Monday, February 02, 2009

The Illusion of the Pay Gap

The first homework of the semester informs my students that one of the reasons men get paid more than women is because women can get pregnant. Because contracts promising to not get pregnant are illegal, hiring women is riskier for employers. An employee could take several months leave with relatively little warning. What a coincidence that President Obama's first bill addresses the same issue, but ignores the explanation.

Insisting that employers pay women less for malevolent reasons, Mr. Obama made it easier to make pay discrimination suits. Notes CNN,
companies will need to meticulously document pay decisions and retain detailed employment records, legal experts say. In this, small companies may be at a disadvantage - few have access to the attorneys and human-resources professionals that will help larger businesses comply with the newly expanded law.
A strange thing to do in rough economic times.

Risk of pregnancy isn't the only reasons for pay gap. Women tend to enter less technical jobs. Part of that is a natural tendency; for whatever reason, women are less likely to enter the sciences and, save economics, more likely to enter the social sciences. Go to any college math class and I bet you'll see mostly men (no going to an all-women school; that's cheating).

The other reason for the draw to the less technical is, you guessed it, pregnancy. Technical jobs have a high degree of obsolescence, meaning you have to stay on top of the latest developments to keep pace. But when you leave the job market for months at a time to care for a newborn, this is really hard to do. Hence women become secretaries, not engineers.

Biology's a sneaky thing and even seeps in after the children are born. Women also tend to work part-time, since women are still the gender that tends to take of kids. Being restricted to part-time adds additional restrictions to where they can work; again, these places that accept part-time employment tend to be paid less. Fringe benefits are usually off the table as well.

The nice thing about this story is it's easily testable. Simply look at groups where the biology issue isn't relevant and control for education. Thomas Sowell from Economic Facts and Fallacies:
Among college-educated, never-married individuals with no children who worked full-time and were from 40 to 64 years old--that is, beyond the child-bearing years--men averaged $40,000 a year in income, while women averaged $47,000. (p70)
Women are paid 17.5% more than men! Maybe we'll see that new legislation end the tyranny of male bosses upon his fellow man.


Anonymous said...

I can't find a problem with your logic but it feels wrong. And offensive. Why?

Anonymous said...

The "logic" of using pregnancy has a few BIIIIG flaws.

1) Since when do businesses let something like "it's illegal" stop them from trying to maximize their profits? Women who can't get pregnant are treated the same as if they could be, ie being paid less. The fact they don't come out with the reports crunching the numbers justifying the "she could get pregnant" excuse tells me it has more to do with old fashion sexism.

2) There's no corresponding penalty for men. A Time magazine poll reports that men are more than twice as likely to take a sick day when they're not sick. Then you also have to include how men are more likely to pad expense reports, embezzle funds, steal office supplies, commit workplace violence, etc. What kind of salary hit should men take for that drain on resources?

3) When a woman gets pregnant they can give more than 6 months advance warning to their employer. That's hardly "short notice". Many companies also allow sabbaticals. How's that different from maternity leave?

4) Being in a technical position myself, I can tell you being out of the loop for a month or two won't kill your career. Technical jobs do not have as high degree of obsolesce as you imply. I was still getting calls from recruiters to work in programming languages I haven't touched for years. There's still a demand for PHP, Cobol, and Fortran programmers for crying out loud! In my experience companies are slow to adopt new technologies. They want someone else to do it first and work out all the bugs (ie Windows Vista). Your skills can be stagnant for 3 years and you can STILL find a job. Plus keeping up on things isn't hard. A mother on maternity leave can easily maintain her skills without much effort.

5) The claim on job loss is misleading. People leave jobs for many reasons. Singling out one as an excuse to reduce the pay of half the population is disingenuous.

6) Let's run the numbers. Over a 40 year working life a normal woman has less than 3 children. Including doctor visits into maternity leave let's give them a total of 3 months off. So we have 9 months out of 40 years to perpetuate the species. 9 months/40 years = 0.01875. That's less than 2%. How is that a justified reason for a pay gap?

7) Oh and Sowell's example. I have found he tends to leave out factors and facts that would disagree with his conclusions. As such, I don't find him to be a trustworthy source of information.


David said...

Jason, let's take these one at a time.

(1) This is an interesting point; women who are infertile have (I assume) no difference in pay to women who are not infertile. But that's a pretty weak argument because a company knowing who can and can't get pregnant is a matter of doctor-patient confidentiality. Also assuming that laws mean nothing to companies is even more absurd. No one's saying there's 100% legal compliance but that's a far cry from suggesting laws which are regularly enforced don't influence behavior. (And given how easy it is to sue people in this country, you can bet this one is well-enforced.)

