Tuesday, December 20, 2011

You're Doing It Wrong. On The Reverse Gender Gap.

A new round of the reverse gender gap arguments! What fun! And just in time for the holidays.

This time it's the Old Gray Lady doing the opinionating:
As the year ends, much of the talk around women — at least in the United States — has moved from empowerment and global gender gaps to the trend of young single women out-earning men and the rise of female breadwinners.
There are so many views and theories out there, some of them driven by independent research and others by personal experience and still others by a chatty blend of both, that we are getting a sometimes confounding, always provocative and occasionally contradictory picture.
For starters, young women today — and not just in the United States — are moving quickly to close the pay gap, or in some cases have closed it already.
They are marrying later and later, or not marrying at all. They no longer need husbands to have children, or want no children (40 percent of births in the United States each year are now to single women).
Women are ahead of men in education (last year, 55 percent of U.S. college graduates were female). And a study shows that in most U.S. cities, single, childless women under 30 are making an average of 8 percent more money than their male counterparts, with Atlanta and Miami in the lead at 20 percent.
Although that study of 2,000 communities was done only in the United States, it points to a global trend.
The emergence of this cohort of high-earning young women and the increasing number of female breadwinners are transforming gender relationships, upending patterns of matchmaking, marriage and motherhood, creating a new conflict between the sexes, redefining the word “breadwinner” and inspiring tracts on the leveling of men’s roles.
It is being called the reverse gender gap.

And then the article quickly goes to the "Oh my god, if women do the work and women do the child rearing, what's left for men to do?" stuff. Which tends to assume that a) women have never done any other work but child rearing in the past and that b) men have never parented at all. Until now. And parenting is yucky and demeaning for men.

But that's not what I want to write about. I want to write about the concepts the "gender gap" and the "reverse gender gap." Because the way that article was set up at the very beginning contains an error of interpretation.

It is this: The proper way to compare the earnings of men and women, to find out if any gender gap (or reverse gender gap) exists, is by trying to compare like with like. This means that good studies hold constant the education level of the individuals, their years of experience, their age, ethnicity, race, marital status and number of small children, the geographic area in which they work (because economic conditions may differ) and so on.

The aim is to compare individuals who are the same in all other relevant characteristics than the one a researcher is looking at. In this case it would be purely gender and nothing else.

So given that background, what are those studies of young men and young women in urban centers failing to do properly? The most important factor is that They. Do. Not. Control. For. Education.

To give you an example of why that matters greatly: Suppose that the average young woman in some imaginary urban center (The Big Banana) has a college degree, and suppose that the average young man in that same imaginary center has a high school diploma. Suppose, finally, that we are able to control for all other differences between that average man and that average woman, except for their gender and their education levels.

If we then find out that the average young woman earns, say, 20% more than the average young man, have we established what the author of the NYT piece calls "a reverse gender gap?"

NO. And the reason is that we don't know how much more a young woman with a college degree earns in the Big Banana than a young woman with just a high school diploma. We also don't know how much more a young man with a college degree earns in the Big Banana than a young man with just a high school diploma. Perhaps that whole 20% is because a college degree pays 20% more than a high school degree?

In that case there would be no reverse gender gap in the properly standardized sense. What those studies would have established is nothing deeper than the fact that education pays. And, of course, that more young women (in urban centers) have college degrees than is the case with young men.

Bear with me because this is important. That the gender gap, not standardized for education, benefits young women does not have to mean that there is no residual gender gap in earnings, the kind of gap which benefits men.

Let's go back the Big Banana and let's properly standardize for everything but gender, including education. Let's THEN compare an average young man with an average young woman, both with, say, college degrees or both with, say, high school diplomas.

What will we learn about the gender gap or the reverse gender gap in this situation?

The studies do not tell us. But here is what I would predict, based on all the other studies I have read in this field: The women with college degrees are likely to earn less than the men with college degrees and the women with high school diplomas are likely to earn less than the men with high school diplomas.

This is what I mean by the gender gap in wages: An actual gap not explained by the other characteristics of the workers, only by gender, and this standardized gender gap is to the detriment of women. You can read more about it in the series found on my website.

The wider problem with the approach the NYT article takes is this:

Looking at what very new workers do gives us a fairly poor prediction of what will happen later. Check out this quote about the study of 2,000 communities:
Here's the slightly deflating caveat: this reverse gender gap, as it's known, applies only to unmarried, childless women under 30 who live in cities. The rest of working women — even those of the same age, but who are married or don't live in a major metropolitan area — are still on the less scenic side of the wage divide.
Bolds are mine.

Now that I read that bit again, I get flabbergasted by the bias in these stories. So we have the majority of women earning less than men, even in that study, and what we discuss is the "reverse gender gap?" Why don't those other women (the majority of women) matter at all?

Is the assumption that the educated young women in urban centers are the harbingers of the future petticoat government?

But men and women don't stay young and single. They get older. Most of them get married and have children. People move from initial positions without much scope to wage differentiation (or discrimination), some are promoted, some are not, some get raises, some do not, some take time off for family reasons, some do not.

And as the above study shows, the old-fashioned raw gender gap emerges then. I see no reason why that would not happen to the current young workers in the future.

There's a strong The-Sky-Is-Falling flavor to these stories about the end of men or the reverse gender gap or the horror of educated women ending up as spinsters with just a cat for company. I have gotten so used to it that I actually had to look at the real numbers to see how biased the debate has become. Take this:
For starters, young women today — and not just in the United States — are moving quickly to close the pay gap, or in some cases have closed it already.
Where is the evidence on that? Sure, the raw wage gap has been very slowly diminishing in several countries but I don't know of any drastic and quick recent reductions in it.

I think the linked comment is intended to provoke that sky-is-falling fear: The upside-down world where women are the rulers and men the ruled. And that's how this piece of news can provoke fear:
Women are ahead of men in education (last year, 55 percent of U.S. college graduates were female).
Perhaps the fear would be less if we were reminded of the fact that 58% of college students in Saudi Arabia are women? And that the percentage of female students in Iran exceeded that of male students until the government there decided to limit women with quotas?

Here's the general point of the Sky-Is-Falling aspect of the end of men and the supposedly coming petticoat regime: It misinforms the readers. It obfuscates rather than clarifies. And it bases its arguments on that hidden idea of the world as a seesaw: They are going to do to you the same thing that was done to them! Be very afraid! Even if you are female, you should be very afraid because your achievements mean that you will never find a partner.

That means the only alternatives are the jockstrap regime or the petticoat regime. So choose carefully.

Which is of course utter crap.
For more on the NYT piece, check out here and here. I especially liked this bit in the latter:
Maybe the dissonance between the mostly grim headlines about American women’s progress and the mild hand-wringing of the women-on-top school is more than just which indicators you think matter (or which statistics you cherry-pick). It’s also about the supposition that what women have managed to gain in the last 40 years adds up to a wild feminist hegemony, which now requires a sober-eyed reassessment. Meanwhile, it’s less fun to point out the many things that are still lagging — but it’s even more frustrating that it’s still necessary.