|
| |
         |
EXPLAINING TRENDS IN THE GENDER WAGE GAP June 1998
A Report by The Council of Economic Advisers I. Introduction II. Trends in the Gender Wage Gap III. Explaining the Trends in the Gender Gap IV. Discrimination V. Conclusion
Thirty five years ago, President Kennedy signed the Equal Pay Act into law, making it illegal to pay men and women employed in the same establishment different wages for "substantially equal" work1. At that time, the ratio of women's to men's average pay was about 58 percent2 . Although the gap in average pay between men and women has decreased since then, the "gender gap" in pay persists. Decreases in labor market discrimination towards women may be partly responsible for these improvements, but continued discrimination may also contribute to the remaining earnings gap. The long-standing differences in the average pay of men and women in the labor market are the result of many forces, including differences in the characteristics (such as average labor market experience) that men and women bring to their jobs, differences in the characteristics of the jobs in which men and women work, and differential and discriminatory treatment of women by employers and co-workers. All of these factors interact in complex ways. Hence it is difficult to determine precisely how much of the difference in female/male pay is due to discrimination and how much is due to differential choices and preferences by female workers. For example, if women have less experience than men, they may choose occupations where extensive experience is less necessary. If women consistently choose different occupations than men, stereotypes about women's abilities may be reinforced and discriminatory behavior by employers may be perpetuated. If employers make it difficult for women to enter certain occupations, women's incentives to invest in training for those occupations may be reduced. Women and men differ, both in terms of the jobs in which they work and in their responsibilities for children (which affect work at home and in the market). Nevertheless, over time, women's skills have become more similar to men's. The occupations and industries in which men and women work have also become more similar. But, as of the late 1980s, the date of the most recent detailed study, there were still large differences between men and women in personal and job characteristics that influenced their relative wages. And even when all of these differences were taken into account, a significant gender wage gap remained. As we discuss below, this suggests that there is continuing discrimination against women in the labor market.
II. TRENDS IN THE GENDER WAGE GAP Trends in the overall gap 
After hovering around 60 percent since the mid-1950s3, the ratio of women's to men's pay began to rise in the late 1970s, surpassing 70 percent by 1990 (see bottom line of figure 1).4 After a pause in the mid-1990s, the gender pay ratio has begun to rise again, reaching more than 75 percent in 1997 (see top line of Figure 1). Although the size of the gender pay gap varies across countries, many industrial economies have seen a narrowing of the gap over the last two decades. 5
In order to compare trends in the gender pay ratio across different demographic groups, a data series is needed that permits such disaggregation. Figure 2 shows the aggregate gender pay ratio in this data series, based on a slightly different group of workers.6 Although the level of the gender pay gap in Figure 2 is slightly different from that in Figure 1, the overall trends are similar. 
Figure 3 shows pay ratios for women workers by race and ethnicity, relative to all male workers. Relative to all male workers, wage gains have been faster for non-Hispanic black and non-Hispanic white women than for Hispanic women. 
Women of all ages and education levels experienced gains in pay relative to their male counterparts. The gender gap varies little across education groups7 but is substantially smaller for younger women (see figure 4). This was not always the case. In 1969, the gender gap was similar for those under and over age 40. But the gap narrowed much faster for younger women. The larger gap among older women may be a generational effect, indicating that the current younger cohort of women has made choices more similar to those of men and/or is not facing as much discrimination as the cohort before them. Alternately, the gap for older women could reflect growing differences in accumulated work experience as women age or a "glass ceiling" effect by which women achieve smaller pay gains over time. Indeed, the bipartisan Federal Glass Ceiling Commission found that "in the private sector, equally qualified and similarly situated citizens are being denied equal access to advancement into senior-level management on the basis of gender, race, or ethnicity." 8 
Traditionally, women with children have had lower wages than childless women. However, gains in average pay have been greatest for married women with children (see figure 5). It is worth noting that this is also the group of women among whom labor force participation has increased the most since 1970. Although their pay relative to men has climbed, single mothers continue to have the lowest average rate of pay. Gender gaps in fringe benefits The male-female difference in wages is also visible in fringe benefits, which currently make up about 30 percent of total compensation. As with wages, some of this gap is related to differences between men and women in human capital and job characteristics, and some remains unexplained. Among younger workers, the gender gap in total compensation appears to be smaller than the gap in wages.9 Much of the female-male gap in pension coverage can be accounted for by differences in their labor market histories and is much smaller among younger workers. In addition, among those who have pensions, the gender gap in benefit levels is largely explained by gender differences in income.10 Therefore, lower wages, and hence lower lifetime earnings, result in lower pension benefits upon retirement. For some women, the lack of coverage or lower benefit levels may not be a problem, since they receive benefits through a spouse. For other women, lack of adequate health or pension benefits from their job is a serious problem. More research is needed to understand the impact of lower fringe benefit coverage on female employees.
