Welcome to the Profit of Education website. Continuing the conversation begun in the book Profit of Education, we discuss the latest economic evidence on education reform.

Majors, Race, Gender, and Money

Everyone knows that members of disadvantaged groups—which I will oversimplify to say simply Blacks, Hispanics, and women—do far less well in the market than do White males. Are the differences explained by what people choose to study? Do we see big wage differences in students who chose a particular major in college versus those who make a different choice?

The answer is, unsurprisingly, “yes;” wage gaps definitely vary by major. There are some noticeable patterns in that wage gaps tend to be larger in technical fields than in people-oriented fields. Over generalizing is a mistake though. In some tech fields wage gaps are quite large, while in other tech fields wage gaps are barely there. What’s more, the situations faced by members of different disadvantaged groups are not the same. When it comes to examining the role played by different college majors, race is not the same as ethnicity is not the same as gender.

Let’s look at education and economics first.

For people who majored in economics, the gender wage gap is about average. The race and ethnicity gaps are much larger than average though. Education majors fare somewhat better, the gender wage gap is less than average and the Black-White gap is clearly below average. But it’s not that the gaps have disappeared, just that they’re smaller. You’ll see below that to there is something of a pattern where people-oriented fields tend to have smaller gaps than do technically-oriented fields.

While different majors are associated with different wage gaps, it is not true that choice of major somehow “explains away” the gaps we see. One explanation offered for earnings discrepancies is that more men than women choose to study technical fields and that this explains higher earnings. In other words, once one controls for area of study, the wage gap is largely explained. This turns out not to be a very good argument. Wage gaps are quite strong within most areas of study. For example, it isn’t that men study (high-paying) petroleum engineering and that women study (not so high-paying) English. In fact, the gender wage gap exists within both petroleum engineering and English—and it is a lot higher in petroleum engineering (44 versus 31 percent).

In our analysis, we’ve drawn the most recent five years data from the American Community Survey, looking at people with a college degree between ages 22 and 65, who report working and report positive income.[1] The average wage gap comparing Black to White is 19 percent; comparing Hispanic to White is 18 percent; and women to men is 43 percent. When you adjust for which majors students choose, the gender gap narrows a little as can be seen in the next Figure; nothing happens to the race and ethnicity gaps.

 

In thinking about the role played by choice of college major, care is needed because we don’t know if the wage gaps are due to something done by people running a particular college major, something in the after-college job market, or something about how students choose what to study. In economics jargon, we don’t know if the role played by majors is “causal.” But knowing which majors have particularly large or small gaps is a first step toward thinking about improving the situation. The second step—you will see why below—is to remember that the challenges faced by different disadvantaged group are not all the same.

The existence of a wage gap is nearly universal across majors. In which majors do women earn more than men? None! While Black graduates earn more than White graduates in 10 of our 174 majors, in no case is the difference statistically significant. Similarly, reported earnings are higher for Hispanics than non-Hispanics in four majors, but again none are statistically significant. Given this, there isn’t much point in asking if there is a wage gap following studying a particular major—since there almost always is. Instead, let’s ask which majors have wage gaps much worse than average or better than average. In the pictures that follow, I show the majors with the largest and smallest gaps (but only among those where the gap was significantly different from the average gap). For reference, each graph also has a horizontal line showing the average wage gap; smaller than average gaps are shown in red and larger than average gaps in blue.

 

While the really big Black/White gaps tend to be in engineering and science areas, that clearly oversimplifies some. One of the largest gaps is in astronomy and astrophysics, while in atmospheric sciences and meteorology and in oceanography Black graduates do well.

The pattern across majors for the Hispanic/White wage gap is broadly similar to what was shown in the previous picture for the Black/White gap. Do notice though, that the gaps for Hispanics are generally less extreme in both the up and down direction than for Black graduates. Thus, the often-needed reminder, that “minorities” is too simple a classification. The challenges faced by diverse groups are, well, diverse.

 

It is striking how different that the details in this gender-gap figure are from what we saw above in the race/ethnicity figures. It’s more than just that the average gender gap is larger. The first difference is that even the smallest wage gaps are pretty large. The second difference is that the technical-vs-people explanation of where the gaps are large doesn’t seem to hold. Aerospace engineering has a (relatively) small gender gap, while the gap for Black graduates was large. Family and consumer sciences has a large gender gap.

We do not know what part of the school-to-market process is responsible for the gaps we see—we do see considerable variability across majors. At the least though, the people responsible for those majors where the gaps are especially bad should be asking themselves why their majors stand-out in this embarrassing way.

The author is grateful to UCSB undergraduate and Gretler Fellow Victor Huang for research assistance.


[1] As is common, we define Black as “non-Hispanic Black” and White as “non-Hispanic White.” Our estimates control for age, which probably doesn’t make much difference.

Share
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *