forecasts these choices, it will choose Alpha at node 1. Bold lines indicate equilibrium choices. ha Alp Airbus ha Alp Bet a Node 2 Boeing 40 Airbus 40 Boeing a Node 1 Be Boeing 100 Equilibrium Airbus 50 h Alp Boeing 25 Airbus 25 ta Airbus Bet a Node 3 Boeing 50 Airbus 100 280 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 296 Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics are reversed (as in the lower-right cell). Payoffs are lower for both rms when they choose different technologies. Backward Induction We analyze this extensive form using backward inductionlooking forward to the nal decision nodes and reasoning backward. To illustrate, continue with the example in Figure 9.6. Begin at the nal two nodes (2 and 3). At node 2, Airbus has an incentive to choose Alpha because it prefers the payoff of 50 to 40. At node 3, Airbus chooses Beta. Now move to the initial node. If Boeing foresees how Airbus will act at the nal nodes, Boeing will see that it is in its best interests to choose Alpha in the initial stage. If it chooses Alpha, Airbus will do so as well and Boeing will receive 100. By contrast if Boeing chooses Beta, Airbus will match that choice and Boeing will receive a nal payoff ANALYZING MANAGERIAL DECISIONS: Lets Make a Deal You work for a company that is frequently involved in negotiations to acquire companies from their current owners. You have been invited to be a guest on the popular television game show Lets Make a Deal. Your boss has agreed that you can participate on the game show during business hours because you might learn something that would be valuable in future business negotiations. You are now on the game show. The following events have occurred to this point: The game show host showed you three doors, labeled Door #1, Door #2, and Door #3 and told you that behind one of the doors is the grand prize of $100,000. The other two doors contain smaller prizes of no more than $20,000. You assume that the host knows which door has the $100,000. You were told to choose one of the doors, and you randomly chose Door #1. The host subsequently opened Door #2 and showed you that it contains a smaller prize of $20,000. He has given you the following choice. You can keep what is behind Door #1 (your initial choice). You can trade it for the $20,000 behind Door #2, or you can trade it for the unknown contents behind Door #3. You want to go for the grand prize of $100,000. Thus you must decide whether to keep Door #1 or trade it for Door #3. 1. You initially selected Door #1. Prior to seeing the contents behind Door #2, what was the probability that Door #1 contained the grand prize? What was the probability that Door #1 did not contain the grand prize (i.e., the grand prize was behind one of the other two doors)? 2. The host subsequently showed you that Door #2 contained $20,000 and gave you the choice to keep what was behind Door #1 or to trade it for what was behind Door #3. What should you do or does it matter (assuming you want to go for the grand prize)? [Hint: Put yourself behind the hosts podium and consider his choice to show you the contents behind one of the doors. It might be useful to view the situation as a sequential game where you choose the rst door at Node 1 and the host subsequently chooses which of the remaining doors to show you at Node 2.] 3. Your boss let you participate in the game show because she thought you might learn something that would be of value to your company. Does your experience on the game show suggest any general principles that might be useful in managerial situations? Explain. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory Chapter 9 281 The McGrawHill Companies, 2009 Economics of Strategy: Game Theory 297 of 50. Thus, the equilibrium outcome is for Boeing to choose Alpha in the rst stage with Airbus matching that choice in the second stage. The equilibrium strategy of a rm given sequential strategic interaction consists of a sequence of its best actions, where the actions are taken at the corresponding nodes. In this problem, Boeings equilibrium strategy is to choose Alpha in the rst stage, whereas Airbuss equilibrium strategy is to choose Alpha in the second stage if Boeing chooses Alpha and Beta if Boeing chooses Beta. The strategies of the two rms are a Nash equilibrium: Neither rm wants to change its strategy given the other rms strategy. First-Mover Advantage In this example, Boeing has a rst-mover advantage: By moving rst, Boeing makes higher prots than Airbus, who moves second. If Airbus moves rst, the advantage is reversed. Knowing Boeings incentive to match the technology in the second stage, Airbus would choose Beta in the rst round and receive higher prots. This example illustrates that managers must consider carefully the order of moves in strategic situations. When Boeing and Airbus make simultaneous choices, there are two potential equilibria. In the sequential game there is only one. The outcome, however, depends on who moves rst. The rst mover does not always have a strategic advantage. In some situations, the follower has the advantage. Consider a rm that cuts its development costs by copying product innovations by pioneering rmsafter all, the second mouse gets the cheese. Strategic Moves Actions that are taken to inuence the beliefs or actions of the rival in favorable ways are called strategic moves. Typically, they require individuals to restrict their own future actions. For example, prior to Boeings choice, if Airbus could make a binding commitment to choose Beta no matter what Boeing did, it then would be in Boeings interest to adopt the Beta technology in the rst year. Credibility Can Airbus offset Boeings rst-mover advantage by simply announcing that it will adopt the Beta technology next year no matter what Boeing does this year? Boeing would be clearly better off to adopt the Beta technology if it really believed that Airbus would carry out its announced plan. Yet Boeing probably should ignore such a statement by Airbus because it is not credible. The statement lacks credibility specically because it would not be in Airbuss self-interest to carry out its announced plan if Boeing MANAGERIAL APPLICATIONS First-Mover AdvantageWal-Mart In Chapter 8, we discussed how Wal-Mart was the rst company to place large discount stores in many small towns in the southeastern region of the United States (for instance, in Arkansas). Even though Wal-Mart has been quite protable at these stores, other companies have been reluctant to place competing stores in the same towns. They realize that each town is big enough to support only one such store. If they did enter, they would have to compete with Wal-Mart and both would lose money. Thus Wal-Mart has a rst-mover advantage in these communities. 282 298 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics adopted the Alpha technology rst. Airbus would be foolish to adopt the Beta technology after Boeing chooses the Alpha technology given its low payoff. For a strategic move to be credible, it must consist of a sufcient commitment to convince the rival to change its beliefs. Merely announcing a planned future action typically is not enoughtalk is cheap. As an example of a potentially credible action, suppose that Airbus signs a contract with the manufacturer of the Beta technology to pay the supplier a large sum of money whether or not Airbus employs its technology in the following year. This contract is legally binding and expensive to renegotiate. Now Boeing has a substantive reason to believe that Airbus will carry through with its announced plan to adopt the Beta technology even if Boeing were to choose the Alpha system. If it does not, it still will have to pay a large sum of money to the supplier and thus would be worse off than if it matches Boeing by choosing the Alpha technology. If Boeing decides this announcement is credible, Boeing should adopt the Beta system in the rst year.12 Managerial Implications Many strategic situations involve a sequence of actions by the rivals. Examples include management negotiations with a labor union and the dissemination of a new product within an industry. To illustrate the management implications of our analysis of sequential strategic interactions, consider Helmut Mueller, the manager who must choose the communication technology for Airbus in the communication technology problem. Helmut should begin by dening carefully the sequence of moves. For example, he may have enough information to know that Boeing will choose a new technology this year, whereas Airbuss production schedule will not allow it to make a choice until the following year. He then should work backward to predict the likely outcome of their interaction. Starting at the end of the process, he will realize that he will have a strong incentive to match Boeings choice in the second period. Moving to the rst node, he should place himself behind a Boeing desk. He should realize that Boeing is likely to understand that Helmut will have strong incentives to match Boeings technology in the second period. Given this belief, Boeing will adopt the Alpha technology. Airbus is less protable with this outcome and thus would prefer the joint adoption of the Beta system. As a nal step, he should analyze whether he could make any strategic moves that would inuence the beliefs of Boeing managers and thus motivate them to adopt the Beta technology. He must realize that Boeings management is unlikely to believe a simple announcement that Airbus will choose the Beta technology in the future. Helmut might conclude that a contract with the Beta manufacturer would provide sufcient commitment to alter Boeings beliefs and choice. After entering this contract, it is important that Helmut make his action known to Boeing. For instance, Helmut might want to report the contract to the nancial press. 12 Note to the technically inclined: If Boeing actually believed Airbuss statement that it would choose Beta in the second round regardless of Boeings rst round choice, Boeing would logically choose Beta in the rst round. The outcome Beta/Beta is a second potential Nash equilibrium to this gameneither party has an incentive to alter its choice of Beta given the other rms action. However, this second potential equilibrium can be ruled out if we require that both parties expect each other to act rationally (in their own interest) at any node in the game (on or off the equilibrium path). Game theorists refer to this criterion for eliminating unlikely equilibria as subgame perfection. We think this selection criterion is quite reasonable and should be employed by managers in their analysis of strategic situations. Using this criterion, there is only one economically reasonable equilibrium to this game, Alpha/Alpha (unless, as we discuss next, Airbus can make an economically credible commitment always to choose Beta in the second round). