Reflections on Goff and the Cost Disease

Brian Goff’s open email on Baumol’s “cost disease” turned out to be one of Bet On It’s all-time most popular guest posts. And it’s no accident; Goff crafts a thoughtful exploration of a plausible story. Still, I think he’s largely incorrect, for reasons I shall now explain. Brian Goff’s in blockquotes; I (Bryan Caplan) am not.

Disclaimer: This is not an April Fool’s essay! As Mirabeau said of Robespierre, “He will go far, for he believes all he says.” (Quoted in Richard Pipes, The Russian Revolution)

---

On medical care the view that prices and expenditures are incredibly high and wasteful overlooks two important points: i) that prices and expenditures need to be adjusted for quality; and ii) income levels and growth accounts for almost all expenditure differences across countries and over the long run.  

I’m puzzled. The idea that the quality of health care has improved is already almost universally accepted. Economists who argue that healthcare expenditures are wasteful are self-consciously challenging this standard view, highlighting surprisingly strong evidence that marginal medicine fails to improve health. See Robin Hanson’s “Cut Medicine in Half” for details.

Point (ii) is similarly non-responsive. Income can statistically explain the growth in expenditure even if 50% of expenditure has always been pure waste, no?

On higher ed, the underlying problem is that economists (myself included) almost always treat it as single product industry (education) with everything else on campus treated as a type of input. For better or worse, higher ed has developed into a multi-product bundle that includes entertainment/recreation (workout facilities, rec teams, collegiate athletics, drama, …), networking, counseling, and just hanging out with peers (and probably other stuff). A useful analogy would be to a year-round version of a (very upscale) summer camp.

You’re not entirely wrong, but the key question is: How many people/families would pay for these “(very upscale) summer camp” add-ons with zero taxpayer subsidies? Remember, moreover, that plenty of professional jobs also offer a camp-like package of activities for workers. So the marginal entertainment/recreation gain is a lot smaller than it looks.

As economists, we all are aware of the problem of quality adjustment in price indexes. For the most part, however, the profession just proceeds like it doesn’t matter.  

I strongly agree as a general matter. But I see CPI Bias throughout the economy, not just in services. In fact, I think the quality of non-services has demonstrably improved more. So the appeal to rising service quality doesn’t really help.

I’ve come to view the assumption that quality of services hasn’t changed that underlies most applications of Baumol’s Cost Disease idea as incorrect. Most services have improved in quality over the past several decades. This is true even for the “go to” example of haircuts. The 15-minute clip & buzz in the barber chair that I grew up with has been replaced by 30-minute wash/condition/rinse/massage and stylish cut, all in a nicer atmosphere. The closest thing to the old barber shop experience would be something like SuperCuts, which is quite cheap. Why did old fashioned barber shops nearly vanish? Because consumers valued the new amenities.

SuperCuts still has more than 2,400 locations nationwide. So cheap haircuts have hardly “vanished.” Cheap healthcare and education, in contrast, have practically vanished, though taxpayer subsidies sometimes preserve the illusion of cheapness.

Nowhere is that lack of adjustment for quality of services more consequential than with health care. This isn’t merely anecdotal. Murphy and Topel’s 2006 JPE article (in what may be the most underappreciated article in the last 50 years) estimates that from 1970 to 2000, the gain in life-years from medical advances added $3.2 trillion PER YEAR to national wealth. While that figure may off, the order of magnitude of their estimate is intuitively in the ballpark, when considering the number of life years gained and the value of a life-year at or above $1 million. Their estimate does not factor in any value from improved quality of life from things like joint replacements, asthma treatments, and so on. Cutler & McClellan (also very under-appreciated) generated related results in a series of articles.  Their 1998 QJE piece indicates that after adjusting for quality, the price of heart treatments actually fell.   

Robin Hanson powerfully argues that all these estimates are question-begging. Despite the veneer of technical sophistication, they casually smuggle in the assumption that better health care is the prime or sole cause of better health, despite strong experimental evidence to the contrary.

The other important data observation that steered me away from my reliance on regulation and supply restrictions as the primary explanation for medical prices and expenditures in the U.S. is that these expenditures per capita are almost completely explained by one of our two most basic variables in economics – income (more precisely disposable income or the OECD’s Actual Individual Consumption). The Random Critical Analysis blog has extensive and compelling material on this.  The upshot is that health care expenditure is, over the long run, a highly income elastic good and that various differences in “systems” are not needed to explain the high U.S. expenditures, income does. The big expenditures in health care tend to come from areas such as heart, cancer, and surgical or intensive care treatments that are (and will be) expensive under any system, but these are treatments that an affluent society wants to make to keep “mom” and “dad” alive and well for 25 more years, “grandma” for 5 more years, and the “pre-me” baby alive 85 more years.

