Education and the Baumol Cost Disease

One teacher, one blackboard, thirty students: a recipe for educational bankruptcy.

Matt Yglesias is right to point to what he calls the “Baumol effect” (which I was taught to call the “Baumol cost disease”) as the central problem in educational finance. But I think he’s wrong to say that our only options are lower relative teacher salaries, larger class sizes, or higher school taxes.

The crucial missing assumption is that education must remain in what Baumol calls the “non-progressive sector.” The way out of the box is a change in the basic technology of teaching.

Baumol’s model, laid out in his classic “Macroeconomics of Unbalanced Growth,” is straightforward. If one part of the economy – widget-making – enjoys productivity growth, and another part – live chamber music – doesn’t, then the non-progressive sector faces a progressive cost squeeze. The price of chamber-music tickets in terms of widgets has to go up, or the pay of chamber musicians relative to widget-makers has to go down.

As long as teaching is stuck in the one-teacher-one-blackboard-thirty-students model, then education is a classic Baumol enterprise. That’s why we’re spending several times as many real dollars per pupil as we did in 1970, and getting by many standards worse results.

And that, in turn, is why the discussions about “educational technology” always seem so bizarre to me. It’s as if vaudeville promoters in 1920 were discussing whether they should start using loudspeakers or disco balls, rather than understanding that radio and the movies were making their whole business model obsolete.

The future of education is for students to educate themselves, individually and in groups, with the help of computers, networks of computers, recorded media (including, of course, the greatest educational innovation of all, the printed book), and the skilled facilitation of a relatively small number of live helpers. If you think it’s impossible to get masses of young people to spend astounding numbers of hours on cognitively-demanding tasks, then how do you explain the success of the video-game industry?

No, I don’t think this will be easy, or that there won’t be real losses. (A movie isn’t a perfect substitute for a stage play, and even a great CD isn’t the same as a live performance.) But there’s no option, so let’s stop complaining about the future and start inventing it.

And yes, that means universities, too. How many people, right now, are preparing to give a lecture tomorrow in introductory economics or organic chemistry? And how many of those lectures have more educational value than would a video of the best such lecture being given tomorrow? Or, better yet, a professionally-produced lecture on DVD, with hot-links to relevant materials?

The problem is that (some of ) the universities also produce new knowledge. Right now, the teaching business cross-subsidizes the research activity.

We’re going to need a new financing model for the knowledge-production part of the university’s mission. But we can’t keep operating the universities as technologically backward enterprises, just to maintain teaching jobs as an excuse for supporting scholars.

Author: Mark Kleiman

Professor of Public Policy at the NYU Marron Institute for Urban Management and editor of the Journal of Drug Policy Analysis. Teaches about the methods of policy analysis about drug abuse control and crime control policy, working out the implications of two principles: that swift and certain sanctions don't have to be severe to be effective, and that well-designed threats usually don't have to be carried out. Books: Drugs and Drug Policy: What Everyone Needs to Know (with Jonathan Caulkins and Angela Hawken) When Brute Force Fails: How to Have Less Crime and Less Punishment (Princeton, 2009; named one of the "books of the year" by The Economist Against Excess: Drug Policy for Results (Basic, 1993) Marijuana: Costs of Abuse, Costs of Control (Greenwood, 1989) UCLA Homepage Curriculum Vitae Contact:

25 thoughts on “Education and the Baumol Cost Disease”

  1. Indeed. A sad and harsh reality, but probably a reality nonetheless. Of course, most adults today think of college as an institution in which 30-100 students sit in a classroom and listent to a live lecture, potentially with the opportunity to ask questions, interact with the instructor, discuss ideas with fellow students, etc. And the loss of that is a shame. But a couple generations before one-on-one tutoring with professors was relatively common, and thos days are (largely) long gone.

  2. At last, we agree on something.

    E.G. West
    Education Without the State

    What is needed is choice in education. School choice has not and will not lead to more productive education because the obsolete technology called ʺschool” is inherently inelastic. As long as ʺschoolʺ refers to the traditional structure of buildings and grounds with services delivered in boxes called classrooms to which customers must be transported by car or bus, ʺschool choiceʺ will be unable to meaningfully alter the quality or efficiency of education.

    Ken Burns (The Civil War, Baseball, Jazz) is probably the best US History teacher who ever lived.