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory Chapter 9 283 The McGrawHill Companies, 2009 Economics of Strategy: Game Theory 299 MANAGERIAL APPLICATIONS Strategic BehaviorNBA There was a major labor dispute between National Basketball Association players and team owners in 1998. Issues included the percentage of revenues devoted to salaries, maximum salary caps, rookie salaries, maximum annual raises, as well as several other items. The players union had delegated negotiating authority to a group of 19 players. The general membership of the union was to vote only on contract proposals endorsed by the negotiating group. The key negotiator on the owners side was David Stern, commissioner of the NBA. The rst half of the season was canceled by a labor lockout as the two sides repeatedly exchanged contract proposals. By January 1999, the dispute had cost the league over a billion dollars in forgone revenue and the players over $500 million in salaries. Given the short life of the typical players career (around ve years) the cost borne by the players was particularly high. The dispute also created signicant ill will among NBA fans. At the beginning of January, Stern informed the players that he would not consider their nal offer, would recommend to the owners that they cancel the rest of the season, and would begin looking for replacement players for the following year. The season, however, would be played if the players accepted the owners nal offer. Sterns threat to end the season was a strategic move to motivate the players to settle the dispute. Was this move credible? After all, canceling the season also would impose signicant costs on the owners. Apparently the players negotiating group thought that the threat had merit. Following Sterns announcement, the negotiating group decided to allow the full 470 members of the union to vote on the owners proposal. This action was something that Stern had wanted all along but the negotiating group had refused. However, Stern did back off on his commitment not to consider new proposals from the union. Prior to the nal vote, a union attorney called the commissioners ofce and indicated that the players were prepared to vote to cancel the remainder of the season. However, a last minute deal was still possible if the NBA owners gave some ground. After an all-night bargaining session, a deal was reached and the labor dispute settled. Source: S. Fatsis (1999), NBA, Players Reach Accord, Saving Season, The Wall Street Journal ( January 7), A3. Repeated Strategic Interaction Many strategic situations involve repeated interaction among rivals. For example, Boeing and Airbus have many opportunities to compete against each other to supply commercial jet aircraft. Indeed many, if not most, companies deal with the same competitors, suppliers, employees, and regulators over extended periods of time. When interaction is expected to occur repeatedly, more equilibria frequently are possible. For instance, recall the situation illustrated in Figure 9.1. In this problem, Boeing and Airbus make bids to produce 10 airplanes. The dominant strategy is for both to bid low, MANAGERIAL APPLICATIONS It Pays to Think Sequentially According to a story that frequently has circulated over the Internet, two college students drove to an out-of-town basketball game the day before an important exam. The students drank too much and were unprepared to take the test. Rather than unk the exam, they decided to tell the teacher that a at tire delayed their return. Being the understanding type, the professor said, No problem; take the test tomorrow. The students arrived the next day to take the exam; they were put into separate rooms. The rst question, worth 5 points, was quite easy. But condence turned to apprehension as each student turned to the next question, worth 95 points, that simply asked, Which tire? If the students had thought ahead more carefully, they would have behaved differently. 284 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 300 Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Boeing and Airbus Accused of Price Collusion On July 21, 1998, Boeing raised prices on its aircraft by 5 percent, its rst price increase in 23 years. Airbus followed with a 3 percent increase in September. Antitrust regulators in Washington and at the European Union in Brussels almost immediately began separate investigations of whether the two companies had engaged in price xing. Both companies vehemently denied the allegations. Game theory implies that a cooperative equilibrium is possible given repeated interaction without explicit communication or collusion. One interpretation of the price increases is that the two rms had moved from cutthroat competition where both rms were losing money to a more cooperative equilibrium. Experience suggests, however, that cooperative equilibria are not always easy to sustain. Consider the difculties experienced by OPEC in trying to restrict output to maintain high oil prices or the American airline industry in avoiding price wars. It will be interesting to see if Boeing and Airbus engage in vigorous price competition in the future. Source: M. Freudenheim and J. Tagliabue (1998), Regulators Investigating Price Increases by Boeing and Airbus, New York Times (November 27), C1. even though they would be better off if both bid higha rivals dilemma. With repeated interaction, the equilibrium still might consist of both rms pricing low. However, other equilibria, such as both rms pricing high, are possible. The basic idea is that with repeated interaction, there is more to consider than the short-run payoffs. The decision maker also has to consider the potential benets of establishing a long-term cooperative relationship. The appendix to this chapter contains an example of how rivals dilemmas (such as in the Boeing/Airbus problem) sometimes can be overcome given repeated interaction.13 The likelihood of a cooperative outcome increases if (1) the long-run gains from cooperating are larger relative to the short-run gains from not cooperating (assuming the other rm does), (2) it is easier for the rms to recognize whether or not cooperation has occurred, and (3) the expected length of the repeated relationship is longer. Discount rates also are important (higher discount rates imply that the decision maker cares more about present payoffs relative to future payoffs). The appendix emphasizes the managerial implications of this analysis. In Chapter 21, we revisit this topic by discussing how reputational concerns can motivate managers to behave with integrity in economic transactions. Strategic Interaction and Organizational Architecture We have focused on competition between rival rms throughout this chapter, but the analysis of strategic interactions has much broader implications and can provide useful implications for managers when interacting with other parties within their own organization. Part 3 of this book examines organizational architecture. As an introductory example, consider the problem in Figure 9.7 between a manager, Kiana Ross, and a lower-level employee, Lenin Steenkamp. Len must decide whether to work or shirk. If he 13 Generally, it is important to have an uncertain ending date in the repeated relationship. If there is a nite ending date, both you and your rival are likely to realize that there is no incentive for either of you to cooperate in the last interaction. Then both of you will have an incentive not to cooperate in the next to the last interaction (since your reputation for cooperating will not matter in the last round). Yet if this is true, there is no incentive to cooperate in the interactions before. The nal result can be no cooperation in the rst interaction. If on the other hand the date of the nal interaction is uncertain, there always can be some reason to cooperate. The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Incentive Game The employee, Len Steenkamp, decides to work or shirk, and the manager, Kiana Ross, decides whether or not to monitor (payoffs in thousands of dollars). Len would prefer to shirk than to work, whereas it is costly for the manager to monitor. As the circle technique shows, there is no pure-strategy equilibrium. If the employee works, the manager does not want to monitor; if the manager does not monitor, the employee shirks. There is a mixed-strategy equilibrium where Len randomizes between working and shirking and Kiana randomizes between monitoring and not monitoring (all with a probability of .5). Economics of Strategy: Game Theory $5 $15 $0 $5 LenShirks 285 301 $15 $10 LenShirks KianaMonitor Figure 9.7 KianaNot monitor Chapter 9 KianaNot monitor II. Managerial Economics LenWorks KianaMonitor BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition $10 $10 LenWorks exerts effort, the value of his expected output is $25 (all numbers in this example are in thousands of dollars); if he shirks, the expected output is $5. Len does not like exerting effort and bears personal costs if he works that he values at $5, but no cost if he shirks. Kiana can monitor Len for a cost of $5 and can tell whether or not he has exerted effort.14 The labor contract pays the employee $10, unless he is caught shirkingin which case, he is red and receives no payment. The circle technique indicates that there is no pure-strategy equilibrium. The reason is that if Len always works, Kiana has no incentive to monitor, whereas if Kiana were never to monitor, Len always would shirk. This problem has an equilibrium in mixed strategies, where Len works some of the time and shirks others, and Kiana monitors some of the time and not others.15 MANAGERIAL APPLICATIONS AuditingA Mixed-Strategy Equilibrium Independent accountants commonly audit rms. One of the reasons audits exist is to help ensure that managers are not embezzling funds. Accountants do not verify all economic transactions when they audit a rm. Rather, they review most of the large transactions and draw a random sample of the smaller ones. If accountants reviewed every transaction, there would be little or no fraud (since managers would know they would get caught if they were fraudulent). However, given the correspondingly small chance of fraud, it would not make economic sense to bear the high costs of auditing every transaction. The result is a mixed-strategy equilibrium. Auditors randomly select accounts to audit and managers sometimes engage in fraud. 