Signaling, Not Learning
A review of Bryan Caplan’s The Case Against Education
“We don’t need no education”
[Pink Floyd, “Another Brick in the Wall, Part 2”]
As an undergraduate, I always crammed for exams and rarely remembered anything I may have “learned” beyond the next week. I certainly did not remember much after a several months, not to mention after a year or two. So long as I received the grade I was shooting for, I had succeeded.
More than four decades of teaching at the university level convinces me that my experiences are no different from those of most students. This observation is at the heart of Bryan Caplan’s new book, The Case Against Education: Why the Education System Is a Waste of Time and Money (Princeton University Press, 2018).
The major point of Caplan’s study is that high school diplomas and undergraduate degrees generally do not show that the students actually learned and retained much that will be useful after graduation (those who go on to teach are, to some extent, exceptions). Rather, completion of high school or the BA signals a student's ability, intelligence, work ethic, and willingness to perform to satisfy teachers’ requirements.
Caplan reviews the compelling evidence that even after taking account of ability, intelligence, and a host of other variables, those students who complete high school on average earn more than those who don’t, and those who complete a BA earn more than those who don’t. Completion of a programme clearly means something in the job market to potential employers and is valuable. The puzzle is: why do those students earn more even though they don’t remember, much less use, most or any of what they studied?
Caplan’s answer: The very fact that they completed high school or received an undergraduate degree is worth a great deal to future employers because completion conveys valuable information about students’ character traits and general abilities, despite telling very little about specific skills and learning. Students spend years and years (and years and years and years) studying things they don’t remember and never use. Why? To generate signals to potential employers that they have what it takes to be productive employees; or, in the case of high school students applying to college or university, that they have what it takes to be successful undergraduate students.
I have been reading these ideas from Brian Caplan for nearly a decade, as he has posted them at the blog Econlog. I was so persuaded by them that several years ago I gave a public lecture at The University of Regina “I Didn’t Learn a Thing as an Undergraduate.” in which I summarized his ideas.
When I first started teaching, every year there was a student or two who would plead for a higher grade, and my reaction was usually, “I won’t certify that you know this material when you don’t.”
Even then, I knew I was in the business of certifying something, but I thought I was certifying the students’ knowledge of economics. According to Caplan, though, I wasn’t – instead I was certifying their intelligence, their ability to buckle down and study enough to get a certain grade, and their ability to “learn” (at least for a brief period) whatever it was I wanted them to learn. I was certifying their ability to generate signals about character traits that employers and educators believe are valuable.
Caplan questions why we need so much education if the primary purpose of education is to certify and to generate signals, but not to foster learning. He points to all the language courses students are forced to take but forget quickly, to the history courses despite which so many people know so little basic history, and even to the math and science courses that so many students take in high school only to signal their suitability for university. If the primary reason students take these courses is to signal character traits, not skills or knowledge, then it seems mighty wasteful for them to be spending so much time and energy on coursework. Surely students would be better off if they spent more time learning skills and less time signaling.
Let’s face it: you don’t need a BA for many, maybe even most, of the jobs that require them, but employers require BAs because students with a BA signal that they have more ability, better work ethics, and greater willingness to conform to the employers’ work standards. The BA hasn’t taught the students anything useful for the job, but it has signaled character traits that many employers are willing to pay for.
While that might be correct overall, it likely does not hold for any given individual. A reasonable student can rightly ask, “How can I send a reliable message to potential employers and universities about my abilities and personality if not by completing my education programme? Doing so is next to impossible.”
As a result, the student quite sensibly decides to signal quality by getting more education. And so do zillions of other students, all contributing to a signaling war with each one trying to send a signal that they are better-suited --- i.e. better credentialed --- than other job or college applicants.
The result is an attempt to generate better credentials than the others have who might be competing for the same position. The students quite properly say, “I know what employers want – diplomas and degrees. If I want to get the job I need a diploma or degree.”
Caplan’s conclusion is clear: all this signaling is costly and wasteful. He sees it as what economists and others refer to as a negative sum game: Everyone devotes more and more scarce resources (time and money) to advancing themselves and/or their progeny in the credential war, trying to create more impressive (albeit more costly) signals.
High school students take courses that signal their willingness and ability to submit to the rigours of college or university coursework. They also join extra-curricular activities to prove they are well-rounded; they participate in or even start their own charity fund-raisers to show they care about others; and they vie for “leadership” positions in local groups to show their leadership potential. Many sign up for SAT training courses or buy books with sample questions and guidance on how to improve their SAT scores. These are all attempts to signal. They have little to do with producing what economists call “human capital”—i.e. skills and useful knowledge.
In my own case, in high school I had a friend who had some SAT study books. I had never heard of studying for the SATs. I naively believed the SATs were truly aptitude tests. But we studied the books, and I felt they helped me increase my SAT scores.
Also, in my last year of high school, I had read that someone who restored an old MG was regarded highly by Yale. As a result, when I was visited by an admissions officer for an elite liberal arts college in the US, I made a point of mentioning that I had built the hi-fi system I had. She seemed impressed, but I don’t know that it mattered. The point is that apparently even then, nearly sixty years ago, I was well-aware of the importance of signals.
Later in my studies, after having been an abysmal student at Carleton College, I was speaking with the admissions director [then-famous labour economist, H. Gregg Lewis] for the economics graduate programme at The University of Chicago. He told me he wouldn’t admit me, despite the lax admission standards there, because I had such a low grade point average from Carleton.
“The best predictor of someone’s likely success in our graduate programme is their undergraduate grade point,” he told me. “The next best predictor is my personal impression from their letters of recommendation.”
“But,” I argued, “I have very high scores on the mathematics and economics portions of the Graduate Record Exams [GREs].”
He said those were horrible predictors of success in graduate school. His predictors were signals. And my low grade-point signaled to him that I would not be a success in the graduate programme at The University of Chicago. If I had wanted to signal my ability to succeed there, I should have buckled down and gotten better grades as an undergraduate.
In many ways, signals are good things. They aren’t perfect, but they are short-hand ways of conveying generally useful information. When someone completes high school, they signal they have the mental ability to grasp the basic material, and they also have the personal ability to withstand boredom, to do as they are told, to jump through hoops. This is an important signal to potential employers. Further, if they get high enough grades and high enough SATs, they signal they probably have the mental ability and the personal characteristics to finish an undergraduate degree.
Similarly, completing a BA tells potential employers the student has even more ability intellectually and more stamina and conformity.
Caplan asks how many college and university graduates actually use the things they studied as undergrads. The answer, overwhelmingly outside certain professions, is very little.
As an example, I know one person who majored in English literature with a minor in Religious Studies at a top-ranked university. He ended up getting a job with a major corporation only because he had become fluently bilingual in French on his own. He has since gone on to head up their business-to-business web sales programme, despite studying zero business or computer science in high school or university. When I asked him how much of his schooling he had ever used on the job, he answered “Four percent. Writing all those essays helped me learn how to organize and structure my time.”
Put differently, his university credentials (coupled with his ability to speak French) signaled a set of abilities valued by the corporation. They indicated a strong probability that he would be highly productive.
This story is repeated everywhere. Students who work to earn diplomas and degrees signal things that are valued by employers, no matter what subjects they study.
Caplan goes on to suggest that from a societal perspective there must be a more efficient way to generate these signals. Having people spend, say, 16-17 years in school to create these signals seems like a very expensive way to do it.
Indeed, some of his critics have argued that if there were a more efficient way to acquire this information, employers and intermediaries would have developed mechanisms for doing so by now. Caplan’s first response to this is weak: he responds that the signals from education are so strongly embedded in our culture and psyches that we don’t trust alternative signals. If that is the only reason, I expect things will change, and fairly quickly over the next few decades.
However, his second response has more strength: we have had a credential explosion as students (encouraged by their parents and by educators) scramble and slave to create better signals to make themselves more attractive to college admission officers and employers. A generation ago, people could and did do jobs with only high school diplomas but for which employers generally seem to require BAs today. People don’t need BAs to do those jobs, but having a BA signals things that employers want: intelligence, solid work ethic, and conformity.
Caplan argues that these signals are relative, in comparison with the applicant pools, and not absolute. If they were measured according to some absolute scale, then a high school diploma today would signal the same thing it did a generation ago (assuming high school standards haven’t changed much; however, see below). But when an employer is faced with a job applicant with stronger signals, like a BA, then the one with the BA tends to win.
