Monday, February 10, 2014

Christina Hoff Sommers: Nope, Women Don't Earn Less Than Men. Part 3 Of 3


This is the third and final post in the series which talks back to Christina Hoff Sommers (of the AEI which should stand for Anti-Feminists Essentially Incorporated but stands for the American Enterprise Institute).  The first two posts can be read here and here.

This last post will be all about engineers.  Honestly.  Well, not just about engineers, but because Hoff Sommers focuses on engineers (as ultimately the reason why the gross gender gap between men and women is to the detriment of women),  I will have to do the same.  Isn't that fun?  We can begin with the engineer vs. lady song (hat tip to Lee Rudolph):

And then you can watch the Volkswagen Superbowl ad about engineers getting their wings:

It's a funny ad, I think, and it seems to have two female engineers, too, except that they don't get their wings.  Yet, anyway.

Funny songs and ads are not evidence of the difficulties women might face in engineering or of the question whether girls just don't like engineering and that's why we have so few women in the field ("choice," again).  But they will make this post a little less dry-and-boring.

The College Majors Data

This is what Hoff Sommers says about the topic of engineers and different life choices by women and men and the gross gender gap in earnings:

Much of the wage gap can be explained away by simply taking account of college majors. Early childhood educators and social workers can expect to earn around $36,000 and $39,000, respectively. By contrast, petroleum engineering and metallurgy degrees promise median earnings of $120,000 and $80,000. Not many aspiring early childhood educators would change course once they learn they can earn more in metallurgy or mining. The sexes, taken as a group, are somewhat different. Women, far more than men, appear to be drawn to jobs in the caring professions; and men are more likely to turn up in people-free zones. In the pursuit of happiness, men and women appear to take different paths.

As I wrote in the second post of this series, "much of the wage gap CANNOT be explained away by simply taking account of college majors."  This is because the majority of American adults do not have a college degree (only 33% have a four-year college degree or higher, and 61.3% have no college degree at all).  But the argument has more problems than that.


To see those, I will reproduce the rankings Hoff Sommers gives us.  Those include the ten most remunerative Bachelor's degree majors and the ten least remunerative Bachelor's degree majors, taken from one study about college majors.  Here are the lists:

Consider, for example, how men and women differ in their college majors. Here is a list (PDF) of the ten most remunerative majors compiled by the Georgetown University Center on Education and the Workforce. Men overwhelmingly outnumber women in all but one of them:

1.   Petroleum Engineering: 87% male
2.   Pharmacy Pharmaceutical Sciences and Administration: 48% male
3.   Mathematics and Computer Science: 67% male
4.   Aerospace Engineering: 88% male
5.   Chemical Engineering: 72% male
6.   Electrical Engineering: 89% male
7.   Naval Architecture and Marine Engineering: 97% male
8.   Mechanical Engineering: 90% male
9.   Metallurgical Engineering: 83% male
10. Mining and Mineral Engineering: 90% male

And here are the 10 least remunerative majors—where women prevail in nine out of ten:

1.  Counseling Psychology: 74% female
2.  Early Childhood Education: 97% female
3.  Theology and Religious Vocations: 34% female
4.  Human Services and Community Organization: 81% female
5.  Social Work: 88% female
6.  Drama and Theater Arts: 60% female
7.   Studio Arts: 66% female
8.   Communication Disorders Sciences and Services: 94% female
9.   Visual and Performing Arts: 77% female
10. Health and Medical Preparatory Programs: 55% female

 Don't those lists look damn persuasive!  I admire Hoff Sommers' ability to make such superficially convincing cases!  I have to go on and on and bore even me.  But she has just summarized the gross gender gap into two neat tables (well, except for arguing that college major choices can explain most of the gross gender gap even though most working adults don't have a college degree at all.)

Would you like to learn more about the problems with those lists?

Here's the first problem:

In one sense those tables are flipped over from what we really needed to know in this context.  They tell us, for example, that petroleum engineering is a well-paid field and that it is 87% male.  But this doesn't mean, obviously, that 87% of all men with a four-year college degree (or more) have that in petroleum engineering.