14 Kiana cannot tell for sure if Len has worked from simply observing his output. Although the expected output is $25 or $5, depending on whether or not Len works, output is also affected by random factors beyond Lens control. Thus output can sometimes be high even if Len has not worked, and low even if he has. 15 In this equilibrium, both Kiana and Len randomize between their two actions, choosing each action .5 of the time. 286 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 302 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Managerial Economics ANALYZING MANAGERIAL DECISIONS: Holland Sweetener versus Monsanto Aspartame is a low-calorie sweetener marketed by Monsanto under the name of NutraSweet. It was a major impetus to the rapid growth of Diet Coke and Diet Pepsi during the 1980s and 1990s. A scientist at the G. D. Searle & Co. rst discovered aspartame in 1965; Searle received a patent for the product in 1970. U.S. regulators did not approve its use in soft drinks until 1983. In 1985, Monsanto acquired Searleand with it a monopoly on aspartame. Monsantos patents expired in 1987 and 1992 in Europe and the United States, respectively. In 1986, Holland Sweetener was formed through a joint venture of Tosoh Corporation and Dutch State Mines. Its sole purpose was to challenge Monsanto in the aspartame market. It began by building a plant in the Netherlands to compete in the European market. The big prize, however, was the U.S. soft-drink market, which was to open up at the end of 1992. Initially, Holland Sweetener was quite optimistic about capturing a large share of the U.S. market. To quote their vice president of marketing and sales in referring to Coke and Pepsi, Every manufacturer likes to have at least two sources of supply. To Holland Sweeteners surprise, they never became a big player in the U.S. market. In 1992, just before Monsantos patent expired, Coke and Pepsi signed long-term contracts with Monsanto for the continued supply of NutraSweet. The big winners in this contract negotiation were Coke and Pepsi, who realized about $200 million a year in savings. Monsanto remained the major supplier to these companies, while Holland Sweetener was left pretty much out in the cold. Envision a pricing problem between Monsanto and Holland Sweetener in 1992 that led to the Monsanto contract. Assume (1) the cost to Holland Sweetener of entering the U.S. market, $25 million, has been incurred; (2) Monsanto and Holland Sweetener simultaneously choose to quote either a high or low price to Pepsi and Coke for aspartame; (3) if both Monsanto and Holland Sweetener quote the same price, Pepsi and Coke contract with Monsanto because customers are familiar with the NutraSweet label Holland Sweetener loses its initial investment; (4) if both rms submit a high price, Monsanto nets $300 million; (5) if both rms submit a low price, Monsanto nets $100 million; (6) if Monsanto prices high and Holland Sweetener prices low, Holland Sweetener nets $100 million (after the initial investment) and Monsanto nets $0. 1. Construct the strategic-form payoff matrix for this strategic pricing problem. Find the Nash equilibrium. 2. Now assume that the interaction is sequential where Holland Sweetener chooses to enter and if so they face the pricing problem in the second stage. Should Holland Sweetener enter? 3. Why do you think Holland Sweetener entered? Were they just dumb or were there other potential considerations? 4. Prior to Holland Sweeteners entry into the U.S. market, Pepsi and Coke began deemphasizing the NutraSweet label on their cans and bottles. Why do you think they did this? 5. Explain how Monsanto had a rst-movers advantage. 6. Pepsi and Coke were the big winners in this case. Explain why. SOURCE: A. Brandenburger and B. Nalebuff (1996), Co-opetition (Doubleday: Garden City, NY). Figure 9.7 indicates that the combined payoff is highest when Len always works and Kiana never monitors. Although this outcome is not a feasible equilibrium if the interaction is not repeated,16 Kiana might be able to promote this equilibrium within an actual rm. For instance, one tool that Kiana has at her disposal is the structure of the labor contract. It might be possible to motivate Len to work in the absence of 16 It is achieved one-quarter of the time in the mixed-strategy equilibrium (.5 .5). BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory Chapter 9 The McGrawHill Companies, 2009 Economics of Strategy: Game Theory 287 303 monitoring by paying him a share of total output, thus offering incentive compensation rather than a xed salary. Altering the labor contract creates a new set of payoffs and potentially a more preferred equilibrium. Part 3 of the book analyzes in detail how managers can use compensation plans and other organizational arrangements to increase productivity and the rms value. This example is but one of many ways that managers might apply these methods to analyze business problems. The chapter appendix provides another example. Summary Game theory is concerned with the general analysis of strategic interaction. It focuses on optimal decision making when all decision makers are presumed to be rational, with each attempting to anticipate the likely actions and reactions of its rivals. These techniques can be employed to study a wide variety of phenomena ranging from parlor games to politics and competitive strategy. In this chapter, we introduce the basic elements of this theory in the context of managerial decision making. In simultaneous-move problems, rms must make decisions without knowledge of the decisions made by their rivals. Nonrepeated means that the game is played only once. We diagram these interactions by showing the payoffs in strategic (normal) form. A dominant strategy exists when it is optimal for the rm to choose a particular strategy no matter what its rival does. A rm should employ a dominant strategy if one exists. Firms do not always have dominant strategies. When dominant strategies do not exist, we employ the concept of a Nash equilibrium to predict the outcome of the interaction. A Nash equilibrium is a set of strategies (or actions) in which each rm is doing the best it can, given the actions of its rival. A problem can have multiple Nash equilibria. Nash equilibria are not necessarily the outcomes that maximize the joint payoff of the rms. The circle technique provides a simple method to identify Nash equilibria. The power of a Nash equilibrium to predict the outcomes of strategic interactions stems from the fact that Nash equilibria are self-enforcingthey are stable outcomes. In many cases it is reasonable to expect that a Nash equilibrium will occur. This is particularly true when the rms have experience with similar problems, when they have information about each other, or when the Nash equilibrium outcome is a natural focal point. Some interactions are competitiveat least one of the rms has an incentive to take actions that benet them at the expense of their rival. Other strategic situations involve coordination rather than competition. Some interactions have elements of both. When a rm chooses one specic strategy or action, it is called a pure strategy. Sometimes it can pay to randomizeto use a mixed strategy. The benet of a mixed strategy comes from the element of surprise; sometimes a rm is at a disadvantage if it is too predictable. In many business situations, managers make decisions sequentially. Sequential interactions are pictured in extensive form. The extensive form displays the sequence of actions and corresponding outcomes. A node indicates a point at which a party must choose an action, and the branches leading from a node display the possible choices at the node. We solve the extensive form by backward inductionlooking forward to the nal decision nodes and reasoning backward. In some interactions it is advantageous to move rst: There is a rst-mover advantage. In other cases it is better to move second. Strategic moves are taken to inuence the beliefs or actions of the rival in favorable ways. Typically, strategic moves involve a rm restricting its own future actions. For strategic moves to work, they must be credible. For a strategic move to be credible, it must include a commitment sufcient to convince its rival to change its beliefs. 288 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 304 Part 2 II. Managerial Economics 9. Economics of Strategy: Game Theory The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Key Managerial Insights from Game Theory Most business problems are more complex than the simple examples used in this chapter. Although technical challenges mount as the number of rivals and the range of potential actions increase, some of the basic insights from elementary game theory are quite robust. Some of the key insights are as follows: Understand your business setting. Identify the relevant set of rivals and the nature of their interaction. What are the potential actions? What information will they have when they choose their actions? What are the consequences of their various actions? Is similar interaction among these rivals likely to be repeated over time or across other markets? Place yourself behind your rivals desk. Consider the entire sequence of decisions that are likely to be made over the course of this interaction. Look forward to the ultimate set of potential outcomes and then reason backward to determine your best strategy. This process identies critical choices that your rivals face and highlights why you should understand the basis for their choices. With a rst-mover advantage, move rst. If the business setting does not naturally permit you to implement your action rst, consider whether you can credibly precommit to a particular action. Effective precommitment, by convincing your rivals of your future actions, can induce them to change their actions. This logic highlights the fact that maintaining exibility undercuts your ability to precommitin this sense, exibility can be quite expensive. With a second-move advantage, avoid moving rst. Delay implementation of actions where possible. Try to reduce the predictability of your actions. Finally, if you have to implement an action, maintain as much future exibility to change your actions as possible. Repetition facilitates cooperation. Some productive form of cooperation is more likely when interaction is expected to be repeated either over time or across markets. Many strategic situations involve repeated interaction among rivals. When an interaction is expected to occur more than once, it is called a repeated interaction. Typically, more equilibria are possible with repeated interaction than in a nonrepeated problem. For instance, cooperation is a potential equilibrium in repeated rivals dilemmas (see the appendix to this chapter for a more detailed analysis). This chapter focuses primarily on competition between rival rms. The analysis of strategic interactions also provides managerial insights for interactions with other parties, such as suppliers and government regulators. It can be particularly useful in analyzing interactions with other employees within the rm. Appendix Repeated Interaction and the Teammates Dilemma17 Prisoners dilemmas occur when it is in the joint interests of the parties to cooperate but individual incentives motivate an equilibrium where they fail to cooperate. Situations of this type occur frequently both between and within business rms. This appendix provides an example of how cooperation sometimes can be achieved when the decision makers are confronted with this same dilemma on a repeated basis. The example highlights the factors that are important in determining whether cooperation will be achieved in a repeated setting. Managerial implications are discussed. 