III. EXPLAINING THE TRENDS IN THE GENDER GAP In this section, we examine changes in the relative characteristics of female versus male workers and in the jobs in which they work, and relate these changes to the decline in the gender gap. The next section considers the role of discrimination. The gender gap has declined both because gender differences in measured characteristics have narrowed, and because the "unexplained" difference in pay (which may be discrimination, as discussed below) has also fallen. Among the measured characteristics, the two most important factors contributing to a narrowing of the gender wage gap between 1970 and 1995 have been improvements in women's relative labor market experience and improvements in their relative occupational status. Changes in unionism also benefitted women relative to men, but have played a smaller role. Although substantial, shifts in employment across industries have had relatively little effect on the gender pay gap. Increases in the returns to skill and increasing wage differences across occupations and industries -- particularly in the 1980s -- dampened women's relative wage gains, since women were disproportionately represented among the less-skilled and in lower-paying occupations. The Role of Human Capital Human capital, which includes education and labor market experience, is often viewed as the most important determinant of wages. An additional year of schooling is estimated to increase wages by 5 to 15 percent.11 An additional 25 years of work experience increases wages by an estimated 80 percent.12 Differences in labor market experience between men and women are far greater than differences in their educational attainment. But in the 1980s, the difference between men and women workers in average labor market experience began to fall. The increase in women's relative experience by itself would have reduced the pay gap by about 3.5 percentage points over the 1980s. Still, as of the late 1980s, when such data were last analyzed, remaining differences in work experience between men and women were found to explain about one-third of the pay gap that existed at that time.13 The Role of Family Status and Children The relationship between family status and pay is different for men and women. While married men, most of whom have children, typically earn more in the labor market than unmarried men, for women the relationship is reversed. Children are associated with lower wages for women but not for men, in part because children tend to reduce women's work experience and time with their employer. The pay premium for married men appears to have shrunk during the 1970s and 1980s.14 However, trends in earnings differences between mothers and women without children are less clear. Out of all women, gains in average pay over the last three decades have been greatest for married women with children (see Figure 5). But there is evidence that for younger women, the wage gains for mothers were much less than for non-mothers over the 1980s. 15 The Role of Occupation Men and women tend to work in different occupations. And wages differ substantially according to the gender composition of the occupation. In particular, men and women who work in predominantly female occupations earn less than comparable workers in other occupations.16 
Figure 6 indicates how men's and women's occupational status have changed over time. Women have increasingly moved into traditionally male occupations. Between 1983 and 1997, the proportion of employed women who worked in managerial and professional occupations increased from 23 to 32 percent, while the proportion of men in these occupations rose only from 26 to 29 percent.17 But women are still much more likely to work in service and clerical jobs than men, while men remain more likely to be in blue collar (craft, operator, and labor) jobs. Occupational segregation by gender began to decline noticeably in the 1970s. Such changes may be due to integration of formerly male or female occupations or to increases in the total employment share of occupations that have traditionally been more integrated. Movement of women into traditionally male occupations was the predominate cause of the decrease in occupational segregation in the 1970s and 1980s. In the 1980s, growth of overall employment in more integrated occupations was somewhat more important than it had been in the 1970s.18 The decline in occupational segregation alone in the 1980s would have reduced the gender gap by about three percentage points.19 Although occupational desegregation has continued in the 1990s, the rate of desegregation through the mid-1990s appears to have been somewhat slower than the rate during the 1970s and 1980s. 20 The Role of Unions Union membership is estimated to boost wages of union members relative to non-union members by 10 to 20 percent.21 Men have traditionally been more likely to be union members than women, which helped increase the gender pay gap. The decline over the last 25 years in the fraction of the workforce that is unionized has raised women's relative pay as fewer men receive union wages. In addition, the share of unionized workers who are female has increased as unions have declined less (or even grown) in certain public sector and service-related occupations that have a greater share of female workers.22 These female union members have benefitted from higher union wages. But, overall, the decline of unions has had a relatively small role in the declining gender pay gap; by itself, it would have caused the gender pay gap to decline by about one percentage point over the 1980s. The Role of Industry Differences in wages across industries are substantial and persistent. For example, wages for similar workers are 37 percent higher than average in the petroleum industry but about 17 percent lower than average in retail trade.23 