Is Education Worthless?

This article was originally published in the Monday Magazine of 3 Quarks Daily (see here).

“How do you get a philosophy major away from your front door? You pay them for the pizza.”

As a doctoral candidate in philosophy people often ask me what I am going to “do” with my degree. That is, how will I get a job and be a good, productive little bourgeois worker. How will I contribute to society, and how will my degree (which of course was spent thinking about the meaning of “meaning”, whether reality is real, and how rigid designation works) benefit anybody. I have heard many variations on the theme of the apparent uselessness of philosophy. Now, I think philosophy has a great many uses, both in itself and pragmatically. Of concern here, however, is whether not just philosophy, but education in general might be (mostly) useless.

If you are like me, then you think education matters. Education is important, should be funded and encouraged, and it generally improves the well-being of individuals, communities, and countries. It is with this preconception that I went head-first into Bryan Caplan’s well written (and often wonderfully irreverent) The Case Against Education, where he argues that we waste trillions in taxpayer revenue when we throw it at our mostly inefficient education system. Caplan does not take issue with education as such, but rather the very specific form that education has taken in the 21st century. Who hasn’t sat bored in a class and wondered whether circle geometry would have any bearing on one’s employability?

As the title suggests, this is not a book that is kind in its assessment of the current state of education. While standard theory in labour economics argues that education has large positive effects on human capital, Caplan claims that its effect is meagre. In contrast to “human capital purists”, Caplan argues that the function of education is to signal three things: intelligence, conscientiousness, and conformity. Education does not develop students’ skills to a great degree, but rather seeks to magnify their ability to signal the aforementioned traits to potential employers effectively.

In essence, Caplan wants to make the case that education is massively overrated. He does this by explicitly arguing for the signaling model of education. According to his view, only a fraction (around 20%) of the extra wages that are given to university graduates are a result of the extra “skills” that they have acquired during the course of their studies. Caplan maintains that the rest (around 80%) is all signaling. School does not make skilled workers, but it does make rich ones. What gives?

Before getting into the meat of Caplan’s argument, we should first consider what he is arguing against, namely, the human capital model (HCM) of education (specifically, what he terms “human capital purists”). According to this theory, education is positively correlated with prosperity by providing students with productive and valuable skills. To note here is that this argument is limited in its scope: in the first half of the book Caplan explicitly wants to argue that education, from the perspective of the individual does not make economic sense. The second half of the book is devoted to the more controversial claim that education, from a societal perspective, doesn’t make economic sense either.

According to Caplan, however, the HCM is too static in its core assumptions. For example, the HCM seems to leave out the fact that students regularly forget what they learn (pg. 39–40). Evidence for this comes from what is termed “summer slide”, which describes the loss of skills and knowledge experienced by students over their vacation months. In the USA it was found that students lose on average a full month of education during their time off. How exactly “human capital” is supposed to be accumulated in the minds of students who have heads like leaky sieves is therefore contested. In addition to this, there is also evidence that adults forget much of what they learnt in high school (pg. 39–50). Such forgetfulness, however, is not enough to take down the HCM. To do the job, Caplan has a few other arguments: the first is that schooling typically does not do what its advocates claim it does. Specifically, that it does not provide students with productive skills that they will then use in the labour market. His second point is the significance of “sheepskin effects”, which severely undermine education’s efficacy.

To the first claim: that the current education system is not about skill building but is largely signaling. If this model is correct then education does not significantly improve our skills but is basically a shortcut for employers to know whether potential candidates have desirable characteristics, making them suitable for the job. These traits are intelligence, conscientiousness, and conformity, as noted above. Basically, these traits tell employers that you are capable of doing long, boring tasks, without getting distracted. Moreover, you probably won’t rock the boat too much, and will be generally agreeable and not disrupt the functioning of the company by challenging authority. You have proved all of these things by spending (wasting, according to Caplan) years in an educational system that seemingly has the goal of turning you into a good little worker bee. Employers like this, and so with each degree you attain, you further increase the probability that you have these underlying traits (and that you don’t have other, more disruptive ones). In a sense, education screens you for potential employers, not for your skills, but for your ability to play nice with others and do as you are told. Moreover, you signal your ability to stick with tasks until they are completed, which brings me to Caplan’s second point.

