Tuesday, September 03, 2013


The Theory Is Too Damn Thin: A Response to Dylan Matthews

Yesterday, following Jennifer Silva, I asked how people define adulthood.  For me, I recall the feeling a couple of times when it occurred to me that I was the grownup, and that I had to step up.  Nobody around me was willing or capable, so I had to put aside my misgivings and do it.

I had that feeling again reading Dylan Matthews’ series of articles under the banner “the tuition is too damn high.”  Matthews writes with the confidence of someone with a Big Idea, and he has some nifty anecdotes for texture.  But the grasp on how colleges actually work is so thin, and so otherworldly, that I feel the adult’s obligation to step up.  Left uncorrected, articles as poorly conceived and widely shared as these could do real damage.

The latest (part VI) states baldly that colleges and universities have no incentive to control their spending, so they don’t.  And it marshals a strange collection of half-truths, category errors, and flat-out whoppers to make an implied case for price controls.

I’ll start with a whopper.  Matthews claims that

there are a number of features of the higher education marketplace that make the effect possible. One is that higher ed is dominated by nonprofits. If revenue increases, it must be spent, or else refunded. It cannot simply be pocketed by the college or university. That means that it’s likelier for increases in revenue to lead to increases in spending than it is for normal companies.

Cannot be pocketed?  Um, they’re called “endowments.”  Or “reserves.”  Colleges and universities are judged, in part, by the size of their endowments or reserves.  Everybody who works in higher ed knows that.

I’m amazed that the Post published that basic a mistake.  To be fair, though, the piece had enough errors in it that an exhausted editor may have just given up in despair.  Here’s another:

The quality of college isn’t completely evident during freshman orientation. People overwhelmingly only get one undergraduate degree, so colleges and universities don’t have to worry as much about satisfying repeat customers. And there are more than two parties to the transaction. There’s a student and a college, sure, but there is also the student’s parents, and the government, and even the student’s future employers. Those all end up mattering for the financing of the product in question. It’s a really, really messed up market.

In the enrollment-driven sector, a retained student is a repeat customer.  We worry about retention all the time.  The idea that a student makes a decision in his senior year of high school and the die is cast from that point on simply does not describe the reality of most non-elite colleges.  

But never mind that.  Let’s deal with a more basic point.  Matthews observes, correctly, that the parties affected by higher education aren’t limited to the student and the school.  (He specifically names parents, the government, and future employers, though the list could go on.)  He concludes that therefore “[i]t’s a really, really messed up market.”

Or a really, really messed up category mistake.  Higher education is a public good.  The error comes in treating it as a private transaction, because as Matthews correctly notes, the benefits to higher education spill over well beyond the immediate parties involved.  If the market isn’t structured to capture those “positive externalities,” it will underinvest.  That’s the idea behind having the government offer financial aid, and it’s the same idea behind subsidizing public colleges.   The error isn’t in imperfect market information.  The error is in treating it as a market in the first place.

Perhaps that ship has sailed, politically.  And it’s certainly fair to say that if public institutions want to maintain their claim on the public purse, they should show that they’re actually serving the public.  So, how should that be done?

First, by recognizing that generalizing from flagship research universities to all of public higher education is ridiculous.  Matthews’ piece starts with a portrait of a new sports center at Purdue.  It sounds pretty opulent, and if Matthews’ portrayal is correct, there may be an argument to be had there.  Go for it.  But quick quiz: what public higher ed institution in Indiana has higher enrollment than Purdue?  Higher, even, than every campus of Indiana University combined?

Ivy Tech, which is the statewide community college system.  But you won’t find that, or any other community college, in Matthews’ article.  

This sector doesn’t compete on prestige.  The “Bowen effect” is absent here.  You won’t find many climbing walls, football teams, lazy rivers, or other symbols of conspicuous consumption here.  Instead, you’ll find faculty and staff laboring under budget cuts stemming, in part, from a political movement to which pieces like Matthews’ give aid and comfort.

Matthews can’t seem to decide what his unit of analysis is.  Is it private research universities, research universities in general, or higher education in general?  His evidence is drawn exclusively from research universities, yet he refers repeatedly to “colleges and universities,” as if he had examined any colleges.  It’s a common mistake among policy people who don’t really know higher ed, but it’s a fundamental error, and it matters.  The economics of a teaching college are not simply those of a university but smaller.  They’re structurally different.

His cavalier dismissal of the Baumol effect is even worse.  There, he notes that total expenses at universities -- again, he ignores colleges -- have risen faster than instructional spending.  From that he concludes that the Baumol effect is irrelevant.

That.  Is.  Not.  How.  Budgets.  Work.

The Baumol effect -- basically, the impossibility in improving “dollars per hour” when you denote your product in units of time, as we mostly do with teaching -- is gradual.  You fall behind the rest of the economy by a couple of percentage points per year, basically forever (or at least as long as the productivity of the rest of the economy is improving).  Over the past, let’s say, forty years, why is it that instructional budgets haven’t risen as quickly as a one-to-one application of Baumol would suggest?

Adjuncts.  Colleges have adapted to the increased cost of full-time faculty by hiring proportionately fewer of them.  They’ve compensated somewhat for the Baumol effect by replacing full-timers with part-timers.  

In other words, college budgets are not static, or black boxes, simply being buffeted by outside forces that we need pundits to explain.  They adapt.  They shift to compensate.  

That’s not to say that every shift is good, or that every decision is right.  I would personally love to see significant changes to public funding of higher education, starting with per-student parity between community colleges and universities.  If research is so lucrative, let it pay for itself; there’s no discernible principled reason that students at inclusive schools should get a fraction of the support that students at exclusive schools get.  Let’s start with that.  Then let’s talk about accountability.