Recent shifts in employment across industries (notably, the decline in the relative employment share of blue-collar jobs, where women are under-represented) have benefitted women relative to men.24 However, the industry shifts had relatively little effect on the gender gap in the 1980s.25 The Role of Changes in the Wage Structure Trends in the above factors have boosted women's wages relative to men's. However, two major trends have worked to widen the gender gap: increases in the pay premium associated with higher "skills" (i.e., higher levels of education and labor market experience) and increased pay differences across industries and occupations.26 This has served to widen the gender gap because female workers continue to have less labor market experience, on average, than male workers, and are, on average, in lower-paying occupations. The rising wage inequality and increasing economic returns to skills slowed women's progress during the 1980s and alone would have increased the gender pay gap by about four to six percentage points.27 The Role of Policy A number of policies in the 1960s targeted gender discrimination in the labor market. The Equal Pay Act of 1963 proscribes gender-based pay discrimination among employees within the same establishment who do "substantially equal" work.28 Title VII of the 1964 Civil Rights Act (and subsequent amendments) proscribes employment discrimination on the basis of sex in a broader set of categories, including hiring, promotion, and other conditions of employment. Executive Order 11246 (issued in 1965 and amended in 1967 to include sex) requires that non-exempt federal contractors and subcontractors take affirmative action in employment. Few studies have examined the effects of these policies on the gender pay gap. One difficulty in making such an assessment is that these policy changes took place at a time of enormous changes in gender roles and expectations. Isolating the effects of a single policy change from these broader social and economic changes is difficult. One study estimated the effects of affirmative action on hiring by comparing hiring in firms that are Federal contractors to those that are not (and therefore are not subject to affirmative action provisions). Employment of women increased somewhat faster in contractor firms.29 Although no such studies of the effect of policies on the gender pay ratio exist, it is clear that a role for these policy changes cannot be ruled out in both the increase in the gender pay ratio and the other changes that have served to increase the gender pay ratio.
What is gender discrimination in the labor force? Gender discrimination may take a variety of forms, from practices that reduce the chances that a woman is hired to differences in pay for men and women who work side by side doing the same tasks equally well. There are a variety of theories about how and why women face discrimination in the labor market. An employer may dislike female employees or underestimate their abilities; customers may dislike female employees or underestimate their abilities; or male co-workers may resist working with women. These attitudes may not be directed toward all workers but may only focus on women in higher status occupations. For instance, male employees may not object to having women work for them but may object when women are their superiors. In addition, employers may engage in what is called "statistical discrimination," meaning that they assume an individual woman has the average characteristics of all women. For example, because women on average have higher turnover rates than men, an employer may assume that a given female job candidate is more likely to leave the firm than a similar male candidate. Statistical discrimination, like other forms of discrimination, is illegal. An employer is required to base hiring or pay decisions on specific information about an individual, not on presumptions based on gender. The unexplained portion of the gender gap Although the gender gap has narrowed since the late 1970s, at 25 percent it is still substantial (see figure 1). And as late as the 1980s, the date of the most recent detailed longitudinal study, a gender pay differential of about 12 percent remained unexplained even after adjustments for gender differences in education, labor market experience, broad occupational and industrial distributions, and union status. According to this study, the gender pay gap ratio in 1988 was 72 percent. Women's lower full-time experience explained roughly one-third of the pay gap, and gender differences in industry, occupation, and union status explained about 28 percent of the pay gap. This left about a 12 percentage point pay gap unexplained. This unexplained differential, after adjustments, declined by half over the 1980s, from about 22 percentage points to about 12 percentage points. This decline alone would have reduced the gender gap by about 10 percentage points. 30 The unexplained portion of the pay gap is often interpreted as the result of discrimination. In this view, once differences between men and women in the relevant determinants of wages are taken into account, any remaining difference in pay must be due to discrimination. But this explanation may be too simplistic. To the extent that discrimination affects women's educational, job, and family choices, the "unexplained" differential will understate the true effect of discrimination. And, to the extent that an analyst cannot adequately measure all the determinants of wages using available data, there may be significant unmeasured labor market skills that differ between men and women. For instance, if women's labor market experience is less likely to be continuous (for example, due to childbearing), then just controlling for years of work may not fully control for the differential effects of experience on male and female