CH. 12: GENDER, RACE, AND ETHNICITY IN THE LABOR MARKET

1
CH. 12 GENDER, RACE, AND ETHNICITY IN THE LABOR
MARKET

• Chapter objectives
• Document levels and trends in earnings
  differentials by gender and race.
• Provide economic theories of wage discrimination.
  
• Review evidence on whether earnings differentials
  are due to discrimination

2
Source 2000 Economic Report of the President
3

• 2000 Economic Report of the President
• A recent study using longitudinal data from the
  late 1980s found that
• about one-third of the pay gap was explained by
  differences in the skills and experience that
  women bring to the labor market.
• about 28 percent of the gap was due to
  differences in the industries and occupations in
  which men and women worked and in their union
  status.
• Accounting for these differences raises the ratio
  of female to male median wages for the late 1980s
  from about 72 percent to about 88 percent,
  leaving around 12 percent unexplained.

4

• 2000 Economic Report of the President
• Even as several beneficial trends have tended to
  boost womens wages relative to mens and helped
  narrow the male-female wage gap, two major trends
  have worked simultaneously to widen it.
• increases in the pay premium for high skills
  (education/experience)
• Rising disparity in earnings across occupations
  (wages in male occupations are rising faster).

5
Measuring Discrimination

• The Blinder-Oaxaca Decomposition
• Y1a1 b1x1
• Y2a2 b2x2
• Y1-Y2 EXP UNEXP
• where EXP (x1-x2)b1
• UNEXP (a1-a2) (b1-b2)x1

6
2000 Economic Report of the President
7
Economic Theories of Discrimination

• Employer Discrimination
• Employee Discrimination
• Customer Discrimination
• Monopsonistic Discrimination

8
Employer Discrimination
Employer behaves as if minority workers true MRP
is MRP-d where d is measure of employer prejudice
against minority. Firm pays for discrimination in
the form of lower profits (BFG).
9
Employer Discrimination

• Non-discriminating employers generate horizontal
  portion of demand curve.
• Discriminating employers generate downward
  sloping portion of demand curve.
• Over time, non-discriminating employers should
  drive discriminating employers out of business.

10
Employer Discrimination
Statistical discrimination. Each workers
productivity is measured with some error true
productivity measured productivity error If
expected value of error depends on demographic
group, can use group status to improve estimate
of productivity. As variance of error within a
group rises, the value of discriminating on basis
of group status is reduced. Profitable to firm
and may persist.
11
Employer Discrimination
Monopsonistic discrimination. If search costs
create upward-sloping labor supply curves to
individual employers, and if discrimination
raises the search costs of certain groups of
workers, monopsonistic behavior creates wage
differentials among otherwise identical workers.
Profitable for firm may persist.
12
Evidence on Monopsonistic Discrimination

• Monopsonistic Discrimination and the Gender Wage
  Gap, by Erling Barth, Harald Dale-Olsen, June
  1999
• Using matched employer-employee data from Norway,
  we investigate the wage structure within and
  between establishments, and present novel
  evidence that the establishments' excess turnover
  of employees is sensitive to the wage premium of
  men, but not to the wage premium of women.
  Furthermore, we show that male turnover is more
  wage-elastic than female turnover.

13
Evidence on Employer Discrimination
American Economic Review, September 2004
14
Employee Discrimination

• Suppose employees from majority work dislike
  working with the minority group.
• Employee discrimination generates supply-related
  behavior that might cause employers to segregate
  their plants by race or sex if possible.
• If not, wage differentials arise as a result of
  the need of employers to retain workers in the
  majority group.
• Could be profitable for firm and may persist.
• Difficult to find empirical evidence on employee
  discrimination.

15
Customer Discrimination

• MRP for minority group is reduced because
  customer is willing to pay higher price for
  product when provided by majority group.
• Customers pay for prejudice in form of higher
  prices.
• Profitable for firm and may persist.

16
Evidence on Customer Discrimination
Quarterly Journal of Economics, August 1998
17
Evidence on Customer Discrimination
Journal of Labor Economics, January 1988.
18
Evidence on Customer Discrimination
of waitstaff that is male rises with of
clientele that is male, but independent of sex
of owners managers
Quarterly Journal of Economics, August 1996
19
Other empirical evidence

• Educational Achievement and black-white earnings
  differences
• Peer effects in the classom
• Gender, occupation choice and the risk of death