1. Labor Market Discrimination

2. Labor market discrimination
treating two individuals with equal qualifications differently for reasons unrelated to their productivity

3. Feedback effects of labor market discrimination
Faced with labor market discrimination that lowers the returns to human capital investments, women have less incentive to undertake them.
Women’s economic outcomes are therefore adversely affected both directly and indirectly by labor market discrimination.

4. Feedback Effects Gender Division of Labor in the Family
Gender Differences in Labor Market Outcomes

5. Similarities & Differences in Characteristics of Male & Female Full-Time Workers
Men and women have similar amounts of education.
Men have more full-time experience and women have more part-time experience.
Men are more likely to be in blue-collar jobs and to work in mining, construction, or durable manufacturing. Women are more likely to be in clerical or professional jobs and to work in the service industry.

6. The Gender Wage Gap
The percentage of the gender wage gap that is attributable to discrimination is usually measured as a residual after productivity factors are taken into consideration.
About 30.8% of the gender gap in wages can be explained by differences in labor market experience.
Differences in educational attainment only explain about 0.3% of the gender wage gap.
About 1.8% of the gender wage gap is explained by differences in the racial composition of the male and female full-time labor force. (Part of the effect of race is due to poorer quality education and part is due to racial discrimination.)
Even when differences in occupations and industries and differences in union/non-union status are taken into consideration, 38% of the gender wage gap remains unexplained. (Given some forms of gender discrimination, taking these factors into consideration is questionable.)

7. Female-Male Wage Ratio
Women earn on average 72% what men earn.
If women had the same education, experience, and race as men, their wages would be 81% of men’s wages.
Even if women had the same industrial and occupational distribution, and the same union coverage, their wages would still only be 88% of men’s wages.

8. Measuring discrimination accurately is problematic, partly because it is not possible to take into consideration all the variables than can affect potential productivity.
If, for example, men are more motivated, then discrimination will be overestimated due to omitted factors.
If women have greater interpersonal skills, then discrimination would be underestimated.

9. Another problem with measuring discrimination occurs when studies control for variables that themselves reflect the direct effects of discrimination. Then, the impact of discrimination on the gender wage gap is underestimated.
For example, women may be excluded from some jobs due to discrimination in hiring or promotion. Therefore, controlling for occupation and industry will lead to underestimating discrimination.

10. Some Experimental Studies
Male & female pseudo-job seekers were given similar resumes and sent to apply for jobs waiting on tables at the same set of high-priced Philadelphia restaurants.
A female applicant’s probability of getting an interview was 40 percentage points lower than a male applicant’s, and her probability of getting an offer was 50 percentage points lower.

11. When a screen was used to conceal the identity of candidates at auditions for musicians for symphony orchestras, the probability that a woman would advance out of the preliminary rounds and be the winner in the final round increased substantially.
The switch to blind auditions was found to explain one quarter of the increase in the percentage female in the top five symphony orchestras in the United States, from less than 5% of all musicians in 1970 to 25% in 2000.

12. Other Findings
In a study of small firms, male employers paid higher wages and employed fewer women.
Women’s chances of being hired and promoted were greater when there was a higher proportion of women at the level of the job being filled.

13. Discrimination Models: How does discrimination operate?

14. Becker’s Model of Taste Discrimination
Tastes for discrimination may exist on the part of
1. Employers
2. Employees
3. Customers

15. Employer Taste Discrimination
First, consider an employer with no taste for discrimination.
The employer hires workers up to the point where the amount that the worker contributes to revenues is equal to the amount the worker adds to costs.
That is where marginal revenue product (MRP) equals marginal resource cost (MRC).

16. Marginal Revenue Product MRP = TR/L
Marginal Resource Cost MRC = TC/L.

17. Marginal Revenue Product
MRP = TR/L
= (TR/L)(Q/Q)
= (TR/Q)(Q/L)
= (MR)(MPP),
where MR is the marginal revenue and MPP is the marginal physical product.
For firms in perfectly competitive product markets, MR equals the product price P.
So MRP would be (P)(MPP) which is the value of the marginal product VMP.

