Enough people have asked me a particular sticky question that I decided I’d take a crack at it.
“How big is the gender pay gap?” (That is, the gap between what women and men are paid in the United States.)
After much research, I have found a decisive answer: “It’s as big as you want it to be.”
I know that’s not a very satisfying answer, but unfortunately I don’t think I can provide a decisive one--and anyone trying to be too decisive about a certain number is likely trying to take you for a ride and push an agenda.
How Can We Cut It?
There are a number of ways we can cut this. The most basic way is this:
How much money does the average man make per week, versus the average woman, in the US as a full-time worker?
According to the BLS in 2012, the average woman made about 81% what the average man makes. So by that measure, it’s 81 cents to the dollar. The Institute for Women’s Policy Research says the gap at 2015 is 79 cents to the dollar.
What about per hour?
Women tend to work fewer hours than men per week, so the gap becomes 88 cents per dollar per hour.
What about for the same jobs?
It may not surprise that many jobs are male-dominated and many jobs are female-dominated. This actually gets hard to analyze precisely, as no single body has a database for what every person gets paid, and their gender. But one can estimate it.
For women and men working the same jobs, the average difference is about 8%, or women getting paid something like 92 cents on the dollar.
What if you control for education and time on job?
In one study by the American Enterprise Institute, controlling for the same job (which is largely affected by women taking time away when they become mothers), the same time in job, and the same education level and relevancy, then the gap shrinks to somewhere between 95 and 99 cents on the dollar.
Does That Mean There’s No Gender Pay Gap?
This is the part where you get to decide what does and does not count in the analysis. If you’re simply asking, “do employers in the US discriminate their pay based on gender for exactly the same work?” Then the answer is, “maybe a little, maybe not.” I’ve heard people put forth the hypothesis that women might not be conditioned not to bargain as hard, and that may explain some of the remaining 0-5%.
A lot of the gap seems to come down to the jobs that women have, how much they work per week, how much time in that job they have, and their education level.
Men, for example, are much more likely to work 50+ hours/week, and the per-hour rate of those jobs is much higher than full-time. White collar men also are more likely to be CEOs and executives, and blue collar men are more likely to be in dangerous or physically taxing jobs that require higher pay to incentivize the work (think “Deadliest Catch”).
You might say, “so there’s no gap.” You may also hypothesize that women go into lower-paying jobs due to workplace discrimination against them even entering the higher-paying jobs, or you might hypothesize that they choose lower-paying jobs for cultural or socialization reasons, or a combination of both. They may work fewer hours out of choice, or there may be active workplace discrimination about this. So controlling for these factors does not necessarily mean that there isn’t a problem, either.
So depending on what factors you want to “bake in” to your analysis, or which ones you want to use as a statistical control, you’re going to get a very different number.
In all, it’s hard to conclude from statistics alone what the cause is of a specific discrepancy. Knowing that women tend to take lower-paying jobs doesn’t imply that the discrepancy is due completely, some, or not at all from a form of workplace discrimination or bad cultural teaching--we simply don’t know based on the statistical analysis.
What You Control For Changes What Problem You’re Trying to Solve… and Whether It’s a Problem
Gender discrimination in the workplace is a very different problem from cultural influences that young boys and girls may face as they grow up, that may cause them to take on different fields of study or value family vs. work differently.
Something like Equal Pay for Equal Work would only solve for any discrepancy (possibly 0-5 cents) within a single job with the same work experience and education.
If you’ve decided it’s your goal to have women and men paid equally on average throughout the country, then you’d make more progress by understanding what influences or discriminatory factors are playing a role in educational selection, job selection, and hours/time worked.
You may also have to decide whether it is a problem that more women choose to be in certain jobs with lower pay but fewer hours, less travel, more security, more safety, or whether that’s fine. You may have to decide whether your goal is to have a fully 50/50 equal split of women and men in different fields like STEM, the humanities, art, medicine, etc, in an effort to get women and men to be paid the same.
Which of these factors is a social or market system you want to change, and which is a matter of personal choice, is something you’ll need to figure out without the statistics.