Despite recent criticism from “The Fact Checker” blog on The Washington Post, there is nothing at all misleading or biased about President Obama’s use of the 77 percent figure as a measure of wage inequality between women and men in the United States. Women’s median earnings for year round, full-time work in 2010 of $36,931 amounted to 77.4 percent of what men’s median earnings for year-round, full-time work were in the same year ($47,715). These numbers come from the Current Population Survey’s Annual Social and Economic Supplement (ASEC) and include the non-institutionalized civilian population who are either self-employed or work for wages or salary and are 15 years of age or older. These data are reported on an annual basis each year in August or September by the Census Bureau based on a household survey they conduct. This particular earnings series—annual median earnings for full-time, year-round workers—has the longest history, most likely explaining why it is the most frequently cited data series. It is the series on which NOW’s famous pin saying simply 59¢ was based, as that represented the wage ratio back in the late 1960s when NOW was founded. Because this data series has the longest history, its wage ratio serves as a well-known index to measure trends over time.
A Variety of Wage Gap Numbers
There are a range of numbers given for the wage ratio or gender wage gap (generally the gap is 100 percent minus the ratio, so with a ratio of 77 percent the gap is 23 percent), stemming from different data sets or different ways of analyzing the data. Each can be correct, depending on what the analyst wants to study. Each data set and methodology yields estimated pay gaps. Each is based on a survey, generally of a sample of all households, though wage data can also be gathered from samples of employers or of administrative records such as unemployment insurance or Social Security earnings records. Generally no data set is complete; all are subject to both sampling and non-sampling errors. Furthermore, different researchers may choose to extract different elements of data. For example, some researchers may restrict the age range of workers to prime age adults, those aged 25 to 54, in order to compare those for whom education is generally complete but who have not yet reached retirement age. To illustrate the burden of inequality faced by women of color, some researchers might compare the earnings of minority women to white men; others restrict the comparison of the earnings of minority women to those of minority men. Both ways can be correct depending on what one wants to illustrate.
In the United States, researchers generally restrict the comparisons to those who work full-time, whether on a weekly basis or an annual basis, but in other countries, for example Canada, total earnings of all workers (both those who work full time year-round and those who work part time or part year) might be compared in a gender wage ratio. If we do that in the United States we get a wage ratio of 72.4 percent and a larger gap of 27.6 percent. Although the Post‘s fact checker claimed that President Obama picked the wage ratio that made gender inequality look the worst, he clearly did not—he could have picked this one.
Another even smaller wage ratio (and larger wage gap) was generated by IWPR in our report entitled Still a Man’s Labor Market (February 2004), where based on a different data set, the Panel Study of Income Dynamics, on a survey of households conducted by the University of Michigan, we calculated that across 15 years, prime age women earned just 38 percent of what prime age men earned, for a staggering gender gap of 62 percent. This ratio is just as valid and just as accurate as others. It is telling us that across a 15-year period the typical woman in the United States earns only 38 percent of what the typical man earns. As the study points out, the reason women earn so much less across 15 years is that they spend more time out of the labor market; women typically work both fewer years and fewer hours per year than do men. No one would take this measure as a measure of discrimination by employers, but as a measure of women’s economic independence or lack of it or of what women contribute to family income across 15 years, this is an excellent measure. This type of life-time measure was used by the United Kingdom under the Labor government.
A larger wage ratio and smaller wage gap is generated by the Bureau of Labor Statistics from the Current Population Survey by looking at median weekly earnings for full-time workers each week of the calendar year and then combining those medians to get an annual median weekly earnings figure. Currently this ratio is a bit higher than the annual ratio released by the Census Bureau, standing at 82.2 percent for 2011. As an IWPR fact sheet shows, in some years, these two ratios are virtually identical, yet the Post fact checker made a big deal of how President Obama chose the lower ratio. Not so, President Obama just chose the most commonly used wage ratio. And, contrary to the fact checker’s claim, there is nothing superior about the weekly measure. It is not, for example, more inclusive: on the one hand it includes some workers who work full-time but not all year, but on the other hand it excludes the self-employed. It also underestimates earnings from annual bonuses—a substantial part of income in some high paying professional jobs and a source of income where a number of law cases show that women lose out.
