1 FeenTayTrade2e_CH07_Layout 8/7/10 2:20 PM Page 199
Offshoring of Goods and Services
7
1 A Model of
Offshoring
2 The Gains from
Offshoring
One facet of increased services trade is the increased use of offshore outsourcing in which a company relocates labor-intensive service-industry functions to another country. . . . When a good or service is produced more cheaply abroad, it makes more sense to import it than to make or provide it
domestically.
Economic Report of the President, 2004, p. 229
3 The Politics and
Future of
Offshoring
4 Conclusions
Increasing numbers of Americans . . . perceive offshoring . . . as an actual or potential threat to
their jobs or to their wages even if they hold onto their jobs.
Jagdish Bhagwati and Alan S. Blinder, 2007, Offshoring of American Jobs
We should . . . encourage . . . businesses to stay within our borders, its time to finally slash the tax
breaks for companies that ship our jobs overseas.
President Barack Obama, State of the Union address, January 27, 2010
I
f you take the battery out of your cell phone to see where the phone was produced, you will likely see several countries listed inside. Motorola, for example, is a
U.S. company that produces some of its cell phones in Singapore using batteries and
a battery charger made in China. Nokia is a Finnish company that produces some of
its American-sold cell phones in the United States using batteries made in Japan and
software that was written in India. Apple produces its iPhone in facilities found in
China, Taiwan, Thailand, Malaysia, Singapore, South Korea, the Czech Republic,
Philippines, and the United States. A vast array of products, including simple toys
like the Barbie doll and sophisticated items like airplanes and personal computers,
consist of materials, parts, components, and services that are produced in numerous
countries. The following excerpt from a New York Times article illustrates this
1
observation:
1
Louis Uchitelle, Why Hasnt a Weak Dollar Slowed Imports? New York Times, April 8, 2005, online
edition.
199
FeenTayTrade2e_CH07_Layout 1 9/21/10 11:27 AM Page 200
200 Part 3
New Explanations for International Trade
General Electric in the United States and Snechma of France . . . jointly manufacture the jet engine for Boeings 737 and Airbuss 320. G.E. makes the hot section at its plant in Cincinnati, while Snechma manufactures the giant fans in
France. They ship these components to each other and each partner does the
final assembly of the engines for its customers. In addition, G.E. makes smaller
jet engines for the commuter planes that Bombardier makes in Canada and Embraer makes in Brazil. The engines are exported to those countries, but 24 percent of the value of the engines comes from components imported from Japan.
The provision of a service or the production of various parts of a good in different
countries that are then used or assembled into a final good in another location is called foreign outsourcing or, more simply, offshoring. We will not worry about the distinction
between these two terms (see Side Bar: Foreign Outsourcing versus Offshoring),
2
but will use offshoring because it has become most commonly used by economists.
Offshoring is a type of international trade that differs from the type of trade analyzed
in the Ricardian and Heckscher-Ohlin models; the goods traded in those models were
final goods. Offshoring is trade in intermediate inputs, which can sometimes cross borders several times before being incorporated into a final good that can be sold domes3
tically or abroad. Offshoring is a relatively new phenomenon in world trade. The
amount of world trade relative to the GDPs of countries was high even in the late nineteenth and early twentieth centuries. But it is unlikely that a good would have crossed
borders multiple times at several stages of production because the costs of transportation and communication were too high. Today, however, these costs have fallen so much
that it is now economical to combine the labor and capital resources of several countries to produce a good or service. Indeed, if you have ever called for help with your laptop, chances are that you have spoken with someone at a call center in India, which
shows just how low the costs of communication have become!
Is offshoring different from the type of trade examined in the Ricardian and
Heckscher-Ohlin models? From one point of view, the answer is no. Offshoring allows
a company to purchase inexpensive goods or services abroad, just as consumers can purchase lower-priced goods from abroad in the Ricardian and Heckscher-Ohlin models.
This is what the quote from the Economic Report of the President at the beginning of the
chapter suggests: with offshoring we import those goods and services that are cheaper to
produce abroad. From another point of view, however, offshoring is different. Companies now have the opportunity to send a portion of their activities to other countries. The
jobs associated with those activities leave the United States, and by paying lower wages
abroad, U.S. firms lower their costs and pass on these savings to consumers. Offshoring
results in lower prices but changes the mix of jobs located in the United States. Higherskilled workers in the United States, engaged in activities such as marketing and research,
will be combined with less skilled workers abroad, engaged in assembling products. In a
sense, offshoring is similar to immigration in that U.S. firms are able to employ foreign
workers, even though those workers do not have to leave their home countries.
2
Both foreign outsourcing and offshoring have been coined recently to describe this new type of international trade. The earliest known use of the word outsourcing is a quotation from an American auto executive in the Journal of the Royal Society of Arts, 1979, who said, We are so short of professional engineers
in the motor industry that we are having to outsource design work to Germany (William Safire, On Language, New York Times Magazine, March 21, 2004, p. 30).
3
There is also the concept of domestic outsourcing, which occurs when a company decides to shift some of its
production activities from one location to another within the same country. In this text, outsourcing always
means foreign outsourcing.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 201
Chapter 7
Offshoring of Goods and Services 201
SIDE BAR
Foreign Outsourcing versus Offshoring
In discussions of foreign outsourcing, we often hear the term
offshoring. The quote from the Economic Report of the
President at the beginning of the chapter combined these
terms as offshore outsourcing. Is there a difference between
foreign outsourcing and offshoring?
The term offshoring is sometimes used to refer to a company moving some of its operations overseas but retaining
ownership of those operations. In other words, the company
moves some operations offshore but does not move production
outside of its own firm. Intel, for example, produces microchips in China and Costa Rica using subsidiaries that it
owns. Intel has engaged in foreign direct investment (FDI) to
establish these offshore subsidiaries.
Mattel, on the other hand, arranges for the production of the
Barbie doll in several different countries. Unlike Intel, however, Mattel does not actually own the firms in those countries.
Furthermore, Mattel lets these firms purchase their inputs (like
the hair and cloth for the dolls) from whichever sources are
most economical. Mattel is engaging in foreign outsourcing as
it contracts with these firms abroad but has not done any FDI.
Dell is an intermediate case. Dell assembles its computers
overseas in firms it does not own, so it is outsourcing rather
than offshoring the assembly. However, Dell exercises careful
control over the inputs (such as computer parts) that these
overseas firms use. Dell outsources the assembly but monitors
the overseas firms closely to ensure the high quality of the
computers being assembled.
In this chapter, we will not worry about the distinction between offshoring and foreign outsourcing and will use the
term offshoring whenever the components of a good or service are produced in several countries, regardless of who owns
the plants that provide the components or services.
The first goal of this chapter is to examine in detail the phenomenon of offshoring and
describe in what ways it differs from trade in final products. We discuss how offshoring
affects the demand for high-skilled and low-skilled labor and the wages paid to those
workers. Since the early 1980s, there has been a significant change in the pattern of wage
payments in the United States and other countriesthe wages of skilled workers have
been rising relative to those of less skilled workers. We will examine whether this change
in relative wages is the result of offshoring or whether there are other explanations for it.
A second goal of the chapter is to discuss the gains from offshoring. We argue that
offshoring creates gains from trade, similar to those seen from the trade of final goods
in the Ricardian or Heckscher-Ohlin models. But having overall gains from trade for
a country does not necessarily mean that every person in the country gains. As the second quote at the beginning of the chapter shows, many workers are fearful that their
jobs and wages are threatened by offshoring. We focus attention on how offshoring affects high-skilled versus low-skilled workers.
A third goal of the chapter is to examine the response to offshoring in the United
States. The final quotation at the beginning of the chapter, from the 2010 State of the
Union address by President Obama, indicates that he proposes to limit tax breaks to
companies engaged in offshoring. Many economists would disagree with this proposal,
and argue instead that offshoring has overall benefits. We examine these arguments
and also discuss the newest trend of inshoring activities back into the United States.
1 A Model of Offshoring
To develop a model of offshoring, we need to identify all the activities involved in producing and marketing a good or service. These activities are illustrated in Figure 7-1.
Panel (a) describes the activities in the order in which they are performed (starting with
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 202
202 Part 3
New Explanations for International Trade
FIGURE 7-1
(a) Activities Ranked by Order in Production
Finish
Start
R&D
Component
production
Assembly
Marketing
and sales
(b) Activities Ranked by High-skilled/Low-skilled Labor
A
Done in Foreign
Done at Home
High-skilled
Low-skilled
Assembly
Component
production
Marketing
and sales
R&D
The Value Chain of a Product Any product has many different activities involved in its
manufacture. Panel (a) lists some of these activities for a given product in the order in which they
occur. The value chain in (b) lists these same activities in order of the amount of high-skilled/lowskilled labor used in each. In panel (b), the assembly activity, on the left, uses the least skilled
labor, and R&D, on the right, uses the most skilled labor. Because we assume that the relative wage
of skilled labor is higher at Home and that trade and capital costs are uniform across activities, there
is a point on the value chain, shown by line A, below which all activities are offshored to Foreign and
above which all activities are performed at Home.
research and development [R&D] and ending with marketing and after-sales service).
For instance, in producing a television, the design and engineering are developed first;
components such as wiring, casing, and screens are manufactured next; and finally the
television is assembled into its final version and sold to consumers.
For the purpose of building a model of offshoring, however, it is more useful to line
up the activities according to the ratio of high-skilled/low-skilled labor used, as in panel
(b). We start with the less skilled activities, such as the manufacture and assembly of
simple components (like the case or the electric cord for the television), then move to
more complex components (like the screen). Next are the supporting service activities
such as accounting, order processing, and product service (sometimes called back-office activities because the customer does not see them). Finally, we come to activities
that use more skilled labor, such as marketing and sales (front-office activities), and
those that use the most skilled labor such as R&D.
Value Chain of Activities
The whole set of activities that we have illustrated in Figures 7-1(a) and 7-1(b) is sometimes called the value chain for the product, with each activity adding more value to
the combined product. All these activities do not need to be done in one countrya
firm can transfer some of these activities abroad by offshoring them when it is more
economical to do so. By lining up the activities in terms of the relative amount of skilled
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 203
Chapter 7
Offshoring of Goods and Services 203
labor they require, we can predict which activities are likely to be transferred abroad.
This prediction depends on several assumptions, which follow.
Relative Wage of Skilled Workers Let WL be the wage of low-skilled labor in
*
*
Home and WH the wage of high-skilled labor. Similarly, let W L and W H be the wages
of low-skilled and high-skilled workers in Foreign. Our first assumption is that Foreign
*
*
wages are less than those at Home, W L < WL and W H < WH, and that the relative wage
*
*
of low-skilled labor is lower in Foreign than at Home, so W L /W H < WL /WH. This
assumption is realistic because low-skilled labor in developing countries receives especially low wages.
Costs of Capital and Trade As the firm considers sending some activities abroad,
it knows that it will lower its labor costs because wages in Foreign are lower. However,
the firm must also take into account the extra costs of doing business in Foreign. In
many cases, the firm pays more to capital through (1) higher prices to build a factory
or higher prices for utilities such as electricity and fuel; (2) extra costs involved in transportation and communication, which will be especially high if Foreign is still developing roads, ports, and telephone capabilities; and (3) the extra costs from tariffs if Foreign
imposes taxes on goods (such as component parts) when they come into the country.
We lump together costs 2 and 3 into what we call trade costs.
Higher capital and trade costs in Foreign can prevent a Home firm from offshoring all its activities abroad. In making the decision of what to offshore, the Home
firm will balance the savings from lower wages against the extra costs of capital
and trade.