Here he claims that if education really does transmit valuable skills, then we should expect to see that the number of years spent in education would, on average, result in equal pay for individuals. In other words, pay should positively correlate with the number of years spent in education. This, however, is just not the case. It turns out that people with academic degrees earn more than those with an equal amount of studying behind them, sans degree (pg. 97–102). This is what is termed the “sheepskin effect”. What is implied here is that most of education’s real benefit come in the final year(s) of study: graduation year is far more valuable than the years that come before it. If education is all about skills and knowledge, then surely one would expect the gains of each year to be consistent? A counter to this might be that for most programs the final year is the most difficult, and so the most essential skills are imparted at this point, and this therefore explains firms’ willingness to fork out more for employees that have completed degrees.

However, the data does not support such a claim. For example, Caplan finds that in the US, grade 9, 10 and 11 each offer a 4% increase in wages, while completing grade 12 raises wages by 16%. Moreover, he finds that the first three years of college each raise potential earnings by about 6%, while graduation years raises them by 30%. Even if the final year of study is the most important, it seems highly unlikely that the skills bequeathed in that fateful year are somehow four times as valuable as the years that came before. Employers basically rely on universities as glorified screening tests.

In light of all of the above, it should be simple enough to foresee Caplan’s conclusion: educational austerity, which proposes cuts to public funding for higher education. He claims that children should spend less time in school, and that certain subjects (particularly in the humanities, such as art, history and music) should not be encouraged. Moreover, he also argues that we should place more emphasis on vocational education, allowing students who do not excel in traditional forms of schooling should start training for jobs as soon as possible.

I find much of what Caplan had to say interesting, and the way that he presents his ideas is straightforward and quite easily digestible. That being said, I do have a few issues with the book. The first is a meta-question regarding the framing of the debate more generally: does it make sense to locate discussions regarding the value of education by focusing only on its use? This is what Caplan seemingly does, although he is even more restrictive, as it is not only use that he focusses on, but rather economic use. On this view it seems that the only way to measure education is by counting its economic costs and benefits. I think this misses the point. Education is not merely about transmitting “useful” content (although this is certainly one of its objectives). Rather, education is about giving students the tools so that they can learn how to think. Skills like being able to evaluate evidence, use logic, and apply ethical principles to practical situations are all incredibly valuable. The usage of some of these may even have adverse economic consequences.

Google’s recent unjust firing of AI ethicist Timnit Gebru illustrates this point. Gebru, a trailblazer in the field, who has previously argued that facial recognition is less accurate at identifying women and people of colour, and that the use of such systems may discriminate against these groups, was fired due to a conflict regarding a paper she was working on. The paper in question outlines the risks associated with large language models. These are AIs that aim to produce “convincing” text and are trained on massive amounts of data. The details are important, but I cannot begin to go into them here. I mention this case because it shows how economic incentives (“use” in Caplan’s case) at times need to be fought against, and that education can aid us in this regard. In Gebru’s case, she was uncovering some inconvenient truths about these large language model systems (at least from Google’s perspective). Subsequently, it seems that she was subsequently fired because the research that she was conducting could have impacted Google’s bottom line negatively. Of concern here is that we risk missing the larger impact of the value of education if the only measure we use in the evaluation of our practices is economic concerns. It might just be that education leads to negative economic consequences, but if the distributional outcomes are fair, then this is surely preferred. If such “bad” economic outcomes entail the wealthiest members of society losing their money to the most vulnerable, then all the better for humanity.

Now obviously this cannot by itself rule out Caplan’s claim that most of education is broadly useless, but it does at least contest it, and shows how some skills transfer during education is indeed possible. Restricting one’s analysis of these skills to their economic use alone is to misrepresent the full range of benefits afforded. Where I certainly agree with Caplan is that a comprehensive, all things considered evaluation of education would include the kind of economic analysis that he provides. Such an analysis, however, does not settle the issue.

