THE WAGE GAP MYTH, PART 1: 
You are probably wrong, but that (in of itself) does not make me a dick

Statistics are tricky. Many people feel that if you can't draw a knee-jerk conclusion from the data, then statistics -- in general -- are no good (e.g. hard alcohol consumption goes up with teacher salaries, so clearly if we give teachers more money they'll just spend it on booze). Or if they've ever seen statistics misapplied (e.g. IQ is correlated to race, and therefore President Trump), they assume that all data can be manipulated and don't trust anyone (which may be exactly the conclusion certain creationist, anti-science, vote-against-your-interests, and believe-both-evils-are-equal politicians want you to draw). The truth is, statistics are only as good as the statisticians wielding them, numeracy in this country is pretty low, and it's the lack of analysis (aka effort) that lets people misrepresent stats, or hold onto convincing (but wrong) data-attributed beliefs. I'm going to make the women of the world like me even less by explaining away the gender-based wage gap (in the USA), one of those things that everyone "knows" to be true, but is not.


There are two issues here: one is that the stat you commonly hear -- which is that women make $.77-.80 on the dollar -- is not comparing males and females doing the same work. Secondly, when you do compare those jobs for people with comparable years of experience, location, education level, etc., males and females generally aren't doing those jobs in the same way.

So how do we explain (away) the gap? I will be the first to admit that even with a math degree, the multi-multivariate separation of cause-and-effect is beyond me, even if I had that kind of time and energy (this is why we have division of labor in the first place). But the kind and number of factors are so deep and wide that I have no trouble buying the mathematicians' conclusions. In fact, I've read several studies explaining the gap to within a few cents of even -- including some of the ones cited below -- but each of those studies used *different* factors to explain the gap; in other words, adjusting for everything, it's quite possible the gap goes the other way (if, for example, business is reluctant to pay a class of individuals their effective "worth" because it appears discriminatory, or you add back in the value of time off, paid or otherwise, or equal insurance rates for unequal insurance costs).

Here's one such conclusion, from a decidedly pro-women's group: The American Association of University Women (AAUW) has now joined ranks with serious economists who find that when you control for relevant differences between men and women (occupations, college majors, length of time in workplace) the wage gap narrows to the point of vanishing. The 23-cent gap is simply the average difference between the earnings of men and women employed “full time.” What is important is the “adjusted” wage gap-the figure that controls for all the relevant variables.

Here are just some of the many causes of the perceived gap (quotes come from references at the end of this article, or can easily be found with a quick Googling):

1. Women on average (and more so historically) enter the workforce later, with fewer job skills, and have less education among older workers.

While the education gap has completely reversed itself, with more women college graduates than men, that's not the composition of the current work force, and that also does not hold for advanced degrees or STEM degrees (where a lot of the big money jobs are):
"...those who always see themselves as going to be mothers (and yes, there’s still a substantial portion of the population who do indeed regard life as being one of those things best run by the traditional gender time allocations, whatever you or I might think about that) tend to invest less in education and careers when they know that they’re intending to drop out of that rat race."

Many of those women entering the workforce later in life are coming off careers as homemakers, which limit the types of jobs they are eligible for, and their expected pay.

2. Women choose different types of jobs than men. Those jobs have lower pay, and, significantly, lower rates of pay increase.

"Jobs themselves seem to have a gender component. Jobs that involve helping and caring are typically female (though many men work in them) and lower paid, while men’s jobs (with many women working in them) involve controlling and managing and are higher paid."

In the older salary surveys, a whole bunch of white-collar jobs get lumped together (including, say, a broad category containing "pro athlete," where pay is commensurate with the dollars those performances bring in), skewing those "buckets" so they were actually apples-to-adam's-apples comparisons. Newer, better analyses (again, doing the work matters in statistics) shows these studies to be flawed, so please don't quote them when commenting (which you are encouraged to do).

3. Women leave the workforce, often permanently, to have kids.

This is going to be the most controversial one, since many women feel they should be able to do just that without affecting their career options. However, companies aren't just hatin' babies (in fact, single folks help subsidize the insurance costs of full families in companies that provide insurance benefits, males now subsidize females, there are several benefits available only to the childfull, etc. ... so one could argue -- but I won't -- that non-breeders that can't host an embryo are the discriminated class here). It's not just that some businesses've evaluated the cost of maternity in the face of current laws, in terms of expected time off (both for leave and future child-related time costs), in the cost of holding open a position that may get vacated, etc. and discounted the average value of a women's contribution to the bottom line, but that time off takes away from anyone's future earnings, and dropping out of the workforce before peak earning years is obviously going to affect the mean income of all females severely.

One studies estimate wage costs at about 5.7% less for each child women have. The gender wage gap for unmarried people with no kids? About $300 out of $47K. [I consider that in "rounding error" territory.]