Our second assumption is that these extra costs apply uniformly across all the activities in the value chain; that is, these extra costs add, say, 10% to each and every component of operation in Foreign as compared with Home. Unlike our assumption about
relative wages in Home and Foreign, this assumption is a bit unrealistic. For instance,
the extra costs of transportation versus those of communication are quite different in
countries such as China and India; good roads for transport have developed slowly,
while communications technology has developed rapidly. As a result, technology in
telephones is advanced in those countries, so cell phones are often cheaper there
than in the United States and Europe. In this case, the higher infrastructure costs
will affect the activities that rely on transportation more than activities that rely on
communication.
S licing the Value Chain Now suppose that the Home firm with the value chain in
Figure 7-1(b) considers transferring some of these activities from Home to Foreign.
*
*
Which activities will be transferred? Based on our assumptions that W L /W H < WL /WH
and that the extra costs of capital and trade apply uniformly, it makes sense for the
firm to send abroad the activities that are the least skilled and labor-intensive and
keep at Home the activities that are the most skilled and labor-intensive. Looking at
Figure 7-1, all activities to the left of the vertical line A might be done in Foreign, for
example, whereas those activities to the right of the vertical line will be done in Home.
4
We can refer to this transfer of activities as slicing the value chain.
4
This term is drawn from Paul Krugman, 1995, Growing World Trade: Causes and Consequences, Brookings Papers on Economic Activity, 1.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 204
204 Part 3
New Explanations for International Trade
Activities to the left of line A are sent abroad because the cost savings from paying
lower wages in Foreign are greatest for activities that require less skilled labor. Because the
extra costs of capital and trade are uniform across activities, the cost savings on wages are
most important in determining which activities to transfer and which to keep at Home.
Relative Demand for Skilled Labor Now that we know the division of activities
between Home and Foreign, we can graph the demand for labor in each country, as illustrated in Figure 7-2. For Home, we add up the demand for high-skilled labor H and
low-skilled labor L for all the activities to the right of line A in Figure 7-1(b). Taking
the ratio of these, in panel (a) we graph the relative demand for skilled labor at Home
H/L against the relative wage WH /WL. This relative demand curve slopes downward
because a higher relative wage for skilled labor would cause Home firms to substitute
less skilled labor in some activities. For example, if the relative wage of skilled labor increased, Home firms might hire high school rather than college graduates to serve on
a sales force and then train them on the job.
*
In Foreign, we add up the demand for high-skilled labor H and for low-skilled labor
*
*
L for all the activities to the left of line A . Panel (b) graphs the relative demand for
*
*
*
*
skilled labor in Foreign H /L against the relative wage W H /W L. Again, this curve
slopes downward because a higher relative wage for skilled labor would cause Foreign
firms to substitute less skilled labor in some activities. In each country, we can add a
relative supply curve to the diagram, which is upward-sloping because a higher relative
wage for skilled labor causes more skilled individuals to enter this industry. For
instance, if the high-skilled wage increases relative to the low-skilled wage in either
country, then individuals will invest more in schooling to equip themselves with the
skills necessary to earn the higher relative wage.
FIGURE 7-2
(a) Home Country
High-skilled/
low-skilled
wage, WH/WL
Home supply
(b) Foreign Country
High-skilled/
low-skilled
**
wage, WH/WL
Foreign supply
A*
A
Home demand
High-skilled/low-skilled
labor, H/L
Relative Demand and Supply for High-Skilled/Low-Skilled
Labor In panel (a), we show the relative demand and supply for
skilled labor at Home, H/L, depending on the relative wage, WH /WL.
The equilibrium relative wage at Home is determined at A. In panel
Foreign demand
High-skilled/low-skilled
labor, H*/L*
(b), we show the relative demand and supply for skilled labor in
Foreign, H* L*, depending on the relative wage, W * /W *. The Foreign
/
H
L
equilibrium is at point A*.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 205
Chapter 7
Offshoring of Goods and Services 205
The intersection of the relative demand and relative supply curves, at points A and
A , gives the equilibrium relative wage in this industry in each country and the equilibrium relative employment of high-skilled/low-skilled workers. Starting at these
points, next we study how the equilibrium changes as Home offshores more activities
to Foreign.
*
Changing the Costs of Trade
Suppose now that the costs of capital or trade in Foreign fall. For example, the North
American Free Trade Agreement (NAFTA) lowered tariffs charged on goods crossing
the U.S.-Mexico border. This fall in trade costs made it easier for U.S. firms to offshore
to Mexico. And even before NAFTA, Mexico had liberalized the rules concerning foreign
ownership of capital there, thereby lowering the cost of capital for U.S. firms. Another
example is India, which in 1991 eliminated many regulations that had been hindering
businesses, communications, and foreign investment. Before 1991 it was difficult for a
new business to start, or even to secure a phone or fax line in India; after 1991 the regulations on domestic and foreign-owned business were simplified, and communication
technology improved dramatically with cell phones and fiber-optic cables. These policy
changes made India more attractive to foreign investors and firms interested in offshoring.
Change in Home Labor Demand and Relative Wage When the costs of capital
or trade decline in Foreign, it becomes desirable to shift more activities in the value
chain from Home to Foreign. Figure 7-3 illustrates this change with the shift of the
dividing line from A to B. The activities between A and B, which used to be done at
Home, are now done in Foreign. As an example, consider the transfer of television
production from the United States to Mexico. As U.S. firms first shifted manufacturing to Mexico, the chassis of the televisions were constructed there. Later on, electronic circuits were constructed in Mexico, and later still the picture tubes were
5
manufactured there.
FIGURE 7-3
A
Done in Foreign
B
Done at Home
Low-skilled
Assembly
High-skilled
Component
production
Marketing
and sales
R&D
Offshoring on the Value Chain As the costs of capital or trade fall in the Foreign country, a Home
firm will find it profitable to offshore more activities. Offshoring shifts the dividing line between Home
and Foreign production from A to B. The activities between A and B, which formerly were done at Home,
are now done in Foreign. Notice that these activities are more skill-intensive than the activities formerly
done in Foreign (to the left of A) but less skill-intensive than the activities now done at Home (to the
right of B).
5
Martin Kenney and Richard Florida, 1994, Japanese Maquiladoras: Production Organization and Global
Commodity Chains, World Development, 22(1), 2744.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 206
206 Part 3
New Explanations for International Trade
How does this increase in offshoring affect the relative demand for skilled labor in
each country? First consider the Home country. Notice that the activities no longer
performed at Home (i.e., those between A and B) are less skill-intensive than the activities still done there (those to the right of B). This means that the activities now done
at Home are more skilled and labor-intensive, on average, than the activities formerly
done at Home. For this reason, the relative demand for skilled labor at Home will increase, and the Home demand curve will shift to the right, as shown in Figure 7-4,
panel (a). Note that this diagram does not show the absolute quantity of labor demanded,
which we expect would fall for both high-skilled and low-skilled labor when there is
more offshoring; instead, we are graphing the relative demand for high-skilled/lowskilled labor, which increases because the activities still done at Home are more skillintensive than before the decrease in trade and capital costs. With the increase in the
relative demand for skilled labor, the equilibrium will shift from point A to point B at
Home; that is, the relative wage of skilled labor will increase because of offshoring.
C hange in Foreign Labor Demand and Relative Wage Now lets look at what
happens in Foreign when Home offshores more of its production activities to Foreign.
How will offshoring affect the relative demand for labor and relative wage in Foreign?
As we saw in Figure 7-3, the activities that are newly offshored to Foreign (those between A and B) are more skill-intensive than the activities that were initially offshored
FIGURE 7-4
(a) Home Country
(b) Foreign Country
As the least skill-intensive activities
are outsourced, Home relative
demand for skilled labor increases.
High-skilled/
low-skilled
wage, WH/WL
Home supply
New outsourced tasks
increase the relative demand
for skilled labor in Foreign
High-skilled/
low-skilled
**
wage, WH/WL
Foreign supply
B*
B
A*
A
Home demand
High-skilled/low-skilled
labor, H/L
Change in the Relative Demand for High-Skilled/LowSkilled Labor With greater offshoring from Home to Foreign,
some of the activities requiring less skill that were formerly done
at Home are now done abroad. It follows that the relative demand
for skilled labor at Home increases, and the relative wage rises
Foreign demand
High-skilled/low-skilled
labor, H*/L*
from point A to point B. The relative demand for skilled labor in
Foreign also increases because the activities shifted to Foreign are
more skill-intensive than those formerly done there. It follows
that the relative wage for skilled labor in Foreign also rises, from
point A* to point B*.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 207
Chapter 7
Offshoring of Goods and Services 207
to Foreign (those to the left of A). This means that the range of activities now done in
Foreign is more skilled and labor-intensive, on average, than the set of activities formerly done there. For this reason, the relative demand for skilled labor in Foreign also
increases, and the Foreign demand curve shifts to the right, as shown in panel (b) of Figure 7-4. With this increase in the relative demand for skilled labor, the equilibrium
*
*
shifts from point A to point B . As a result of Homes increased offshoring to Foreign,
then, the relative wage of skilled labor increases in Foreign. The conclusion from our
model is that both countries experience an increase in the relative wage of skilled labor
because of increased offshoring.
It might seem surprising that a shift of activities from one country to the other can
increase the relative demand for skilled labor in both countries. An example drawn from
your classroom experience might help you to understand how this can happen. Suppose
you have a friend who is majoring in physics but is finding it difficult: he is scoring
below average in a class he is taking. So you invite him to join you in an economics
class, and it turns out that your friend has a knack for economics: he scores above average in that class. How does your friends transfer from physics to economics affect the
class averages? Because your friend was performing below average in the physics class,
when he leaves the class, his departure raises the class average (computed now using the
students still there, not including him). Because he performs better than average in the
economics class, his arrival raises the class average there, too (computed using everyone, including him). Thus, this students move from one class to another raises the average in both classes.
This result is just like the logic of the offshoring model: as activities in the middle
of the value chain are shifted from Home to Foreign, they raise the relative demand for
skilled labor in both countries because these activities are the least skill-intensive of
those formerly done at Home but the most skill-intensive of tasks done in Foreign. That
is why the relative demand for skilled labor increases in both countries, along with the
relative wage of skilled labor. This result is one of the most important predictions from
our offshoring model and it would not occur in our earlier models of trade, such as the
6
Heckscher-Ohlin model. We now turn to evidence from the United States and Mexico to see whether this prediction is borne out.
APPLICATION
Change in Relative Wages across Countries
Since the early 1980s, the wages of high-skilled workers have risen relative to those of
low-skilled workers in many countries. The relative wage of skilled workers in industrial
countries (such as the United States, Australia, Canada, Japan, Sweden, and the United
Kingdom) and in developing countries (such as Hong Kong, Chile, and Mexico) has increased. Our offshoring model predicts that the relative wage of skilled workers will rise
in both the country doing the offshoring and the country receiving the new activities. At
first glance, that prediction seems to be consistent with the change in wages that has actually occurred. Let us dig more deeply, however, using evidence from the United States
and Mexico, to see what the change in wages has been and whether it is due to offshoring.
6
The Heckscher-Ohlin model tells us that the factor prices in the two countries will move toward equality
when they open trade. So the wage relative to the capital rental will move in different directions in the two
countries due to the opening of trade, not in the same direction.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 208
New Explanations for International Trade
Change in Relative Wages in the United States
To measure the wages of skilled and low-skilled workers, we can use data from the manufacturing sector on production and nonproduction workers. As their name suggests, production workers are involved in the manufacture and assembly of goods,
whereas nonproduction workers are involved in supporting service activities. Firms are
required to report wages for both types of workers. We could also call these two types
of workers blue collar and white collar. Generally, nonproduction workers require
more education, and so we will treat these workers as high-skilled, whereas the pro7
duction workers are treated here as low-skilled workers.
