Caplan’s ‘case against education’ (1)

Photo: Matthew T Rader / Unsplash

It’s a pitty that Bryan Caplan’s extremely interesting book The Case Against Education: Why the Education System Is a Waste of Time and Money 1, published in 2018, has not deserved more careful attention in the public and permanent debate over education (see here an early review in Spanish by Juan Ignacio Pérez Iglesias). I think part of the reason is that the book’s conclusions are so extreme that the most than reasonable elements of its analysis and of its message are easily discarded tout court as merely a part of a silly proposal. Caplan’s well known ultra libertarianism of the Randian type does not help to make him, or his works, amiable in a area (education economics, or education in general) that has traditionally leant towards the left, or at least towards a strong ideological submission to the state, left- or right-winged), not only in the United States (the country whose system is studied in the book, and where its advices are basically addressed to), but also in the rest of the world. After all, public education, both in the sense of universal, free and compulsory education for children and adolescents, and of a strongly subsidised university system, is taken to be one of the fundamental pillars of the modern welfare state and of the democratic systems. Education is universally seen as a first-rate social investment, if not the best one, and lagging behind other countries in terms of the GNP share of education expenditure is systematically a reason for shame in the political sphere. A voice that claims that we should spend not only less, but quite a lot less, in education at basically all levels will expectably be received with incredulity, or worse, with nothing but silence, no matter how relevant the arguments and data behind the proposals are.

Caplan claims in his introduction that his book “does not try to change your mind about brute facts, it tries to chance your mind about the best way to interpret facts you’ve known for ages”. In a parallel fashion, my aim here is trying to show that Caplan’s new interpretation can only be partially accepted, in a way that his most radical political proposals are hardly the best way to attack the flaws that his own interpretation of the problems more ore less rightly identifies. I mean, I shall accept a big part of his diagnostic, but will argue for a different treatment of the patient. Since this review is necessarily short, I can only summarize very briefly the core of the book’s points, and will leave for the next entry my comments on Caplan’s practical proposals.

In a way, the only thing that Caplan does in the book (as many good books do) is to apply a simple and logical idea to a field that has been reticent to derive the consequences that follow from it: the idea is just that education diplomas, and education attainment in general, basically have a signalling role (José Luis Ferreira commented it here a few years ago); this simply means that employers use the education levels of candidates to a job as a useful information about the quality of the people, without necessarily assuming that the content of the things learned during that education are particularly useful for the job in question. Since the field of information economics started to bloom fifty years ago, ‘education as signalling’ has been a commonplace for its practitioners, though, as Caplan complains, this (platitude?) hasn’t had the mildest influence on the field of economics of education, not to say on education policy. These latter fields are dominated, instead, by what Caplan calls ‘the Human Capital model’: the assumption that education consists in creating ‘human capital’, i.e., useful abilities that are the ones that employers are looking for when selecting an employee. According to this orthodox model, there is just a linear causal relation between the amount of resources (money, time, effort…) employed in the education system, and the economic productivity of a country (not to mention other socially desirable ‘by-products’ of the education years, like a more enlightened and less violent citizenry). This orthodoxy accepts that some signalling exist in the labour market (after all, what better measure of the human capital accumulated by a person than her own CV?), but denies that this may have any significant effect on the quality and quantity of the education provided and of the human capital created.

Caplan, instead, accumulates fact upon fact to show the shaky basis of this orthodoxy, though many of the stubborn statistical facts described in the book do little more than convincing you that you already knew the essentials through your own direct experience of the education system. And what are these essentials? I will comment here the most important two ones. First, that most of what is taught from the school to the university is basically useless for the immense majority of the jobs most people end performing. The enigma (from the theoretical economist’s point of view) is not whether those ‘contents’ may or may not serve for other things apart from those jobs, but why employers are willing not only to use the diplomas as signs of ‘worker quality’, but to offer, as they actually do, more than substantial salary increments for each additional education level (for example, the mean salary for people with a Master’s degree is around 90.000; for people with a Bachelor's degree 70.000; with a High School degree, 40.000; and without any of those, around 31.000; US data from 2012).