17 This appendix modies and extends an example in G. Miller (1992), Managerial Dilemma: The Political Economy of Hierarchy (Cambridge University Press: Cambridge), 184186. It requires knowledge of basic statistics. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Chapter 9 Figure 9.8 Economics of Strategy: Game Theory 289 305 Payoffs to Two Members in a Single-Period Setting $1000 $1000 AnneShirks AnneShirks The payoff on the upper left in each cell is the payoff for Anne, and the payoff on the lower right is for Bert. If Bert and Anne both shirk, they receive salaries of $1,000. If they both work hard, they receive a bonus. However, they experience disutility from working hard. The payoff, net of this disutility, is $2,000. If Bert shirks and Anne works hard, they meet their production target and receive a payoff of $3,000. Bert also receives a cash bonus. However, being the only one to exert effort, Anne incurs a back injury. Her net payoff is $0. The opposite payoff occurs if Bert works and Anne shirks. The Nash equilibrium in this single-period setting is for both to shirk (shaded cell). Given the payoffs, it is always in their individual interests to shirk. $0 $3000 BertShirks $0 BertWorks AnneWorks AnneWorks BertShirks $3000 $2000 $2000 BertWorks Much of the analysis in this chapter has focused on interaction between rival rms. This example focuses on the interaction between two employees assigned to a team and is chosen to illustrate the wide applicability of the analysis of strategic interactions. The Example Anne van Gastel and Bert Dijkstra work on a production team. They want to agree to a contract that both will work hard so that they can earn a bonus. They face a problem in that the contract is not legally binding and effort is not contractible. This problem is similar to the prisoners dilemma introduced in Chapter 6 as well as several examples in this chapter (for example, the pricing problem between Boeing and Airbus). Figure 9.8 displays the possible payoffs for this teammates dilemma. If Anne and Bert both shirk, they receive salaries of $1,000. If both work hard, they receive a bonus. However, they experience disutility from the additional effort. The payoffs, net of this disutility, are $2,000 each. If Bert shirks and Anne works hard, they meet their production target and receive bonus payments. Bert, however, experiences no disutility from working hard and receives a payoff of $3,000. Anne also receives a cash bonus. However, being the only one to exert effort, she incurs a back injury. Her net payoff is $0. The opposite payoffs occur if Bert works and Anne shirks. The Nash equilibrium in a single-period setting is for both to shirk. Given the payoffs, it is always in their individual interests to shirk. If Anne works, Bert is better off shirking since he receives $3,000 rather than $2,000. Similarly, if Anne shirks, Bert would rather receive the $1,000 from shirking than the $0 payoff from working. This same logic holds 290 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 306 Part 2 Figure 9.9 II. Managerial Economics The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Managerial Economics Generalized Payoffs for Two Members in a Multiperiod Setting $1000 1p BertAlways shirks $1000 + $1000 1p $2000 + $1000 1p BertAlways shirks AnneAlways shirks $1000 1p AnneTit-for-tat AnneTit-for-tat AnneAlways shirks Anne and Bert each consider two options in choosing their effort levels. One is to follow the strategy of always shirking every period. The other strategy, known as tit-for-tat, is to work hard for the rst period and thereafter mimic their teammates previous choice. For instance, if Bert works hard in the rst period and Anne shirks, Bert will shirk in the second period. If one person shirks in the rst period, then in all future periods both people shirk (the person who selects to shirk in the rst period has chosen the strategy of always shirking). The upper-left payoff in each cell is Annes, and the lower-right payoff is Berts. In a given work period, there is a probability p that they will work together in the next period. $2000 + $1000 1p $1000 + $1000 1p BertTit-for-tat $2000 1p $2000 1p BertTit-for-tat for Anne. This equilibrium outcome is not efcient. Both Bert and Anne would prefer the outcome where they both workthere, payoffs are $2,000 instead of $1,000. The problem is that they cant observe each others effort until after the work is complete. Now suppose that Bert and Anne expect to work together in the future. In particular, suppose that in a given period there is a probability p that they will work together in the next period.18 Thus, the probability that they will work together through n periods is p n 1. Now suppose that Bert and Anne each consider only two options in choosing their effort levels. One is to follow the strategy of always shirking every period. If both Bert and Anne select this strategy, the expected sum of each persons future earnings will be:19 E(future earning) $1,000 $1,000p $1,000p2. . . (9.1) This expression is an innite sum with a value of $1,000 [1 (1 p)]. The other strategy, known as tit-for-tat, is to work hard the rst period and thereafter mimic their teammates previous choice. For instance, if Bert works hard in the rst period and Anne shirks, Bert will shirk in the second period. If one person shirks in the rst period, then in all future periods both people shirk (the person who selects to shirk in the rst period has chosen the strategy of always shirking). Both Bert and Anne want to maximize the expected sum of their individual future earnings from the working relationship. Figure 9.9 shows the generalized payoffs in the 18 In each period, Bert and Anne choose an effort level, observe output, and receive compensation from the rm. 19 For simplicity, we ignore discounting future cash ows. Also, the formulation assumes that Bert and Anne have the possibility of living forever. Neither of these simplications is crucial for our analysis. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 291 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Chapter 9 Economics of Strategy: Game Theory 307 Figure 9.10 Payoffs for Two Team Members of a Production Team When the Likelihood of Working Together in the Future Is Low ( p 1 3) $1500 BertAlways shirks $500 $3500 AnneAlways shirks $1500 AnneTit-for-tat AnneTit-for-tat AnneAlways shirks Bert and Anne each consider two options in choosing their effort levels. One is to follow the strategy of always shirking every period. The other strategy, known as tit-for-tat, is to work hard for the rst period and thereafter mimic their teammates previous choice. For instance, if Bert works hard in the rst period and Anne shirks, Bert will shirk in the second period. If one person shirks in the rst period, then in all future periods both people shirk (the person who selects to shirk in the rst period has chosen the strategy of always shirking). The payoff on the lower left in each cell is Annes payoff, while the payoff on the lower right is Berts. In a given work period, there is a probability p that they will work together in the next period. In this example, the probability of working together in the future is relatively small (1 3). The circle technique indicates that the Nash equilibrium is for both to shirk (shaded cell). $3500 $500 BertTit-for-tat $3000 BertAlways shirks $3000 BertTit-for-tat setting. Unlike the single-period case, it does not always pay a person to shirk. Clearly, if Anne expects Bert to shirk, she will shirk as well, since $1,000 (1 p) is always greater than $1,000 $1,000 (1 p).20 However, if Anne thinks that Bert is going to choose the tit-for-tat strategy, it can be in her interest to do so as well. Her expected payoff from selecting tit-for-tat is $2,000 (1 p). Thus, she will select tit-for-tat whenever p is greater than 1 2since $2,000 (1 p) is greater than $2,000 $1,000 (1 p). Symmetric logic holds for Berts choice. Figure 9.10 and Figure 9.11 on page 308 show the payoffs for Bert and Anne for the cases where p 1 3 and p 3 4. When p 1 3, the Nash equilibrium is for both to shirkthe probability of repeated interaction is not large enough to promote cooperation. This example illustrates the general point that reputational concerns are unlikely to promote cooperation when the relationship is expected to be short-term. In the second case, two equilibria are possible: mutual shirking or mutual cooperation. The existence of multiple equilibria when p 3 4 suggests that Berts and Annes initial expectations are important. For instance, if Bert expects Anne to shirk, he will shirk as well. However, if he expects her to choose tit-for-tat, it makes sense for him to 20 The payoff from tit-for-tat is [$0 ($1,000p $1,000p2 . . .)] $1,000 $1,000 (1 p). $1,000p2 . . .)] [ $1,000 $1,000 ($1,000p 292 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 308 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Managerial Economics Figure 9.11 Payoffs for Two Team Members of a Production Team When the Likelihood of Working Together in the Future Is High ( p 3 4). $4000 BertAlways shirks $3000 $6000 BertAlways shirks AnneAlways shirks $4000 AnneTit-for-tat AnneTit-for-tat AnneAlways shirks Bert and Anne each consider two options in choosing their effort levels. One is to follow the strategy of always shirking every period. The other strategy, known as tit-for-tat, is to work hard for the rst period and thereafter mimic their teammates previous choice. For instance, if Bert works hard in the rst period and Anne shirks, Bert will shirk in the second period. If one person shirks in the rst period, then in all future periods both people shirk (the person who selects to shirk in the rst period has chosen the strategy of always shirking). The payoff on the lower left in each cell is Annes payoff, and the payoff on the upper right is Berts. In a given work period, there is a probability p that they will work together in the next period. In this example, the probability of working together in the future is relatively high (3 4). The arrow technique indicates that two Nash equilibria exist. One is mutual shirking; the other mutual tit-for-tat (both cells are shaded). $6000 $3000 BertTit-for-tat $8000 $8000 BertTit-for-tat select the same strategy. Anne has