  3. Ah, in this post is the difference between liberals and others. Mark just went against his interests, and the interests of his tribe, in the name of the common good. Could Brett or Redwave or other non-liberal commenters on this site please point to any of their comments in which they argue against their own interests?

  4. MK: “Ken Burns (The Civil War, Baseball, Jazz) is probably the best US History teacher who ever lived.”

    Per Mark Kleiman’s analysis, he might qualify as one of the more productive in some important sense, his efforts amplified by his medium, but “best”?? “Best” implies a level of quality, accuracy and depth, he never obtains. I enjoyed his Civil War and recommend it, but, knowing some Civil War history and historigraphy, I grind my teeth at the inaccuracies and the lack of depth, that comes from avoiding analytical conflict and extended interpretation.

    @Ebenezer Scrooge

    Completely unfair, as “conservative” is virtually defined the defense of vested interest. What Mark does here, practically defines the liberal progressive. Now, we just need a conservative to pervert his idealism into some rent-seeking scam, and, voila!

  5. The whole point of Baumol’s cost disease is that paying more for things like education is the kind of problem you want to have! I just means that everything else is getting cheaper.

  6. Argh…. my economics bone is hurting. I would agree with chrismealy, and then ask this question: where is the morality of shared interest in this economic equation? There are plenty of things in life that are good, valuable services yet which will likely, in 1000 years, be about as productive. Like parenting, for instance. The husband or wife who chooses to stay home and raise the babies is doing something invaluable. So how do we value it? Mark might suggest we find ways of increasing the productivity of parenting – maybe involving videoscreens?

    We’ve become so used to business-speak about “growth” being this magic genie that grants all things good, that we have lost sight of what really matters in society. Once we discovered that there were millions of starving people in the 3rd world willing to live on slave wages (because at least slaves get to eat), we sold our neighbors out faster than you can get to page 6 of the WSJ.

    Now, I’m not suggesting I have any grand answer to trade and productivity. But I do think that we’ve bent so far over the altar of *quantified productivity* and *growth outcomes*, that we’re forgetting the real value of the real work that real people do. Just because Mark Zuckerburg figured out how to print gold, it doesn’t mean that schoolteachers, firefighters, waiters, taxi drivers, professors, mechanics, cashiers and the rest of us now have to have our god-damned stock diluted.

    OK, I took a breath. You want productivity? A school teacher who turns out good students prints gold. A taxi driver who can drive you home from a bar at 2am prints gold. A plumber who can fix the gushing pipe in your basement prints gold.

    Whether we as a society choose to value them, that’s up to us. Maybe the “market” decides they’re all better off with no unions, shitty working conditions, and measly pay. But we can make sure they get affordable housing. We can pay for their dialysis. We can keep the lights on in the libraries for them and the mail delivered on time. That, we can do.

    …wow, and I didn’t even address the crap assumptions about educational pedagogy and the realities of the achievement gap. Not bad.

  7. Besides “introductory economics or organic chemistry”, probably most freshman/sophomore level medium or large size lecture course (e.g. Calculus, Lin Alg, Diff Eqn, etc.) course could be replace with high quality recorded productions. Maybe the textbook industry can do this. Labs are another matter. Also I’m guessing the first two years of med school (basic sciences) can be also be easily done on video as well (especially since those students already have the requisite maturity and focus). Maybe the education test prep industry can do this (e.g. Kaplan). Again the anatomy labs would be an issue.

    Though we have the Fermi Paradox: If this future, why aren’t we already delivering course like this?

  8. (Bruce): “…’conservative’ is virtually defined the defense of vested interest…
    You mean, like, defense of Post Office subsidies and its statutory monopoly on first-class mail? Defense of the NEA/AFT/AFSCME cartel’s exclusive position in receipt of the taxpayers’ pre-college education subsidies?
    (Bruce): “What Mark does here, practically defines the liberal progressive. Now, we just need a conservative to pervert his idealism into some rent-seeking scam, and, voila!
    Horace Mann beat you to it by about 170 years.

  9. (AC): “If this future, why aren’t we already delivering course like this?
    1. The NEA and its subsidiaries collect hundreds of millions in dues and lobbies stridently to maintain its exclusive position in receipt of the taxpayers’ K-12 education subsidies.
    2. College professors are (a) articulate, (b) well-paid, and (c) have a lot of free time.

  10. Compare teachers to waiters; schools to restaurants.

    Fancy restaurants command high prices. Partly, it’s because they feature platoons of tuxedoed waiters. A cafeteria could, in principle, serve equally tasty and nutritious food, prepared by master chefs, out of vending machines. The nation’s fine-dining bill could be substantially reduced by such efficiency. Right?

    Oh, sure: “fine dining” is different from “education” in all sorts of ways. Knowledge is not like food, whose “value” increases when it is served up with personal attention. Knowledge is a commodity, which can be mass-produced and is best served by the intellectual equivalent of vending machines. Fine dining is about status, which is worth spending money on. Education is a mere necessity, so the cheaper we can get it the better off we are. Personalized service is good for grown-ups, not schoolchildren. By spending less on teachers, we will have more money to spend on waiters, in The Economy.

    I don’t know what fraction of the GDP is spent on “fine dining”. I do know that one of the reasons people like being rich is that they can afford to eat at elegant restaurants where the waiters are numerous, attentive, and expensively upholstered . I suspect that if more Americans were eating more of their meals at expensive restaurants (instead of in cafeterias or at home) then GDP would be higher than it is now. And not just by definition. As we measure GDP, “inefficient” — i.e. costly — waiters probably do more to “add value” to a slab of meat or a bushel of spuds than ranchers or farmers do. “Adding value” is a Good Thing, no? So: should we be glad, or sorry, if the fraction of the GDP spent on fine dining were to increase? Is it not essentially a tautology that a nation which spends more in expensive restaurants is a wealthier nation?

    Note, please, that GDP is, in fact, SPENDING. A growing GDP is considered desirable by all right-thinking people. But you can’t grow GDP by cutting spending in EVERY sector of The Economy. Cut spending on education; cut spending on health care; cut spending on energy, or research, or litigation, or the arts; and you have to INCREASE spending on something ELSE if you want to “grow the economy”. What’s left? Pornography?


  11. “How many people, right now, are preparing to give a lecture tomorrow in introductory economics or organic chemistry? And how many of those lectures have more educational value than would a video of the best such lecture being given tomorrow?”

    Pretty much all of them, actually, since the evidence from studying higher education suggests that recorded lectures produce worse educational outcomes than live ones. Lectures, even with hyperlinks, are no good for internet education; you need interactive courses online with constant feedback, ideally with live tutors available. (Video games, remember, are interactive.)

    The best college outcomes, by a long way, seem to be produced by a mixture of interactive online courses with quick, regular online feedback, supplemented by live tutorial help plus live lectures. But that won’t save any money.

  12. “Could Brett or Redwave or other non-liberal commenters on this site please point to any of their comments in which they argue against their own interests?”

    Well, I am a professional engineer, a skilled profession. It is unambiguously in my self interest that immigration policy promote the immigration to this country of unskilled workers: Doing so will bid down the labor of numerous sorts of services I might avail myself of. For the exact same reason it’s in my interest that our immigration laws obstruct the immigration of skilled professionals who might compete with, and bid down, MY labor.

    And yet, I steadfastly advocate the exact opposite immigration policy, wide open to the skilled professionals who compete with me, and slam the door and start deporting illiterate unskilled laborers. Why? Because that’s what I think would be best for the country, and most of it’s citizens. We’re not going to become wealthier by loading up on unskilled labor. We’re just going to keep the people who DON’T have highly sought after skills poor.

  13. (Dominic): “The best college outcomes, by a long way, seem to be produced by a mixture of interactive online courses with quick, regular online feedback, supplemented by live tutorial help plus live lectures.

    I wonder. That’s not how university education works at Oxford or Cambridge, I understand. I recall a story about a visitor to the London School of Economics observing a lecturer addressing a nearly empty hall, with an audience of four or five scribes with tape recorders. He asked: “Where are the students” and his guide replied: “They do not have time to attend lectures.”

    To hear a string quartet live you need four musicians. To hear a hundred string quartets, you could buy a CD player and 25 CDs, much cheaper than 30 concert tickets (okay, I’m a dinosaur. Steal an mp3 file). Books do the same for lectures and college courses.

    Anyway, Richard Arkwright was homeschooled. Thomas Heighs and James Hargreaves were minimally schooled. Cyrus McCormick was homeschooled. Thomas Edison was homeschooled and started work at 13. Hiram Maxim left school at 13 and apprenticed. Robert FitzRoy attended the Admiralty school for twenty months between age 12 and 14 and went to sea at 14. Sam Colt went to sea at 16 and carved a wooden model for his revolver while underweigh to Bombay. The Wright brothers did not finish high school. Bertrand Russell was homeschooled before university and then entered that tutorial system, where you read assigned works and discuss with your tutor once a week. Yehudi Menuhin was homeschooled.

  14. Malcom
    1.) The quote about lectures at Oxford and Cambridge was a joke. The average student was from a higher class background than the average lecturer. The lecturers were beneath the students, socially. The quip was not a commentary on pedagogy.

    2.) There are physical vibrations which you feel in your whole body during a live musical performance. Recorded performances are dead in comparison. Listening to CD’s allows students to compare how a large number of musicians approach a piece and allows people to have music in their homes every night. That is great. But, it is not adequate substitute for going to live performances.

    3.) They were “homeschooled”, passive voice. It just spontaneously happened with no labour on anyone’s part? No, tutoring and home schooling involve one-on-one or very small group instruction with a person. It would be great if every child could have that. But, both methods are more labour intensive than what have come to be traditional class rooms, not less.

  15. I found lectures to be an enormous waste of my very scarce time in college. Much better to just assign a good textbook and have professors available to answer questions.

  16. to be naieve about this, it seems to me that teaching is always an activity that is rooted to some degree in a teacher to student model. it is hard to break too far from the 1 teacher, one chalkboard and 30 student model. to me it is more analogous to parenting. it is hard to break away from each child needing approximately 2 parents. How does the Baumol effect apply there?

  17. It is not like we don’t have extensive data and studies about the effectiveness of using this technology. There is a reason why Universities were rushing to embrace distance learning in the early 2000s and have backed off it in the past decade. And there is also reason why for-profit schools are under so much scrutiny now for their abysmal educational outcomes compared to everyone else.

    If Mark Kleinman has proven pedagogy for how to use all this technology effectively, I recommend that he publish it. Or even better yet, go into business.

  18. Hypothesis: In theory reading a text book, watching a lecture (live or taped), doing HW, and taking tests should be sufficient to learn most material. In practice, for many undergraduates, interacting with experts (profs and/or TA’s) and peers in a school setting increases the rate of absorption & retention of the material, plus it also helps with the requisite motivation needed to do the cognitive heavy lifting. For many of us learning difficult new concepts is difficult and every little bit helps (even a sympathetic comment from a teacher can sometimes help a student stay the course).

    Prediction: Even if the above hypothesis is true, the electronic delivery of education will accelerate simply because the current style of teaching at the scale needed would be cost prohibitive (college is the new high school). Lectures alone are not as effective as Oxbridge style tutorials, yet most US universities rely on lectures alone (office hours not withstanding) because presumably we cannot afford to educate undergraduates in the tutorial style.

    Anecdote: Female relative with a BA from Rutgers in a social science (B average or maybe better), got married, started raising kids, got divorced, got a job (in an agency which places temp works), got laid off (in the great recession), searched for another job (in vain for about a year), decided to get trained as a speech therapist, started taking pre-req courses (in Bay Area community colleges). After taking Intro Bio, it took her a year before she could grovel her way into an oversubscribed Human Anatomy course. Miles to go before she sleeps.

    Proposal: Have a central source in the UC System produce quality video lectures of the standard undergraduate courses and have them piped into all the community colleges in CA. Have the teachers in the community colleges clarify the lectures, answer questions, proctor exams, and run labs. Will this help?

  19. Ebenezer, it was nice of Brett to provide a specific example, but I don’t blog about my work, not even on my own blog! There are good reasons, not the least of which is to allow for my continuing employment. So you will just have to take my word for it – I am pretty good at compartmentalizing and the political and economic views I express are sometimes in conflict with what would be best for me, my industry, and especially my investments. On the other hand, why would I promote policy choices that are in direct conflict with my beliefs, and those seldom, if ever, conflict with the above.

    As for Mark, I am not so sure his post is such a demarcation from his interests. It may be that he simply recognizes that the current educational (business) model is unsustainable.

  20. What this suggests to me is that we are already way past the point of diminished returns in our simple-minded attempts to industrialize education. There are a few (maybe hundreds, maybe thousands) great lecturers, and attending their performances is an inspiration. (I don’t know whether they’re as good recorded as live, maybe, maybe not.) Everybody not privileged to take courses from those people is going to need more attention rather than less (and saying students should give it to one another is going to be about as uniformly successful as getting people to pay to generate content on the intertubes).

