[An updated version of this article can be found at Gender Gap in the 2nd edition.]
When economists speak of the "gender gap," these days they usually are referring to systematic differences in the outcomes that men and women achieve in the labor market. These differences come in the percentages of men and women in the labor force, the types of occupations they choose, and the difference in the average incomes of men and women. These economic gender gaps have been a major issue in the women's movement and a major issue for economists.
The gender gap in U.S. labor force participation has been eroding steadily for 100 years (see chart 1). In 1890 the percentage of married white women who reported an occupation outside the home was extremely low—just 2.5 percent for the entire United States. The figure increased to 12.5 percent by 1940, 20.7 percent by 1950, and then by about 10 percentage points for every decade since then. By 1990 the labor participation rate for all married women had climbed to almost 60 percent, versus 78 percent for married men. (By 1990 women made up 45 percent of the total labor force.) In the forties and fifties, increases were the greatest for older married women, and then for younger married women in the seventies and eighties. And the eighties witnessed an increase in labor force participation of the sole group that had resisted change in previous decades—women with infants.
Chart 1. Labor Force Participation Rates of Men and Women, 1890-1990
SOURCES: Men: 1890 to 1970, U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970. Government Printing Office, 1975; and 1980 to 1990, U.S. Department of Labor, Bureau of Labor Statistics, Employment and Earnings.
Women: C. Goldin, Understanding the Gender Gap: An Economic History of American Women, table 2.1., 1990.
Enlarge in new window
The gender gap that gets the most attention, however, is in earnings. Although no comprehensive data exist for the period before 1950, evidence for certain sectors of the economy suggests that the gender gap in earnings narrowed substantially during two earlier periods in U.S. history. Between about 1820 and 1850, the era known as the industrial revolution in America, the ratio of female to male full-time earnings rose from about 0.3, its level in the agricultural economy, to about 0.5 in manufacturing. That is, women's earnings rose from, on average, about 30 percent of what men made to about 50 percent. From about 1900 to 1930, when the clerical and sales sectors began their rise, the ratio of female to male earnings rose from 0.46 to 0.56. In neither period did married women's employment expand greatly. Yet between 1950 and 1980, when so many married women were entering the labor force, the ratio of female to male earnings for full-time, year-round employees was virtually constant at 60 percent.
What accounts for the differences in earnings between men and women? Observable factors that affect pay—such as education, job experience, hours of work, and so on—explain no more than 50 percent of the wage gap. The remainder—termed the residual—cannot be explained by observable factors. This residual could result from workers' choices or, alternatively, from economic discrimination. Surprisingly, only 10 percent to 33 percent of the difference in male and female earnings can be explained by the differing occupations of men and women. The remainder is due to differences within occupations, and part of that is due to the observable factors.
Many observers have noted the paradox that as married women entered the labor force in steadily increasing numbers between 1950 and 1980, their earnings and occupational status relative to men did not improve. Yet that is not so paradoxical as it might seem. Indeed, with so many new female entrants to the labor force, an economist would expect women's wages to fall (relative to men's) because of the huge increase in supply. In other words, the pay of women relative to men probably stayed constant not in spite of, but because of the increase in the female labor force.
As more and more women entered the labor market, many of the new entrants had very little job market experience and few skills. If women tend to stay in the labor force once they enter it, the large numbers of new entrants will continually dilute the average labor market experience of all employed women. Various data demonstrate that the average job experience of employed women did not advance much from 1950 to 1980 as participation rates increased substantially. Economists James P. Smith and Michael Ward found that among working women aged forty, for example, the average work experience in 1989 was 14.4 years, hardly any increase at all over the average experience of 14.0 years in 1950. Because earnings reflect the skills and experience of the employed, it is not surprising that the ratio of female to male earnings did not increase from 1950 to 1980.