Relative Wage of Nonproduction Workers Figure 7-5 shows the average annual
earnings of nonproduction workers relative to production workers (analogous to the
ratio of high-skilled to low-skilled wages, or WH /WL) in U.S. manufacturing from 1958
to 2006. We see that relative earnings moved erratically from 1958 to 1967, and that
from 1968 to about 1983, relative wages were on a downward trend. It is generally accepted that the relative wage fell during this period because of an increase in the supply of college graduates, skilled workers who moved into nonproduction jobs (the
increase in supply would bring down the nonproduction wage, so the relative wage
FIGURE 7-5
Relative 1.8
wage
1.75
1.7
1.65
1.6
1.55
20
05
20
00
19
95
19
90
19
85
19
80
19
75
19
70
1.5
19
65
19
60
208 Part 3
Relative Wage of Nonproduction/Production Workers, U.S. Manufacturing This
diagram shows the average wage of nonproduction workers divided by the average wage of
production workers in U.S. manufacturing. This ratio of wages moved erratically during the 1960s
and 1970s, although showing some downward trend. This trend reversed itself during the 1980s
and 1990s, when the relative wage of nonproduction workers increased until 2000. This trend
means that the relative wage of production, or low-skilled, workers fell during the 1980s and
1990s. In more recent years the relative wage has become quite volatile, falling substantially in
2002 and 2004, then rising in 2005 and 2006.
Source: Annual Survey of Manufacturers and National Bureau of Economic Research (NBER) productivity database, updated from
U.S. Bureau of the Census.
7
This distinction is far from perfect, however. Nonproduction workers include clerical and custodial staff,
for example, who may be less skilled than some production workers.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 209
Offshoring of Goods and Services 209
would also fall). Starting in 1983, however, this trend
reversed itself and the relative wage of nonproduction
workers increased steadily to 2000. Since that time the
relative wage has become quite volatile.
Relative Employment of Nonproduction Workers
In Figure 7-6, we see that there was a steady increase in
the ratio of nonproduction to production workers employed in U.S. manufacturing until the early 1990s.
Such a trend indicates that firms were hiring fewer production, or low-skilled workers, relative to nonproduction workers. Then during the 1990s the ratio of
nonproduction to production workers fell. Most recently, the ratio has increased slightly.
The increase in the relative supply of college graduates from 1968 to early 1983 is consistent with the reduction in the relative wage of nonproduction workers, as shown in Figure 7-5, and with
the increase in their relative employment, as shown in Figure 7-6. But after 1983 the
story changes. We would normally think that the rising relative wage of nonproduction
workers should have led to a shift in employment away from nonproduction workers, but
it did not; as shown in Figure 7-6, the relative employment of nonproduction workers
continued to rise from 1980 to about 1990, then fell until 1997. How can there be both
FIGURE 7-6
Relative 0.5
employment
0.45
0.4
0.35
05
20
00
20
95
19
90
19
85
19
80
19
75
19
70
19
65
19
19
60
0.3
Relative Employment of Nonproduction/Production Workers, U.S. Manufacturing This
diagram shows the employment of nonproduction workers in U.S. manufacturing divided by the
employment of production workers. There was a steady increase in the ratio of nonproduction to
production workers employed in U.S. manufacturing until the early 1990s. That trend indicates firms
were hiring fewer production workers relative to nonproduction workers. During the 1990s there was a
fall in the ratio of nonproduction to production workers, and then a rise again after 2000.
Source: Annual Survey of Manufacturers and NBER productivity database, updated from U.S. Bureau of the Census.
David Horsey. Represented by Tribune Media Services (http://www.Comicspage.com).
Chapter 7
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 210
210 Part 3
New Explanations for International Trade
an increase in the relative wage and an increase in the relative employment of nonproduction workers? The only explanation consistent with these facts is that during the 1980s
there was an outward shift in the relative demand for nonproduction (skilled) workers,
which led to a simultaneous increase in their relative employment and in their wages.
This conclusion is illustrated in Figure 7-7, in which we plot the relative wage of nonproduction workers and their relative employment from 1979 to 1990. As we have already noted, both the relative wage and relative employment of nonproduction workers
rose during the 1980s. The only way this pattern can be consistent with a supply and
demand diagram is if the relative demand curve for skilled labor increases, as illustrated.
This increased demand would lead to an increase in the relative wage for skilled labor and
an increase in its relative employment, the pattern seen in the data for the United States.
Explanations What factors can lead to an increase in the relative demand for skilled
labor? One explanation is offshoring. An increase in demand for high-skilled workers,
at the expense of low-skilled workers, can arise from offshoring, as shown by the rightward shift in the relative demand for skilled labor in Figure 7-4(a). The evidence from
the manufacturing sector in the United States is strongly consistent with our model
of offshoring.
There is, however, a second possible explanation for the increase in the relative demand for skilled workers in the United States. In the 1980s personal computers began
to appear in the workplace. The addition of computers in the workplace can increase
the demand for skilled workers to operate them. The shift in relative demand toward
FIGURE 7-7
Nonproduction/
production wage
Supply
Demand, 1979
1.65
1990
1989
1.60
1984
1980
1.55
1979
1.50
0.2
0.25
0.3
0.35
1983
1981
0.4
1988
1987
1985
Demand, 1990
1986
1982
0.45
0.5
0.55
0.6
Nonproduction/production labor
Supply and Demand for Nonproduction/
Production Workers in the 1980s This
diagram shows the average wage of nonproduction
workers divided by the average wage of production
workers on the vertical axis, and on the horizontal
axis the employment of nonproduction workers
divided by the employment of production workers.
Both the relative wage and the relative employment
of nonproduction, or skilled, workers rose in U.S.
manufacturing during the 1980s, indicating that
the relative demand curve must have shifted to
the right.
Source: NBER productivity database.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 211
Chapter 7
Offshoring of Goods and Services 211
skilled workers because of the use of computers and other high-tech equipment is called
skill-biased technological change. Given these two potential explanations for the
same observation, how can we determine which of these factors was most responsible
for the actual change in wages?
Answering this question has been the topic of many research studies in economics. The approach that most authors take is to measure skill-biased technological
change and offshoring in terms of some underlying variables. For skill-biased technological change, for example, we might use the amount of computers and other
high-technology equipment used in manufacturing industries. For offshoring, we
could use the imports of intermediate inputs into manufacturing industries. By studying how the use of high-tech equipment and the imports of intermediate inputs have
grown and comparing this to the wage movements in industries, we can determine the
contribution of each factor toward explaining the wage movements.
The results from one study of this type are shown in Table 7-1. The goal of this
study was to explain two observations. First, it sought to explain the increase in the
share of total wage payments going to nonproduction (high-skilled) labor in U.S.
manufacturing industries from 1979 to 1990 (part A). Because wage payments equal
the wage times the number of workers hired, this statistic captures both the rising relative wage and the rising relative employment of skilled workers. Second, the study analyzed the increase in the relative wage of nonproduction labor in particular over the
8
same period (part B).
The study considered two possible explanations for the two observations: offshoring
and the use of high-tech equipment such as computers. Offshoring was measured as the
intermediate inputs imported by each industry. For example, the U.S. auto industry
builds seats, dashboards, and other car parts in Mexico and then imports these for assembly in the United States. In addition, high-technology equipment can be measured
in two ways: either as a fraction of the total capital equipment installed in each industry or as a fraction of new investment in capital that is devoted to computers and other
high-tech devices. In the first method, high-tech equipment is measured as a fraction
of the capital stock, and in the second method, high-tech equipment is measured as a
fraction of the annual flow of new investment. In the early 1980s a large portion of the
flow of new investment in some industries was devoted to computers and other hightech devices, but a much smaller fraction of the capital stock consisted of that equipment. In Table 7-1, we report results from both measures because the results differ
depending on which measure is used.
Using the first measure of high-tech equipment (as a fraction of the capital stock), the
results in the first row of part A show that between 20 and 23% of the increase in the
share of wage payments going to the nonproduction workers can be explained by offshoring, and between 8 and 12% of that increase can be explained by the growing use
of high-tech capital. The remainder is unexplained by these two factors. Thus, using
the first measure of high-tech equipment, it appears that offshoring was more important
than high-tech capital in explaining the change in relative demand for skilled workers.
8
To illustrate the distinction between parts A and B, consider the following example. If nonproduction workers earn $25 per hour and 5 are hired, and production workers earn $10 per hour and 20 are hired, then the
total wage payments are $25 5 + $10 20 = $325, and the relative wage of nonproduction labor is $25/$10
= 2.5. In contrast, the share of total wage payments going to nonproduction workers is then $125/$325 =
38%, which equals the actual, average share of wages going to nonproduction workers in U.S. manufacturing from 1979 to 1990.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 212
212 Part 3
New Explanations for International Trade
TABLE 7-1
Increase in the Relative Wage of Nonproduction Labor in U.S. Manufacturing,
19791990 This table shows the estimated effects of offshoring and the use of high-technology
equipment on the wages earned by nonproduction (or skilled) workers. Part A focuses on how these
two variables affect the share of wage payments going to nonproduction workers. Part B shows how
these two variables affect the relative wage of nonproduction workers.
PERCENT OF TOTAL INCREASE EXPLAINED BY EACH FACTOR
Offshoring
High-Technology Equipment
Part A: Share of Wage Payments Going to Nonproduction Workers
Measurement of high-tech equipment:
As a share of the capital stock
As a share of capital flow
(i.e., new investment)
2023
812
13
37
Part B: Relative Wage of Nonproduction/Production Workers
Measurement of high-tech equipment:
As a share of the capital stock
2127
As a share of capital flow
(i.e., new investment)
12
2932
99
Source: Robert C. Feenstra and Gordon H. Hanson, August 1999, The Impact of Outsourcing and High-Technology Capital on Wages:
Estimates for the United States, 19791990, Quarterly Journal of Economics, 114(3), 907940.
The story is different, however, if we instead use offshoring and the second measure
of high-tech equipment (a fraction of new investment), as shown in the second row of
part A. In that case, offshoring explains only 13% of the increase in the nonproduction
share of wages, whereas high-tech investment explains 37% of that increase. So we see
from these results that both offshoring and high-tech equipment are important explanations for the increase in the relative wage of skilled labor in the United States, but
which one is most important depends on how we measure the high-tech equipment.
In part B, we repeat the results but now try to explain the increase in the relative
wage of nonproduction workers. Using the first measure of high-tech equipment
(a fraction of the capital stock), the results in the first row of part B show that between
21 and 27% of the increase in the relative wage of nonproduction workers can be explained by offshoring, and between 29 and 32% of that increase can be explained by the
growing use of high-tech capital. In the second row of part B, we use the other measure of high-tech equipment (a fraction of new investment). In that case, the large
spending on high-tech equipment in new investment can explain nearly all (99%) of
the increased relative wage for nonproduction workers, leaving little room for offshoring to play much of a role (it explains only 12% of the increase in the relative wage).
These results are lopsided enough that we might be skeptical of using new investment
to measure high-tech equipment and therefore prefer the results in the first rows of
parts A and B, using the capital stocks.
Summing up, we conclude that both offshoring and high-tech equipment are important explanations for the increase in the relative wage of nonproduction/production
labor in U.S. manufacturing, but it is difficult to judge which is more important because
the results depend on how we measure the high-tech equipment.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 213
Chapter 7
Offshoring of Goods and Services 213
Change in Relative Wages in Mexico
Our model of offshoring predicts that the relative wage of skilled labor will rise in both
countries. We have already seen (in Figure 7-1) that the relative wage of nonproduction (skilled) labor rises in the United States. But what about for Mexico?