In many countries an hourly wage ratio is used to avoid the measurement problem of full-time male workers working slightly more time than full-time female workers. Since, in the United States, some workers are paid on an hourly basis and others on a weekly basis, using either measure requires calculating a consistent wage measure, and the BLS does not routinely generate an hourly wage rate for all workers.
Response to Criticisms of the Wage Gap Measure
The most frequent criticism I hear of the wage gap is that it is comparing apples and oranges—it’s not comparing women and men in the same jobs or women and men who have the same education or same college major or whatever, and therefore the whole gap cannot be considered the result of pay discrimination. Interestingly, I don’t know of any individual or group who claims the whole pay gap is due to discrimination, so I don’t know why so much hot air is spent saying that it isn’t all due to discrimination. Many economists, sociologists, and other researchers have spent years trying to identify how much of the gap can be explained by factors that might reasonably affect wages, such as work experience, education, and so on. Generally in these analyses what cannot be explained by reasonable factors is considered possibly or likely the result of discrimination. Several comprehensive literature reviews that have been published in peer reviewed scholarly journals conclude that about 25 to 40 percent of the wage gap remains unexplained. But most of these studies do not assess whether some of the differences observed between women and men that might help explain the gender wage gap, like college major, are themselves the result of discrimination or of limited choice sets faced by women and men. In a world where most social workers are women and most engineers are men, few women and men may consider training for occupations that are nontraditional for their gender.
Much is also made of women’s choice to bear children and to spend some time out of the labor market as a result. But is that just a woman’s choice, or is it also a societal necessity? Years after that labor market absence should women still be suffering a wage penalty for that societal necessity? Or should society try to equalize the playing field by providing paid parental leave, encouraging fathers to share equally in child rearing, and providing subsidized, high quality child care to facilitate both parents’ return to the labor market?
The Case for Government Action
As Rachel Maddow recently pointed out on her news show, the existence of the wage gap should not be in dispute—the gap is there as measured in all the data sets released by federal government agencies. What is being argued about is whether that gap is meaningful; whether, if we can explain it by several reasonable factors, we don’t have to worry about it; whether we can pretend it isn’t really there. Conservatives, as she pointed out, tend to argue there is no gap, at least no gap that can be attributed to employer discrimination and therefore no gap that government policy needs to address. Liberals, in contrast, tend to argue there still is employer discrimination (with several horrendous cases of it coming to light each year as women bring legal actions against a wide variety of employers, despite the difficulty of doing so), and that, furthermore, a case can also be made for minimizing the negative economic effects of child bearing, particularly on women.
If we generally believe that women and men are equally talented and work equally hard on the job, that they tend to value the same things about work (such as making money and having some flexibility on the job), then they ought to be able to find opportunities in the labor market that pay them about the same. Yet while the evidence suggests that women and men generally do have equal ability and work equally hard and have equal value preferences, the evidence also suggests that they do not find labor market opportunities that tend to pay them about the same. In my view, this makes the case for government intervention.
The gender wage gap is a good measure of the lack of equal earnings between women and men in the labor market. Many women and men believe the gap should be smaller, that such a large gap as we have in the United States is unfair and reflects an unfair tendency for women to get paid less for what they do than men get paid for what they do. Moreover, such unequal pay inevitably leads to the misallocation of our human resources and a general reduction in U.S. productivity. Not only do women and their families suffer from unequal pay, but our society as a whole suffers as well, a circumstance that furthers the case for government intervention.
Heidi Hartmann is the President of the Institute for Women’s Policy Research.
To view more of IWPR’s research, visit IWPR.org