18. For nondiscriminatory firms in perfectly competitive labor markets, the marginal resource cost (MRC = TC/L) is simply the wage.

19. Consider now an employer who has a taste for discrimination against women. He/she acts as if there is a nonpecuniary cost of hiring women. This cost or discrimination coefficient can be called dr.
To the discriminatory employer, the cost of employing a man will be his wage wm, but the cost of employing a woman will be wf + dr.

20. The discriminatory employer will hire a woman only if her cost (wf + dr) is no higher than that of a man (wm). The employer will be indifferent between hiring a man and a woman if wf + dr = wm, or wf = wm- dr.
Therefore, if men and women are equally productive, and men are paid their marginal revenue product, women are paid less than their productivity.

21. If there are a relatively large number of nondiscriminatory firms or there are relatively few women seeking employment they may all be absorbed by the nondiscriminatory firms.
Then there will be no discriminatory pay differential based on gender even though some employers have tastes for discrimination against women.

22. However, if discriminatory tastes are widespread or there are relatively many women seeking employment, some women will have to find jobs at discriminatory firms.
If the labor market is competitive, all employers pay the same market wage for workers of a particular gender.
In equilibrium, the market wage differential between men and women must be large enough so that all women who are looking for employment obtain it.
So the more prevalent and the stronger employers’ discriminatory tastes against women and the larger the number of women seeking employment, the larger will be the market wage gap (wm - wf ) between men and women.

23. Consider two employers, one is a discriminator and one is not.
The non-discriminator operates solely to maximize profits, without regard to tastes for different employees. He/she will hire the profit-maximizing amounts of employees of all types.
The discriminator, acting as though women are less productive hires fewer women, and is therefore not hiring the profit- maximizing number of workers.
So, the non-discriminator will have greater profits than the discriminator. The discriminator is giving up some profits to indulge his/her discriminatory taste.

24. If a discriminatory employer operates in a competitive product market, where long run economic profits are zero (that is, there are only normal accounting profits), the employer will be driven out of business by other firms with no taste or less taste for discrimination.

25. If, however, an employer operates in a product market which is not competitive, the firm will be able to sacrifice some profit in order to discriminate.
There have been studies that have found evidence of more discrimination in monopolized product markets.

26. Employee Taste Discrimination
If a male employee has tastes for discrimination against women, he will act as if there were nonpecuniary costs of working with women.
This cost or discrimination coefficient can be called de. This is the premium he must be paid to induce him to work with women.

27. A discriminatory man, who would receive wm if he did not work with women, would only be willing to work with women if he received at least wm + de.
One way that an employer might deal with this situation is to hire a sex-segregated work force.
If all employers responded this way (but had no taste for discrimination themselves), male and female workers would work in segregated settings and be paid the same wage rate.

28. If, however, discriminatory employee tastes are widespread or there is a relatively large number of women seeking jobs, some of the women will have to work with discriminating male workers.
Those males will require higher compensation to induce them to work with women.
The result will be a wage differential between male and female workers, on average, since some males will receive this higher pay, and women may be paid less to compensate.

29. There will be more variation in male workers’ wage rates than would otherwise be the case.
Discriminating male workers who do not work with women do not need to be paid a wage premium, nor do nondiscriminatory males, regardless of whether or not they are employed with women.

30. Employee discrimination may also adversely affect the morale and productivity of discriminating male workers who are forced to work with women.
This would make employers reluctant to hire women.
If employers did hire women under such circumstances, they would pay them less to compensate for the reduction in the productivity of the discriminating male employees.
In a sense, a woman’s marginal productivity is lower than a man’s because adding her to the work force causes a decline in the productivity of previously employed male workers. Adding another male causes no such decline in output.

31. Employee discrimination may also reduce the productivity of women in comparison with men, if discriminating male employees are reluctant to teach their new female coworkers important skills. As a result the women would learn less and be less productive.

32. Customer Taste Discrimination
If customers have tastes for discrimination against women, they will act as if there were nonpecuniary costs of purchasing a good or service from a women.
This cost or discrimination coefficient can be called dc.