Even if women just took maternity leave, and were willing and able to raise a baby without impacting their job, in many high-paying careers, *any* time off is a massive income hit:
"Specifically the penalties for time out of the office — regardless of gender — are high among those with MBAs and JDs."

The timing of this departure, when permanent, is a huge driver in the wage gap. Women's pay growth stops *outstripping* men's (not just matching, but surpassing) in their 30's, which is when, on average, college-educated women start having children.

Interesting aside: There is clearly a significant pay gap between women who have children and those who don't. However, there's a wage *bonus* for fathers over single men, comparable in size, so *families* don't end up earning much less even when women do. I don't know what accounts for that bonus, but I suspect it's a division of labor + having responsibilities thing.

There's always going to be the question of how a *family* maximizes its earnings and minimizing daycare expenses, and for a family with at least one parent in a high-wage, high-demand job, that’s generally having one full-time-plus worker and one part-time-or-less caregiver, except in those cases where both parents are both such highly paid workers that they just farm out child rearing:
The “motherhood penalty” — the relative decline in wages for women when they have children — has disappeared or even reversed for highly paid, highly educated women.
This would also apply in the increasingly popular but still relatively rare cases of the stay at home dad, which brings us to ...

4. Women have different life preferences, somewhat (but not entirely) based on expectations of life roles:

I'll touch on this in depth next time, but the bottom line is women *are* nurturers (whether by nature or nurture it's hard to tell, but you can see a feedback cycle here if you are a women and were raised by a nurturing woman, and that is your dominant role model, because your daddy is at work all the time). On the balance, women also have different life priorities (or nurturing responsibilities) than males, which limit their commitment to certain professions (I am NOT saying this is true of ALL women, only that's it's measurably more true for women than men, and will probably continue to be so until I too can lactate out my nipples).

Being a nurturer limits your available time in total: 
Women are more likely to spend time away from the workforce and are more likely to work truncated schedules as they try to balance both professional and personal priorities, such as caring for children or parents.

In the latest stats I can find 19% of men and 7% of women worked more than 55 hours a week. And women not working those long hours is spreading the unadjusted wage gap the "wrong way" (which implies that there are still advances being made in the absolute gap, ignoring job types, in spite of the appearance of stagnation):
"In the past, the tendency of men to work longer hours would not actually have contributed much to the wage gap, because the payoff for doing so was negligible. In 1979, workers who chalked up more hours actually earned less per hour than those who worked full time. The average man working a typical full-time job, 35 to 49 hours a week, now earns about $26 an hour. But the man working 50 hours a week or more now earns close to $33 an hour."
"Men make up a bit more than half the full-time workforce, but they account for more than 70 percent of those working 50 hours a week or more. So as wage gains have gone disproportionately to people working long hours, they have also gone disproportionately to men."
Note that is more hours AND more per-hour, which is an N-squared contributor, PLUS a bonus for promotions going to the most "dedicated." Also interesting: women who work long hours have seen even faster gains than men.

Women are also less likely to give up their life and sanity to claw their way to the top (we address why next week). One study saw this attrition highest in the most grueling (and highest paid) professions: "the penalty for M.B.A.s is higher than in any other profession she’s looked at. A high percentage leave the highest paying jobs after just a few years."

This issue also affects the simple comparison of same-job-different-sex in a manner not factored in previously (i.e. under point #2, wherein the gap was still a few cents after accounting for ONLY "identical" jobs, experience, title, etc.):
"For instance, a woman with young children may have a job with a title that reflects her seniority but opts to work at a company that pays somewhat less in exchange for more flexibility. "

5. The economic supply-and-demand argument (or, "it must be so because it makes sense superficially, not because I have irrefutable evidence" ... this tack doesn't work for me, but if you're the kind of person who falls asleep during these discussions and just want the Karl-Rove-style "there's this simple argument that appeals to many plain-ol'-folk and makes any convoluted truth sound faggy*, elitist, and wrong" then you can have this nugget, which completely defeats the point of this post):

If women were truly cheaper, and did the same work as males in every way, a business trying to increase its bottom line would hire more of them, and you'd see a lower unemployment rate for females (ceteris paribus, which they always ain't). Although the female unemployment rate was much lower after the last crash -- possibly due to everyone losing earnings potential and men taking a longer time to get over their entitlement, and women convinced they're going to get paid less, thus taking the lower paying jobs first -- that's evened up again, which is business as usual, and I'm a bit surprised how consistently about-equal these rates are. There's pretty much no other axis and metric along which you can slice up the work force and come up with roughly equal numbers on both sides of the fence.


* With apologies to Idiocracy, as well as my gay friends

Some references (and generally good reading if you're a 538-type-of-person):

http://www.theatlantic.com/business/archive/2016/02/whats-really-behind-the-gender-wage-gap/462363
http://www.thefiscaltimes.com/2015/11/05/5-Charts-Explain-Gender-Pay-Gap
http://www.payscale.com/gender-lifetime-earnings-gap
https://www.bls.gov/web/empsit/cpseea10.htm