Tuesday, January 10, 2012

Are STEM Degrees Already Subsidized More?

Alex Tabarrok argues in his ebook, Launching the Innovation Renaissance, that graduates with STEM (science, technology, engineering, math) degrees are more likely to create innovations that benefit the rest of society – and therefore they are relatively more deserving of educational subsidies than students in other disciplines. Here’s how Alex puts it:
Most importantly, graduates in the arts, psychology and journalism are less likely to create the kinds of innovations that drive economic growth. Economic growth is not a magic totem to which all else must bow, but it is one of the main reasons we subsidize higher education.

The potential wage gains for college graduates go to the graduates — that’s reason enough for students to pursue a college education. We add subsidies to the mix, however, because we believe that education has positive spillover benefits that flow to society. One of the biggest of these benefits is the increase in innovation that highly educated workers theoretically bring to the economy.

As a result, an argument can be made for subsidizing students in fields with potentially large spillovers, such as microbiology, chemical engineering, nuclear physics and computer science. There is little justification for subsidizing sociology, dance and English majors.
I think Alex is right; if we’re going to subsidize education, we should subsidize education that generates external benefits for society at large.

But I’m wondering if, in fact, we might already subsidize STEM degrees more than other degrees. Consider the following three factors that make STEM courses more costly to teach:

1. STEM professors are typically paid higher salaries. See, for example, this report from the Chronicle of Higher Education. The last table shows salaries by discipline, as a percentage of the average salary of English professors. Across all disciplines, the average salary is 13.4% higher than an English professor’s. But Engineering professors earn 25.2% more, Computer & Information Sciences 28.4% more. Mathematics is below average at 7.2%, but overall, STEM professors appear to get paid a good bit more than the average. Meanwhile, Fine Arts, Education, Communications, Philosophy, and Psychology are all below the average. (This makes sense, because STEM professors probably have better outside job opportunities and thus a higher opportunity cost.)

2. It’s easier to teach non-STEM courses in large lecture halls, whereas STEM courses often require smaller class sizes to be taught effectively. (I don’t know this with certainty, but I’ve been told as much by university administrators.)

3. When STEM courses are taught in large lecture halls, they require a larger number of teaching assistants to give the students the attention they need. (Again, I don’t know this with certainty, but it’s what I’ve been told.)

Putting 1-3 together, it seems pretty likely that STEM education is more costly to produce. And yet colleges and universities typically charge all students the same tuition regardless of major. True, STEM students may be charged nominal lab fees, but I doubt such fees make a large difference in percentage terms.

So when we consider how much students are charged relative to cost, it looks like STEM students might be getting the larger subsidy. Of course, I don’t know how high the optimal subsidy would be, so it’s possible the current subsidy isn’t large enough.

It’s also worth noting that professors in Law and Business Administration earn the highest pay differentials of all (59.5% and 50.9% above the average English professor), which would imply that these fields are getting among the highest relative subsidies. Econ professors also earn a high differential of 41.2%. Again, this is presumably driven by outside job opportunities. Unless we believe Law, Business Admin, and Econ generate strong positive externalities, maybe we should be charging students more to major in those fields.

Is there some factor I’m missing that would diminish the relative subsidy to STEM (or at least STE) degrees?


Brad said...

Regarding point #1, I agree with you that STEM professors largely earn more because the opportunity cost of taking a teaching job is much higher. I'd also suggest that STEM professors (especially at larger institutions; I went to Purdue) are recruited as much for their research promise as for their teaching ability. There is status associated with the research side of the house, and attracting the talent there is not easy.

Point #2 is largely correct. Although I think class sizes get smaller as class level increases, the entry-level engineering classes I took were undoubtedly taught in smaller lecture halls than the entry-level gen-ed classes.

For Point #3, I'm not sure that more TA's are needed simply due to lecture hall size, but I would definitely say that sometimes TA's are necessary because the professors are researchers first and teachers second. And where I went to school, the professors often wrote the textbook being used (another signalling/status symbol in professorship), which meant that they thought they were the most competent authors in the world and shouldn't need to explain anything in the book. So often the TA's had to reteach what the professor should have taught during class. And this was particularly hard, because a very large proportion of TA's were not native English speakers -- something I'm sure is not the case in the humanities.

But I would point out an exception to both points #2 and #3. I minored in Philosophy* to go along with my Elec. Engineering degree, and those classes were typically quite small -- especially the 300-level and above. Some of the other gen-eds, like Psychology, included large lecture components but also featured class breakouts of 20-30 students for discussion. And my 100-level English gen-ed class was also very small. I think that while 100-level Psych/Soc/etc classes might be taught in large lecture halls, the same was true of 100-level physics, Bio/Chem, and math courses. Then, when the higher-level "discussion-based" humanities classes went to MUCH smaller class sizes, many of my 200-400 level engineering classes were still presented in 80-100 people lecture halls.