33. The discriminating customer behaves as if the full price of the good or service is p + dc, if provided by a woman but only p if sold by a man.
Then at the going market price, women will sell less.
So customer discrimination can result in potentially equally productive women being less productive (in terms of revenue) than comparable men.
Women are therefore less desirable employees and will receive less pay.
If such customer discrimination exists in some areas but not in others, occupational segregation may also result.

34. Statistical Discrimination
Employers are constantly faced with the need to make decisions under conditions of incomplete information and uncertainty.
They never know for certain how individuals will perform on the job or how long they will stay with the firm after being hired.
Mistakes can be costly, especially when there are substantial hiring and training costs.

35. Suppose the firm believes that, on average, women are less productive or less stable than men. They realize that some women are more productive or more stable than some men.
However, it may be very costly to invest the resources needed to distinguish the more productive or more stable workers.
So the firm may only hire men or only hire a woman at lower wages.
Thus, statistical discrimination against individual women occurs.

36. Productivity Distributions of Men as Perceived by a Hypothetical Employer
males
Productivity
avg prod of males

37. Productivity Distributions of Women as Perceived by a Hypothetical Employer
females
Productivity
avg prod of females

38. Productivity Distributions of Men & Women as Perceived by a Hypothetical Employer
males
females
Productivity
Distribution
avg prod of females avg prod of males

39. Feedback Effects of Statistical Discrimination
Suppose a firm believes that women are less stable workers.
They, therefore, invest less firm-specific training in the women and assign them to jobs where turnover costs are minimized.
Since the women do not receive the benefits of firm-specific training, they have less incentive to stay and are therefore more likely to leave.
Since the employers perceptions have been confirmed, they see no reason to change their discriminatory behavior.
If the firm had placed the women in positions that rewarded stability, the women might well have been stable workers.

40. Studies indicate that most of the difference in quit rates between men and women is explained by the types of jobs women are in and other individual characteristics.
These findings suggest that when women are confronted with the same incentives to remain on the job in terms of wages, advancement opportunities, etc., a woman is no more likely to quit than a comparable male worker.

41. Overcrowding
Regardless of the reason for gender segregation (socialization, personal preferences, labor market discrimination, etc.), the consequence may be a male-female pay differential. This will occur if demand (job opportunities) in the female sector is small relative to the supply of workers available for work.

42. Overcrowding Wage Wage Sfo Smo W0 W0 Df Dm Lf0 Labor Lm0 Labor
F Occupations
M Occupations
Suppose that men and women are equally qualified for occupations F and M. Then the wages for those jobs will be the same (W0).

43. Overcrowding Wage Wage Smd Sfo Smo Wmd W0 W0 Df Dm Lf0 Labor Lmd Lm0
F Occupations
M Occupations
If women are excluded from the M occupations, the reduced supply in those jobs results in higher wages.

44. Overcrowding Labor Wage Sfo Sfd Df Lf0 Lfd F Occupations W0 Wfd Wage
Smd
Smo
Wmd W0 Dm
Lmd Lm0
Labor
M Occupations
In the F occupations, where women can get jobs, the supply of workers is increased and the wages fall.

45. Effect of Overcrowding on Productivity
In the low-wage F sector, employers substitute labor for capital, producing in more labor-intensive ways.
In the higher-wage M sector, employers economize on labor and produce in more capital-intensive ways.
Since the workers in the F sector have less capital with which to work, their productivity is lower.
Thus, potentially equally productive workers in the two sectors have different wages and different productivity levels.

46. Institutional Models
Some firms hire workers from the outside labor market for entry level jobs.
The remainder of the jobs are allocated internally by the firm as workers progress along well-defined promotional ladders, by acquiring job-related skills many of which are firm-specific.
This situation is called an internal labor market.
If it is believed that different groups of people (men/women or whites/blacks) differ in their productivity-related characteristics (such as quit or absenteeism rates), statistical discrimination is likely to result in their being channeled into different jobs.

47. According to the dual labor market model, there are primary jobs and secondary jobs.
Primary jobs have high levels of firm-specific skills, and thus pay high wages, have good promotion opportunities, and long-term attachment between workers and firms.
Secondary jobs have low levels of firm-specific skills, lower pay, fewer promotion opportunities, and high turnover rates.
Men are more likely to be in primary jobs and women in secondary ones.