The gender gap in earnings decreased substantially during the eighties. By 1989 the ratio of female to male earnings for those who work full-time throughout the year had climbed by about 8 percentage points to 68 percent. Thus, in the nine years from 1980 to 1989, 20 percent of the preexisting gender gap in pay had been eliminated. Moreover, the size of the gender gap has been overstated. That is because women working full-time work about 10 percent fewer hours than men. Per hour worked, women now earn about 75 percent of what men earn.
According to economists June O'Neill and Solomon Polachek, the ratio of women's to men's pay increased for virtually all ages, all levels of education, and all levels of experience in the labor market. For workers with less than a high school degree, they found, the increase was 6.1 percentage points. For those with at least a high school diploma but no college degree, it was 5.3 percentage points. For those with at least a college degree, it was 7.2 percentage points. (These statistics are for whites twenty-five to sixty-four years old during the period from 1978 to 1987.) What is more, the gains occurred across all age groups. Although women in their thirties had the greatest gains relative to men their own age, the pay of older women relative to older men rose almost as much.
In this sense the move to greater gender equality in the eighties was remarkable. It was not merely a reflection of increased opportunities for younger or more educated women in relation to comparable groups of men. Moreover, the increase did not occur only at the point of initial hire. It is not surprising, therefore, that conventional methods of explaining the decrease in the gender gap in earnings—those that rely on changing composition of the female work force by education, potential job experience, occupational skill, and industry—can account for, at most, 20 percent of the increase.
Just as the stability of the earnings gap between 1950 and 1980 was probably due to the large influx of inexperienced women into the labor force, the narrowing of the gap in the eighties may owe to the fact that female participation rates are now exceedingly high. Because a larger proportion of women currently employed were previously in the labor force, their skills and experience cannot be greatly diluted by those of new entrants.
Other changes also account for the decrease in the earnings gap. Educational advances, particularly among the college-educated, have placed more women on par with men. College-educated women now major in subjects that are very similar to those chosen by men. Whereas in 1960 male college graduates outnumbered female by five to three, by 1980 the number of female and male college graduates was equal. In the sixties, for every hundred male recipients of professional degrees, there were fewer than five female recipients. By 1990 almost sixty females earned professional degrees for every hundred males. Young women are now forming more realistic expectations of their own futures than was the case twenty years ago. In 1968 only 30 percent of fifteen- to nineteen-year-old women said that they would be in the labor force at age thirty-five; by 1979 more than 70 percent thought they would be. Because the 1968 group vastly underestimated their future participation rate, they may have "underinvested" in their skills by taking academic courses that left them less prepared to compete in the job market.
To what extent has legislation narrowed the gender gap? One piece of legislation is Title VII of the Civil Rights Act of 1964, which forbids discrimination on the basis of sex in hiring, promotion, and other conditions of employment. The other is affirmative action. There is only scant evidence that either law has had any effect on the gender gap in earnings or occupations, although not enough research on this has been done to justify strong conclusions one way or the other (see the discussion in Ehrenberg and Smith, 1988, p. 577).
The gender gap in employment, earnings, and occupations has narrowed in various ways during the twentieth century, but with increasing significance, it seems, in the eighties. Whether or not the gap will continue to narrow and eventually disappear is uncertain, and probably depends on the gender gap in time spent in child care and in the home.
Claudia Goldin is the Henry Lee Professor of Economics at Harvard University and program director and research associate at the National Bureau of Economic Research in Cambridge, Massachusetts.
Ehrenberg, Ronald G., and Robert S. Smith. Modern Labor Economics: Theory and Public Policy, 3d ed. 1988.
Goldin, Claudia. Understanding the Gender Gap: An Economic History of American Women. 1990.
O'Neill, June, and Solomon Polachek. "An Analysis of Recent Trends in the Male-Female Wage Gap." Unpublished manuscript, 1991.
Smith, James P., and Michael Ward. "Women in the Labor Market and in the Family." Journal of Economic Perspectives 3 (Winter 1989): 9-23.