similar incentives. Thus, the efcient outcomefor both to workwill occur when there is a mutual expectation that both will work hard. This example suggests that businesses might promote cooperation by fostering particular expectations among employees. For instance, suppose the company publicizes in a credible manner that its employees have a long record of mutual cooperation. Given this corporate culture, it is reasonable for Bert to expect that Anne will select tit-for-tat. Anne will have similar expectations. Both will select tit-for-tat, and the corporate culture is reinforced. We have shown that the expected length of the relationship is important in determining the level of cooperation. So are the expected payoffs. For instance, if the payoffs from mutual shirking are increased to $1,500, the probability of working together in the next period must be 2 3 to promote cooperation. Alternatively, if the payoffs from mutual cooperation are $2,500, the required probability falls to 1 3. These examples illustrate the general principles that reputational concerns will work best at resolving incentive problems when the short-term gains from cheating are small and when the gains from continued cooperation are large. Another factor that is important is the likelihood of being caught shirking. In this example, shirking is observed perfectly before the next period. If Bert and Anne do not BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 9 293 The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Economics of Strategy: Game Theory 309 know for certain that the other person has shirked, they might continue to cooperate even if the other person has shirked. In this case, the temptation to shirk will be greater. Managerial Implications Recently, managers have delegated more work assignments to teams. Problems like the one in this example can reduce team output and the rms value. Managers can limit these problems and promote cooperation among team members by structuring rewards (for example, bonuses) that are high if team members cooperate and low if they do not. Also, managers must be careful not to change team composition too frequently: Incentive problems are larger when team members do not expect to work together in the future. Analyzing Managerial Decisions: Restructuring The BQM Company frequently restructures. Employees regularly are transferred among departments and given different job assignments. Management argues that this action promotes a better trained and more responsive workforce. Do you see potential problems with this type of frequent restructuring? Does this mean that BQM is making a mistake? Explain. Suggested Readings Self-Evaluation Problems A. Dixit and B. Nalebuff (1993), Thinking Strategically (W.W. Norton & Company: New York). J. McMillan (1996), Games, Strategies, & Managers (Oxford University Press: New York). 91. Time Magazine and Newsweek are two competing news magazines. Suppose that each company charges the same $5.00 price for their magazines. Each wants to maximize its sales given the $5.00 price. Each week, there are two potential cover stories. One is on politics. The other is on the economy. Sales of both companies are affected by the decisions on which story to place on their covers. The two magazines make this decision independently and at the same time. The resulting sales for the two companies are given in the following table: Time Cover Newsweek Cover Politics Politics Economy Economy Politics Economy Economy Politics a. b. c. d. Time Sales ($000s) Newsweek Sales ($000s) 400 700 200 300 150 200 150 700 Construct a diagram that shows the payoffs to the two rms in strategic (normal) form. What is the Nash equilibrium in this game? Does either or both of the magazines have a dominant strategy in this game? Suppose that both magazines are owned by the same publishing company that maximizes the combined prots of the magazines. Will the company make the same choice as in the noncooperative game? 92. a. Suppose that Newsweek chooses and announces its cover story before Time Magazine chooses its cover. Construct a diagram that shows the payoffs of the game in extensive form. b. What is the Nash equilibrium in this game? c. Is there a rst-mover advantage or rst-mover disadvantage in this game? The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Managerial Economics Solutions to Self-Evaluation Problems 91. Simultaneous Moves, Nonreported Interaction a. Below is the strategic form of the game where the two magazines make simultaneous moves (note that the circles and shaded quadrant are for part b of this problem). Time Politics $400 $150 Newsweek Politics $300 $700 Newsweek Politics $700 $200 Newsweek Economy Time Economy Part 2 II. Managerial Economics Time Politics 310 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Time Economy 294 $200 $150 Newsweek Economy b. The Nash equilibrium can be found by the circle technique. When Newsweek chooses Politics (rst column), Time does best by choosing Politicscircle the $400 payoff. When Newsweek chooses Economy (second column), Time does best by choosing Politicscircle the $700 payoff. When Time chooses Politics (rst row), Newsweek does best by choosing Economycircle the $200 payoff. When Time chooses Economy (second row), Newsweek does best by choosing Politicscircle the $700 payoff. The Nash equilibrium is for Time to choose Politics and Newsweek to choose Economy where both payoffs are circled. This is the shaded quadrant in the diagram. c. Time has a dominant strategy in this game. It wants to choose Politics no matter what Newsweek does. Newsweek does not have a dominant strategy. Its optimal choice between Politics and Economy depends on what Time chooses to do. d. The publishing company would place Economy on Times cover and Politics on Newsweeks cover. The combined prots are $1,000 versus $900 in the noncooperative game. 92. Sequential Interaction a. Below is the extensive form of the game, where Newsweek moves rst. Newsweek 150 Time 400 s ic olit P cs Time ti oli Eco nom y P Newsweek Newsweek 200 Time 700 s ic olit P Ec ono my Time Eco Newsweek 700 Time 300 nom y Newsweek 150 Time 200 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 9. Economics of Strategy: Game Theory Chapter 9 The McGrawHill Companies, 2009 Economics of Strategy: Game Theory 295 311 b. The Nash equilibrium (in bold/underlined in the gure) is for Newsweek to choose Economy and Time to choose Politics. This equilibrium is found as follows. Newsweek looks forward to Times choice and reasons backward. If Newsweek chooses Politics, Time will choose Politics and Newsweek will obtain 150 subscriptions. If it chooses Economy, Time will choose Politics and Newsweek obtains 200 subscriptions. Therefore it is in Newsweeks interest to choose Economy. Time will subsequently choose Politics. c. There is no rst-mover or second-mover advantage in this game. Time has a dominant strategy always to choose Politics. Given this choice, Newsweek wants to choose Economy. Therefore the equilibriums in the simultaneous and sequential games are the same. Review Questions 91. Some manufacturers that contract with the United States government have most favored nation clauses in their contracts. This provision makes the rm sell to the government at the lowest price it charges to any other customer. On the surface this provision seems to be advantageous to the government because it ensures them the lowest price charged to any customer. Others argue, however, that the clause gives manufacturers more power in bargaining with other buyers. Explain how this increased bargaining power might occur. 92. Suppose Microsoft can produce a new sophisticated software product. However, it wants to do so only if Intel produces high-speed microprocessors. Otherwise, the software will not sell. Intel, in turn, wants to produce high-speed microprocessors only if there is popular software on the market that requires high-speed processing. Is this a game of competition or coordination? What is the equilibrium? 93. What is the relation between a dominant strategy and a Nash equilibrium? 94. In this chapter we gave an example of coordination problems in the market for HDTVs. Show the game in strategic form using hypothetical payoffs of your choice. Use the arrow technique to identify the equilibria. 95. Some foolish teenagers play chicken on Friday nights. Two teenagers drive their cars at each other at high speeds. The rst to swerve to the side is the chicken and loses. If both swerve out of the way, they are both chickens and both lose. Neither of the drivers wants to get into an accident. It causes a signicant loss in utility (possibly death). However, both do not want to be known as a chicken. This causes some loss in utility. What is the equilibrium of this game? Do you think the two drivers will necessarily produce an equilibrium outcome? Do you think the chances are better or worse for achieving an equilibrium outcome if the two players know each other? Explain. Do you think it matters whether the two players have played the game before? Explain. 96. Two basketball players, Barbara and Juanita, are the best offensive players on the schools team. They know if they cooperate and work together offensivelyfeeding the ball to each other, providing screens for the other player, and the likethey can each score 12 points. If one player monopolizes the offensive game, while the other player cooperates, however, the player who monopolizes the offensive game can score 18 points, while the other player can only score 2 points. If both players try to monopolize the offensive game, they each score 8 points. Construct a payoff matrix for the players that captures the essence of the decision of Barbara and Juanita to cooperate or monopolize the offensive game. If the players play only once, what strategy do you expect the players to adopt? If the players expect to play in many games together, what strategy do you expect the players to adopt? Explain. 97. General Electric has frequently placed managers together to work on teams. Often the work assignment is only for a short period of time. General Electric makes sure that the quality of an employees performance on a given assignment is recorded and shared with future teams. Why do you think they do this? 98. Some managers commit undetected fraud in producing nancial statements. Presumably, if the auditors were really diligent and the penalties for fraud were high enough, there would be no fraud. Does this mean that the accounting rms are not doing a good enough job in auditing? Explain. 296 312 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics The McGrawHill Companies, 2009 9. Economics of Strategy: Game Theory Managerial Economics 99. A labor leader has announced that her union will go on strike unless you grant the workers a signicant pay raise. You realize that a strike will cost you more money than the pay raise. Should you concede to the wage increase? Explain. 