Education has an air that seems to demand religious reverence. Who could be against it? Well, there’s Bryan Caplan, an economist at George Mason University. In his new book, The Case Against Education: Why the Education System Is a Waste of Time and Money, Dr. Caplan confronts the justifications for what he argues is a terrible investment.
Jacobite: In The Case Against Education, you make that the argument that the current levels of investment in education are a bad deal for society — that it’s a systemic problem rather than an individual problem. But what are some bad arguments against education, ones that you don’t make?
Bryan Caplan: The most common naive argument against education is that there’s this PhD that you personally know who works at a bookstore, and therefore education doesn’t pay. It is true that there are people who get an advanced education who are not getting a job, but all of the evidence points to education, on average, paying very well as long as you finish. And then the other big one is that colleges are these giant left-wing indoctrination centers. In one sense there’s something to it, because the humanities and social sciences are overwhelmingly left-wing. But on the other hand, the actual change in opinions that come from attending college tend to be very small.
Jacobite: So your thesis is that college signals something about the pre-existing attributes of an individual rather than imbuing them with new, useful attributes. Tell us a little about that.
Caplan: The main thing that’s going on, rather than gaining skills that are useful on the job, is a process of students jumping through hoops that result in diplomas that makes them stand out from the crowd.
Jacobite: You’ll often hear people draw the distinction between the humanities and “hard” technical disciplines. Don’t those fields make students better at useful things?
Caplan: I would say that almost every subject gives you the ability to do some things that you weren’t able to do before, but the difference is that some are likely to lead to a job and others are not. But what’s funny is that even with STEM majors that are technically demanding and highly vocational, like engineering and computer science, you still barely ever use what you learn in school in the job market. Like the stuff that you learned doing a math degree: nobody ever gets a job proving theorems unless you’re a math professor. One of the most lucrative things you can do with a math or physics degree is just to go to Wall Street, and then you’re not going to use much of what you learn. You do use math, but not the math that you spent a majority of class time learning.
Jacobite: What if we put public money into like STEM fields and defunded everything else? Even though most of the classes aren’t directly useful for a career, wouldn’t society be better off with more scientists and engineers?
Tough call. Eighty percent of STEM majors don’t actually work in STEM jobs, and most college students couldn’t hack STEM anyway. But maybe.
Jacobite: A core argument in your book is that society overvalues and over-invests in education by a very large margin. But if markets efficient and rational, how can this be?
Caplan: First of all, a trillion dollars in annual government subsidies goes towards education in the U.S. So the current system doesn’t pass the market test by any stretch of the imagination. Government is pouring money into one approach to education, year after year. In terms of employers incorrectly valuing education, I don’t think that they are doing that. What they’re doing is looking at the applicants that they have in the context of the subsidized system that we’ve got, and they take advantage of it, at least in a way, by using it to help select the right candidate. So I’m not saying that employers are making a mistake, and I wouldn’t advise employers to be more open-minded. If they’re more open-minded then there’s going to be high costs. You’d have to go and interview more people, which is a pain in the neck, of course.
Jacobite: It seems, then, that this is a problem that there aren’t individual solutions to this problem. You hear people saying things like, “college is a scam! Just go become a welder and you’ll be better off than those people who waste four years.” Is that misguided?
Caplan: It’s mostly misguided. In the book, I point out how we overvalue education from a social point of view. But I also do point out that higher education is a bad idea, as far as their own interests are concerned, for the subset of people who are unlikely to finish, like people who didn’t do well in high school. With so many people now going to college, the population of attendees includes people who were mediocre or worse high-school students.
Jacobite: There are capable individuals who want to have high-paying white-collar jobs but don’t want to be part of the education-credentials arms race. How can they opt out of this wasteful system without sabotaging themselves? Do all alternative strategies signal defectiveness?
Caplan: Alternative path usually do signal defectiveness, but you can try to work through cracks in the armor. At small firms, for example, you might be able to bypass the arms race if you have a close friend willing to vouch for you to the boss.
Jacobite: You’re in the camp that believes in the signaling theory of education. The opposing idea is the human capital theory of education, which holds that education makes people better at doing things. How dominant is the human capital theory? Why is it as dominant as it is?
Caplan: Within the sub-field of education economics, the human capital theory is overwhelming. Looking at Google Scholar citations, human capital beats signaling by a ratio of 100 to 1 or more. That doesn’t mean that they endorse it — I suppose someone could talk about something all the time without thinking it’s important, though that would be odd. If you talk about something a hundred times more often than something else, you probably think it’s a lot more important than the other thing. For economists specializing in labor and education, signaling accounts for a very small share of their explanation of things, maybe ten percent. But what’s really crazy is when you ask economists in general, they put the share of at one-third, or even half. There was a survey of econ bloggers, and it found that they had a much higher belief in signaling than specialists.
