Read The Case Against Education

Yesterday was the Kindle publication date for my colleague Bryan Caplan’s new book The Case Against Education. The hardcover publication date is in nine days. It is an excellent book, on an important topic. Beyond such cheap talk, I offer the costly signal of having based an entire chapter of our new book on his book. That’s how good and important I think it is.

The most important contribution of Caplan’s book is to make very clear how inadequate “learn the material, then do a job better” is as an explanation for school. Yes, the world is complex enough that it must apply sometimes. Which is why it can work as an excuse for what’s really going on. After all, “the dog ate my homework” only works because sometimes dogs do eat homework.

So what is really going on? Caplan offers plausible evidence that school functions to let students show employers that they are smart, conscientious, and conformist. And surely this is in fact a big part of what is going on. I’ve blogged before one, and in our book we discuss, some other functions that schools may have served in history, including daycare, networking, consumption, state propaganda, domesticating students into modern workplace habits.

But I should be clear that students today don’t need nearly as much school as they get to serve these other functions; showing off to employers is likely the main reason kids get so much school today. Our world would be better off with less school, such as would happen if we cut school subsidies.

I see Caplan’s book as nicely complementing ours. As I said recently:

The key problem is that, to experts in each area, no modest amount of evidence seems sufficient support for claims that sound to them so surprising and extraordinary. Our story isn’t the usual one that people tell, after all. It is only by seeing that substantial if not overwhelming evidence is available for similar claims covering a great many areas of life that each claim can become plausible enough that modest evidence can make these conclusions believable. That is, there’s an intellectual contribution to make by arguing together for a large set of related contrarian-to-experts claims.

GD Star Rating
Tagged as:
Trackback URL:
  • Robert Koslover

    Re: “Beyond such cheap talk, *I offer the costly signal of* having based an entire chapter of our new book on his book.” What a remarkable thing to say, that is, to “offer the costly signal…” And yet this phrase also makes perfect sense in the context of this forum. Robin, I think you may actually be influencing our terminology and culture. Bravo.

  • Robert Koslover

    Un-schooling / home-schooling pioneer John Holt had some important things to say about our systems of education that I suspect may have some relevance here. See and

  • Yoram Gat

    But if employers like schooling as evidence of desirable employee qualities (whatever these are), then reducing school subsidies would make it harder for poor people to obtain such evidence and therefore to find good jobs. Since allowing poor people to attain good jobs is a major declared goal of school subsidies, then it seems they are working as intended.

    (By the way, as far as I can remember, Caplan’s “The Myth of the Rational Voter” was a book-length version of the “people don’t vote as economists think they should therefore they are irrational” argument: poorly argued elitist self-congratulation. But maybe I missed something?)

    • If your focus is on helping the poor, then you want to give the poor a larger subsidy than others. You could even tax the schooling of others at a higher level, relative to the poor. Giving everyone a uniform subsidy doesn’t help the poor or everyone.

      • Yoram Gat

        So, is Caplan arguing in favor of taxing education for, say, the top 50% of households, and subsidizing education for the bottom 50%? I suspect his audience wouldn’t be half as happy with this idea as with the proposal to eliminate school subsidies.

      • sanxiyn

        Actually, yes. Quoting from the page 276 of the book:

        “For elementary and secondary school, means-tested vouchers. Taxpayers fund education for poor children. Otherwise, parents pay.”

      • Yoram Gat

        Thanks for the pointer. I note that there is no mention of taxing education for the rich. Also, how much education would those vouchers pay for? Is there a guarantee (or even an aspiration) that the poor would have access to similar-quality education as the rich?

        In context, Caplan’s position seems much as I would have suspected it to be: let people take care of their own education. He is essentially unconcerned about the obvious fact that this means increasing educational disparities (and, by implication, income disparities, replicating those disparities across generations).

      • sanxiyn

        The book discusses taxing education from the page 218, “Why Not Tax Education?”. Choice quote:

        “Taxing education doesn’t just butt against conventional presumptions in favor of education and the status quo. It runs afoul of the libertarian presumption in favor of leaving people alone.”

        Libertarianism aside, Hanson’s reply was that you need to subsidize/tax poor and rich differently to help the poor. Current school subsidies mostly don’t. Caplan is in favor of treating poor and rich differently.

      • Yoram Gat

        Yes – the choice quote sums it up. The guiding light here is the “libertarian presumption”. All else is trappings.

        It’s actually very simple. Obviously, as things stand now the rich consume more education than the poor, so the natural, most effective place to cut on over-education is with the rich. Taxing the rich for education should therefore be the natural way forward for any self-respecting economist advocating reducing over-education. And so we are shocked to discover that this is not at all what Caplan is proposing.

        [BTW, to state the obvious, of course subsidizing education affects the rich and the poor differently. It allows the poor to consume much more education than they
        otherwise would, while it makes little difference for the rich. Thus it reduces educational disparities.]

        Could I also point out the irony of you, Hansen et al., who are so busy identifying hidden motives in the world, taking Caplan’s claim as being concerned with the inefficiency of over-education at face value while there are rather transparent different motives which explain his actions much better?

      • Paul

        In general, the case for “stop subsidizing X” is significantly stronger than the case for “and furthermore, start taxing X.” So it’s a pretty natural first step.

        Taxing X for the rich and subsidizing it for the poor is almost certainly less efficient than taxing the income or consumption of the rich and giving money to the poor.