But it’s a universally known fact (repeatedly brought to light by scrupulous social surveys, if direct acquaintance were not enough proof) that most of the people don’t learn most of what their teachers explicitly taught them, and forget very soon most of what they have happened to learn. More sophisticated evidence is needed (and Caplan does provide it) to show that the efficiency with which tacit abilities (like, e.g., verbal competence, critical thinking, reasoning strategies, a taste for high culture…) are implicitly taught and learned is not better than that of the explicit curriculum, but usually still worse. Hence, if all this is so, why do employers prefer (and are ready to pay more than double for) hiring a youngster with a Master degree instead of one of similar age with a mere High School degree?

Caplan suggests that the ‘signal’ displayed by the diploma is not relevant as information about the skills supposedly contained in the syllabus backing the diploma, but merely about the candidate’s general intelligence, conformity, and perseverance (or, to use Caplan’s term, ‘conscientiousness’): the more steps in the education system you have surpassed, the more of these three things you needed to have to begin with. The question is, will these capacities of a particular individual have been created by the education process (as a sculptor that creates a statue out of a marble block)? Or will they be pre-existing, and the only role of the school or the college will have been to select the individuals having them, and certify their having them (more or less as an art expert hired by a museum to know whether a particular statue is original or fake)? The Human Capital model assumes the first answer is right, giving only a marginal role to the second. The, so to say, Pure Signalling model assumes the opposite. Caplan concedes that the truth must lie somewhere in between; his own estimate, through more or less painful statistical work, and with lots of generous assumptions for the orthodox alternative, is that education is about an 80% signalling and only a 20% human capital.

This, of course, leads to the second well known fact: grade inflation. For many jobs that decades ago were competently performed by people that, on average, had a certain education level, now it is required to have a much higher degree. This inflation is one of the main reasons to fear about the most dangerous mistake committed, according to Caplan, by defenders of the Human Capital model: the fallacy of composition. It is obviously true that, for a single individual, it is enormously profitable to reach the highest diploma she can attain: at the individual level, education pays; the effort, the time and the money invested in attending the school and the university are compensated enough (on average) by the extra gains this education will grant you in the future (though, I would add, in countries like US, with its system of student loans, this may be near to cease to be the case in many cases). But the social gains a country enjoys thanks to education are not the simple sum of the individual gains, because the latter are essentially competitive: the job you get is not given to the one with the second best CV. To see the reason, imagine that by art of magic everybody in a country wakes up tomorrow with an additional diploma; the more these diplomas are useful because the signal relative competence, and the less they are useful because they signal real new abilities (i.e., the more right is the Pure Signalling model, and the less right is the Human Capital model), the less will the economy grow (for workers productivity will have increased little), but the more replete will your CV have to be in order to warrant you a decent job.

Hence, there is no immediate, linear connection between the growth in salary that more education warrants an individual, and the growth in productivity the country has thanks to the sum of all that additional education. The last fact we will mention in this first part of our review refers, precisely, to another way of looking at this difference. For individuals, the interest rate of investment in education is, on average, about 9% (it’s a really good investment). In contrast, national investment in education seems to contribute to GNP with just an interest of 1-3%. In both cases, correlation is not the same as causation: after all, both one family and one country can invest more on education, not because this will be produce more rent in the future, but because they are more wealthy now. But statistics allow to discriminate the causal direction more easily in the individual than in the national case, and so we can take the 9% figure at face value, but be a little more sceptic about the 1-3% figure (i.e., perhaps wealthy countries are not more wealthy because they have spent more on education, but spend more on education because they are more wealthy, and education is what economists call a ‘luxury good’).

In the next entry, I shall comment on the (dramatic) political consequences Caplan derives from facts like these.


  1. Caplan, B., 2018, The Case against Education: Why the Education System Is a Waste of Time and Money. Princeton. Princeton University Press.

Written by

Leave a Reply

Your email address will not be published.Required fields are marked *