Now as to why the specialists are so dismissive of signaling, it’s not because they’ve done the work and proven it wrong, because there’s hardly been any work done on it. There’s been very successful papers published on signaling, but almost all of them are in the realm of pure theory. Michael Spence won a Nobel Prize for his work on this, but he’s got really no empirics. But I think people could be dismissing it for a bunch of reasons. In the minds of a lot of academics who work in this area, education is the salvation of the economy and the salvation of disadvantaged individuals. And it’s actually bipartisan, because Democrats can like education for egalitarian reasons and Republicans can like it as a meritocratic system. The sheer dismissal with which people treat signaling even though it makes so much sense to anyone who hears it makes me think that something else is going on.
Jacobite: Many people who support the conventional wisdom on education will argue that, more than teaching any specific skill, schooling “teaches students how to learn.” How much truth is there to this?
Caplan: In general, not much. Based on recent studies on educational psychology, learning one thing has almost no benefit for learning an unrelated or distantly related subject. With closely related subjects, yeah, if you practice one type of writing you’ll probably improve in other types of writing, and practicing one type of math will probably improve your abilities in other types of math. But even that’s hardly a given; you probably know a bunch of mathematicians who screw up arithmetic.
The psychologist Edward Lee Thorndike heard countless people defend the teaching of Latin in school by arguing that it orders the mind for other kinds of learning. This is an important example because people taught the Latin language based on the theory that it’s a uniquely logical language. Thorndike compared the performance of students who took Latin that of ones who didn’t, and it turned out that the theory was completely wrong.
Jacobite: Steel-man your opponents for a moment. What’s the absolute best argument against the signaling theory of education?
Caplan: I debated education economist Eric Hanushek a few weeks ago, and he’s probably got the best argument against mine. He concedes that throwing money at education isn’t good and that levels of education doesn’t predict much about national prosperity. But he makes the point about the importance of test scores, particularly math and science test scores. And so he argues that the best case for education is that if it’s not working well now, we should redouble our efforts and find effective ways to get math and science test scores up.
Jacobite: A lot of folks will argue that tech, perhaps the country’s most productive industry, is more meritocratic and less credentialist than the rest. You’ll hear about wealthy companies hiring people who don’t have a degree but win a coding competition. Is the high-tech industry able to cut through this wall of signaling in ways that other industries aren’t? Are they onto something?
Caplan: I don’t have the data — that’s probably proprietary — but the people who I talk to in the industry really give the opposite story. People who get hired at places like Google have degrees from top schools with great grades. While it’s true that they hire people through these unconventional channels, that’s a tiny minority of the workforce. I was speaking with one person at Google — and keep in mind that this is from memory and this is hearsay — and I asked what fraction of people hired at Google had conventional degrees and what fraction did things like win programming contests. And he said that under one percent were hired through the non-conventional avenues. So you have to be completely awesome to get hired without having to jump through the usual hoops.
Back in the 70s and 80s, tech companies were more open-minded about hiring people without the usual credentials, but they became more credentialist over time. Back in the 70s and 80s, there were so many qualified people that hadn’t gone to college that it would stupid not to consider then. But now, there’s very few people who would be great programmers who didn’t come from regular families who would insist on them going to college. So there just isn’t that much of a loss anymore from being credentialist. What I say in the book is that if there’s a lot of great diamonds in the rough, you want to be able to mine them. But if there’s just a few, then your time is too valuable to worry about them all that much. If you say, “in this stack of 100 uncredentialed resumes that you’re throwing away, there’s going to be five great workers.” Yeah, but then you have to interview 100 people. Then you might find out that some of them aren’t any good after you hire them. You could find some good ones, but at what cost?
Jacobite: Fifty years ago or one-hundred years ago, we had much lower rates of college attendance. Would you argue that a system of lower college attendance is a better one?
Caplan: Definitely. You could get jobs back then with a lot less education. Over time the economy has changed, and there are more intellectually demanding jobs, but that only explains about a fifth of the rise in education. The other four-fifths is that more people are getting more education, and so to keep up, you need to get more, too. It means that now to be a waiter or a bartender you might need a college degree. There’s little evidence that we’re any better at these jobs than we were — these are all jobs we learn by doing. Yet if you’re running a fancy restaurant, there’s a big pool of college graduates that you could hire. Fifty years ago if you would only consider waiters with college degrees, good luck running a restaurant. As I say in the book, the more educated the workers are, the more they need to be considered employable.
