Do women earn less money than men for equal work?

Economics professor Nancy Folbre says they do, in this blog post that the New York Times' website published yesterday. At the top of her post, she claims:

Tuesday is the day on which women’s wages catch up to men’s wages from the preceding week. On average, female workers have to put in more than six days of paid work to earn what men earn in five.

Among those who usually worked full time during the first quarter of 2009, women’s median weekly earnings were 79 percent those of men. That implies that the catch-up clock for them rings at about 10:38 a.m. on Tuesdays (assuming a standard five-day week and eight-hour day starting at 8 a.m.).

Some women earn less than men because they choose less lucrative occupations or take more time out from employment. But a 2003 Government Accountability Office study controlling statistically for these factors showed that women’s average pay between 1983 and 2000 flat-lined at about 80 percent of men’s over the entire period.

You'll notice that her link on the word "study" goes to this PDF of the 2003 U.S. government report. Well, if you actually click the link and read the report, you'll find that her summary is literally correct but misleading. The authors conceded that while they tried to control for confounding factors, they might have failed to adequately consider others:

[W]omen have fewer years of work experience, work fewer hours per year, are less likely to work a full-time schedule, and leave the labor force for longer periods of time than men. Other factors that account for earnings differences include industry, occupation, race, marital status, and job tenure....

Even after accounting for key factors that affect earnings, our model could not explain all of the difference in earnings between men and women. Due to inherent limitations in the survey data and in statistical analysis, we cannot determine whether this remaining difference is due to discrimination or other factors that may affect earnings. For example, some experts said that some women trade off career advancement or higher earnings for a job that offers flexibility to manage work and family responsibilities.

In conclusion, while we were able to account for much of the difference in earnings between men and women, we were not able to explain the remaining earnings difference. It is difficult to evaluate this remaining portion without a full understanding of what contributes to this difference. Specifically, an earnings difference that results from individuals’ decisions about how to manage work and family responsibilities may not necessarily indicate a problem unless these decisions are not freely made. On the other hand, an earnings difference may result from discrimination in the workplace or subtler discrimination about what types of career or job choices women can make. Nonetheless, it is difficult, and in some cases, may be impossible, to precisely measure and quantify individual decisions and possible discrimination. Because these factors are not readily measurable, interpreting any remaining earnings difference is problematic.

Now, maybe you think 80% is such a huge discrepancy that it couldn't plausibly be explained by unmeasured factors.

But that's an old report -- it only looked at data up to 2000. The GAO released a new report yesterday that looks at data from as recently as 2007 and suggests that we should be much more optimistic:

GAO used data from the Office of Personnel Management's (OPM) Central Personnel Data File (CPDF)--a database that contains salary and employment data for the majority of employees in the executive branch. GAO used these data to analyze (1) "snapshots" of the workforce as a whole at three points in time (1988, 1998, and 2007) to show changes over a 20-year period, and (2) the group, or cohort, of employees who began their federal careers in 1988 to track their pay over a 20-year period and examine the effects of breaks in service and use of unpaid leave....

The gender pay gap--the difference between men's and women's average salaries--declined significantly in the federal workforce between 1988 and 2007. Specifically, the gap declined from 28 cents on the dollar in 1988 to 19 cents in 1998 and further to 11 cents in 2007. For the 3 years we examined, all but about 7 cents of the gap can be explained by differences in measurable factors such as the occupations of men and women and, to a lesser extent, other factors such as education levels and years of federal experience. The pay gap narrowed as men and women in the federal workforce increasingly shared similar characteristics in terms of the jobs they held, their educational attainment, and their levels of experience. For example, the professional, administrative, and clerical occupations--which accounted for 68 percent of all federal jobs in 2007--have become more integrated by gender since 1988. Some or all of the remaining 7 cent gap might be explained by factors for which we lacked data or are difficult to measure, such as work experience outside the federal government. Finally, it is important to note that this analysis neither confirms nor refutes the presence of discriminatory practices.

GAO's case study analysis of workers who entered the workforce in 1988 found that the pay gap between men and women in this group grew overall from 22 to 25 cents on the dollar between 1988 and 2007. As with the overall federal workforce, differences between men and women that can affect pay explained a significant portion of the pay gap over the 20-year period. In particular, differences in occupations explained from 11 to 19 cents of the gap over this period.

Prof. Folbre acknowledges this study and blandly recites a few of its findings, but she buries it under her sweeping, sensationalistic announcement that women need to work till "10:38 a.m." on Tuesday of a second week to catch up to what men make in one week.

She also asserts that "the pay gap is narrower among federal employees than in the work force as a whole." Her theory for why this is the case: "Job descriptions are more standardized in government employment, and salaries are a matter of public record."

But is that what's really going on? Is federal government hiring more enlightened than hiring in America as a whole?

Or is it simply easier to effectively control for variables when you're looking at a narrower slice of the whole job market?

A commenter on her post ("Milton Recht") gives some specific reasons to think so:

[T]he reason that non-government job studies show a greater gender wage difference that the government sector is that the statistics on private sector wages do not include the employer cost of benefits as part of the wage. All government jobs provide about equal benefits.

Studies are published that show women on average will choose a job with better benefits over higher salary and men on average choose jobs with higher salaries over better benefits. The same studies show that men on average will choose a job without health benefits for higher salaries and women on average will not.

When private sector benefits are adjusted to include employer cost of benefits, the private sector difference shrinks to about the same seven percent difference as government wages....

[T]here are probably valid non-observed, non-discriminatory variables for the pay gap. Otherwise, any employer would be foolish not to hire the cheaper labor if it were comparable in every other aspect except gender.

I realize he's referring to "studies" without mentioning or linking to them, so I'm skeptical of his specific claims. But his theory seems more plausible to me than Prof. Folbre's conclusions, even though she links to several empirical studies.

[UPDATE: Since I wrote that, Milton Recht dropped by in the comments of this post and added details: "A January 2009 Department of Labor study (link below) that studies of gender wage discrimination do not include total compensation." He quotes from the study:

Specifically, CONSAD’s model and much of the literature, including the Bureau of Labor Statistics Highlights of Women’s Earnings, focus on wages rather than total compensation. Research indicates that women may value non-wage benefits more than men do, and as a result prefer to take a greater portion of their compensation in the form of health insurance and other fringe benefits.]

More citations, links, and figures don't necessarily add up to a better analysis. For instance, Prof. Folbre's link to support her claim that women overall catch up on Tuesday at 10:38 a.m. doesn't seem to make any attempt to control for confounding factors. The report she links to simply says:

Women who usually worked full time had median earnings of $649 per week, or 78.9 percent of the $823 median for men.

That's essentially meaningless if we don't know a lot more details about what kinds of jobs they had, what their credentials were, and how much they worked. (Perhaps the study secretly controlled for these factors, but if so, it's not apparent from the link.) Of course, Prof. Folbre emphasizes these findings over the GAO reports.

It's nice to link to multiple studies and invoke the principle of controlling for variables. But if you ultimately cherry-pick the gloomiest-sounding figures you can find, and -- whoops! -- by the way, forget about those pesky variables, context, and alternate explanations based on factors other than sexism ... then you've given up any pretense to empirical validity.


UPDATE: Comments over here.