        The normal justification for inefficient redistributive policies is that they may be more politically feasible. But this is probably both inefficient and a political non-starter.

      • Yoram Gat

        Even if the fact that the pretensions of this community to rationality are really little more than a cover for the articles of libertarian faith was not unexpected, the speed at which this discussion shed those pretensions was somewhat of a disappointment.

      • brianholtz

        Caplan’s opposition to the inefficiency of the education signaling arms race seems quite independent of his libertarianism. By contrast, your priority seems to be to use this arms race as yet another argument for more redistribution — even if that extra redistribution has little prospect of getting us out of the current inefficient equilibrium in education signaling. Thus your talk of “pretensions” seems to be a case of projection.

      • Yoram Gat

        It you manage to actually address my point above, do let me know.

        On the odd chance that a restatement of my point could help you along, here is one:

        Reducing over-education is more efficiently done by incentivizing the most over-educated (the rich) to consume less education rather by incentivizing the least over-educated (the poor) to consume less education.

      • brianholtz

        Any “self-respecting economist” knows that the most over-educated (in the sense of educated beyond what is efficient) will tend to be those who pay for their education with Other People’s Money. A more subtle issue here is that much of the OPM problem in higher education in America is because retail tuition prices are only charged to higher-income families. I’m curious how/whether Caplan addresses that issue.

      • Yoram Gat

        So if we believe your argumentation, although the rich are more educated than the poor, the poor are more over-educated than the rich. In other words, the rich should be educated, the poor should not. It’s good that we have that clear.

      • Yoram Gat

        The solution now seems obvious: educational subsidies should be converted to cash grants. Then everybody would use the money efficiently. This way, we keep the distribution of money unchanged.

      • I can easily support that as better than the status quo, & I expect Caplan would as well.

      • Yoram Gat

        But would he, or you, support this as being as good as his proposal? According to his reasoning, it is just as good at promoting his declared goal (and differs only by not having the redistributional effect).

      • Seems that you are one of those folks who want to make every topic an occasion to argue about redistribution.

      • Yoram Gat

        As the author of a blog on “why we do what we do and why we pretend otherwise”, it seems you are very incurious about why Caplan is proposing what he is proposing, and why he pretends otherwise.

        Maybe the answer is too obvious to merit inquiry.

      • brianholtz

        On the contrary, Caplan discuss in his book at length about how his thesis/prescription is independent of — and in tension with — standard libertarianism. Meanwhile, Yoram’s comments are explainable as redistributionist mood affiliation.

      • Paul

        I think you misunderstand. The case for “change the price by $1” is almost always stronger than “change the price by $2.” Costs usually grow superlinearly while benefits grow sublinearly, so the (cost/beneift) ratio usually gets smaller as the proposed changes get larger.

        There is a further benefit to going from $1 to $0 subsidy (since you remove the administrative expense) and a further (larger) cost to going from $0 to $1 of tax (since you introduce a new administrative expense), but these aren’t important to my point.

    • JW Ogden

      If employers learn that lack of money is keeping children from poor families from achieving a higher level of school completion, wouldn’t they be able to compensate for that?

      There is also the issue that amount of spending per student year can vary greatly with little effect. That is poor children could go to much much cheaper schools for the same amount of time and get out with as good a signal. Also charities seem to love to subsidize education (and health care BTW).

      • brianholtz

        Also, note that if investing in education for poor students has such great personal ROI, then not only would charities be lining up to subsidize, but evil capitalists would be picking up all those $100 bills laying on the sidewalk — i.e. making tuition loans to capture some of that ROI.

        This is yet another instance of anti-market people not answering the question: if you’re right, why aren’t you (or people who think like you) rich?

  • Pingback: Rational Feed – deluks917()

  • Ari T

    This is a book I’ve been waiting for.

  • Ari T

    Now this is a book I’ve been waiting for.

  • brianholtz

    Perhaps the the biggest problem with Caplan’s book is that his diagnosis is too convincing. Having skimmed his discussion of his (admirable!) prescriptions, I don’t see a convincing explanation of how they would break us out of the current signaling-arms-race equilibrium. If everyone in the audience is standing to see better, it’s not enough to remove the subsidy for platform shoes.

    • So people shouldn’t point out problems unless they can offer a complete and convincing solution?

      • brianholtz

        On the contrary, problems still seeking a solution are the most important to write about. With my oldest kid in her senior year of high school here in Palo Alto, I can report from the front lines that the signaling-arms-race is even worse than one might conclude from Caplan’s book.

      • Then how is it a “problem” with Caplan’s book that he points out a problem without giving, to you, sufficient solutions?

      • brianholtz

        Asked and answered. Maybe if Caplan gave more historical data about how the arms race arose, it would be more obvious how to unwind it. And maybe this is a topic more for your new book than Caplan’s, but I wonder if there are historical examples of successfully unwinding other kinds of signaling arms races.

      • Cowboydroid

        Maybe it’s not a problem that will be unwound by grand managerial efforts. Maybe that’s the point.

        Decay will define the end of government education and education subsidies, as it does most other big government initiatives and big government itself.

  • Nate Spears

    Robin – I heard you on Sam Harris’ podcast and you mentioned that you had previously done research on how to improve education. Where can I read about your work there? Also, have you evaluated Direct Instruction methods, and if so, what do you think about them?