I’m glad to see higher education get serious public attention, but I’m consistently mystified that the voices that get the most attention are people who literally -- and I mean this without malice -- do not know what they are talking about.  I’m frustrated by airline pricing, but I don’t do extended pieces on it because I don’t know the industry well enough.  If I were asked to do an extended series on it, I might try, I don’t know, asking some of the people who actually make the pricing decisions why they do what they do.  I wouldn’t stop there -- first-person testimony is flawed in any number of ways -- but I certainly wouldn’t skip the step.  It’s pretty basic.

I’ll make Dylan Matthews -- and all the major news outlets -- a deal.  If you want some insight into how people who actually make these decisions make them, call.  I’ll take the call, and talk as long as you want.  But for heaven’s sake, don’t postulate from some Hayekian theology and call it reportage.  We matter too much for that, even if you wouldn’t know it from our budgets.

Thanks for being the "grown up"! I've been really annoyed reading the Matthews pieces and many of the equally ill-informed comments. Glad to hear a reasonable and informed person take Matthews to task!
Nicely argued, particularly the point about the size of Ivy Tech and the rest of higher ed in Indiana put together. And I was really glad to see you pick up the point about unequal funding of freshman and sophomore classes by states when sometimes they are being taught by the same adjuncts!

Two thoughts:

1) I'd REALLY like to see some longitudinal data about adjuncts, counted on the basis of sections taught or students taught, not simply numbers. Do you have that for your college, going back a few decades? It hasn't changed much at my CC for the last 15 years, but I don't know what it was like in the 80s or earlier.

It can be tricky. Back in a previous century I was the instructor of record for my sections of a large lecture class (i.e. I signed and submitted the grade sheet) because that is how it was managed, and grad students easily outnumbered faculty by numbers of sections. Yet more than half of a student's time was in a lecture led by a full prof that looked on paper like part of my section. But that was at a university, and they could have complicated matters by changing how they treat those same sections today. CCs should be easier because I don't know of many that follow that model.

2) Category errors cut both ways. That actually is how budgets work at a big R1, where the staff-to-student ratio is positively scary and is all driven by research and indirect cost recovery rather than teaching. But Dylan doesn't know this, so he makes the huge mistake of asserting that those staff have something to do with teaching and correctly draws your ire.

Universities are where you find the Prize Winner Prof who teaches a handful of grad students while Instructors generate tuition dollars, not colleges. And although I agree that CCs cut close to the bone, we are not without our flaws. I've seen some really irrational choices made over the years between core functions and what look to be money losing efforts.
I disagree that research is lucrative and should pay for itself. I don't think that is a consensus argument. Research needs support from the public because it provides a public good knowledge that can be used in some contexts for profit, but in many contexts is merely foundation for future research or applications in unprofitable enterprises.
The problem as I see it is that only the most advanced students have the skills to be at the cutting edge of knowledge (at least as I see it in science). In the olden days that edge was at a less advanced level.

So I think what's going to happen is that Higher Ed is going to morph into two parts - Lower Higher Ed and Upper Higher Ed.

Lower Higher Ed teaching is going to turn into something more like High School (with the pay that goes with it) because most of the knowledge at that level can come out of text books rather than from subject matter specialists.

Upper Higher Ed will turn into something more like the Higher Ed of old - where subject matter specialists lecture and do small group teaching e.g. in this case to PhD and post-Docs. I suspect a post-Doc will go from a paid position to a pay-to-learn position.

Has Anonymous@12:19AM ever written a research proposal and noticed the addition of about 50% to the bottom line that you never see? (The rate is 56.5% at the University of California at Berkeley.) IMO, that is where you see the Bowen effect because everything the university adds that can be construed as supporting research is reimbursed in this way.

Meghan@3:05 makes an interesting point, but fails to explain what institution will bridge between the new HS-like LowerEd and grad-school-like HigherEd. Junior and Senior year classes in your major?

Maybe it would help if "olden days" was defined. Freshman physics content has not changed in more than 50 years. We still don't teach Hamiltonian and Lagrangian mechanics to wannabe engineers or biologists or physicists, and we still teach those topics along with the quantum versions to upper undergrads.
he Baumol effect -- basically, the impossibility in improving “dollars per hour” when you denote your product in units of time, as we mostly do with teaching -- is gradual. You fall behind the rest of the economy by a couple of percentage points per year, basically forever (or at least as long as the productivity of the rest of the economy is improving). Over the past, let’s say, forty years, why is it that instructional budgets haven’t risen as quickly as a one-to-one application of Baumol would suggest?
As you note adjuncts answers the question but wouldn't the bigger answer be an overwhelming number of graduate students?

In terms of the sciences, there have been gains in productivity/graduate student. In my field what used to earn someone a PhD in 5 years can be done by an undergraduate working part in lab in one semester. What my PhD covered almost a decade ago was significantly more than what my equivalent did a decade before that. Those earning an equivalent PhD now, are doing more than I did. These gains in productivity allow students to learn more than they would have otherwise.

For major research university, science faculty devote most of their teaching to training graduate students. Per graduate student they are teaching more and they are able to supervise more students as well.

At the undergraduate level because of the advances in technology, students can do more in lab courses in a set amount of time and can be more productive in undergraduate research. In other words there are gains in learning productivity. It isn't the same as you find with the graduate students, so there is institutional variability.

As your post demonstrates why thinking about institution type matters but what also needs to be considered is discipline.
It's very Informative Thanks for sharing with us Famous Play Schools in Hyderabad
Not that it matters, but education is a private good with large positive externalities. I'd education really were a public good, then MOOCs would have put you out of business years ago. Education is a private good often heavily subsidized by the state because of well known externalities,
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?