An “affluent society” or “affluent consumers”? Government-funded health care routinely pays massive sums for trivial health gains, but that’s a prime example of waste. You can object that private insurance is even more profligate, but as I’ve explained before, when government supplies a product for free, private alternatives must provide even higher quality to win customers. If government subsidies ended, we would almost certainly discover that most consumers are not willing to pay hundreds of thousands of dollars to extend their lives by a few weeks. Nor would they pay market rates for insurance that covered such treatments.

On the growth of higher ed expenditures, for years, I bought into the lagging productivity view as forwarded by Richard Vedder and others. In fact, William Poole, while still president of the St. Louis Fed, delivered a speech on this topic to a small luncheon group at WKU with our university president sitting right there. It was fairly humorous seeing our president’s reaction. However, I realized that in assessing productivity, Vedder, Poole, and others (like myself) routinely treated education as the sole product/service… For all my grousing about the lack of student interest and “why are they even here,” the puzzle existed only because of my underlying, single-product assumption.

No, the puzzle is still very real. If the summer camp is the main reason they’re on campus, why waste all this money on “we pretend to teach, they pretend to learn”? And why are there zero highly selective schools for the many students who barely appreciate on-campus activities? Which probably includes most STEM majors, because they have so little free time.

There is a large demand for the consumption aspects of the higher ed “experience.” The U.S. is a rich society, and students (and their families through intra-household economics) want to consume the college product/service bundle — the “college experience” as even this verbiage began to reflect the shift toward non-educational, “summer camp”/ consumption.

The question, again, is: How many students/families would pay for these “summer camp” add-ons out of their own pockets? Some, yes. But many? No.

It is related to what we have observed at lower levels of education where the reason there isn’t more learning has little to do with teaching methods or school systems but the lack of interest by students and their parents in the educational component. For many parents, school is mainly about cheap day care. That’s a point that I make in my 1999 Spoiled Rotten: Affluence, Anxiety, and Social Decay in America.

Agree, but then you should grant that much (most?) of the instructional spending is waste!

Yes, $1.5+ trillion in student loan guarantees helps bolster this demand, but $1.5 trillion in loan guarantees is not the same as handing out $1.5 trillion in direct subsidies (grants), at least not yet. During most of this growth period in demand, the majority of students/parents expected to repay these loans. That may have changed more recently.

Government subsidizes education in many ways. Sure, federal student loans. But there are also direct government subsidies to state schools, plus large implicit subsidies via initial real estate purchases. The best way to ballpark the total state taxpayer subsidy, in my view, is to look at the net tuition gap between in-state and out-of-state students. At my own GMU, for example, list in-state tuition is $13,812, versus $37,976 for out-of-state. Of course you need to adjust that for financial aid to find net tuition, but that will barely alter the implication that Virginian taxpayers ultimately cover about 2/3rds of the cost for Virginian students.

A couple of additional items support this view. For one, students and their families have lower cost alternatives available, but most don’t utilize them. They could live at home. They could attend lower priced schools. Do we really think that all of these students and families are being duped? Do we really think that the guy getting a degree in art history is doing so because of how great the university is marketing the job market for that degree?

I agree, subject to the proviso that taxpayer (and donor) subsidies largely destroy the market for highly-selective, no-frills education. If government sells B+ education for a heavily-subsidized price, that kills off all quality levels below B+.

Both my daughters went to a private university in Nashville that isn’t at Ivy League expense levels but much more expensive than attending WKU. They wanted to go there, and, in all honesty, I wanted them to go there, not so much for the different educational value (there is some), but for the different peer group. My well-off friends here in town send their kids to Vandy but also UK, IU, UT, Ole Miss, and various state schools that are substantially more expensive than WKU.  They do so not because of any loans they are receiving, but because that’s the “experience” that they or their kids desire.

I agree the some wealthy families do so, but this is a modest share of the market.

Maybe my (and the profession’s) reluctance to accept quality adjustments to medical care or income growth and consumer demand as the primary contributors to medical and educational prices because these explanations don’t feed a political agenda of the interventionists or non-interventionists.  In any case, I hope that you will give these ideas a chance.

You make a strong case, Brian, but not strong enough. Seriously:

1. If all government support for medicine ended, how much do you think consumption would fall? How much budgetary “fat” would the surviving medical providers manage to unearth? And how much would life expectancy or other objective health measures actually decline?

2. If all government support for higher education ended, how much do you think enrollments would fall? How much budgetary “fat” would the surviving schools manage to unearth? And how much would worker productivity actually decline?