910. Suppose you are one of two producers of tennis balls. Both you and your competitor have zero marginal costs. Total demand for tennis balls is P 60 Q where Q the sum of the outputs of you and your competitor. a. Suppose you are in this situation only once. You and your competitor have to announce your individual outputs at the same time. You expect your competitor to choose the Nash equilibrium strategy. How much will you choose to produce and what is your expected prot? b. Now suppose that you have to announce your output before your competitor does. How much will you choose to produce? What is your expected prot? Is it an advantage or a disadvantage to move rst? Explain. 911. You are considering placing a bid over the Internet in an eBay auction for a rare oriental rug. You are not a dealer in these rugs, and you do not have a precise estimate of its market value. You do not want to buy the rug for more than its market value. However, you would like to buy it if you can get it below the market value. You expect that many people will participate in the auction (including rug dealers). eBay asks that you give them the maximum bid you are willing to make. They will start low; whenever you are outbid, they will raise your bid just enough to lead the auction. eBay quits bidding on your behalf once your maximum price is reached. Your best guess at the market value is $1,000. What should you bid? 912. Formulate the following situation as an extensive form game (using a game tree) and solve it using backward induction. Bingo Corporation and Canal Corporation are the only competitors in the electronic organizer industry. Bingo Corporation is considering an R&D investment to improve its product. Bingo can choose from three levels of investment: High, Medium, and Low. Following Bingos investment, Canal Corporation will have to choose between continuing to compete by selling its current product or undertaking an R&D project of its own. Canal can only choose one level of investment, so its choices are Invest or Not Invest. The net payoffs to Bingo if it invests High, Medium, or Low given that Canal chooses to Invest would be $50, $40, and $30, respectively, and the corresponding net payoffs to Canal would be $5, $10, and $15. On the other hand, the net payoffs to Bingo if it invests High, Medium, or Low given that Canal chooses to Not Invest would be $100, $80, and $60, respectively, and the corresponding net payoffs to Canal would be $0, $15, and $20. What will Bingo choose to do in equilibrium, and what will Canals response be? BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 10. Incentive Conflicts and Contracts The McGrawHill Companies, 2009 297 Incentive Conicts and Contracts CHAPTER 10 CHAPTER OUTLINE Firms Incentive Conicts within Firms Owner-Manager Conicts Other Conicts Controlling Incentive Problems through Contracts Costless Contracting Costly Contracting and Asymmetric Information Postcontractual Information Problems Precontractual Information Problems Implicit Contracts and Reputational Concerns Incentives to Economize on Contracting Costs Summary O ne of the largest takeovers in history occurred in 1988the purchase of RJR-Nabisco by Kohlberg, Kravis, Roberts & Company. Public accounts report lavish expenditures and decisions of questionable merit by RJR executives preceding the takeover. For example, Burrough and Helyar in their best seller Barbarians at the Gate write, It was no lie. RJR executives lived like kings. The top 31 executives were paid a total of $14.2 million, or an average of $458,000. Some of them became legends at the Waverly for dispensing $100 tips to the shoeshine girl. [Ross] Johnsons two maids were on the company payroll. No expense was spared decorating the new headquarters, highlighted by the top-oor digs of the top executives. It was, literally, the sweet life. A candy cart came around twice a day dropping off bowls of bonbons at each oors reception areas. Not Baby Ruths but ne French confections. The minimum perks for even lowly middle managers was one club membership and one company car, worth $28,000. The maximum, as nearly as anyone could tell, was Johnsons two dozen club memberships and John Martins $105,000 Mercedes. In addition, it appears that major investment decisions at RJR often were driven by the preferences of managers rather than by value maximization. For instance, Ross Johnson, chief executive ofcer of RJR, reportedly continued to invest millions of dollars in developing a smokeless cigarette long after it was obvious that the project would never be protable. 298 314 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition Part 2 II. Managerial Economics 10. Incentive Conflicts and Contracts The McGrawHill Companies, 2009 Managerial Economics More recently other executives have been charged with even more extreme examples of excessive behavior. For example, in July 2002 the U.S. Justice Department charged that John Rigas and his two sons looted Adelphia on a massive scale and used it as a personal piggy bank. Over the past ve decades Rigas had built a small cable company in rural Pennsylvania into Adelphia Communicationslisted on Nasdaq and the sixth largest U.S. cable company. The governments complaint charged the Rigas family with using Adelphias assets to buy company stock, timberland, New York condos, and a National Hockey League team; to build a golf course on family-owned property; to bankroll movies produced by a Rigas daughter; and to pay for automobiles on behalf of a Rigas-owned dealership.1 The behavior of RJR and Adelphia executives raises at least four interesting issues: In previous chapters, we assumed that managers always maximize prots. Apparently, they do not. To understand management problems within the rm, we need a richer characterization of the rm and managerial decision making. Both RJR and Adelphia suggest that material conicts of interest can exist between owners and managers: Shareholders are interested in the rms value, whereas the managers are interested in their own utility. What other conicts of interest exist within rms? These cases suggest that owner-manager conicts can result in reduced productivity and waste. Unchecked, such conicts of interest can destroy a rm. How do rms limit unproductive actions to enhance value and avoid failure? If techniques to limit unproductive actions exist, why did the owners (shareholders) at RJR and Adelphia allow the managers to engage in such dysfunctional behavior? In this chapter, we examine these and related issues. We begin by enriching our understanding of the denition of a rm. We then use this more explicit understanding to discuss various conicts of interest that exist within rms. Next, we examine how contracts help reduce or control these conicts. We focus particular attention on the problems created by costly information. Finally, we discuss how reputational concerns can control incentive conicts within rms. Firms In Chapter 3, we characterized the rm in terms of administrative decision making: Markets use prices to allocate resources; rms use managers. This prompted a discussion of the relative efciency of rms and markets. Thus far in the book, we have treated the rm as if it had one central manager who acts to maximize the rms value. This characterization is employed widely in economics and has proved quite useful in explaining production and pricing decisions. The actual decision-making process within rms, however, is extremely complex and differs from this simple characterization in at least three ways. First, there are many decision makers within rms. In large corporations, the board of directors makes major policy decisions such as naming the CEO. The CEO, in turn, retains certain important 1 D. Lieberman and G. Farrell (2002), Five Former Adelphia Arrested, USA Today ( July 24); and R. Grover (2002), Adelphia vs. Deloitte in a Game of Blame, BusinessWeek Online ( June 27). Rigas subsequently was sentenced to 15 years in prison. See R. Farzad (2005), Jail Terms for 2 at Top of Adelphia, New York Times ( June 21), C1. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics The McGrawHill Companies, 2009 10. Incentive Conflicts and Contracts Chapter 10 Incentive Conicts and Contracts 299 315 Figure 10.1 The Firm as a Focal Point for a Set of Contracts The rm is a creation of the legal system that has the standing of an individual in a court of law. The rm serves as one party to the many contracts that make up the rm. Employees Suppliers Labor unions Insurance providers The firm Stockholders Bondholders Banks Customers decision rights while delegating many operating decisions (for instance, pricing, production, and nancing decisions) to lower-level managers. Even the lowest-paid employee in the rm usually has some decision-making authority. Second, the primary objective of most of these decision makers is not to maximize the value of the rm: The investment behavior of the RJR executives certainly suggests interests in things other than value maximization. Third, rms often use internal pricing systems (transfer prices) to allocate internal resources. Analyzing organizational issues within the rm requires a richer concept of the rm. Several useful denitions have been developed by economists.2 We focus on one denition that is particularly useful for our purposes:3 The rm is a focal Denition point for a set of contracts. This denition focuses on the fact that the The rm is a focal point for a set of rm ultimately is a creation of the legal system; it has been granted the contracts. legal standing of an individual (it can enter contracts, sue, be sued, and so on). The term focal point indicates that the rm always is one of the parties to each of the many contracts that constitute the rm. Examples of these contracts are employee contracts, supplier contracts, customer warranties, stock, bonds, loans, leases, franchise agreements, and insurance contracts. This contract view of the rm is illustrated in Figure 10.1. Some contracts are explicit legal documents, whereas many others are implicit. And even within a relationship that has been formalized with an explicit contract, there is a broad array of aspects of the relationship that are not spelled out within the written agreementthey are implicit. An example of an implicit contract is an employees understanding that if a job is done well, it will result in a promotion. Implicit contracts are often difcult to enforce in a court of law. Later in this chapter, we discuss how reputational concerns can help ensure that individuals honor implicit contracts. 2 O. Hart (1989), An Economists Perspective on the Theory of the Firm, Columbia Law Review 89, 17571774. M. Jensen and W. Meckling (1976), Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, Journal of Financial Economics 3, 305360. 