In Figure 7-8, we show the relative wage of nonproduction/production labor in Mexico from 1964 to 1994. The data used in Figure 7-8 come from the census of industries
in Mexico, which occurs infrequently, so there are only a few turning points in the
graph. We can see that the relative wage of nonproduction workers fell from 1964 to
1985, then rose until 1994. The fall in the relative wage from 1964 to 1985 is similar
to the pattern in the United States and probably occurred because of an increased supply of skilled labor in the workforce. More important, the rise in the relative wage of
nonproduction workers from 1985 to 1994 is also similar to what happened in the
United States and illustrates the prediction of our model of offshoring: that relative
wages move in the same direction in both countries.
After 1994 the change in the relative wage of nonproduction workers in Mexico depends on whether we look at the maquiladora sector, which are the industrial plants near
Mexicos border with the United States engaged in offshoring, or the non-maquiladora
plants. For the maquiladora sector, the evidence in Figure 6-8(b) in Chapter 6 showed
that real monthly income (including production workers and nonproduction workers)
rises faster than real wages (for production workers). That evidence means there is a
continuing rise in the relative wage of nonproduction (skilled) workers, at least in the
FIGURE 7-8
Relative 3.1
wage
2.7
2.3
1.9
94
92
19
90
19
88
19
86
19
84
19
82
19
80
19
78
19
76
19
74
19
72
19
70
19
68
19
66
19
19
19
64
1.5
Relative Wage of Nonproduction/Production Workers, Mexico Manufacturing This
diagram shows the wage of nonproduction workers in the manufacturing sector of Mexico divided by
the wage of production workers. After declining during the 1960s and 1970s, this relative wage
began to move upward in the mid-1980s, at the same time that the relative wage of nonproduction
workers was increasing in the United States (see Figure 7-5). The relative wage in Mexico continued
to rise until 1994, when NAFTA began.
Sources: Robert C. Feenstra and Gordon H. Hanson, May 1997, Foreign Direct Investment and Relative Wages: Evidence from Mexicos
Maquiladoras, Journal of International Economics, 4, 371393; and Raymond Robertson, December 2004, Relative Prices and Wage
Inequality: Evidence from Mexico, Journal of International Economics, 64, 387409.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 214
214 Part 3
New Explanations for International Trade
maquiladora sector. For the non-maquiladora plants in the rest of Mexico, however, the
evidence is that the relative wage of nonproduction workers fell after 1994.
The year 1994 is special because that is when the North American Free Trade Agreement (NAFTA) was formed, establishing free trade between the United States, Canada,
and Mexico. Tariffs between the United States and Mexico were reduced starting in
1994: the tariffs in some products were eliminated that year, and other tariff reductions
were phased in over five or ten years. According to one research study, the tariff reductions in Mexico tended to favor the import of skill-intensive goods from the United
9
States. So after 1994 the prices of those goods fell in Mexico, which might explain
why the relative wage of nonproduction labor also fell in the non-maquiladora plants.
Summing up, the changes in relative prices in the United States and Mexico match
each other during the period from 1964 to 1985 (with relative wages falling) and during the period from 1985 to 1994 (with relative wages rising in both countries). Offshoring from the United States to Mexico rose from 1985 to 1994, so the rise in relative
wages matches our prediction from the model of offshoring. After 1994 the relative
wages move in opposite directions in the maquiladora and non-maquiladora plants in
Mexico, which appears to be due to the start of NAFTA.
2 The Gains from Offshoring
We have shown that offshoring can shift the relative demand for labor and therefore
raise the relative wage for skilled workers. Because the relative wage for low-skilled
workers is the reciprocal of the high-skilled relative wage, it falls in both countries.
High-skilled labor gains and low-skilled labor loses in relative terms. On the other
hand, the ability of firms to relocate some production activities abroad means that their
costs are reduced. In a competitive market, lower costs mean lower prices, so offshoring
benefits consumers. Our goal in this section is to try to balance the potential losses
faced by some groups (low-skilled labor) with the gains enjoyed by others (high-skilled
labor and consumers). Our argument in previous chapters on the Ricardian and
Heckscher-Ohlin models has been that international trade generates more gains than
losses. Now we have to ask whether the same is true for offshoring.
One answer to this question comes from a surprising source. The Nobel laureate
Paul Samuelson has been among the foremost proponents of global free trade, but in
10
2004 he had the following to say about the gains from foreign outsourcing:
Most noneconomists are fearful when an emerging China or India, helped by their
still low real wages, outsourcing and miracle export-led developments, cause layoffs from good American jobs. This is a hot issue now, and in the coming decade,
it will not go away. Prominent and competent mainstream economists enter in the
debate to educate and correct warm-hearted protestors who are against globalization. Here is a fair paraphrase of the argumentation that has been used. . . .
Yes, good jobs may be lost here in the short run. But still total U.S. net national
product must, by the economic laws of comparative advantage, be raised in the long run
(and in China, too). The gains of the winners from free trade, properly measured,
9
Raymond Robertson, 2004, Relative Prices and Wage Inequality: Evidence from Mexico, Journal of International Economics, 64, 387409. Notice that this pattern of tariff cuts violates the assumption that we made
in the model of offshoring, that the costs of trade (or changes in these costs) are uniform across all activities.
10
Paul Samuelson, Summer 2004, Where Ricardo and Mill Rebut and Confirm Arguments of Mainstream
Economists Supporting Globalization, Journal of Economic Perspectives, 18(3), 135146.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 215
Chapter 7
Offshoring of Goods and Services 215
work out to exceed the losses of the losers. . . . Correct economic law recognizes
that some American groups can be hurt by dynamic free trade. But correct economic law vindicates the word creative destruction by its proof that the gains of
the American winners are big enough to more than compensate the losers.
Does this paraphrase by Samuelson sound familiar? You can find passages much like
it in this chapter and earlier ones, saying that the gains from trade exceed the losses. But
listen to what Samuelson says next:
The last paragraph can be only an innuendo. For it is dead wrong about [the]
necessary surplus of winnings over losings.
So Samuelson seems to be saying that the winnings for those who gain from trade
do not necessarily exceed the losses for those who lose. How can this be? His last statement seems to contradict much of what we have learned in this book. Or does it?
Simplified Offshoring Model
To understand Samuelsons comments, we can use a simplified version of the offshoring
model we have developed in this chapter. Instead of having many activities involved in
the production of a good, suppose that there are only two activities: components production and research and development (R&D). Each of these activities uses high-skilled
and low-skilled labor, but we assume that components production uses low-skilled labor
intensively and that R&D uses skilled labor intensively. As in our earlier model, we assume that the costs of capital are equal in the two activities and do not discuss this factor. Our goal will be to compare a no-trade situation with an equilibrium with trade
through offshoring, to determine whether there are overall gains from trade.
Suppose that the firm has a certain amount of high-skilled (H) and low-skilled (L)
labor to devote to components and R&D. It is free to move these workers between the
two activities. For example, scientists could be used in the research lab or could instead
be used to determine the best method to produce components; similarly, workers who
are assembling components can instead assist with the construction of full-scale models in the research lab. Given the amount of high-skilled and low-skilled labor used in
total, we can graph a production possibilities frontier (PPF) for the firm between components and R&D activities, as shown in Figure 7-9. This PPF looks just like the production possibilities frontier for a country, except that now we apply it to a single firm.
Points on the PPF, such as A, correspond to differing amounts of high-skilled and lowskilled labor used in the components and R&D activities. Moving left from point A to
another point on the PPF, for example, would involve shifting some high-skilled and
low-skilled labor from the production of components into the research lab.
Production in the Absence of Offshoring
Now that we have the PPF for the firm, we can analyze an equilibrium for the firm, just
as we have previously done for an entire economy. Suppose initially that the firm cannot engage in offshoring of its activities. This assumption means that the component
production and R&D done at Home are used to manufacture a final product at Home:
it cannot assemble any components in Foreign, and likewise, it cannot send any of its
R&D results abroad to be used in a Foreign plant.
The two production activities are used to produce a final good. To determine how
much of the final good is produced, we can use isoquants. An isoquant is similar to a
consumers indifference curve, except that instead of utility, it illustrates the production
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 216
216 Part 3
New Explanations for International Trade
FIGURE 7-9
R&D
No-Trade Equilibrium for the Home
Firm The PPF shows the combinations
Home firm
PPF
Relative price of
components = (PC /PR)A
No-trade Home firm
equilibrium
QR
A
Y1
of components and R&D that can be
undertaken by a firm with a given
amount of labor and capital. In the
absence of offshoring, the firm produces
at A, using quantities QC of components
and QR of R&D to produce amount Y0 of
the final good. The line tangent to the
isoquant through point A measures the
value that the firm puts on components
relative to R&D, or their relative price,
(PC /PR)A. Amount Y1 of the final good
cannot be produced in the absence of
offshoring because it lies outside the PPF
for the firm.
Home firm
isoquants
Y0
QC
Components
of the firm; it is a curve along which the output of the firm is constant despite changing
combinations of inputs. Two of these isoquants are labeled as Y0 and Y1 in Figure 7-9.
The quantity of the final good Y0 can be produced using the quantity QC of components
and the quantity QR of R&D, shown at point A in the figure. Notice that the isoquant
Y0 is tangent to the PPF at point A, which indicates that this isoquant is the highest
amount of the final good that can be produced using any combination of components
and R&D on the PPF. The quantity of the final good Y1 cannot be produced in the absence of offshoring because it lies outside the PPF. Thus, point A is the amount of components and R&D that the firm chooses in the absence of offshoring or what we will call
the no-trade or autarky equilibrium for short.
Through the no-trade equilibrium A in Figure 7-9, we draw a line with the slope of
the isoquant at point A. The slope of the isoquant measures the value, or price, that the
firm puts on components relative to R&D. We can think of these prices as internal to
the firm, reflecting the marginal costs of production of the two activities. An automobile company, for example, would be able to compute the extra labor and other inputs
needed to produce some components of a car and that would be its internal price of
components PC . Similarly, it could compute the marginal cost of developing one more
prototype of a new vehicle, which is the internal price of R&D or PR. The slope of the
price line through point A is the price of components relative to the price of R&D,
A
11
(PC /PR) , in the absence of offshoring.
Equilibrium with Offshoring Now suppose that the firm can import and export its
production activities through offshoring. For example, some of the components could
be done in a Foreign plant and then imported by the Home firm. Alternatively,
11
Recall that the slope of a price line is the relative price of the good on the horizontal axis, which is Components in Figure 7-9.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 217
Chapter 7
Offshoring of Goods and Services 217
some R&D done at Home can be exported to a Foreign plant and used there. In either case, the quantity of the final good is no longer constrained by the Home PPF.
Just as in the Ricardian and Heckscher-Ohlin models, in which a higher level of utility (indifference curve) can be obtained if countries specialize and trade with each
other, here a higher level of production (isoquant) is possible by trading intermediate
activities.
We refer to the relative price of the two activities that the Home firm has available
W1
through offshoring as the world relative price or (PC /PR) . Let us assume that the
world relative price of components is cheaper than Homes no-trade relative price,
W1
A
(PC /PR) < (PC /PR) . That assumption means the Home firm can import components at a lower relative price than it can produce them itself. The assumption that
W1
A
(PC /PR) < (PC /PR) is similar to the assumption we made in the previous the section,
*
*
that relative wage of low-skilled labor is lower in Foreign, W L /W H < WL /WH .
With a lower relative wage of low-skilled labor in Foreign, the components assembly
will also be cheaper in Foreign. It follows that Home will want to offshore components, which are cheaper abroad, while the Home firm will be exporting R&D (i.e.,
offshoring it to Foreign firms), which is cheaper at Home.
