Yesterday was Equal Pay Day and my Facebook reposting of my Learn Liberty video on the gender wage gap generated a whole bunch of discussion both on the thread and privately. The assertion that the gap is mostly not the product of discrimination always generates much heat. I thought I might take an opportunity here to say a few things about this issue and how it gets talked about among libertarians and their interlocutors.
One point of confusion is not understanding that the issue, for economists, is about discrimination in labor markets and wages. When we say that wage differentials between men and women are mostly not due to discrimination, what we are saying is the following: men and women with identical labor market characteristics and preferences about jobs will get paid nearly the same. Our concern is whether wages are related to those variables rather than the employer’s like or dislike of one group or another, regardless of their job characteristics etc.. Put differently, economists tentatively conjecture that discrimination exists if any part of the pay differential between men and women cannot be explained by human capital and job preferences.
Notice that this says nothing about whether there is sexism elsewhere in the economy. All that is says is that we only presume that employers are discriminating when they don’t pay people the same who have otherwise the same job market characteristics (aside from gender).
Those who argue, as I do, that most of the gender wage gap is not due to discrimination are often accused of just making a “libertarian” point, as if we think the market magically makes discrimination disappear. In truth, the claim that most of the gap is not due to discrimination is widely accepted in the economics literature in general, libertarians or not. The general consensus is that once the other economic factors are accounted for, only about 5 cents on a dollar remains unexplained. In other words, of the gap of 23 cents (77%), about 75% is explained by those factors. This is not libertarian priors. This is the consensus of the literature. Even a recent studey by the American Association of University Women (hardly libertarians) of 2008 college grads one year out found that there was a raw pay gap of 18 cents, but that only one-third of that was not explained by the usual factors. In other words, the gap conjectured to be due to discrimination was 6 cents. Arguing that discrimination explains only a fraction of the gender pay gap is a well-established result in economics.
If I had it to do over again, I would have softened the language in that video, which is phrased too much as “it’s not discrimination.” The more accurate statement is that it’s “mostly” not discrimination. Of course even the finding of an unexplained portion of the gap is subject to refutation and much of the literature on this topic involves economists trying to make sure they’ve accounted for all the human capital factors to see if they can explain more of the gap. The debate continues.
As I noted earlier, even if there were no unexplained part of the gap, and that economists found no discrimination in labor markets, it doesn’t mean there isn’t sexism elsewhere. For example, even if markets paid everyone exactly according to their human capital and job preferences, that human capital and those preferences could well be the result of sexism elsewhere in society. Suppose we had a law that prohibited women from attending college. It’s at least possible that gender wage gap studies would find no discrimination because women’s lower pay would be the result of their lower human capital. All that would show is that employers don’t discriminate, but it would hardly be cause for celebrating the end of sexism. The point people like me are making about the gender wage gap is that most of it is not the fault of sexist employers and that markets generally do pay people in accord with their skills and preferences. That’s all we're saying.
For libertarians, that seems to be an important result, especially if those arguing the other side think they can improve matters with public policy. Trying to eliminate the pay gap through wage setting like comparable worth or anti-discrimination laws is not going to help and may well backfire to the extent such policies raise the cost of hiring women. Instead, libertarians might want to join their feminist friends in working on the cultural side of the issue. Some of the best things we could do to equalize men and women’s pay involve making sure that we aren’t discouraging girls from entering majors and acquiring skills that lead to better paying jobs. Telling little girls that math isn’t for them, for example, contributes to pay differentials by affecting human capital. Similarly, we could do more to encourage boys to enter the supposed "feminine" occupations. We, and by “we” here I mean men, could do more to shoulder the responsibility of household production. To the extent that it remains mostly women who interrupt their careers to care for children, or who are responsible for more of the “second shift,” it will remain the case that their wages are lower. We can also use social pressure to encourage firms to offer more flexible employment arrangements that make this possible. All of these are ways to work outside the state to reduce the role of sexism in determining wages even if we think labor markets are not a source of discrimination.
Finally, as libertarians we might want to challenge some of our feminist friends with the following consideration: if patriarchy is real and men have disproportionate power over all of society’s major institutions, why should a feminist trust the government to be the solution to problems like the gender wage gap? Even without assuming patriarchy, given the track record of government in the 20th century and before, why should we believe it will not only care about women’s interests but be able to effectively pursue them? To simultaneously complain about how Congress is controlled by men and still think that the federal government is the solution to men’s oppression of women seems… problematic. And if our feminist friends agree that the state is not the solution, then it would seem we’re all on the same side.
So arguing that labor market discrimination plays a very small role in explaining the gender pay gap does not mean that sexism has nothing to do with it. Sexism matters, but far more in the ways human capital is acquired and preferences are formed than in the wage decisions made by employers.