(2) Maybe there is a corresponding penalty for men but even with it there's still a gap. I have no data on all the ways men are more likely to be unproductive or women more likely to be unproductive, but we have no counter factual, either. You can't say it's not included unless you can compare it to a time when all other things are equal but that difference doesn't exist and the gap is the same. And since there's an incentive to include it, the burden of proof rests on the theory that Time magazine noticed it, but for some reason, no one else did.

(3) You and I have no idea what "short notice" means, as it differs from company to position to industry to circumstance. And at the time of hire and pay offers, a maternity leave will be under conditions that, at the time, will be largely unknown. Also not your use of the word "can." Just because they could give ample notice doesn't mean they will, and the employer knows this.

(4) There are many technical positions, Jason. Some have higher levels of obsolescence than others. And a field can be benign most of time with occasional sudden advances and still dissuade some women from joining it. Moreover, women are more likely to want to stay at home and raise the child (which makes sense because they gave birth to the infant) so we have to include not just a chance of maternity leave but a chance of a much longer leave (or quitting altogether).

(5) Yes people leave them for many reasons, but those reasons occur with relatively equal reason for both genders. (I assume, but I can't imagine why a man, would, for example, be more likely to quit to take care of his suddenly sick parents.) Pregnancy, however, is going to occur only in women (with one famous exception) and women pretty overwhelming stay to take care of the kids.

(6) This is naive math. It ignores the chance to switch to part time work or not working all together. It assumes uniform chance across 40 years (i.e. makes no adjustment for child-bearing vs non-child-bearing years; or between married and unmarried women). Moreover, I argued pregnancy is only part of the explanation (another big one is the tendency to be interested in very different subjects, and yes, I'm sure there is some sexism involved but not nearly as explanatory as people claim).

(7) Saying Sowell can't be trusted is not only a non-argument but a logical fallacy. For the record, that particular factoid is from Warren Farrell's Why Men Earn More: The Startling Truth Behind the Pay Gap -- and What Women Can Do About It (though admittedly since I don't have the book, I don't know where Farrell got it).

At the heart of the "discrimination as the key reason for the gap" argument is a fundamental paradox. If, as you noted at the beginning of your comment, businesses' key goal is to maximize profit, and women are being paid less for no good reason, why don't business hire only women to save money? Why doesn't an entrepreneur start a company to challenge the established industries? A 25% reduction in salary (which is a huge part of the budget of any company) is enormous. If women are being paid (far) less than they are worth, then hiring only women is a smart business decision. But we don't see that and firms are greedier than they are sexist. And sexism is very expensive.

Anonymous said...

1) It still doesn't explain the lack of a clear report. There are still too many unknowns to justify the widespread differences in pay.

2) Have you tried LOOKING for data? I found my example doing a quick google search. You should do the same. Also Time wasn't the only one who noticed it. I simply chose to list one example.

3) If you can't say that 6 months is short notice, then why did you in your initial post? What research have you done to show that it's NOT enough time in the majority of careers? In my experience in the working world, I have not encountered any career where 6 months would be "short notice". And even if you were to find some (and list them here), how does that justify making a statement about all jobs, including the jobs I know that 6 months would be plenty notice. Consider this, how many jobs give you a 6 month notice that you're being fired?

4) # of positions Jason listed where you can be out of circulation for a few months and get back in the game: 1. # of positions David listed where you can be out of circulation for a few months and get back in the game: 0. Hmm very interesting.

5) I couldn't help but notice the lack of data siting how many women quit their jobs to rear children and the lack of data showing how many men quit their jobs during a mid-life crisis.

6) You missed the point. What I was trying to explain that over a career, a small percentage of that time will be missed due to pregnancy-related reasons. That does not justify penalizing a woman for all her jobs. If there was a way to easily tell visually who does and who does not like alcohol, would a company be justified in giving a lower salary to the people who do because there's a chance they'll get in a drunk driving accident?

7) If you don't have the book and don't know how Farrell arrived at his conclusions, why do you think he can be trusted? Isn't that same non-argument and logical fallacy you accuse me of?

8) You really missed the point didn't you. This isn't about comparing the pay of John in career A with Jane in career B. It's about a difference in pay for people in the same type of company with the same qualifications overall and often within the same company itself. An all-women company wouldn't have anything to compare it to. More importantly, such a company couldn't exist due to anti-discriminatory laws. That leaves you with a strawman's argument.