The Home equilibrium with offshoring is illustrated in Figure 7-10. The world
relative price of components is tangent to the PPF at point B. Notice that the world relative price line is flatter than the no-trade relative price line at Home. The flattening
FIGURE 7-10
Offshoring Equilibrium
for the Home Firm In the
R&D
World relative price of
components = (PC /PR)W1
No-trade relative price of
components = (PC /PR)A
Activities done by the Home
firm in offshoring equilibrium
B
Home firm
exports of
R&D
A
C
Activities used as inputs to
Home firm with offshoring
Y1
Y0
Gains from
offshoring
Components
Home firm imports
of components
presence of offshoring,
the Home firm will do more
R&D and less component
production, at point B. The
Home firm then exports
R&D activities and imports
components at the world
relative price of components,
which allows it to produce
the amount Y1 of the final
good. The difference between
Y0 and Y1 represents the
gains to the Home firm from
offshoring.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 218
218 Part 3
New Explanations for International Trade
of the price line reflects the lower world relative price of components as compared with
the no-trade relative price at Home. As a result of this fall in the relative price of components, the Home firm undertakes more R&D and less component production, moving from point A to point B on its PPF.
Starting at point B on its PPF, the Home firm now exports R&D and imports
components, moving along the relative price line to point C. Therefore, through
offshoring the firm is able to move off of its PPF to point C. At that point, the isoquant labeled Y1 is tangent to the world price line, indicating that the maximum
amount of the final good Y1 is being produced. Notice that this production of the
final good exceeds the amount Y0 that the Home produced in the absence of
offshoring.
Gains from Offshoring within the Firm The increase in the amount of the final
good producedfrom Y0 to Y1is a measure of the gains from trade to the Home firm
through offshoring. Using the same total amount of high-skilled and low-skilled labor
at Home as before, the company is able to produce more of the final good through its
ability to offshore components and R&D. Because more of the final good is produced
with the same overall amount of high-skilled and low-skilled labor available in Home,
the Home company is more productive. Its costs of production fall, and we expect that
the price of its final product also falls. The gains for this company are therefore spread
to consumers, too.
For these reasons, we agree with the quote about offshoring at the beginning of
the chapter: When a good or service is produced more cheaply abroad, it makes
more sense to import it than to make or provide it domestically. In our example,
component production is cheaper in Foreign than in Home, so Home imports components from Foreign. There are overall gains from offshoring. That is our first conclusion: when comparing a no-trade situation to the equilibrium with offshoring, and
assuming that the world relative price differs from that at Home, there are always gains
from offshoring.
To see how this conclusion is related to the earlier quotation from Samuelson, we
need to introduce one more feature into our discussion. Rather than just comparing the
no-trade situation with the offshoring equilibrium, we need to also consider the impact
of offshoring on a countrys terms of trade.
Terms of Trade
As explained in Chapters 2 and 3, the terms of trade equal the price of a countrys
exports divided by the price of its imports. In the example we are discussing, the
W1
Home terms of trade are (PR /PC) , because Home is exporting R&D and importing
components. A rise in the terms of trade indicates that a country is obtaining a higher
price for its exports, or paying a lower price for its imports, both of which benefit the
country. Conversely, a fall in the terms of trade harms a country because it is paying
more for its imports or selling its exports for less.
In his paper, Samuelson contrasts two cases. In the first, the Foreign country improves its productivity in the good that it exports (components), thereby lowering the
price of components; in the second, the Foreign country improves its productivity in
the good that Home exports (R&D services), thereby lowering that price. These two
cases lead to very different implications for the terms of trade and Home gains, so we
consider each in turn.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 219
Chapter 7
Offshoring of Goods and Services 219
F all in the Price of Components Turning to Figure 7-11, let the Home country
start at the equilibrium with offshoring shown by points B and C. From that situation,
suppose there is a fall in the relative price of component production. That price might
fall, for instance, if the Foreign country improves its productivity in components, thereby
lowering the price paid by Home for this service. Because components are being imW2
ported by Home, a fall in their price is a rise in the Home terms of trade, to (PR /PC ) .
Let us trace how this change in the terms of trade will affect the Home equilibrium.
Because of the fall in the relative price of components, the world price line shown
in Figure 7-11 becomes flatter. Production will shift to point B , and by exporting R&D
and importing components along the world price line, the firm ends up at point C .
Production of the final good at point C is Y2, which exceeds the production Y1 in the
12
initial equilibrium with offshoring. Thus, the Home firm enjoys greater gains from
offshoring when the price of components falls. This is the first case considered by
Samuelson, and it reinforces our conclusions that offshoring leads to overall gains.
FIGURE 7-11
R&D
World relative price of
components = (PC /PR)W2
World relative price of
components = (PC /PR)W1
B
No-trade relative price of
components = (PC /PR)A
B
Home firm
exports of
R&D
C
Y2
A
C
Y1
Y0
Home gains
from trade when
components
relative price falls
Components
Home firm imports
of components
Fall in the Price of Component If the relative price
of components falls from (PC /PR)W 1 to (PC /PR)W 2, then
the Home firm will do even more R&D and less
components production, at point B rather than B. The
12
increase in the terms of trade allows the Home firm to
produce output Y2 at point C , and the gains from trade
are higher than in the initial offshoring equilibrium
(points B and C ).
Notice that the fall in the relative price of components leads to an increase in the amount of components
imported but that the amount of R&D exported from Home does not necessarily increase. You are asked to
explore this case further in Problem 10 at the end of the chapter.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 220
220 Part 3
New Explanations for International Trade
F all in the Price of R&D We also need to consider the second case identified by
Samuelson, and that is when there is a fall in the price of R&D services rather than components. This is what Samuelson has in mind when he argues that offshoring might allow
developing countries, such as India, to gain a comparative advantage in those activities in
which the United States formerly had comparative advantage. As Indian companies like
Wipro (an information technology services company headquartered in Bangalore) engage in more R&D activities, they are directly competing with American companies exporting the same services. So this competition can lower the world price of R&D services.
In Figure 7-12, we reproduce Figure 7-10, including the Home no-trade equilibrium
at point A and the Home production point B with offshoring. Starting at point B, a fall in
the world relative price of R&D will lead to a steeper price line (because the slope of the
price line is the world relative price of components, which increases when PR falls). At the
W3
new price (PC /PR ) , Home shifts production to point B and, by exporting R&D and
importing components, moves to point C . Notice that final output has fallen from Y1 to
Y3. Therefore, the fall in the price of R&D services leads to losses for the Home firm.
To explain where the losses are coming from, notice that Home is exporting R&D
and importing components in the initial offshoring equilibrium (points B and C), so its
FIGURE 7-12
R&D
World relative price of
components = (PC /PR)W1
World relative price of
components = (PC /PR)W3
No-trade relative price of
components = (PC /PR)A
B
Home firm
exports of R&D
B
A
C
Home losses as relative
price of R&D falls
C
Y1
Y3
Y0
Home gains as compared
to no-trade
Components
Home firm imports
of components
A Fall in the Price of R&D A fall in the relative price of
R&D makes the world price line steeper, (PC /PR)W 3. As a
result, the Home firm reduces its R&D activities and increases
its components activities, moving from B to B along the
PPF. At the new world relative price, the Home firm faces a
terms-of-trade loss and can no longer export each unit of
R&D for as many components as it could in the initial
offshoring equilibrium. The final good output is reduced from
Y1 to Y3 at point C . Notice that the final good output, Y3, is
still higher than output without trade, Y0. After the fall in
the relative price of R&D, there are still gains from trade
relative to no-trade (point A) but losses relative to the
initial offshoring equilibrium (points B and C).
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 221
Chapter 7
Offshoring of Goods and Services 221
terms of trade are the price of R&D divided by the price of components (PR/PC). With
the fall in the price of R&D, the Home terms of trade have worsened, and Home is
worse off compared with its initial offshoring equilibrium. Samuelsons point is that
the United States could be worse off if China or India becomes more competitive in, and
lowers the prices of, the products that the United States itself is exporting, such as
R&D services. This is theoretically correct. Although it may be surprising to think of
the United States being in this position, the idea that a country will suffer when its
terms of trade fall is familiar to us from developing-country examples (such as the
Prebisch-Singer hypothesis) in earlier chapters.
Furthermore, notice that final output of Y3 is still higher than Y0, the no-offshoring
output. Therefore, there are still Home gains from offshoring at C as compared with the notrade equilibrium at A. It follows that Home can never be worse off with trade as compared with no trade. Samuelsons point is that a country is worse off when its terms of
trade fall, even though it is still better off than in the absence of trade. With the fall in
the terms of trade, some factors of production will lose and others will gain, but in this
case the gains of the winners are not enough to compensate the losses of the losers.
Our simple model of offshoring illustrates Samuelsons point.
The offshoring that occurred from the United States in the 1980s and the 1990s often
concerned manufacturing activities. But today, the focus is frequently one of the newer
forms of offshoring, the offshoring of business services to foreign countries. Business
services are activities such as accounting, auditing, human resources, order processing,
telemarketing, and after-sales service, like getting help with your computer (see Side
Bar: Offshoring Microsoft Windows). Firms in the United States are increasingly
SIDE BAR
Have you ever been frustrated in your attempt to install a new
version of Microsoft Windows on your computer? Well, you are
not alone. It just so happens that Microsoft Corporation, located in Seattle, Washington, also avoids loading upgrades of the
Windows software onto its own computers. Who loads these
upgrades? It turns out that Wipro, an Indian high-tech firm,
manages Microsofts computers in the nighttime hours in
Seattle, which is during the daytime hours in Bangalore, India.
Called infrastructure outsourcing, the remote management
of computer resources is just one example of service outsourcing. Wipro performs infrastructure outsourcing for Microsoft and
also for Royal Dutch Shell, which is one of the largest oil companies in the world. During the nonoperating hours in those
companies, Wipro remotely accesses their computers, performs
routine maintenance, troubleshoots for viruses, repairs corrupt
files, and checks for other problems that can arise. As the sun
rises in Seattle, the computers at Microsoft are better than
when the employees left the night before, which makes those
employees more productive in their own computer work. In the
AP Photo/Gautam Singh
Offshoring Microsoft Windows
They could be talking to you.
future, perhaps universities in the United States will also use
infrastructure outsourcing to maintain their computer systems
so that the computers in your dormitory, too, will be maintained during the nighttime hours by programmers in India!
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 222
New Explanations for International Trade
transferring these activities to India, where the wages of educated workers are much
lower than in the United States. This is the sort of competition that Samuelson had in
mind when he spoke of China and India improving their productivity and comparative
advantage in activities that the United States already exports. The next application discusses the magnitude of service exports and also the changes in their prices.
APPLICATION
U.S. Terms of Trade and Service Exports
Because Samuelsons argument is a theoretical one, the next step is to examine the evidence for the United States. If the United States has been facing competition in R&D
and the other skill-intensive activities that we export, then we would expect the terms
of trade to fall. Conversely, if the United States has been offshoring in manufacturing,
then the opportunity to import lower-priced intermediate inputs should lead to a rise
in the terms of trade.
Merchandise Prices To evaluate these ideas, we make use of data on the terms of
trade for the United States. In Figure 7-13, we first show the terms of trade for the
13
United States for merchandise goods (excluding petroleum), which is the gold line.
FIGURE 7-13
Annual 1.30
average
1.25
terms of
trade 1.20
1.15
Air travel
1.10
1.05
1.00
Merchandise goods
0.95
Terms of Trade for the United States,
19872009 Shown here are the U.S. terms of trade
for merchandise goods (excluding petroleum) and for
air travel services. The terms of trade for merchandise
goods fell from 1987 to 1994 and then rose to 2008,
with a slight dip in 2009. The terms of trade for air
13
09
20
07
20
05
20
03
20
01
20
99
19
97
19
95
19
93
19
91
19
19
89
0.90
87
19
222 Part 3
travel services are more volatile, falling from 1995 to
2002, rising until 2005 and falling erratically until
2009. In sum, we do not see a pattern of declining
terms of trade in either of these sectors.