Indeed, using the same data source Hoff Sommers did*, I find that engineering overall was the major of 8.2% of the respondents in the study used as data here,  that the eight types of engineering listed in the "ten best" list were the college majors of 3.89% of the respondents,  and that the percentage of petroleum engineering majors out of all college majors was just 0.082%.

The point?  That if you look at the number of people possessing each of the twenty college majors those two lists give us,  you find that most of them are not popular majors.  The two exceptions are "social work," a female-dominated field among the low-pay majors and "pharmacy, pharmaceutical science and administration," a gender-integrated field among the high-pay majors.

When I describe these two tables as being flipped over, this is what I mean.  They tell us that certain college majors are predominantly male and others predominantly female, but they don't tell us what the college majors of most students, male or female, might be.  The same source Hoff Sommers uses does give a table on the ten most popular college majors, this**:

Some of the majors listed in that table show wide gender differences, others not.  It's worth noting that the three most popular majors, all business related, are roughly gender-integrated.  At the same time, nursing and elementary education (which is not the same as early childhood education in Hoff Sommers' table of low-paid majors)  are, indeed, very segregated, and several others are somewhat gender-segregated.

I introduce this table because it provides a more realistic picture of what might be going on in the "choice"of a college major.  Not all the "choices" students make exhibit strong gender segregation and that includes the most popular "choices."

The second problem:

What is it that Hoff Sommers wants to tell us with those tables of the best-paying and worst-paying college majors?  If it is to explain the overall gender gap in earnings, well, we have to remember that a) most adults don't have a college degree at all, and, b) that the best-paying college majors may not be terribly common college majors.  These work against her argument that the "choice" of a college major could somehow explain away much of the gross gender gap in wages.

Still, let's interrogate the data a little bit further.  Hoff Summers appears to say that if only women chose majors such as those eight engineering majors in the "ten best" list, they'd be earning every bit as much money as men, and if men chose majors such as those in the "ten worst" list, they'd be earning every bit as little money as women.

Luckily, the data source she and I use here, does provide us income data separately for men and women for a few*** of the eight types of engineering degrees in that list.  Here's what we find (p.114 onward):

The median earnings of male chemical engineers in the data set are $92,000 p.a., the median earnings of female chemical engineers in the data set are $72,000 p.a..  The median earnings of male electrical engineers in the data set are $86,000 p.a., the median earnings of female electrical engineers in the data set are $70,000 p.a..

What about the college majors in the "ten worst" list?  Overall, we observe the same pattern (with one exception****):  Among those who have the same college major, men, on average, earn more than women, on average.  For instance, the median earnings of female social workers in the data set are $38,000 p.a., those of male social workers $48,000 p.a., the median earnings of women with a major in "theology and religious vocations" are $33,000 p.a., the median earnings of men with a major in "theology and religious vocations" are $40,000 p.a..  And so on.

What can we conclude from this?   That Hoff Sommers' argument about college major "choices" explaining "much" of the general gross gender gap in wages is weakened even further:

Most people don't have college degrees, the best-paying and worst-paying college degrees aren't necessarily the most numerous, and there is a gross gender gap between female and male full-time workers who majored in the same college field.

The Third Problem:

It's worth one final visit to the Happy Choice Land of conservative thinking (much more on this in the second post of the series).  Hoff Sommers uses "choice" (possibly in combination with biological determinism) to explain why girlz don't want to do engineering and boyz do:

The sexes, taken as a group, are somewhat different. Women, far more than men, appear to be drawn to jobs in the caring professions; and men are more likely to turn up in people-free zones. In the pursuit of happiness, men and women appear to take different paths.
People-free zones?  The military is free of people?  Well, it might be when a nuclear war is over, but a career in the military forces (a very male-dominated field) is very much about how to cooperate and manage people.  Priests, ministers, rabbis and mullahs are supposed to care about the people they shepherd.  Politics is pretty clearly about people and so on.