Jacobite: A common argument in favor of education has nothing to do with employability, but argues that schooling creates “well-rounded individuals.” People making this argument say that education isn’t supposed to be corporate trade-school that gives you skills for a specific career path. Isn’t it socially desirable to have a populace that understands civilization?
Caplan: It isn’t clear why employers would want to pay for that. Business is business, it’s not a grand civilizing effort. If schools said, “by the way, this won’t help you get a better job,” there would be a dramatic fall in the number of students who go. There’s been some good surveys asking college students why they attend, and by far the leading reasons are “to make more money” and “to get a better job.” This is striking because almost everyone will downplay their materialism, and yet almost everyone will admit it.
The aspiration toward knowledge and culture for its own sake is noble, and I share it. But just having good intentions doesn’t mean you’re accomplishing anything. If you say that we’re trying to instill an appreciation of civilization, then great: let’s go and see how much appreciate for civilization college graduates typically have. There’s a lot of kinds of data on this saying that the upper bound on how much civilization-instilling schools accomplish is next to nothing. Like most people, college graduates have near-zero knowledge of history, civics, philosophy, and culture. Maybe they know it for the final exam, but it doesn’t stay with them, so what good is it?
Jacobite: But it’s not just college. You question the social value of secondary education.
Caplan: Individually speaking, high school is a great deal. At that age your job options are pretty limited, so you’re not giving up much income by going to high school. And of course it’s free, so that’s another big plus. But socially speaking, I say that if you really look at the curriculum, most of what students have to do in school is stuff that they will never need to know again. If you’re the kind of person who absolutely does not like, you resent it, it would be better to go onto something that you like or are good at. That’s why I have a chapter on the need for more vocational education. Remember, there’s a lot of students who never write articles or books about their experiences in school, who thought it was the most pointless thing. Kids like this, especially boys, tend to drop out of school and into jail. If we could find something else that they could do that would make them independent adults, that would be great, and vocational education seems to fit the bill.
Jacobite: It doesn’t seem like secondary education effectively imbues students with civic knowledge. Very few American adults know that U.S. Senators used to be appointed by state legislatures, for example. What kind of system would work to make people understand such issues in a sophisticated manner?
Caplan: That sounds pretty much impossible to me. But we could do a lot better. Instead of bothering to teach this stuff in school, my proposal is to have a national civics test three weeks before every election. If you take it and do well, you get money! This serves two purposes. First, it encourages people to learn by whatever means are most effective for them, and sometimes that isn’t sitting in a classroom. But secondly, it encourages retention. There’s overwhelming evidence that if people learn something and don’t use it afterwards, they forget it. Having a yearly check for passing a civics test would be a great reason for people to actually retain the information.
Jacobite: What’s your 20-year prediction on trajectory of education in the U.S.?
Caplan: I don’t think there will be much change. It seems that there will be a modest retrenchment in higher education. While I don’t think that alternative credentials will destroy the system, they may pare off a modest share of people who will take another path. There also probably is a growing awareness among people who are unlikely to finish that it’s not a very good deal. With law school, for example, there was a big crash in enrollment because students realized they weren’t getting jobs. But the main thing I think will happen is that the government will keep pouring money into the system and it pretty much stays in place.
Jacobite: What country do you think has the best contemporary education system?
Caplan: I haven’t done a full study of every country on Earth, but my tentative answer is Switzerland. It combines two things: a low percentage of people going to get bachelor’s degrees, and a high percentage of people getting a vocational education.
Jacobite: What is the best prospect for moving toward a less wasteful system? Is it political reform — activism, lobbying, and trying to get the right candidates to run? Or is it building alternative ways to credibly signal value?
Caplan: What’s most likely to work, all things considered, is creating alternative systems for credentialing people. The system is very stacked against these unconventional arrangements, but still they’re happening. And the good thing about building alternative systems is that you don’t have to persuade a majority of people that they’re a good idea. Cutting government spending would be much more effective if actually done, but it so goes against popular emotions that I’m not optimistic about it happening. Possibly, there will a move for austerity in education if it loses out to an even more popular demand like old-age spending.
Jacobite: What do you say to skeptics of the signaling theory of education?
Caplan: Imagine if you could either have a Princeton education with no diploma or a Princeton diploma without the education. Which would you prefer? If you have to think about it at all, you already agree with me. This means that you realize that a diploma has a lot of value in terms of career opportunities independent of whatever you’ve learned. This is in direct contrast to another thought experiment. Suppose you were on a desert island: would you want knowledge of boat-building without a boat-building degree? You would absolutely want the knowledge. But in a modern labor market it’s not clear which one you want. And if it’s not clear, then you must think signaling is pretty important.