Source: Bureau of Labor Statistics.
Merchandise goods include agriculture, mining, and manufacturing. We have excluded petroleum because
its world price is determined by conditions such as shortages and wars and behaves quite differently from the
prices of other merchandise goods.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 223
Chapter 7
Offshoring of Goods and Services 223
The terms of trade for goods fell from 1987 to 1994 but have been rising since then,
with a slight dip in 2009. The improvement in the merchandise terms of trade shows
that we are able to import intermediate inputs (and also import final goods) at lower
prices over time. This rise in the terms of trade means that there are increasing gains
from trade in merchandise goods for the United States.
Service Prices For trade in services, such as finance, insurance, and R&D, it is very
difficult to measure their prices in international trade. These services are tailored to the
buyer and, as a result, there are not standardized prices. For this reason, we do not have
an overall measure of the terms of trade in services. There is one type of service, however, for which it is relatively easy to collect international prices: air travel. The terms
of trade in air travel equal the price that foreigners pay for travel on U.S. airlines (a service export) divided by the price that Americans pay on foreign airlines (a service import).
In Figure 7-13, we also show the U.S. terms of trade in air travel that are available since
1995. The terms of trade in air travel are quite volatile, falling from 1995 to 2002, then
rising to 2005, and falling erratically again afterward. For this one category of services,
the terms of trade improvement from 2002 to 2005 indicates growing gains from trade
for the United States, the same result we found for merchandise goods. Summing up,
there is no evidence to date that the falling terms of trade that Samuelson is concerned
about have occurred for the United States.
Service Trade What about other traded services?
Although standard prices are not available, data on
the amount of service exports and imports for the
United States are collected annually. These data are
shown in Table 7-2 for 2008. The United States
runs a substantial surplus in services trade, with exports of $525 billion and imports of $364 billion.
Categories of service exports that exceed imports
include several types of business, professional, and
technical services (but in computer and information services, the United States now runs a deficit);
education (which is exported when a foreign student studies in the United States); financial services; travel; and royalties and license fees (which are
collected from foreign firms when they use U.S.
patents and trademarks or are paid abroad when we
use foreign patents). The fact that exports exceed
imports in many categories in Table 7-2 means that
the United States has a comparative advantage in
traded services. Indeed, the U.S. surplus in business, professional, and technical services is among
the highest in the world, similar to that of the
United Kingdom and higher than that of Hong
Kong and India. London is a world financial center
and competes with New York and other U.S. cities,
which explains the high trade surplus of the United
Kingdom, whereas Hong Kong is a regional hub
for transportation and offshoring to China. The
TABLE 7-2
U.S. Trade in Services, 2008 ($ millions) This table shows U.S.
exports and imports in the major categories of services trade for 2008.
Exports
Imports
12,599
26,942
17,139
7,942
16,139
21,565
14,885
958
Computer and information services
Management and consulting services
R&D and testing services
Operational leasing
Other business, professional, and
technical services
Total business, professional, and
technical services
48,901
22,736
$113,525
$76,284
Education
Financial services
Insurance services
Telecommunications
Total other private services
17,796
60,190
10,756
9,163
$233,529
5,204
19,143
42,939
7,193
$153,267
Travel
Passenger fares
Other transporation
Royalties and license fees
Other services
110,090
31,623
58,945
91,599
22,099
79,743
32,597
72,143
26,616
2,505
$525,786
$364,366
Total private services
Source: U.S. Bureau of Economic Analysis.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 224
New Explanations for International Trade
FIGURE 7-14
Trade surplus 120
(US$ billion)
100
United Kingdom
80
60
40
20
United States
India
0
08
20
06
20
04
02
20
00
20
98
20
96
19
94
19
92
19
90
Trade Surplus in Business Services This figure
shows the combined trade surplus in computer and
information services, other business services, and
financial services for the United States, the United
Kingdom, and India from 1980 to 2008. The U.S.
surplus in these categories of services has been
growing steadily since about 1985 and up until 1995
19
88
19
86
19
84
19
19
19
82
20
80
19
224 Part 3
shows a very similar pattern to the trade surplus from
the United Kingdom, its chief competitor. Since 1995
the surplus of the United Kingdom has grown faster,
mainly due to its exports of financial services. Indias
surplus began growing around 1995 and derives entirely
from its exports of computer and information services.
Source: International Monetary Fund.
combined trade balance in computer and information services, other business services,
and financial services for the United States, the United Kingdom, and India since 1980
are graphed in Figure 7-14.
The U.S. surplus in these categories of services has been growing steadily since
about 1985 and up until 1995 shows a very similar pattern to the trade surplus
from the United Kingdom, its chief competitor. Since 1995 the surplus of the United
Kingdom has grown faster, mainly due to its exports of financial services. Indias
surplus began growing around 1995 and derives entirely from its exports of computer and information services. In 2008 the combined U.S. surplus in computers,
other business services, and financial services ($66 billion) was 50% greater than
that in India ($46 billion), but 50% smaller than in the United Kingdom ($99
billion).
What will these surpluses look like a decade or two from now? It is difficult to
project, but notice that as the Indian surplus began growing in Figure 7-14, the
rise in the U.S. surplus slowed down. Indian net exports of computer and information services have grown by $46 billion since 1997, while U.S. net exports in this
category have fallen by $7.5 billion. Although it is too soon to tell whether such
trends will continue, it is at least possible that in a decade or two, Indias overall
surplus in service exports could overtake that of the United States. Only time will
tell whether the United States will eventually face the same type of competition
from India in its service exports that it has already faced for many years from the
United Kingdom.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 225
Chapter 7
Offshoring of Goods and Services 225
3 The Politics and Future of Offshoring
Offshoring is controversial and is often the topic of political debate. In February 2004
the first quote at the beginning of this chapter appeared in the Economic Report of the
President. The writer of that sentence, Harvard economist N. Gregory Mankiw, who
was chairman of the Council of Economic Advisors, also said that outsourcing is just
a new way of doing international trade. More things are tradable than were tradable in
the past, and thats a good thing. Those comments were widely criticized by the Democrats and Republicans alike, and Professor Mankiw later apologized in a letter to the
House of Representatives, writing, My lack of clarity left the wrong impression that
I praised the loss of U.S. jobs.
In the Democratic primary elections of 2007 and in the presidential campaign of
2008, this topic came up again. Senators Barack Obama and Hillary Clinton both
promised that, if elected, they would end tax breaks for companies earning profits
14
overseas:
Obama, Nov. 3, 2007: When I am president, I will end the tax giveaways to
companies that ship our jobs overseas, and I will put the money in the pockets of
working Americans, and seniors, and homeowners who deserve a break.
Clinton, Nov. 19, 2007: And we are going to finally close the tax loopholes and
stop giving tax breaks to companies that ship jobs overseas. Enough with outsourcing American jobs using taxpayer dollars.
To what tax breaks were Mr. Obama and Ms. Clinton referring? The United States
taxes corporate profits at 35%, a high rate when compared with the corporate tax
rates in other countries. Profit earned by overseas subsidiaries of U.S. companies,
however, goes untaxed by the U.S. government provided that such a subsidiary stays
overseas and does not appear on the books of the parent company in the United
States. That tax provision gives U.S. multinational companies an incentive to use
these funds overseas for further investment in the subsidiary, but it does not necessarily lead them to move jobs overseas in the first place. Eventually, when these profits are moved back to the parent company in the United States, they are taxed at the
normal rate.
President Obama recently announced that he would follow through on his campaign pledge to end the tax break on overseas profits of multinational firms, as indicated
by the quotation at the beginning of this chapter: We should . . . encourage . . . businesses to stay within our borders, its time to finally slash the tax breaks for companies
that ship our jobs overseas. That change in policy does not have much support from
economists. One strongly worded response comes from Matthew Slaughter, Professor
at Dartmouth College, in Headlines: How to Destroy American Jobs. He cites evidence that U.S. multinationals have added roughly as many jobs in the United States
as they have added abroad, and argues that these jobs in the United States depend on
the ability of the multinationals to offshore other jobs.
In addition to the employment statistics, Professor Slaughter also cites evidence that
U.S. multinational firms conducted nearly 90% of all private-sector R&D in the United
States, and that these firms account for the majority of U.S. productivity gains.
14
These quotations and some material that follows are taken from http://www.factcheck.org/askfactcheck/
what_kind_of_tax_breaks_does_the.html.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 226
226 Part 3
New Explanations for International Trade
HEADLINES
How to Destroy American Jobs
This article argues that offshoring by multinational companies supports an increase
in jobs at home, and that these jobs would be lost by policies to restrict offshoring.
Deep in the presidents budget released
Monday [February 1, 2010] appear a set of
proposals
headed
Reform
U.S.
International Tax System. If these proposals are enacted, U.S.-based multinational firms will face $122.2 billion in tax
increases over the next decade. This is a
natural follow-up to President Obamas
sweeping plan announced last May [2009]
entitled Leveling the Playing Field:
Curbing Tax Havens and Removing Tax
Incentives for Shifting Jobs Overseas.
The fundamental assumption behind
these proposals is that U.S. multinationals expand abroad only to export jobs
out of the country. Thus, taxing their
foreign operations more would boost tax
revenues here and create desperately
needed U.S. jobs. This is simply wrong.
These tax increases would not create
American jobs, they would destroy them.
Academic research, including most recently that done by Harvards Mihir
Desai and Fritz Foley and University of
Michigans James Hines, has consistently found that expansion abroad by U.S.
multinationals tends to support jobs
based in the U.S. More investment and
employment abroad are strongly associated with more investment and employment in American parent companies.
When parent firms based in the U.S.
hire workers in their foreign affiliates,
the skills and occupations of these workers are often complementary; they arent
substitutes. More hiring abroad stimulates more U.S. hiring. For example, as
Wal-Mart has opened stores abroad, it has
created hundreds of U.S. jobs for workers
to coordinate the distribution of goods
worldwide. The expansion of these foreign affiliateswhether to serve foreign
customers, or to save costsalso expands the overall scale of multinationals.
Expanding abroad also allows firms to
refine their scope of activities. For example, exporting routine production
means that employees in the U.S. can
focus on higher value-added tasks such
as R&D, marketing and general management. The total impact of this process is
much richer than an overly simplistic
story of exporting jobs. But the ultimate
proof lies in the empirical evidence.
Consider total employment spanning
1988 through 2007 (the most recent
year of data available from the U.S.
Bureau of Economic Analysis). Over that
time, employment in affiliates rose by
5.3 millionto 11.7 million from 6.4
million. Over that same period, employment in U.S. parent companies increased
by nearly as much4.3 millionto 22
million from 17.7 million. Indeed, research repeatedly shows that foreign-affiliate expansion tends to expand U.S.
parent activity. . . .
The major policy challenge facing the
U.S. today is not just to create jobs,
but to create high-paying private-sector jobs linked to investment and trade.
Which firms can create these jobs?
U.S.-based multinationals. Theyalong
with similarly performing U.S. affiliates
of foreign-based multinationalshave
long been among the strongest companies in the U.S. economy.
These two groups of firms accounted
for the majority of the post-1995 acceleration in U.S. productivity growth, the
foundation of rising standards of living for
everyone. They tend to create high-paying
jobs27.5 million in 2007. . . . And
these firms also conducted $240.2 billion
in research and development, a remarkable 89.2% of all U.S. private-sector R&D.
To climb out of the recession, we need
to create millions of the kinds of jobs
that U.S. multinationals tend to create.
Economic policy on all fronts should be
encouraging job growth by these firms.
The proposed international-tax reforms
do precisely the opposite.
Source: Matthew J. Slaughter, How to Destroy American Jobs, Wall Street Journal, February 3, 2010, p. A17.
Direct evidence on the positive impact of offshoring on productivity comes from
another source, a 2005 study of the offshoring of material inputs and services by U.S.
15
manufacturing firms in the 1990s. Over the eight years from 1992 to 2000, that
study found that service offshoring can explain between 11 and 13% of the total
15
Mary Amiti and Shang-Jin Wei, 2005, Service Offshoring, Productivity, and Employment: Evidence from
the United States, International Monetary Fund, IMF Working Paper 05/238; and 2006, Service Offshoring and Productivity: Evidence from the United States, National Bureau of Economic Research
(NBER) Working Paper No. 11926.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 227
Chapter 7
Offshoring of Goods and Services 227
increase in productivity within the U.S. manufacturing sector. In addition, the
offshoring of material inputs explains between 3 and 6% of the increase in manufacturing productivity. Combining these effects, offshoring explains between 15 and
20% of overall productivity growth in the manufacturing sector. Evidence of this
type makes economists reluctant to impose additional taxes on U.S. companies that
engage in offshoring, because of the possible adverse effects on productivity here in
the United States.
On the other hand, changes in economic conditions that lead firms to voluntarily bring some activities back home would be viewed favorably by most economists.
There is some evidence that economic conditions have changed in that direction, as
described in Headlines: Caterpillar Joins Onshoring Trend. A combination of
stimulus spending in the United States, lower wages, and a weak U.S. dollar (which
makes it more expensive to operate overseas) has led some companies to onshore
their activities back to the United States. This trend has also occurred because companies are finding that communication with overseas suppliers can be slow and
costly.
The Future of U.S. Comparative Advantage
Just as in our model of this chapter, the recent onshoring trend shows that companies usually avoid offshoring all activities from the United States: the extra
communication and trade costs involved need to be balanced against the lower foreign wages to find the right amount of offshoring. Most often companies find it
HEADLINES
Caterpillar Joins Onshoring Trend
Some American companies have found it advantageous to take activities they had
previously shifted overseas and move them back home, in what is called onshoring.
Caterpillar, Inc. is considering relocating some heavy-equipment overseas
production to a new U.S. plant, part of
a growing movement among manufacturers to bring more operations back
homea shift that will likely spark
fierce competition among states for
new manufacturing jobs. The trend,
known as onshoring or reshoring, is
gaining momentum as a weak U.S. dollar makes it costlier to import products
from overseas. Manufacturers are also
counting on White House jobs incentives, as well as their ability to negoti-
ate lower prices from U.S. suppliers
who were hurt by the downturn and
willing to bargain.
After a decade of rapid globalization, economists say companies are
seeing disadvantages of offshore production, including shipping costs,
complicated logistics, and quality issues. Political unrest and theft of intellectual property pose additional
risks. If you want to keep your
supply chain tight, its hard to do that
with a 16-hour plane ride from
Shanghai to Ohio, said Cliff Waldman,
an economist with the Manufacturers
Alliance/MAPI, a public policy and
economics research group in Arlington,
Virginia.
General Electric Co. said last June it
would move production of some water
heaters from China to its facility in
Louisville, Kentucky, starting in 2011. A
GE spokeswoman said a 2005 labor agreement under which new employees would
be paid $13 an hour, [instead of the]
nearly $20 an hour [they once made],
enabled us to be more competitive. . . .
Source: Kris Maher and Bob Tita, Caterpillar Joins Onshoring Trend, Wall Street Journal, March 11, 2010, p. A17.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 228
228 Part 3
New Explanations for International Trade
advantageous to keep some activities in the United States (such as those using more
highly skilled labor or relying on close communication with customers) and move
other activities abroad (using less skilled labor and involving more routine activities). The fear sometimes expressed in the popular press that offshoring threatens the
elimination of most manufacturing and service jobs in the United States is overstated. The ability to offshore a portion of the production process allows other activities to remain in the United States.
A good example to illustrate this point is the offshoring of medical services. The
transcription of doctors notes from spoken to written form was one of the first service activities offshore to India. Since then, other types of medical services have also
been offshored, and a New York Times article in 2003 identified the reading of X-rays
or radiologyas the next area that could shift overseas: It turns out that even American radiologists, with their years of training and annual salaries of $250,000 or more,
worry about their jobs moving to countries with lower wages, in much the same way
that garment knitters, blast-furnace operators and data-entry clerks do. . . . Radiology
16
may just be the start of patient care performed overseas.
It turns out, however, that the types of radiology jobs that can potentially be
17
transferred overseas are very limited. Radiology is a high-paying profession precisely because the reading of X-rays is difficult and takes years of training and practice to perfect. X-rays are normally analyzed in the same hospital where the patient
is being treated. In a few cases of specific diseases, such as the reading of mammograms for breast cancer, it is possible that the work can be outsourced (i.e., performed outside the hospital), either domestically or offshore. Firms known as
nighthawks already provide some outsourcing services to hospitals, principally
during nighttime hours. Nighthawk firms are headquartered in the United States but
have radiologists at offshore sites, including Australia, Israel, Spain, and India. These
nighttime services allow smaller hospitals that cannot afford a full-time night radiologist to obtain readings during evening hours, and allow the nighthawk firms
to keep their radiologists fully employed by combining the demand from multiple
hospitals.
The outsoucing to nighthawk firms is a natural response to the round-the-clock demand for hospital services but less-than-full-time demand for radiologists on-site.
Often these nighttime services are used only for preliminary reads, leading to immediate treatment of patients; the X-ray image is then read again by the staff radiologist
in the United States the next day. That is, in many cases, the services being outsourced
are not directly competing for the daytime jobs but, instead, are complements to these
U.S. jobs.
Radiology is under no imminent threat from outsourcing because the profession involves decisions that cannot be codified in written rules. Much of the radiologists
knowledge is gained from reading countless X-rays with complex images and shadows,
and the ability to recognize patterns cannot easily be passed on to another person or
16
Andrew Pollack, Whos Reading Your X-Ray: Jobs in Medical Care, Too, Can Be Outsourced Overseas,
New York Times, November 16, 2003, section 3, pp. 1, 9.
17
The material in the following paragraphs is drawn from Frank Levy and Ari Goelman, Offshoring and
Radiology, presented at the Brookings Institute Trade Forum May 1213, 2005.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 229
Chapter 7
Offshoring of Goods and Services 229
firm. It follows that the work cannot be outsourced except for the nighttime activities
of nighthawk firms, which actually work in conjunction with the daytime activities in
major hospitals.
In every profession there will always be jobs that cannot be performed by someone
who is not on-site. For many of the service activities listed in Table 7-2, the United
States will continue to have comparative advantage even while facing foreign competition. In many manufacturing industries, the United States will continue to maintain
some activities at home, such as R&D and marketing, even while shifting a portion of
the production process abroad. Finally, we should recognize that the ability to offshore
to Mexico or India ultimately makes the U.S. companies involved more profitable and
therefore better able to withstand foreign competition.
In this chapter, we have studied a type of trade that is becoming increasingly important: offshoring, by which we mean the shifting of some production activities to another country, while other production activities are kept at Home. Rather than
trading final goods, like wheat for cloth as in the Ricardian model of Chapter 2, or
computers for shoes as in the Heckscher-Ohlin model of Chapter 4, with offshoring
each good can be produced in stages in several countries and then assembled in a
final location.
In the model of offshoring we presented, because low-skilled labor is relatively
cheap abroad, it makes sense for Home to offshore to the Foreign country those activities that are less skill-intensive, while keeping at Home those activities that are
more skill-intensive. Slicing the value chain in this way is consistent with the idea
of comparative advantage, because each country is engaged in the activities for which
its labor is relatively cheaper. From both the Home and Foreign point of view, the
ratio of high-skilled/low-skilled labor in value chain activities goes up. A major finding of this chapter, then, is that an increase in offshoring will raise the relative demand (and hence relative wage) for skilled
labor in both countries.
In a simplified model in which there are
only two activities, we found that a fall in
the world price of the low-skilled and laborintensive input will lead to gains to the Home
firm from offshoring. But in contrast, a fall in
the price of the skilled labor-intensive input
would lead to losses to the Home firm, as
compared with the prior trade equilibrium.
Such a price change is a terms-of-trade loss for
Home, leading to losses from the lower relative price of exports. So even though Home
gains overall from offshoring (producing at
least as much as it would in a no-offshoring
equilibrium), it is still the case that competition in the input being exported by Home will
make it worse off.
Dan Wasserman. Represented by Tribune Media Services (http://www.Comicspage.com).
4 Conclusions
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 230
230 Part 3
New Explanations for International Trade
We concluded the chapter by exploring offshoring in service activities, a topic that
has received much attention in the media recently. Offshoring from the United States
to Mexico consists mainly of low-skilled jobs; offshoring from the United States to
India consists of higher-skilled jobs performed by college-educated Indians. This new
type of offshoring has been made possible by information and communication technologies (such as the Internet and fiber-optic cables) and has allowed cities like Bangalore, India, to establish service centers for U.S. companies. These facilities not only
answer questions from customers in the United States and worldwide, they are also engaged in accounting and finance, writing software, R&D, and many other skilled business services.
The fact that it is not only possible to shift these activities to India but economical
to do so shows how new technologies make possible patterns of international trade
that would have been unimaginable a decade ago. Such changes show globalization at work. Does the offshoring of service activities pose any threat to whitecollar workers in the United States? There is no simple answer. On the one hand,
we presented evidence that service offshoring provides productivity gains, and therefore gains from trade, to the United States. But as always, having gains from trade
does not mean that everyone gains: there can be winners and losers. For service offshoring, it is possible that skilled workers will see a potential reduction in their
wages, just as low-skilled labor bore the brunt of the impact from offshoring in the
1980s. Nevertheless, it is still the case that the United States, like the United Kingdom and other European countries, continues to have a comparative advantage in
exporting various types of business services. Although India is making rapid progress
in the area of computer and information services, there are still many types of service activities that need be done locally and cannot be outsourced. One likely prediction is that the activities in the United States that cannot be codified in written rules
and procedures, and that benefit from face-to-face contact as well as proximity to
other highly skilled individuals in related industries, will continue to have comparative advantage.
KEY POINTS
1. The provision of a service or the production
of various parts of a good in different countries
for assembly into a final good in another location is called foreign outsourcing or offshoring.
2. We can apply the same ideas that we developed
for trade in final goods among countries to the
trade of intermediate offshored activities. For instance, if low-skilled labor is relatively inexpen-
sive in the Foreign country, then the activities
that are least skill-intensive will be offshored
there, and Home will engage in the activities that
are more skill-intensive.
3. We can also predict what happens to relative
wages of skilled labor when there is a change in
trading costs and more offshoring. Our model
predicts that the relative demand for skilled labor
increases in both countries. This result helps to
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 231
Chapter 7
explain the observation that relative wages have
been increasing in the United States and in other
countries at the same time.
4. In an overall sense, there are gains from offshoring, because the specialization of countries
in different production activities allows firms
in both countries to produce a higher level of
final goods. That increase in output represents a
productivity gain, and the gains from trade.
Offshoring of Goods and Services 231
5. With service offshoring, it is possible that a
country like India will have rising productivity in activities in which the United States has
comparative advantage, such as R&D. Rising
productivity in India would lead to a fall in the
price of R&D, which is a terms-of-trade loss for
the United States. For that reason, the United
States could lose due to service offshoring,
though it still gains as compared with a situation of no offshoring at all.