48. Once men and women are channeled into different types of entry jobs, the normal operation of the firm will lead to gender differences in productivity, promotional opportunities, and pay. This is called institutional discrimination.

49. Government Policy & Labor Market Discrimination
Government policies to combat labor market discrimination can be justified on at least two grounds.
Equity or Fairness
Efficiency

50. Equity
Assuring equal treatment for all participants in the labor market regardless of gender, race, ethnicity, etc.

51. Efficiency
Unequal treatment on the basis of gender, etc. may result in an inefficient allocation of resources.
Prices do not serve as accurate indicators of social costs.
Consider for example, the overcrowding model. Since female labor is underpriced, firms will produce too much of the outputs produced by that labor. Since male labor is overpriced, firms will produce too little of the outputs produced by that labor.
In addition, since women are deterred from investing in their human capital by discrimination, society loses a valuable resource.

52. The Equal Pay Act of 1963
Outlaws separate pay schedules for men and women for work requiring similar skills and performed under the same working conditions.
Did not mandate equal employment opportunity.

53. Civil Rights Act of 1964, Title VII
Prohibits all forms of sex discrimination in employment including hiring, firing, promotion, training and fringe benefits.
Established an enforcement agency, the Equal Employment Opportunity Commission (EEOC) , with the responsibility to bring civil actions against private firms.

54. The Civil Rights Act also prohibits employment discrimination on the basis of race, color, religion, and national origin.
It does NOT prohibit employment discrimination on the basis of sexual orientation or gender identity (transgender status).

55. There have been attempts to amend Title VII to include discrimination based on sexual orientation and gender identity.
However, the bills have been unable to pass through Congress.

56. Discrimination based on Gender Stereotyping & Nonconformity
In 1989, Price Waterhouse v. Hopkins, the Supreme Court decided that illegal discrimination on the basis of sex had occurred when a woman was denied partnership because she was not feminine enough, that is, she did not conform to stereotypes of how women should act.
After Price Waterhouse v. Hopkins, some cases by gay, lesbian, and transgender individuals have been argued on the grounds of gender stereotyping and nonconformity.

57. Most states have no state law explicitly prohibiting discrimination based on sexual orientation.
Of the states that do prohibit discrimination based on sexual orientation, not all of them prohibit discrimination based on gender identity. 

58. This map shows which states have these laws:

59. In states that prohibit discrimination on the basis of sexual orientation, what do the laws cover?
In most states protecting sexual orientation, sexual orientation includes heterosexuality, homosexuality, and bisexuality.
In some states, it is only necessary that the alleged harasser believes that the victim is heterosexual, homosexual, or bisexual; the actual orientation of the victim is of no consequence.

60. Most states that offer protection on the basis of sexual orientation have created a statutory exemption for religious & charitable organizations.
These groups are free to discriminate in hiring on the basis of an individual’s sexual orientation where such action furthers the beliefs or principles of the organization.

61. Equal Employment Opportunity Commission (EEOC) Perspective
Although the discrimination prohibition of Title VII of the Civil Rights Act of 1964 does not explicitly include sexual orientation or gender identity, the EEOC has stated that discrimination on the basis of sex includes discrimination based on sexual orientation and gender identity.
This EEOC claim has been supported by some court cases.
The EEOC, therefore, investigates complaints of employment discrimination based on sexual orientation or gender identity, acts to stop the discrimination, and seeks remedies for the individuals filing the complaints.

62. Returning to gender discrimination …
Title VII of the Civil Rights Act permits exceptions to its ban on gender discrimination, when sex is found to be a bona fide occupational qualification (BFOQ).
Example – 1977 Dothard v. Rawlinson case:
Supreme Court allowed the hiring of males only for the position of guard in Alabama’s maximum security male prisons.
Due to the nature of the prison population and the atmosphere of the prison, women were expected to be particularly subject to sexual assault, which would interfere with their job performance.