All this should be taken as a single anecdote, of course... I went to Purdue, which although it is a very large public state university, is VERY heavily biased to STEM and a few other fields (Agriculture, Nursing, etc). The School of Liberal Arts is pretty much an afterthought. So it's possible that my experience is not representative of a school with a good STEM school but large enough to focus elsewhere as well (such as a school like University of Michigan or similar).

Ran said...

My experience (at Case Western Reserve University in Cleveland, a STEM-heavy school) does not accord with #2; quite a few of our low-level STEM classes were taught in large lecture halls. Granted, usually there would also be a lab and/or a "recitation" (a weekly smaller-group meeting led by a TA), but my understanding is that large non-STEM classes would also have such outside sections.

(Disclaimer: aside from gym, the only low-level non-STEM classes I took were micro- and macroeconomics, so most of my knowledge of that is secondhand. It's possible that I just never realized how big the non-STEM lectures were!)

Will said...

The per professor cost isn't as important as the total cost of professors. I suspect the value of salary times department size times number of departments is much higher for non-STEM.

Bob said...
This comment has been removed by the author.
Bob said...

I suspect that the high salaries of the law and business professors were heavily weighted by grad school professors...and tuition for both law and graduate business programs are typically much more expensive than other majors. They tend to be cash cows for colleges and universities (ever hear a commercial for a English grad program ?), certainly not subsidized.

To further add to the other anecdotal evidence, my alma mater, Tulane, dropped engineering (essentially) as it restructured following Katrina. Their rationale: engineering programs are expensive.

Unknown said...

While we're doing anecdotes:

Where I go (University of Chicago) my sense is that the STM (we don't really have engineering, though they're in the process of adding some) tend to be the bigger classes - at least to a certain degree. That's due in part, I would think, to the fact that succeeding in STEM often has more to do with understanding material than critically and individually approaching it. In the fields that require the latter (and it's certainly not all of the non-STEM fields) smaller class sizes are necessary. That's not, of course, to say that they're always achieved - partially because of the subsidy issue.

(By the way, the above analysis shouldn't be taken as critical of STEM. I'm not trying to say that STEM doesn't require real thinking. In fact, I think my higher-level work in mathematics has been more intellectually challenging and deep in many ways than higher-level work in philosophy. They're just very different ways of approaching material.)

But anyways, I've found that even accelerated math courses tend to be taught in classes of about 30 people -- and this at a school that has many humanities seminars with 10 students, a professor, and a TA.

Before I cut this off I'd just like to nitpick something: in the first comment, Brad seems to be contrasting "STEM" with "gen-ed" as the two bits of higher education. That doesn't seem like a fair or useful approach. STEM students may approach the other disciplines as "gen-ed" -- the same way that humanities students approach their STEM classes -- but the fields are just as specialized.

Brad said...


It wasn't intended to disparage, although those of us in STEM do look at those classes as general stuff that we take to fulfill the requirements of a degree instead of something that will actually be meaningful to our careers. For us, the classes are mostly "general stuff we don't care about but have to get out of the way." I would assume that when a typical Liberal Arts student has to take a physical sciences course for graduation requirements, they really don't give a crap whether it's Biology, Chem, Physics, etc. They just want to get it out of the way.

That doesn't mean they're general classes -- as a philosophy minor*, obviously I took a liking to one aspect of the humanities, and I know that Philosophy is as different from Psychology as Number Theory is from Circuit Design.

* As an aside, one of my reasons for being a philosophy minor [in addition to enjoying philosophy] was that I thought a minor in one of the disciplines in Liberal Arts would be a great place to meet girls. If you've taken any higher-level philosophy classes in your life, I assume you're sitting there laughing your ass of at me right now.

Unknown said...

That's fair. I definitely understand that from a STEM perspective those fields are 'gen-ed', I just didn't think it was good terminology to use when taking a more general perspective.

As for the last bit...I'll just not comment. :P

Unknown said...

By the way: I'd challenge the basic premise of the post. Whatever your view of the overall benefits to society by STEM versus the humanities (and that's an argument I won't get into) the relevant question under this view of subsidies is how well those benefits can be internalized. I'd argue that in the US, even assuming there are more social benefits to STEM advancement, the system of at times excessive copyright/trademark/etc protection does a(n overly) good job of internalizing those benefits.

The social benefits of strength in the humanities, however (more informed/critical citizenry in a democracy, increased social awareness, cultural sophistication, etc) are much more difficult to internalize.

Glen Whitman said...

Will: "The per professor cost isn't as important as the total cost of professors. I suspect the value of salary times department size times number of departments is much higher for non-STEM."