KEY TERMS
foreign outsourcing, p. 200
offshoring, p. 200
value chain, p. 202
skill-biased technological change,
p. 211
total wage payments, p. 211
isoquants, p. 215
business services, p. 221
PROBLEMS
1. Consider an offshoring model in which the
hours of labor used in four activities in the
United States and Mexico are as follows:
Note that labor hours in Mexico are twice
those in the United States, reflecting Mexicos
lower productivity. Also note that the ratio of
high-skilled to low-skilled labor used in each
activity increases as we move to the right,
from 1/5 in assembly, to 10/1 in R&D.
Suppose that the wage of U.S. low-skilled
workers is $10 per hour and that of highskilled workers is $25 per hour, and that the
wage of Mexican low-skilled workers is $1 per
hour and that of high-skilled workers is $5 per
hour (these values are made up to be convenient, not realistic). Also suppose that the trade
costs are 25%, 30%, or 50%, which means
that an additional 25%, 30%, or 50% is added
to the costs of offshoring to Mexico.
Hours of Labor Used in Each Activity (per unit of output):
Assembly
Low-skilled labor
High-skilled labor
High-skilled/low-skilled ratio
Component Production
Office Services
R&D
Mexico: 20
U.S.: 5
Mexico: 4
U.S.: 1
Mexico: 12
U.S.: 3
Mexico: 4
U.S.: 1
Mexico: 8
U.S.: 2
Mexico: 8
U.S.: 2
Mexico: 4
U.S.: 1
Mexico: 40
U.S.: 10
1/5
1/3
1/1
10/1
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 232
232 Part 3
New Explanations for International Trade
a. Fill in the blank cells in the following table
by computing the costs of production of
each activity in each country (two cells are
filled in for you):
Assembly
Mexico
Component Production
Office Services
R&D
$40
United States
Imported by United States
from Mexico, Trade Costs = 25%
Imported by United States
from Mexico, Trade Costs = 30%
$52
Imported by United States
from Mexico, Trade Costs = 50%
b. With trade costs of 50%, where is the value
chain sliced? That is, which activities are
cheaper to import from Mexico and which
are cheaper to produce in the United
States?
c. With trade costs of 30%, and then 25%,
where is the value chain sliced?
2. Consider an offshoring model in which Homes
skilled labor has a higher relative wage than
Foreigns skilled labor and in which the costs of
capital and trade are uniform across production
activities.
a. Will Homes offshored production activities
be high or low on the value chain for a given
product? That is, will Home offshore production activities that are skilled and laborintensive, or low-skilled and labor-intensive?
Explain.
b. Suppose that Home uniformly increases its
tariff level, effectively increasing the cost of
importing all goods and services from
abroad. How does this affect the slicing of
the value chain?
c. Draw relative labor supply and demand diagrams for Home and Foreign showing the
effect of this change. What happens to the
relative wage in each country?
3. Consider a U.S. firms production of automobiles, including research and development and
component production.
a. Starting from a no-trade equilibrium in a
PPF diagram, illustrate the gains from offshoring if the United States has a comparative advantage in component production.
b. Now suppose that advances in engineering
abroad decrease the relative price of research
and development. Illustrate this change on
your diagram and state the implications for
production in the United States.
c. Does the U.S. firm gain from advances in
research and development abroad? Explain
why or why not.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 233
W 10
WL 40
5% = H
W 100 + W 100 ,
H
L
for components
Hint: We follow a procedure similar to that
used in Chapter 4 when calculating the change
in factor prices in the Heckscher-Ohlin model.
Use the cost shares and price change data in
these formulas to get
a. In which factor(s) is components intensive?
In which factor(s) is R&D intensive?
b. Suppose that due to the opening of trade,
the price of components fall by PC/PC =
10%, while the price of R&D remains unchanged, PR/PR = 0. Using the hint below,
calculate the change in the wage of skilled
and low-skilled labor.
P
WH WH HR
WL WL LR
0.5 R =
+
P
W P Q
W P Q ,
R
H
R
R
LR
R
for R&D
Share of total costs paid to lowskilled labor = 10/100 = 10%
P
WH WH HC
WL WL LC
0.5 C =
+
P
W
P Q
W P Q ,
H C
C
LC
C
C
for components
Share of total costs paid to highskilled labor = 40/100 = 40%
Dividing the equations by (PC QC) and
(PR QR), respectively, we can rewrite the
equations as
Earnings of capital =
R KR = 50
0.5(PR QR) = WH HR + WL LR,
for R&D
Earnings of low-skilled labor =
WL LR = 10
0.5(PC QC) = WH HC + WL LC,
for components
Earnings of high-skilled labor =
WH HR = 40
Taking the change in these equations:
R&D: Total costs of R&D =
PR QR = 100
0.5(PR QR) = WH HR + WL LR,
for R&D
Share of total costs paid to lowskilled labor = 40/100 = 40%
0.5(PC QC) = WH HC + WL LC,
for components
Share of total costs paid to highskilled labor = 10/100 = 10%
Because we assume that 50% of costs in
either components or R&D is always paid
to capital, then R KC = 0.5(PC QC) and
R KR = 0.5(PR QR), so we can rewrite the
above two equations as
Earnings of capital =
R KC = 50
PR QR = R KR + WH HR + WL LR,
for R&D
Earnings of low-skilled labor =
WL LC = 40
PC QC = R KC + WH HC + WL LC,
for components
Earnings of high-skilled labor =
WH HC = 10
First, write the total costs in each activity as
consisting of the payments to labor and capital:
Components: Total costs of production =
PC QC = 100
Offshoring of Goods and Services 233
4. Consider the model of a firm that produces
final goods using R&D and components as inputs, with cost data as follows:
Chapter 7
WH 40
WL 10
0 =
+
,
WH 100 WL 100
for R&D
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 234
234 Part 3
New Explanations for International Trade
Now solve these two equations for the change
in the high-skilled wage (WH /WH), and the
change in the low-skilled wage (WL /WL).
c. What has happened to the relative wage of
high-skilled/low-skilled labor? Does this
match the predictions of the offshoring
model in this chapter?
5. Consider the model of a firm that produces
final goods using R&D and components as inputs, with cost data as follows:
Components: Total costs of production =
PC QC = 100
Earnings of high-skilled labor =
WH HC = 25
Earnings of low-skilled labor =
WL LC = 25
Earnings of capital =
R KC = 50
ponents stays unchanged, PC /PC = 0. Calculate the change in the relative wage of
high-skilled and low-skilled labor.
c. What has happened to the relative wage of
high-skilled/low-skilled labor? How does
this result compare to Problem 4, and explain why it is similar or different.
6. The following diagram shows what happened
to the relative wage and relative demand for
skilled labor in the U.S. manufacturing sector
during the 1990s. These points are plotted
using the data from Figures 7-5 and 7-6.
a. What must have happened to the supply and
demand curves to explain this change in relative wage and relative employment?
b. Why do you think the supply and demand
curves shifted this way? Hint: Think about
where the workers who leave manufacturing
might be going.
Share of total costs paid to highskilled labor = 25/100 = 25%
Share of total costs paid to lowskilled labor = 25/100 = 25%
R&D: Total costs of R&D =
PR QR = 100
Earnings of high-skilled labor =
WH HR = 30
Earnings of low-skilled labor =
WL LR = 20
Earnings of capital =
R KR = 50
Share of total costs paid to highskilled labor = 30/100 = 30%
Share of total costs paid to lowskilled labor = 20/100 = 20%
a. In which factor(s) is components intensive?
In which factor(s) is research intensive?
b. Suppose that due to the opening of trade,
the relative price of R&D increases,
PR/PR = 10%, whereas the price of com-
7. Read the following excerpt, and using what you
have learned in this chapter, discuss how offshoring creates opportunities for the countries
involved.
Sudhakar Shenoy, chief executive of Information Management Consultants (IMC)
in Reston, makes an effective pitch for
offshoring.
Several years ago IMC saw a market developing for software that would allow
biotech companies to make better and faster
use of the new human genome research.
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 235
Chapter 7
Doing it here, Shenoy calculated, would
cost several million dollars, which he figured would have priced the product too
high for most customers. But by having a
small group of engineers at IMCs Indian
subsidiary do much of the coding work, he
was able to bring the project in at
$500,000. The result: IMC now has a
thriving line of business in bioinformatics,
with major clients and a growing payroll of
six-figure PhDs here. And there are more
engineers than eversix here for every one
in India.
But thats only part of the good-news
story. In Pune, where IMCs Indian operations are located, an airport under construction will require lots of U.S.
engineering, design and electronics. At the
same time, IMCs Indian engineers, who
earned annual salaries of $3,500 a decade
ago, now command up to $12,000
enough to buy all manner of imported consumer goods.
Source: Excerpted from Steven Pearlstein, Still Short
of the Offshoring Ideal, Washington Post, March 12,
2004.
8. The quote from the 2004 Economic Report of the
President at the beginning of the chapter generated a lot of controversy that year, as discussed
at the beginning of section 3 here. The chairman of the Council, N. Gregory Mankiw, made
the following additional comments in a speech
while presenting the report: Outsourcing is
just a new way of doing international trade.
More things are tradable than were tradable in
the past, and thats a good thing.
Those statements quickly led to reactions
from both Democratic and Republican members of Congress. Tom Daschle, then the
Democratic Senate minority leader, said, If
this is the administrations position, they owe
an apology to every worker in America.
Dennis Hastert, then Republican Speaker of
the House, said, Outsourcing can be a problem for American workers and the American
economy. John Kerry, the 2004 Democratic
Offshoring of Goods and Services 235
presidential candidate, referred to businesses
that offshored as Benedict Arnold corporations. In response, Mankiw clarified his earlier comments: My lack of clarity left the
wrong impression that I praised the loss of
U.S. jobs.
Although you might feel that these statements just represented a squabble between
politicians trying to score points during a
presidential campaign, it is still worth trying
to sort out who gains and who loses from
offshoring.
a. Why does Mankiw say that outsourcing is
a good thing Who is it good for in the
United States? Are there overall gains for
the United States? Explain with a diagram.
b. Later in this chapter, Paul Samuelson is
quoted as saying that there is no necessary
surplus of winnings over losings due to offshoring. Use Figure 7-12 to carefully explain
why Samuelson says this.
9. In Figure 7-11, we saw that a fall in the relative
price of components leads to an increase in the
amount of components imported but that the
amount of R&D exported from Home does not
necessarily increase. To explore this further,
complete the following:
a. Let the relative price of components continue to fall in Figure 7-11, and show in a
graph what happens to the equilibrium point
on the isoquant for the final good.
b. Now draw another graph that has the relative price of components on the vertical axis
and the imports of components on the horizontal axis. Start at the no-trade relative
price of components, where imports are
zero. Then label the various world relative
prices of components on the vertical axis,
and graph the quantity of imports at each
price. Can we be sure that the import demand curve slopes downward?
c. Now draw a new graph that has the relative
price of R&D on the vertical axis and the exports of R&D on the horizontal axis. Start at
FeenTayTrade2e_CH07_Layout 1 8/7/10 2:20 PM Page 236
236 Part 3
New Explanations for International Trade
the no-trade relative price of R&D, where
exports are zero. Then label the various
world relative prices of R&D on the vertical
axis, and graph the quantity of exports at
each price. When the relative price of R&D
is high enough, what do you notice about
the export supply curve?
10. Why might it be relatively easier for a developing country like India to export service
activities through offshoring than to participate
in the global economy by producing manufacturing components?
11. It is widely noted that even though China is
the favored destination for manufacturing offshoring, it is far behind India in the business
of offshored services. What differences between these two countries might account for
this observation?