wages. In this case, the "unexplained" differential will overstate the true effect of discrimination, because it includes the effect of relevant unmeasured factors that influence the relative productivity of male and female employees. The decline in the "unexplained" portion of the pay gap over the 1980s could either be due to a relative improvement in women's unmeasured labor market skills or a decline in discrimination. Both explanations are plausible. Women's unmeasured skills may have improved relative to men's over the 1980s. (For instance, women have entered elite private universities -- many of which were closed to women before the mid-1970s -- at an increasing rate in recent decades, perhaps increasing the quality of their schooling.) But reductions in discrimination may also have played a role in reducing the "unexplained" difference between men's and women's wages. For example, as women increased their commitment to the labor force and improved their job skills, statistical discrimination against them may have diminished. In addition, reduced discrimination could also have contributed to the decline in the explained portion of the pay gap if earlier anti-discrimination efforts encouraged women to invest more in labor market skills and to move into traditionally male occupations. Studies of gender pay discrimination Gender pay discrimination under the Equal Pay Act is said to occur when male and female workers employed in the same establishment receive different pay for "substantially equal" work.31 A small but growing area of research attempts to collect direct evidence on gender pay discrimination in the labor market in two ways. While none of these studies alone is definitive, taken together they present compelling evidence of the continued existence of gender discrimination in the labor market. First, a handful of studies have examined whether female workers earn less than comparable male workers within the same establishments and narrowly defined occupational categories. Although evidence is mixed, some studies find substantial pay differences between men and women working in the same narrowly defined occupations and establishments. One recent and thorough study of this kind takes advantage of a unique data set that matches workers in the 1990 Census to information on the establishments in which they work, using the U.S. Census Bureau Standard Statistical Establishment List. Unlike previous studies of this type, the data include workers from all sectors of the economy. The study decomposes the gender gap into a part that is the result of the concentration of women in particular occupations, industries, establishments, and "jobs" (an occupation within an establishment) and a portion due to differences in pay between men and women working in industries, occupations, establishments, and jobs with a similar percentage of female workers. The authors find that a substantial portion -- at least one-quarter -- of the pay gap is the result of differences in pay between men and women working in similar jobs and establishments.32 Second, an even smaller set of studies attempts to measure productivity of female and male workers directly to determine whether gender pay differences can be directly linked to productivity differences. This approach is rarely implemented because it requires information on wages and a reliable (and non-gender-biased) measure of productivity such as physical output per hour or sales. But a recent study of this sort found that pay differences between comparable women and men are too large to be explained by productivity differences: Using a large linked employer-employee data set, it concluded that "at the margin" women were 85 to 96 percent as productive as men but were paid only 66 to 68 percent as much as men.33 Finally, studies of discrimination in hiring offer additional, albeit indirect, evidence related to the gender pay gap. For example, the introduction of "screens" to conceal the identity of candidates from juries in auditions for symphony orchestras markedly increased female musicians' chances of success in such competitions and raised their odds of being hired.34 An audit study -- in which male and female candidates submitted essentially identical resumes for restaurant jobs -- found that female applicants were less likely than their matched male counterparts to be interviewed or hired as wait-staff in high-price restaurants (this was not true in low-price restaurants).35 There is some informal evidence that earnings tend to be higher in high-price restaurants; therefore this difference in hiring could contribute to gender differences in pay.
There is both good news and bad news with regard to gender pay differences. The bad news is that there remains a significant differential between women's and men's pay. On average, women earn about 75 percent of what men earn. Even after controlling for differences in skills and job characteristics, women still earn less than men. While there are a variety of interpretations of this remaining "unexplained" differential, one plausible interpretation is that gender wage discrimination continues to be present in the labor market. This interpretation is buttressed by other more direct studies of pay discrimination, which also show continuing gender differences in pay that are not explained by productivity or job differences. The good news is that these differences have decreased in recent decades. This is true not only for the "raw" gap in average female/male pay, which has decreased from about 40 percent to about 25 percent over the past two decades, but it is also true for the "unexplained" difference in female/male pay once factors that affect pay are controlled for. This suggests both that women's skills and job choices are becoming more similar to those of men, and that discrimination may be lessening as well.
