4-3-06 - 151 lecture Gender and Race in the Labor Market...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 151 lecture Gender and Race in the Labor Market Interesting and important puzzle: We can understand how discrimination and inequality can persist in conditions of legal segregation—like South Africa under apartheid—or when a caste system is so strong—India—that its enforcement is the social norm, breaking it is very rare. But what about the case of a capitalist economy, where competition among employers should lead to freedom and equality for all? This is exactly the challenge posed by Milton Friedman and taken up by Gary Becker (1957). I. Patterns of inequality A. 1. Gender Earnings trends—for full-time workers Ratios for women relative to men: 1970: 50; 1980: 59; 1999: 65. 2. Causes- or correlates—via earnings decomposition controlling for differences in quantities Age and education: expect improvement for younger cohorts and those investing in skills for better-paying professions. But gaps remain even for youngest women with the most education. Occupation, experience, hours. These variables all together still do not “explain” – or account—for the earnings differences by gender. Residual could be result of women’s preferences, or women’s comparative skills in working in the home, or the result of past discrimination that still creates differences in the labor market, or current labor market discrimination. 1 3. Discrimination: concepts and definition Unequal treatment of people who are equal versus much of discussion about inequality for women concerns different pay for comparable work, not for the same work: for example, librarians and childcare teachers have more skills, compared to car washers or gardeners, but are paid less. Occupational segregation by gender: “women’s work.” Has declined some but still many occupations are dominated by one gender or the other. Female-dominated occupations are correlated with lower pay. As text says, occupational segregation is associated with 8-10 percent decline in women’s earnings. But is occupational segregation the result of labor market discrimination (e.g. by personnel managers) or gender preferences (e.g., for caring or indoor work) or pre-market discrimination (e.g., steering by high school counselors, parents)? Blau, using 1988 data finds that of 28 percent M-F wage gap, 16 percent accounted for by productive attributes like experience. 12 percent is the residual. B. 1. Race Earnings trends, for F-T, FY workers, black to white ratios, separately for men and women. Trends show gains in 1970s (actually but not shown in text) in 1965 to 1975, with very little improvement since in earnings ratios. 2. Employment trends: employment to population ratios are much lower for blacks- especially for black males, which means they are falling even faster than for whites. This true at younger ages, so not a phenomenon of earlier retirement—as it is for white males. How much of earnings inequality is pure wage discrimination? 3. 2 Blau and Kahn, cited in text, find a residual of 11 percent, similar to genderassociated residual, but note that their calculation includes controls for education, hours, occupation, firm size, and some of these differences are likely to reflect discrimination as well—e.g., discrimination not just ins starting pay, but also in patterns of promotion. Some of the differences may reflect not labor market discrimination but premarket entry influences, like maternal health care, early childhood development, school systems, that are unequal by race. And these can interact with residential segregation and housing discrimination. AFQT: some economists have used data on these test scores to show that they account for the entire residual, but others point out that the scores are themselves “predicted” by other factors that involve racial inequality. Another, more direct approach to studying discrimination: audit studies using pairs that are similar except for race, ethnicity or gender. Sometimes difference is just in whether first name is more common among African Africans. Audit studies find substantial differences in getting a taxi, price for a car, getting an apartment or mortgage, landing an interview for a job, getting a job. These audits were done in a variety of cities in the 1990s and suggest that inequality is continuing to be reproduced via present-day discrimination. Another study involves blind hiring: musicians who audition from behind a screen. This practice has led to significant increases in hiring of female orchestra players. How can we explain the persistence of discrimination? II. A. 1. Theories of Discrimination “Personal prejudice” theories Becker’s employer discrimination model: lower profits for “psychic gain” of not associating physically with blacks. Not as appropriate for gender. In any case, “d” should be forced to zero with competition, so predicts the disappearance of the phenomenon it is trying to explain. 3 2. 3. Similar remarks apply to sustainability of customer discrimination. Employee discrimination: predicts segregation by workplace but not wage inequality. Statistical discrimination theory B. Basic idea: treatment of individuals as members of a group, which can involve prejudice by a personnel manager. For example, if you believe average Stanford graduate is less productive than an average Cal graduate, rather than trying to obtain information about them as individuals—GPA, extracurriculars, test scores, whatever. Or belief that women are more likely leave the company to have children. Statistical discrimination can be compatible with competition if there are personnel “frictions”: such as hiring and separation costs, so it is profitable to keep the existing workforce rather than to replace them all. Then differential won’t be competed away. Moreover, such treatment can feed back into the education system—a lower return to schooling for blacks, for example, can then lead to less investment—in attainment as well as in learning. These models depend heavily, though, on costly information. We know that firms experience some turnover and that they spend heavily on testing, interviewing, and most importantly, on 90 day or 180 day or more probationary periods, when letting a worker go is very easy. So by the end of the probation firms know a lot more about each worker than their gender or race. C. 1. Noncompeting models Crowding models These take the patterns of crowding as given the pattern of crowding—via gender typing, or residential segregation. 2. Dual labor market models: 4 Some groups disproportionately represented in the secondary labor market. Show that limits to competition and mobility lead to difference by race and gender that can be perpetuated. 3. Monopsony – “job search” costs. Suppose African-Americans and women and Latinos, unlike white males, know that some firms will discriminate but do not know which ones will So they will engage in longer process of job search, which implies they have upward-sloping labor supply schedules. This is sufficient to have discrimination via discriminating monopsonists— as in Fig. 12.8. Get a wage differential that is consistent with profit-maximizing behavior. However, it requires that supply curves of female or black labor are less elastic than white male labor, when in fact the opposite is the case. 4. Bargaining power models II. A. Government policy to end inequality and discrimination Anti discrimination policy— law since 1965, enforcement since1970. What accounts for improvements in 65-75: improvements in education for minorities—quantity and quality. 20 percent and 20 percent But note also dropouts from labor force lead to spurious improvements via selection effects. 10 percent. Half of improvement remains: results of enforcement efforts, especially in the old segregationist South. B. Affirmative action: “ The Mouse That Roared.” Was never very important and mostly ended in early 1980s. 5 Different from anti-discrimination policies. Note that ban on affirmative action in California (since 1996) does not affect employment policies, because Federal law pre-empts state law. The ban does apply to admissions policies in higher education. 6 ...
View Full Document

This note was uploaded on 06/06/2010 for the course ECON 151 taught by Professor Staff during the Spring '08 term at Berkeley.

Ask a homework question - tutors are online