63. Disparate Treatment vs. Disparate Impact
Disparate treatment is differential treatment of women and minorities with the intention to discriminate and is a clear violation of Title VII of the Civil Rights Act.
Disparate impact arises when a firm’s apparently neutral hiring or promotion practices have a disproportionately adverse effect on women or minorities. (For example, a minimum height or weight requirement screens out more women than men.)
After a disparate impact has been demonstrated, the burden of proof is shifted to the employer to show that the practice is a matter of business necessity or that the requirement is job related. Otherwise, the practice is discriminatory.

64. Executive Order 11246 issued in 1965 & amended in 1967 to include sex
prohibits federal contractors from discriminating in employment based on race, color, religion, sex, national origin, disability or veteran's status.
requires affirmative action for classes of workers disadvantaged by past discrimination. Note that affirmative action does NOT impose quotas on employers.
enforced by the Office of Federal Contract Compliance.

65. Eight Factors to be Considered by Federal Contractors in an Affirmative Action Plan
The minority population in the “immediate labor area” (the geographic area from which employees may reasonably commute);
The availability of women seeking employment in the immediate labor area or in the area from which a contractor can reasonably recruit (which may include the region or the nation in some occupations);
The size of the minority or female unemployed labor force in the immediate labor area;
The percentage of minorities and women in the workforce in the immediate labor area;

66. The availability of minorities and women with the requisite skills in the immediate labor area from which the contractor can reasonably recruit;
The availability of minorities and women in the contractor’s organization who are promotable;
The existence of training institutions capable of training persons in the area with the requisite skills; and
The degree of training the contractor is reasonably able to undertake.

67. Comparable Worth
Comparable worth extends the idea of equal pay for equal work to the broader concept of equal pay for work of comparable worth within the firm.
Some opponents of comparable worth argue that determining appropriate values is not possible to do. While job evaluation is difficult, it is widely used by governments and many large firms.

68. Job Evaluation
Step 1. Decide on a set of factors that are believed to determine pay differences across jobs (for example, education, training, experience). For each factor, there is a scale representing increasing levels of worth, and each level is given a certain number of points.
Step 2. Rate each job in the firm based on the factors.
Step 3. Combine the ratings to create a score for each job.
Step 4. Use the scores to help determine wages.

69. Sometimes weights are assigned according to the judgment of experts constructing the scale.
Frequently, the statistical technique called multiple regression is used to discern the value of the different factors to the existing pay structure.

70. Multiple Regression Approach
We have information on all employees on annual income and factors of the job that result in that income. Using statistical analysis, we estimate the relation between the factors (X’s) and the income Y.
Y = a + b1 X1 + b 2 X2 + b 3 X3 + …
The technique estimates the a and the b’s so that the value obtained for Y when the X’s are plugged into the equation comes the closest to the actual values of the Y’s.

71. Regression – Simple Case: Y = a + b X Determine the values of a & b (the Y-intercept & the slope), so the line best fits the data.
Y (income)
X (experience)

72. Regression – Simple Case: Y = a + b X Determine the values of a & b (the Y-intercept & the slope), so the line best fits the data.
X (experience)
Y (income)
Y = a + b X

73. Regression Approach
In the simple example graphed above, suppose that the equation estimated is Y = X, where Y is annual income in thousands of dollars and X is years of experience at the firm.
The equation implies that the firm’s expected starting salary (0 years of experience) would be 15 thousand dollars and the expected salary of an individual with 1 year of experience at the firm would be 17 thousand dollars.

74. Job Evaluation – Problems
Both existing pay structures and the judgments of individuals may be tainted by inequalities in the economy and in society.
Also, the job evaluation approach is more difficult to implement in small firms than in large firms, where there is more information with which to work.

75. What happens when women’s wages increase?
As employers shift to less labor-intensive methods of production, the demand for labor drops, pushing employment down.
The increased wage costs are likely to be passed on to the consumer in the form of higher prices. Then demand for the products/services would decline, again pushing employment down.
However, the higher income resulting from the increased wages would increase demand for goods and services, which would increase the demand for labor and at least partially offset the decrease in employment mentioned above.

76. All we can conclude currently is that the use of job evaluation to determine pay rates is neither an impractical dream nor a sure cure for discrimination.
Any policy to correct discrimination must consider both the supply and demand sides of the labor market in order to attain the desired outcome.