In addition, I take issue with the potential inequitable consequences associated with Caplan’s proposal. He argues that any reduction in education spending should be welcome and celebrated, given that the greatest benefit it provides is “wasteful signaling”. There is reason to believe, however, that the downstream effects of such a policy would be structurally unfair. Discouraging students from pursuing education, and indirectly incentivizing less motivated students to drop out of school (through the higher cost of learning resulting from a lack of government assistance in funding) is advocated by Caplan, as he argues that university is not useful for these students in any case, and that they should seek vocational educational as a better suited alternative

However, such an outcome would discourage those who are economically less well off from further education. Caplan claims the poor suffer most from the current educational regime, and that educational austerity would cause the price of education to rise so much that we could return to a world where costly education would no longer be a barrier to success, as in such a scenario we would be able to focus on increasing economically valuable skills. I don’t buy this. I think that in a scenario of poorly subsidised and expensive education, the rich would continue to reap the benefits of the system, while the poor would be further systematically disadvantaged. Caplan therefore does not sufficiently address the distributional impacts of a reduction in education. There are already diversity issues in higher education, and it seems to me that educational austerity would only further exacerbate these.

I do think, however, that there is a deeper reason for this omission. My sense is that Caplan has a fairly deterministic outlook with respect to human nature. That is, he sees genetic inheritance as playing a major role in the economic success of individuals. For example, he claims that “the genes your parents give you at conception have a much larger effect on your success than all the advantages your parents give you after conception” (pg. 259). He also supports behavioural genetics, which he asserts “consistently finds strong, persuasive effects of nature, and weak, sporadic effects of nurture” (pg. 259). Thinking of human development in binary terms such as these is not productive (or correct), and it is therefore strange that Caplan indulges in such an argument.

However, the fact that he does so explains some of his views. His strong arguments on the relationship between genetic endowment and success lends support to his claims that underplay the importance of education. If it’s all in our genes, then education will be very ineffective at sorting the wheat from the chaff. Those who are “born with it”, will be successful, and those who aren’t, won’t. Simple. However, such deterministic presuppositions regarding human development are false. Mammals (like ourselves) show an enormous sensitivity to environmental influences, and these influences are mediated by our genes. A cooking metaphor will be helpful to elucidate this. When cooking, there are certain raw ingredients, which, when combined in a particular way, produce an outcome that is heavily dependent on how these ingredients are put together, when they are combined and what the raw ingredients are composed of. Furthermore, these ingredients must be cooked, and both the way in which they are cooked and the length of time they are cooked for play important roles in the final product. The raw ingredients represent environmental and genetic influences, and the cooking process represents the biological and psychological processes of development. Therefore, both genetic lineage and environmental influences are important parts of who we become, and over time we can cultivate developmental processes which can guide us in becoming better. In other words, our nurture will influence our nature, and vice versa. There is much more to be said on this topic, but I will leave it at this.

Then, lastly, I think Caplan has managed to get things almost exactly backwards. The poor educational outcomes that he cites as evidence for his austerity measures are exactly the result of too little investment in education. As Joshua Kim puts it:

“If Caplan is driving across a bridge that is poorly designed for optimal traffic, is riddled with potholes, and is potentially unsafe for drivers then I doubt his argument would be to knock down the bridge. Or to abolish all bridge building. Instead, Caplan would say that we need to build better bridges. Building better bridges may be expensive, but the alternatives (unsafe or no bridges) are far worse”

This is the current state of education. By investing more in education we ensure improved design and implementation of educational resources. Learning is difficult, and catering to each student’s needs more so. Caplan’s solution, to give up, is no solution at all. Tertiary education is far from perfect but cutting funding would only further increase inequality between those who can afford education and those who cannot. The proposals Caplan offers would strip the copper from the walls of our education infrastructure.

To return to the joke in my introduction: I am a philosopher by training, and while folk wisdom decries my field of study as nonsense, the data presents a different picture. Philosophy majors, between 2015–2018, had the highest number of standard deviations above the mean for the GRE test. Additionally, those with philosophy majors also have the highest mid-career salaries of non-STEM majors. There is value in philosophy. So even if you’re for educational austerity, you might want to keep philosophy programs going. Caplan’s contribution to the debate on the value of education is an important one: to my mind, it is a healthy sign when arguments which cut against the grain can be published. I find some of what he says provocative and interesting, but at the end of the day, I still think we should be spending more on education, not less.