Never mind.  We are going to keep one eagle eye on the scarcity of girlz among engineers and another eagle eye on the idea that women just don't want to be engineers:

For the past few decades, untold millions of state and federal dollars have been devoted to recruiting young women into engineering and computer technology. It hasn’t worked. The percent of degrees awarded to women in fields like computer science and engineering has either stagnated or significantly decreased since 2000. (According to Department of Education data, in 2000, women earned 19 percent of engineering BA’s, and 28 percent in computer science; by 2011, only 17 percent of engineering degrees were awarded to females, and the percent of female computer science degrees had dropped to 18.) All evidence suggests that though young women have the talent for engineering and computer science, their interest tends to lie elsewhere. To say that these women remain helplessly in thrall to sexist stereotypes, and manipulated into life choices by forces beyond their control, is divorced from reality—and demeaning to boot.  If a woman wants to be a teacher rather than a miner, or a veterinarian rather than a petroleum engineer, more power to her.

I have trouble with this quote partly because I'm not sure if Hoff Sommers interprets all change in the percentages of college degrees awarded as a sign of "some people just don't want to do that" or if this only applies to women and STEM careers.  In the very first post of this series I mentioned that Hoff Sommers has been an advocate of social activism when it comes to the dropping number of young men who "choose" to go to college.

So which is it?  Is the decreasing percentage of college graduates who are men a problem we need to act on, but the decreasing percentage of computer science graduates who are women is not a problem we need to act on?

I'm also confused, because I'm not sure how this weird conservative combination of "free choice" and "innate biological gender differences" uses a change in gender percentages to denote stability in the case of women and STEM careers. I can see how absolutely unchanging numbers could be interpreted in that light.

Never mind.  The biggest problem I have with the above quote is that it abstracts away from everything except the programs attempting to increase female participation in STEM careers and women's own "choices,"  such as the "bro culture" in coding, the problems of a few isolated women in a still male-dominated occupation, and so on.  It also ignores data from other countries:

The low numbers of women in STEM (science, technology, engineering, and mathematics) fields isn’t unique to the U.S. either. Numbers are similarly low in the United Kingdom, Canada, and Australia.
However, many Asian nations have a better record of producing female engineers: in China, 40 percent of engineers are female, and the number of female engineering graduates in India has doubled in recent years. Over the same period, the number of female engineers in North America has dropped. The U.S. is also behind 13 Muslim countries in the percentage of women graduating with STEM degrees, including Bahrain, Brunei, Lebanon, Qatar, Turkey, and Morocco.
What those international differences in the popularity of engineering among women mean can be debated, but I would argue that culture matters.  In the Anglo-Saxon culture engineering is coded male.  That may not be the case in all other countries.

Finally, because Hoff Sommers abstracts away from both cultural forces and possible gender discrimination in the STEM careers, her quote says nothing about this:

So why aren’t more women choosing engineering education and careers? It’s not due to a lack of ability. Female graduates of engineering programs tend to have grades as good as or better than those of male graduates. While a few decades ago, girls did not typically take the advanced math and physics classes in high school that were a prerequisite for entrance into college engineering programs, this is no longer the case, and the gap between males and females in math classes has disappeared. 
Instead, the reasons seem largely social. Perceptions of engineering as a male domain may be one factor. There is also a misconception that engineering involves tedious or manual labor. Sexism may be a factor as well, in addition to a dearth of female engineering role models and a lack of encouragement from parents, teachers, and school counselors. 
More women enter engineering programs than finish (the same is true of men, though not at quite the same rate). Even women who graduate from engineering schools frequently discontinue working in the profession, often citing overt or covert sexism and patronizing attitudes from male colleagues and bosses.


This is the end of the series.  I'm exhausted and you probably deserve a few chocolate truffles, too.  But what I have written here barely scratches the surface of the economic theories which are relevant in analyzing differences between men's and women's earnings.  That should alert you to be skeptical of very short and punchy posts about this topic or very glib statements about "what everyone knows" concerning the gender gap in earnings.  
* I have not scrutinized the data set Hoff Sommers uses in a more general sense, because I'm using it here to point out her selective choices from a source which has much more.

** This table supplements the one in the body of the post:

 *** The remaining engineering majors in the top-ten list had too small samples for gendered earnings data, and "mathematics and computer science" had too few people altogether.  The median earnings of men with a pharmacy degree are $110,000 p.a. in the data set, the median earnings of women with a pharmacy degree are $100,000 p.a. in the data set.

**** This exception is "visual and performing arts" where the median earnings of women ($40,000 p.a.) are higher than the median earnings of men ($36,000 p.a.)