I don't think that's the right measure. The total salary paid by a department is a function of how many professors are teaching in that field, which is driven primarily by how many students are taking courses in that field. It reflects neither the average cost nor the marginal cost of educating students in that field. (And multiplying by the number of non-STEM departments is even stranger; it would be high simply because there are more non-STEM than STEM fields.)

Caveat B said...

I graduated with a STEM degree. #1 may be partly a demand/scarcity issue. As a Manhattan parent in a top school, I have noticed less than half the parents are still able to help their 6th graders with their math homework (almost SAT level), and these parents all have advanced degrees. #2 is not necessarily true. I took all my calculus, physics, computer science, and even grad-level discrete model simulation in large-group situations. #3 may be the most heavily weight-able. My alma mater (Cornell Engineering) seemed to have lots of TA candidates, a mix of PhD, Masters, and even seniors. In hindsight, I think that was a feature, not a bug.

Michael E. Marotta said...

Last week, Virginia Postrel had this in Bloomberg.
"How Art History Majors Power the U.S. Economy."

Moreover, we know that subsidies distort markets. Public support (and further public underwriting) of STEM actually produces hidden losses, by Bastiat's Window. While libertarian and Objectivist icon T. J. Rodgers earned his masters and doctorate (at private schools) by inventing new technology, Steve Jobs and Bill Gates famously dropped out. There is no touchstone for success. That is why markets (people) must be left free of political interference.

Unknown said...

"Moreover, we know that subsidies distort markets."

That's tautologically true, in that a subsidy is a distortion of the market that would have occurred absent the subsidy. But contra baseline libertarian assumptions (and yes, I know many libertarians are more informed about economics than this) the hypothetical market that would exist absent the subsidy is not necessarily more efficient. That is, "free" markets != efficient markets. Subsidies, if done well (to help internalize benefits, as talked about in the OP) can reduce market inefficiencies, "distorting" the market back in the correct direction.

Philip Whitman said...

First of all, I agree that if we’re going to subsidize education, we should subsidize education that generates benefits for society at large.

Second, I note that Science, Technology (incl. Computer & Information Services, etc.), Engineering, and Math are all mutually overlapping sets. I am distrustful of study results that rank salaries by arbitrary groupings, as the boundaries for such groupings must necessarily be subjectively determined.

I am not surprised that STEM professors are typically paid higher salaries than non-STEM professors. However, I think the Math professors’ below-average score could possibly be reflective of the mutually-overlapping-sets definition problem.

I think the importance of knowing (and teaching) English (and linguistics) is grossly undervalued. Without communication, there is nothing.

Finally, I have no objection at all to receiving any subsidies that any reader might feel compelled to send me.

Glen Whitman said...

Dad, I'm not sure I see the problem. It's true that people in STEM majors will often take courses in multiple STEM areas (e.g., Engineering majors will take Math classes). But I'm not sure how that creates a problem for the salary statistics for professors. Extremely few professors hold positions in multiple departments, and extremely few teach courses outside of their departments. And while there is sometimes a degree of arbitrariness (for instance, I know an economist who switched from the Econ department to the Finance department), for the most part departments hire people with degrees in the given field. If the statistics show that professors in some fields get paid more than professors in others, we can debate about the reasons, but I don't see any particular reason to doubt the data itself.

Fearsome Pirate said...

I don't see why STEM degrees should be subsidized at all. Yes, they provide external benefits. This manifests itself in the form of higher salaries for STEM majors. High salaries already provide a sufficient incentive for students to attempt to get STEM degrees. At your typical state school, around half of the STEM enrollees will fail out by the end of their freshman year (either calc, chemistry, or physics gets 'em).

All the subsidies do is give the schools more cash to slosh around and drive up the price of tuition.

Bob said...

"Fearsome Tycoon" is right, except that that is true of all students. STEM may seem under-represented (though who's do say what the "right" numbers are, other than the market), because college eduction in general is so heavily subsidized in general. There would be fewer non-STEM students if in a free market, and there may or may not be more STEM's.

Unknown said...

Fearsome Tycoon is wrong on the economics. The whole point of externalized benefits is that they *can't* be internalized in the form of higher salaries. Absent subsidies or concerted effort, markets can't reward externalized benefits.

Anonymous said...

"That is, "free" markets != efficient markets. Subsidies, if done well (to help internalize benefits, as talked about in the OP) can reduce market inefficiencies, "distorting" the market back in the correct direction."

This is nice in theory, but the fact remains that "'distorting' the market back in the correct direction" assumes that the ones doing the distorting knows what the correct direction is. And that's assuming a bit too much!

For what it's worth, I'm a mathematician, who decided not to become a professor; I am currently working in computer programming, but I have an intense desire to do something more mathematical.

As for distorted markets, I'm still suffering greatly from the distortion that occurred in my life, where student loans were made a little too easy for me to get. Yes, the distortion is partly my fault--but Gov't hasn't exactly been an innocent party in this distortion itself!