1 Schultz v. Wheaton Glass Co., 421 F.2d 259 (3rd Cir. 1970). 2 Based on calculations from the March 1964 Current Population Survey (Bureau of the Census). Data are for weekly wages of full-time, year-round workers aged 25-64. 3 Goldin, Claudia. 1990. Understanding the Gender Gap. New York: Oxford University Press. 4 The data shown in the top line of the figure are ratios of median weekly wages for full- time workers aged 25 to 54, published quarterly by the Bureau of Labor Statistics since 1979. The bottom line shows the ratios of median weekly wages of full-time, year-round workers aged 25 to 54, calculated from the March Current Population Survey data files. Unfortunately, the former series is not available before 1979, and the latter series is not currently available after 1996. 5 Waldfogel, Jane. 1998. "Understanding the `Family Gap' in Pay for Women with Children." Journal of Economic Perspectives 12(1):137-56. 6 Data in figures 2 through 5 are based on tabulations from various years of the March Current Population Survey (Bureau of the Census). The data refer to full-time, year-round workers aged 25 to 64. 7 Blau, Francine. 1998. "Trends in the Well-Being of American Women, 1970-1995." Journal of Economic Literature 36: 112-65. 8 U.S. Department of Labor Glass Ceiling Commission. 1995. "Good for Business: Making Full Use of the Nation's Human Capital," March. Washington DC. 9 Solberg, Eric and Teresa Laughlin. 1995. "The Gender Pay Gap, Fringe Benefits, and Occupational Crowding." Industrial and Labor Relations Review, 48(4): 692-708. 10 Even, William and David Macpherson. 1994. "Gender Differences in Pensions," Journal of Human Resources, 29(2):555-587. 11 Council of Economic Advisers. 1996. "Promoting Economic Growth: Background Briefing Paper." July. 12 Freeman, Richard and Lawrence Katz. 1994. "Rising Wage Inequality: The United States vs. Other Advanced Countries." In Richard Freeman, ed. Working Under Different Rules. New York: Russell Sage Foundation. 13 Blau, Francine and Lawrence Kahn. 1997. "Swimming Upstream: Trends in the Gender Wage Differential in the 1980s." Journal of Labor Economics 15(1, Part 1): 1-42. 14 Blackburn, McKinley and Sanders Korenman. 1994. "The Declining Marital Status Earnings Differential." Journal of Population Economics 7:247-70. 15 Waldfogel, op. cit. 16 Blau, op. cit. 17 Data begin in 1983 due to a major change in the occupational classification system implemented in that year. 18 Blau, op. cit. 19 Blau and Kahn, op. cit. 20 Wootton, Barbara, 1997. "Gender differences in occupational employment." Bureau of Labor Statistics, Monthly Labor Review, April, pp.15-24. 21 Freeman, Richard and James Medoff. 1981. "The Impact of Collective Bargaining: Illusion or Reality?" in U.S. Industrial Relations 1950-1980: A Critical Assessment, IRRA. 22 Blau and Kahn, op. cit. See also Even, William and David Macpherson. 1993. "The Decline of Private Sector Unionism and the Gender Wage Gap." Journal of Human Resources 28(2):279-96. 23 Krueger, Alan and Lawrence Summers. 1988. "Efficiency Wages and the Inter-Industry Wage Structure." Econometrica 56(2): 259-93. 24 O'Neill, June and Solomon Polachek, 1993. "Why the Gender Gap in Wages Narrowed in the 1980s." Journal of Labor Economics 11(1): 205-228. 25 Blau and Kahn, op cit.. 26 Blau, op. cit. 27 Blau and Kahn, op. cit. 28 Schultz v. Wheaton Glass Co. op cit. The Equal Pay Act, 29 U.S.C. ' 206(d) (1), makes it illegal to pay men and women employed in the same establishment different wages for "equal work on jobs the performance of which requires equal skill, effort, and responsibility and which are performed under similar working conditions." 29 Leonard, Jonathan. 1984. "The Impact of Affirmative Action on Employment," Journal of Labor Economics, October, 2, 439-463. 30 Blau and Kahn, op. cit. 31 Schultz v. Wheaton Glass Co., op cit. 32 Bayard, Kimberly, Judith Hellerstein, David Neumark, and Kenneth Troske. 1998. "New Evidence on Sex Segregation and Sex Differences in Wages from Matched Employee-Employer Data." Paper prepared for the U.S. Bureau of the Census= International Symposium on Linked Employer-Employee Data, Washington DC, April. 33 Hellerstein, Judith , David Neumark, and Kenneth Troske. 1996. "Wages, Productivity and Worker Characteristics." National Bureau of Economic Research Working Paper 5626. June. 34 Goldin, Claudia and Cecilia Rouse. 1997. "Orchestrating Impartiality: The Impact of `Blind' Auditions on Female Musicians." National Bureau of Economic Research Working Paper No. 5903. January. 35 Neumark, David, Roy Bank and Kyle Van Nort. 1996. Sex Discrimination in Restaurant Hiring: An Audit Study. Quarterly Journal of Economics, August.
President and First Lady | Vice President and Mrs. Gore |