3 300 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 316 Part 2 II. Managerial Economics 10. Incentive Conflicts and Contracts The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Enforceability of Implicit Contracts While implicit contracts are common, some lawyers argue they arent worth the paper they are not written on. Even though such contracts are hazy, most companies strive to avoid breaches because of their threat to productivity and the resulting employee turnover such breaches can engender. Lawyers advise that if you are offered an implicit contract, beware of managers promising something they cannot guarantee, such as lifetime employment. Also, try to document meetings where expectations are discussed and implicit promises are made. Such records can prove a useful paper trail if future litigation results. Source: M. Culp (1998), Implicit Contracts: Not Worth the Paper They Are Not Written On, Democrat and Chronicle (August 16), 1G. Incentive Conicts within Firms4 Economic theory characterizes individuals as creative maximizers of their own utility. Thus, the collection of individuals that contract with the rm are not likely to have objectives that are automatically aligned. The owners of the rm have title to the residual prots (what is left over after other claimants are paid) and are likely to be interested in maximizing the present value of these prots. Other individuals within the rm do not share this goal necessarily. We now discuss some of the more important incentive conicts that arise within rms. We then discuss how contracts can be used to reduce and control these conicts. Owner-Manager Conicts Owners often delegate the management of rms to professional managers. For instance, in large corporations, the residual prots are owned by shareholders who delegate signicant decision authority to top executives. At least ve sources of conict arise between owners and managers: Choice of effort. Additional effort by managers generally increases the value of the rm, but since the managers expend the effort, additional effort reduces their utility. Perquisite taking. It is in the interests of owners to pay sufcient salaries and bonuses to attract and retain competent managers. However, owners do not want to overpay managers. In contrast, managers are likely to want not only higher salaries but also perquisites such as exclusive club memberships, lavish ofce furniture, luxurious automobiles, stimulating day care for children, and expensive French confections. Differential risk exposure. Managers typically have substantial levels of human capital and personal wealth invested in the rm. This large investment can make managers appear excessively risk-averse from the standpoint of the owners, who 4 This section draws on M. Jensen and C. Smith (1985), Stockholder, Manager, and Creditor Interests: Applications of Agency Theory, in E. Altman and M. Subrahmanyam (Eds.), Recent Advances in Corporate Finance (Irwin Professional Publishers: Burr Ridge, IL), 95131. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 10. Incentive Conflicts and Contracts Chapter 10 The McGrawHill Companies, 2009 Incentive Conicts and Contracts 301 317 MANAGERIAL APPLICATIONS The Spectrum of Organizations The rm can be viewed as a focal point for a set of contracts. One particularly important feature of these contracts is the distribution of the residual prots. Organizations vary remarkably along this dimension. In a sole proprietorship like Espteins Deli, the owner/manager is the residual claimant. In a partnership like the law rm of Nixon-Peabody, the claims are shared by the partners. In a large public corporation like Amazon.com, these claims often are held by thousands of shareholders who take little direct interest in managing the company. In a mutual like the Prudential Insurance Company, ownership and customer claims are merged. In a cooperative like Ocean Spray, supplier and ownership claims are merged. In an employee-owned rm like United Airlines, the claims are owned by the employees. Finally, in a not-for-prot institution like the American Red Cross there are no owners of the residual cash ows. In most of these large organizations, management authority is delegated to professional managers, who often have small or no ownership positions in the organization. (See the Web appendix to Chapter 18, http://www.mhhe.com/brickley5.) According to Coase, individuals have incentives to select the form of organization that minimizes total contracting costs (see Chapter 3). Our discussion of conicts between owners and managers suggests that problems arise in public corporations because of the separation of ownership and control. These problems are costly to resolve. Given these costs, what are the offsetting benets that promote the prominence of large corporations? One of the most signicant benets is that capital is raised from many investors who share in the risk of the company. Individual shareholders place only a small amount of their wealth in a given company, and thus avoid placing all their eggs in one basket. This diversication makes risk-averse investors (see Chapter 2) willing to supply capital to corporations at a lower cost. This benet, however, comes at the cost of having to control the incentive conicts between managers and owners. Thus, in smaller operations, where raising large amounts of capital is less of an issue, one should expect to nd sole proprietorships and small partnerships (where there is less separation of ownership and control). Indeed, this is what is observed.* *Note: There are also tax-related reasons that affect the choice of organizational form. See M. Scholes, M. Wolfson, M. Erickson, E. Mayden, and T. Shevlin (2008), Taxes and Business Strategy (Prentice-Hall: Englewood Cliffs, NJ). (at least in a large public corporation) typically invest only a small fraction of their wealth in any one rm.5 Hence, managers might forgo projects that they anticipate would be protable simply because they do not want to bear the risk that the project might fail and lead to a reduction in their compensation. Differential horizons. Managers claims on the corporation generally are limited by their tenure with the rm. Therefore, managers have limited incentives to care about the cash ows that extend beyond their tenure. Owners, on the other hand, are interested in the value of the entire future stream of cash ows, since it determines the price at which they can sell their claims in the company. Overinvestment. Managers can be reluctant to reduce the size of a rm, even if it has exhausted available protable investment projects; they prefer to empirebuild. Also, managers often are understandably reluctant to lay off colleagues and friends in divisions that are no longer protable. Managers who re their colleagues bear personal costs (disutility), whereas shareholders receive most of the benets. 5 To be more precise, we do not assume that the underlying preferences (utility functions) of owners and managers differ. Rather we focus on the fact that the risk of owners claims on public rms can be managed more easily through diversication than can those of managers. The most valuable component of most managers wealth is their human capital, and managers typically have but one job. 302 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 318 Part 2 II. Managerial Economics 10. Incentive Conflicts and Contracts The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Buyer-Supplier Conicts Immediately after the United States invaded Iraq, the U.S. military hired Halliburton to supply the U.S. troops in Iraq with supplies. One contract involved the provision of gasoline. The contract terms stated that Halliburton would be reimbursed for its costs plus receive a 1 percent prot margin. The problem with such cost-plus contracts is that they provide little incentive for the supplier to control costs, or to pick the most competitive subcontractors, since the higher the costs, the higher the prots to the supplier (Halliburton in this case). The Pentagon investigators charged that Halliburton purchased gasoline at almost double the market value after only receiving bids from two other gasoline suppliers. In 2005, the Defense Contract Audit Agency of the Pentagon and the U.S. Congress investigated the charges without coming to a denitive resolution. Source: N. King, Jr. (2004), Halliburton Says Two Employees Took Kickbacks for Iraq Contracts, The Wall Street Journal ( January 23), A1, A5; R. Klein (2005), Audit Questions $1.4b in Halliburton Bills, Boston Globe ( June 28), www.boston.com. Other Conicts Similar types of incentive conicts are likely to arise among most contracting parties in the rm. For example, top managers worry about effort and perquisite-taking problems with lower-level employees. The rms creditors and shareholders can have disputes over the optimal dividend, nancing, and investment policies of the rm. Firms can have incentives to default on warranties with customers. Managers often quarrel with labor unions. For example, Alcatel-Alsthom SA, a French conglomerate, was unable to divest any of its low-margin plants without engendering an uproar from its unions.6 Owners of rms would like to acquire high-quality inputs at low prices, whereas owners of supplying rms would like to provide inexpensive inputs at high prices. This tension produces conicts between buyers and suppliers. Supplying rms worry about buying rms demanding price concessions, and buying rms worry that suppliers will either shirk on quality (to reduce cost) or raise prices. In Chapter 19, we provide a detailed analysis of buyer-supplier relationships. Incentive conicts also arise with joint ownership. For example, in a large accounting rm, the actions of each partner affect the prots of the organization, which are shared among the partners. This arrangement can motivate partners to free-ride on the efforts of others. Each partner hopes the other partners will work diligently to keep the rm protable. However, each partner has an incentive to shirk: Individual partners gain the full benet of their shirking but bear only part of the costs (their share of the reduced prots). Free-rider problems are common in most group activities and, if left unchecked, greatly reduce the output of teams. (We shall refer to such free-rider problems throughout this book.) Controlling Incentive Problems through Contracts What keeps these incentive conicts from undermining cooperative undertakings and destroying all organizations? For example, might the fear that managers will use all company resources for their personal benet dissuade owners from delegating operating 6 D. Lavin (1998), Union and Regulators Restrain Alcatels Restructuring, The Wall Street Journal (August 7), A8. BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics 303 The McGrawHill Companies, 2009 10. Incentive Conflicts and Contracts Chapter 10 Incentive Conicts and Contracts 319 MANAGERIAL APPLICATIONS Experimental Evidence on Free-Rider Problems More than 50 years ago a German scientist named Ringelmann asked workers to pull as hard as they could on a rope attached to a meter that measured the strength of their efforts. Subjects worked alone and in groups of two, three, and eight. While the total amount of force on the rope increased as group size rose, the amount of effort by each person seemed to drop. While one person pulling alone exerted an average of 63 kg of force, this dropped to about 53 kg in groups of three and was reduced to about 31 kg in groups of eight. The greater the number of people performing the task, the less effort each one expended. The impact of any social force directed toward a group from an outside source (for example, a manager) is divided among its members. Thus, the more persons in the group, the less the impact such force will have upon each. Because they are working with others, each group member feels [that others] will take up any slack resulting from reduced effort on their part. And since all members tend to respond in this fashion, average output per person drops sharply. Source: A. Furnham (1993), Wasting Time in the Board Room, Financial Times (March 10). authority to managers?7 Fortunately, there are mechanisms that help control incentive conicts. Among the most important are contracts.8 Contracts (both implicit and explicit) dene the rms organizational architecture its decision right, performance evaluation, and reward systems. This architecture provides an important set of constraints and incentives that helps resolve incentive problems. For instance, if a contract species that Erin OMalley, the rms chief nancial ofcer, will receive an annual salary of $200,000, she can be red if she unilaterally pays herself more: She does not have the decision right to set her own compensation.9 If Erin is evaluated on rm prots and rewarded with a large bonus for good performance, she has incentives to care about the rms prots. Costless Contracting Under some circumstances, contracts can resolve incentive problems at low cost. As an example, consider Jerold Concannon, CEO of the Bagby Printing Company. Jerry gains utility U, from both his monetary compensation C and perquisites P such as company expenditures on luxury cars and club memberships: U f (C, P ) (10.1) If the rm provides no perquisites to Jerry, it must pay him a salary S in cash compensation; otherwise, he will work for another rm. The owners of the rm are willing to pay Jerry S if he consumes no perquisites. However, as CEO, Jerry has numerous 7 In fact, some authors suggest that this concern ultimately will cause the collapse of the public corporation. See A. Berle and G. Means (1932), The Modern Corporation and Private Property (Macmillan: New York). 8 Other important mechanisms are the market for corporate control and the product market. Managers have incentives to increase a rms prots because rms with inefcient managers can be taken over by other rms and the management team replaced. Indeed, this is what happened at RJR-Nabisco. Also, inefcient rms eventually go out of business in a competitive market. 9 Restricting an agents decision-making authority can reduce incentive problems. However, it also can mean that authority has not been granted to the individual with the best knowledge to make the decision. This tension is a fundamental concern in designing organizational architecture and is a central focus of Chapter 11. 304 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 320 Part 2 II. Managerial Economics The McGrawHill Companies, 2009 10. Incentive Conflicts and Contracts Managerial Economics MANAGERIAL APPLICATIONS Incentive Conicts throughout the World Incentive conicts are not just an American business phenomenon, nor do they occur only in private rms. Rather, these conicts exist throughout the world in both the private and public sectors. For example, government ofcials taking bribes is an example of a basic incentive conict between government ofcials and the people they represent. In 1999, Indonesian President B. J. Habibies political party, Golkar, allegedly siphoned off $66 million from a nationalized bank, PT Bank Bali, to help fund the presidents reelection campaign. After the allegations were made, the government hired PriceWaterhouseCoopers to audit the transactions. They cited numerous indications of fraud.a A similar event occurred in China where a bureaucrat confessed to using a safe-deposit box in Hong Kong to stash away more than $1.2 million in cash bribes from companies attempting to get pieces of real-estate deals she controlled. Meanwhile, in the banking sector, the premier of China is attempting to change the common practice by bankers of ignoring credit standards and awarding loans to people with connections. In one case, eight bankers received stiff sentences, including the death penalty, for accepting bribes in return for loans.b a J. Solomon (1999), Bali High Jinks: In Indonesia, Crisis and Corruption Create Financial Vigilantes, The Wall Street Journal (September 21), A1. b J. Barnathan (1993), A CrackdownOf Sorts, BusinessWeek ( July 26), 47. opportunities to consume company resources. These opportunities present an incentive problem: Jerry wants to spend company resources on himself, whereas the owners do not want Jerry to reduce the rms value by consuming excessive perquisites. As we will see, some amount of perquisites actually increases value. But beyond this level of perquisite consumption, value falls. Suppose for now that the owners of the rm have precise knowledge of the maximum potential prot of the rm, M (if Jerry is paid S and consumes no perquisites). In this case, realized prots of the rm, R (if he is paid S and consumes perquisites, P ) are the difference between maximum prots and Jerrys excess perk consumption: R M P (10.2) With this information, the owners can solve the potential incentive problem by offering Jerry the following compensation contract: C S ( M R) (10.3) This contract, which reduces Jerrys salary by the difference between realized and maximum prots, is equivalent to charging Jerry the full cost of his perquisitesthat is,C S P. Figure 10.2 displays Jerrys choice of perquisites. His objective is to maximize his utility subject to the constraint that he is paid according to his compensation plan. Jerry chooses the combination (C *, P *), which occurs at the tangency point between his indifference curve and the compensation constraint. This combination places Jerry on the highest indifference curve possible given the compensation plan. This choice is Pareto-efcient. The owners are indifferent to Jerrys choice: They always pay the equivalent of S (they pay him S P in cash and P in perquisites). Jerry, however, is better off being able to choose the combination of salary and perks that he prefers. For example, Jerry had a back injury and prefers a more expensive, ergonomically designed desk chair to the companys standard ofce furniture, even though he knows that his compensation is reduced to reect the additional expense. Also, Jerry might BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition II. Managerial Economics Chapter 10 Figure 10.2 Optimal Perquisite Taking Incentive Conicts and Contracts 321 $ S Cash compensation The manager is paid a cash salary S, as long as the manager maximizes prots. If the manager fails to maximize prots by taking perquisites (for example, too many club memberships, paying excessive salaries to top subordinates, or buying expensive company cars), the owners reduce the compensation by the amount of the lost prots. Given this compensation scheme, the manager chooses the combination (C*, P*). This choice is Pareto-efcient. 305 The McGrawHill Companies, 2009 10. Incentive Conflicts and Contracts Compensation plan C* Slope = 1 $ P* S Expenditures on perquisites prefer a combination of salary and perquisites to a pure salary because perks frequently are untaxed.10 Note how the compensation plan aligns Jerrys and the owners incentives. Jerry is given the decision right to choose how much the rm spends on his perquisites. The contract, however, charges Jerry the full cost of his perk consumption. In essence, he is rewarded for consuming fewer perquisites. This reward structure gives him private incentives to limit his perk consumption. In this analysis the owners establish the rms total compensation expense associated with Jerrys employment, S, Jerry chooses his optimal level of perk consumption, P *, and this determines his cash compensation, C * ( S P *). This mechanism probably seems far-fetched. Yet such an outcome can be approximated closely through negotiation. If at the beginning of the period Jerry negotiates for a perquisite level of P *, the owners recognize that their total compensation expense will be the sum of his cash compensation plus his perk consumption. Thus, they will set his cash compensation at C * ( S P *). This example suggests that some perk taking by managers is likely to be efcient (from the standpoint of both the managers and the rm) because some perks increase productivity and because of the differential tax treatment for perquisites and salary. 10 This tax effect is reected in Jerrys indifference curves. Over some range, Jerry is willing to trade more than a dollar of cash for a dollars worth of perquisites because on an after-tax basis, he is better off. Over this range, the slope of the indifference curve has a slope with an absolute value greater than one. The optimal choice of salary and perks occurs at the point where the indifference curves slope is 1. 306 BrickleySmithZimmerman: Managerial Economics and Organizational Architecture, Fifth Edition 322 Part 2 II. Managerial Economics 10. Incentive Conflicts and Contracts The McGrawHill Companies, 2009 Managerial Economics MANAGERIAL APPLICATIONS Jack Welchs Perquisites In 1996, General Electric entered into an employment agreement with its then CEO, Jack Welch, to keep him working until 2000. Rather than taking a special one-time payment of tens of millions of dollars, Mr. Welch accepted an agreement that gave him lifetime access to G.E.s planes, cars, ofces, apartments, and nancial planning. This is an example of a CEO trading off cash compensation and perquisites. Since retiring Mr. Welch has used a palatial Manhattan apartment complete with wine, owers, cook, housekeeper and other amenities, as well as access to General Electrics Boeing 737 jets, helicopters and a car and driver for Mr. Welch and his wife. Also included were tickets for the couple at a number of top sporting events and the opera. Securities and Exchange Commission rules requ