The Gender Wage Gap

There has been an interesting discussion over at Andrew Sullivan's blog on individual choice and the gender wage gap, so I figured I'd add some context and some further thoughts.

There is no doubt that the gender wage gap has narrowed over the past several decades. Full-time working women now make 80 percent of what their male counterparts make -- a figure that is likely to increase as male-dominated manufacturing jobs become increasingly obsolete.

In spite of this progress, though, there remains an enormous income disparity between the sexes. What could be causing this inequality?

The challenge for social scientists has been determining what portion of the 20 percent wage gap is due to discrimination, and what portion is due to individual life decisions. A simple, side-by-side comparison doesn't tell you much about discrimination. It only tells you that there is a gender disparity, which could be explained by any number of things.

One way to address the discrimination question is to use a statistical technique called regression analysis that holds other variables constant in considering the impact of gender on wages. This technique is sometimes called gender-wage"decomposition." When social scientists use regression to control for other relevant job factors -- the number of hours worked, for example -- the wage gap between men and women narrows dramatically, but does not disappear.

The most reasonable conclusion to draw from this is that many individual-level choices contribute to wage differences between men and women in the aggregate. Discrimination may account for some small portion of that difference, but it certainly not all of it.

Of course, this doesn't quite settle the controversy.

A far more complicated question for policymakers is whether women's choices are really independent, or simply a function of the social pressures and cultural dynamics that unwittingly drive many of our actions. Do women who choose to work fewer hours so that they can spend more time at home with their children make this choice freely, or does society compel them to make it?

There are no easy answers to this question. People are is complex, and "society" is far from monolithic. We can't quantify the millions of subtle forces that influence our decisions and bias our perceptions. We don't know what causes some men to choose work over family and some women to choose family over work. We also don't know what causes some women to choose a work over family and some men to choose family over work.

One thing that we do know, however, is that men will never have to endure the months of physical and emotional misery associated with pregnancy. Men will never have to experience the pain of childbirth, or the stress of waking up night after night to breast-feed a newborn baby. And because of this -- ironically and unfairly -- men will always be paid higher wages.

No matter how much society progresses, the gender wage gap will never vanish completely. There will always be some women who choose to have children, and these women will have to make sacrifices for that child that a father will never have to make. This doesn't mean that men shouldn't take on more responsibility for raising children. But, in the aggregate, full-time working women who become pregnant will miss more workdays, be less productive, and see a marginal decline in wages as a result.

I'm not sure whether this is fair or not, but I simply can't see any fair way to correct the problem. We can give women the option not to have children, and attempt to remove any stigma that may be attached to the "career woman" or the stay-at-home dad. Beyond that, I don't know how to proceed without simply generating further inequity.

Any thoughts?

Does Flextime Work for Workers?

A few days ago, the First Lady came to my job to talk about, among other things, work-life balance. She highlighted the success of flextime programs, which give most federal employees broad discretion in setting their own hours.

Without citing any specific studies, the First Lady exclaimed that we now have the "evidence" that flextime "works." That comment received thunderous applause. But what does it really mean for a program like this to be effective?

Studies of flextime typically find positive effects in terms of individual-level productivity, but to my knowledge there is still a limited body of research examining the impact of flextime programs on overall workplace efficiency.

As a full-time student, I'm pretty happy to work on a flextime schedule. If I have to run errands or finish up some schoolwork, I can always come into work early and leave early. If I'm feeling tired one day, I can come in later without having to endure a passive-aggressive interrogation from my boss. The Department of Labor also allows its employees to work overtime and build up "credit" hours, which will likely be extremely helpful in the weeks prior to final exams. Does all of this increase my productivity? Maybe. But I tend to doubt that flextime programs actually enhance workplace efficiency.

In my experience, there are clear trade-offs. While my job doesn't require a lot of interpersonal interaction, I do work with a "team" and I do sometimes need some guidance from coworkers. Flextime can make this much more difficult. For managers, it is often challenging to administer a group of employees with dramatically different schedules. It's also hard to monitor abuses and direct joint activity. My division still uses sign-in sheets, and it's easy to consistently shave a few hours off of your workday.

More importantly, I think, flextime programs are very difficult to repeal if they are not working. Employees come to see these programs as a sort of fringe benefit, rather than an a way for the company to enhance productivity and promote loyalty.

I'm sure that these criticisms aren't new, and I know that studies of flextime have been ongoing for decades. What I'd like to see is a more comprehensive assessment of flextime programs. I haven't been able to find a meta-analysis of the various studies on flextime and worker productivity, but I'm sure that someone has tried to do a systematic review of the literature. I'd love to read it.

There are many other issues related to flextime. Does it make families stronger? Does it make workers happier? Does it improve employee health? Unfortunately, serious selection problems (and Hawthorne effects) make studies like this less than compelling.

I think it's only fair to say that the jury is still out on flextime. With programs like this, the benefits are often readily apparent, while the costs are less visible. There are still a number of concerns that have yet to be considered.

Any thoughts?