    I think chrismealy’s point is important: imagine what a world we would have if the vampire squid in today’s economy were education rather than finance.

  21. “…the skilled facilitation of a relatively small number of live helpers…”

    There’s an assumption based in technological utopianism, here: that a small number of helpers would sufficiently support the rich technological and non-personal provision of information that Kleiman proposes, and that high-quality pedagogy can be sustained with few such helpers.

    Speaking as a university pedagogue in the middle of trying to improve the efficient use of resources in our curriculum (in a public research university facing major funding issues), I wonder. One of our proposals is to introduce a course earlier in the curriculum that focuses explicitly on the intellectual tools common in the discipline involved. (That we have not had such a course in the past is a crying shame, an oversight driven by inertia, institutional friction, and other causes). It seems like a no-brainer.

    And in search of ‘best practices’ and models, we looked around the country. One very attractive and effective model comes from a little private college in Cambridge, MA. Without going into details, the course involves heavy use of books, online resources, with “a small number of helpers (aka: teaching assistants)” to guide students as they acquire familiarity with the resources and tools involved. In fact, the Cambridge MA ratio of such helpers to students is one to four. In our much more fiscally-limited circumstances, we’re envisioning one such helper for every 50 undergraduates. Does Kleiman imagine that our model will provide better outcomes than the one in fair Cambridge? (It actually will, for some students and some assistants, but not on average, no matter what whiz-bang technologies we deploy — which, of course, the high-labor version will also deploy).

    This gets us to the quite helpful restaurant analogy: the food in expensive restaurants is generally better (though not always that much better) than in cheap diners, but the cost difference results also from intense investment in service — and there’s little doubt that attentive service is also valued by many diners and makes the dining experience better. But the result is that some people have the choice of the expensive restaurant (which may or may not have better food), while others can only afford the diner. You can still make food better at diners, and that’s well worth doing. But there are various reasons (from branding and social/cultural capital acquisition, to networking opportunities, to use of scarce or high-quality ingredients, to the seductive pleasure of having people wait on you) that high-end restaurants will continue to exist, and be limited to the wealthy, and will play their own small part in perpetuating wealth, too. And the food at high-end restaurants will be better, too, consistently even if not always.

  22. If you look inside the restaurant analogy, you find some perhaps-salient bits. Many of the common people I know who dine at fine restaurants — comfortably-retired foodies, food writers and so forth — tend to complain about the obtrusiveness of the service at many expensive places. At some point you just want to enjoy the meal, and staff who know when to stay out of the way are coiners of gold. Also, the real costs are going to be in the kitchen — not just in the ingredients, but in having far more people preparing the plates of a given number of customers as you have in a diner. Some dishes require constant attention, others must be prepared to order from the beginning, yet others suffer terribly if served even a couple minutes after their proper moment. Diners limit their selection (usually) to things that will be good even if not given careful attention from minute to minute.

    Which suggests to me that ultimately educators are going to have to look a lot more closely at the strengths and weaknesses of different kinds of presentation, and be sure they have the big guns to bring out when they’re needed. Or simply not offer certain things at certain places.

  23. “If you think it’s impossible to get masses of young people to spend astounding numbers of hours on cognitively-demanding tasks, then how do you explain the success of the video-game industry?”

    This is simply willful stupidity. How about: if you think it’s impossible to get masses of young people to spend astounding numbers of hours becoming pharmacists, then how do you explain the ubiquity of their marijuana use?


    if you think it’s impossible to get masses of young people to spend astounding numbers of hours learning electronics engineering content, then how do you explain their incessant use of cell-phones?


    if you think it’s impossible to get masses of young people to spend astounding numbers of hours learning to play the violin, then how do you explain the number of page views of Justin Bieber videos on youtube?

    willful stupidity

  24. One solution to lowering the cost of college tuition is to have volunteer instructors. Like so many protected institutions (thin Mancur Olsen), the barriers to entry for college instruction insure that only lifelong academics can play. Do I want to take a class in architecture from someone who started and managed a firm through thick and thin for 20 years or an academic? These days, if you do not have a PhD you cannot teach yet if you have a PhD you probably have not spent the better part of your life surviving in a darwinistic world. Imagine a university where all the teachers (or some) were volunteer professors from the private sector? Name a truly successful person that wouldn’t cherish the opportunity to give back.

Comments are closed.