Wednesday, March 10, 2010

Some research I have read on the female-male wage gap

One of the striking empirical facts about the wages of women is that women have consistently made less than men. Even in Biblical times, the rate of pay for women was around 60 percent of men's wages.

One might be tempted to brand this difference in compensation as unfair, as there is no apparent difference between the ability of women and men to perform tasks in a modern economy. Clearly, American politics and women's rights should have made some progress. So, how much progress was made between Biblical times and 1980s in the United States? It turns out: not much. According to a paper in the 1988 Journal of Economic Perspectives by Smith and Ward:
Further, virtually all U.S. government sources indicated that women's wages had been fixed at roughly 60 percent of those of men throughout the post-World War II period.
By many accounts this difference in the compensation of women relative to men has diminished since the 1980s, and it did so in a predictable way (Smith and Ward's paper is an exercise in forecasting what would happen to this gap; they correctly foresaw a converging of wages). By the 1980s, women had already begun acquiring much more useful labor market skills: better training in high-paying occupations, longer tenures and better work experience.

These investments in the labor market have big payoffs, and it was foreseeable that women's labor market prospects would improve from the 1980s. Indeed, they have by most standard measures. Broadly speaking, women no longer make 60 percent of what men make, but 75 to 85 percent, depending on how that gap is measured (within occupation, within education group, within age group, etc.)

In fact, a big debate in labor economics is not whether the compensation-gap has closed significantly since the 1980s, but questions about what this means.

One side of the debate -- espoused in a Quarterly Journal of Economics paper by Casey Mulligan and Yona Rubinstein -- argues that women have had precisely the same opportunities to succeed in the labor market over this time of perceived improvement. Mulligan and Rubeinstein attribute the perceived growth in women's wages relative to men as a difference in the composition of the types of women who enter the labor force.

Here's their argument in three pieces.

First, Mulligan and Rubinstein observe that there are two types of women who do not work. (A) Women whose labor market skill set is poor, but they have better opportunities to work as the head of the household. (B) Women whose labor market skill is excellent, but are married to someone who has even better labor market skill. [There are other types, but this stark contrast is useful to see their point]

Second, M&R note that since the 1980s, the variance in the earnings distribution has skyrocketed for men and women. Low skill workers make a little more than they did in the 1980s, but high skill workers make significantly more. This has increased dramatically the returns to having excellent labor market skill.

Third, M&R observe that more women from group (B) work now than before. The composition of the types of women in the labor force has transformed dramatically toward the higher skilled women who didn't find it optimal to work before. Now, they do and this change has altered the female-male wage gap.

This is an interesting and important paper. Mulligan and Rubinstein tell a story that allows for the possibility that the market changed, but the women didn't. Their paper is interesting because it points out a channel through which women's opportunities relative to men didn't change, but their choices did.

On the other hand, it is not clear that this is the only direction of causation. It is plausible that women changed their human capital investment or family structure choices in response to this dramatic market change. It would be surprising if they didn't. That said, it's useful to have a clear and compelling theory for how the world works, and caveats aside, this paper is both clear and compelling.

*As I mentioned earlier in the post, this is a multiple-sided debate. I have some reading to do to understand the points made by the other side, but once I do, I'll probably include an update in a future blog post.

1 comment:

  1. It is impossible to effectively discuss this issue without focusing on a specific jobs within the labor market. For example, if we assume investment bankers earn significantly more than your average accountant or waiters, they would have the potential to distort an average made up of the three professions.

    What % of applicants for each segment are women, what percentage of employees are women etc. There clearly cases where women of a similar experience level have a lower pay grade than men, but there are also situations where there are less women applicants and their application is actively encouraged to "balance" the work environment.

    Then again, any discussion wouldn't be complete without discussing how we are all different... yes, I know in these days of blinkered "equality" it's sometimes difficult to have a discussion about this, but men still can't have babies and the two genders have been programmed differently... on average men who've traditionally been hunter gatherers "on average" have different wiring to female in home-making role. None of this should be seen as a limit on individual potential, to the man who wants to be a manny or the women who wants to be a hunter... but it just proves that averages are a poor means of comparison.


Please feel free to share your ideas about this post in the open forum. Be mindful that comments in this blog are moderated. Please keep your comments respectful and on point.