According to Caplan, there are two big problems with this credential war that is set off when we use schooling to signal ability, work ethic, and conformity. The first is that the war, like many wars, is a negative sum game: it is beneficial for each individual to play the game but it wastes society’s scarce resources when everyone plays the game. It is to each student’s advantage to get a high school diploma (or to complete a BA), but only because they are competing with other students who are also spending time and money in the credential war.
Not all signaling wars are bad, though. For example, when firms advertise just their company names, they are signaling that they intend to have a good enough product that they wouldn’t want to have wasted money on the general, non-specific advertisements [see, for example, Klein and Leffler]. This type of signaling war encourages firms to produce reliable products and the competition is beneficial to all of society. Caplan argues, though, that using schooling to signal abilities and character traits has no social benefits and hence generates far too much waste, with students taking courses they don’t want and don’t need, only to create acceptable signals.
I like Caplan’s case that signaling is a major part of schooling and education. But there are some positive aspects to the signaling war that Caplan either ignores or dismisses too easily. The longer students are in school, the more they tend to learn and practice good work habits and conformity to teachers’ and professors’ assignments. Caplan’s response is that they can learn and demonstrate these same skills on the job if they go into the workforce; they don’t need to spend so much unproductive time in school to learn and demonstrate these skills. My sense is that the two are not the same, but Caplan does make a strong case.
Interestingly, throughout much of Canada students applying to university can pretty much blow off their time in the lower grades, even in high school. All they need is a high overall average in six different grade twelve courses [see the details here for Ontario]. What is more, they can retake courses to raise their averages, and apparently admissions officers tend to just look at the averages of the best six, as generated by a computer programme.
Here the important signal is grade point. Essays on “why I want to attend your university” aren’t required, nor are SATs or records of charitable work. And so at least some of the wasteful efforts in the signaling war are less valuable here.
The credential war has another pernicious effect however: it induces grade inflation. High school teachers feel pressure to ease up just a bit so students who might be borderline are admitted to college or university. Similarly university professors are under pressure from students to be lenient so that students can get into professional schools. And department chairs put pressure on the hard-ass professors to ease up so the students don’t migrate to other departments to take their courses.
There are numerous true stories about grade inflation and signaling. I’ll share a couple.
When I moved to Canada after being an instructor in graduate school at Iowa State University, I was informed that 80 is an A here and 50 is a pass. In most schools in the US, 90 is an A, 80 is a B, and you need a 60 to pass. The result was that I just adjusted my marking scheme. The numbers were far from absolute.
During my first few years, I had no qualms about giving marks that ended in a “9”. Several years later, a decree came from the department chair that we were no longer to give marks ending in a “9” because students were appealing them and chewing up too much important administrative and faculty time. We were to round everything ending in _8.5 to the next highest grade. That edict cut down on appeals considerably … initially…. And even in the long run to some extent. But of course many students soon figured out that all they needed was a 78.50 to get an 80 (an A), and so there were still appeals, just not as many.
For decades, the University of Western Ontario economics department had a rigid policy that in first year courses there was to be a set grade distribution according to guidelines we all agreed to. For the most part, these guidelines were strictly enforced to maintain inter-section equity. The guidelines were comparatively difficult for the students, though. As a result, enrolments in economics began to tail off over the years as students sought out easier courses in other departments that were more likely to give them the averages they needed to get into a top-ranked business school, law school, or other professional programme. Eventually the economics department caved in to the grade inflation taking place elsewhere in the university and eased up on its grading guidelines. Students wanted to generate grade signals, and we were under pressure to contribute to the signaling and credential wars they waging by being more lenient ourselves.
If students are increasingly studying less, as seems to be the case, and receiving higher grades, then the signals are being diluted. To that extent, students who want to impress employers and admissions officers need to generate additional and stronger signals. Many will choose to further their studies. Others will look for other signals. For example, students seeking admission to top business school programmes now are expected to do something to demonstrate their entrepreneurial drive and talents such as starting a business or organizing a fund-raising campaign. When I talk with these students, it is often the case that they don’t really want to do these things, but “it’ll look good on the application forms”; i.e. it’s part of the signaling war as they try to distinguish themselves from all the other applicants who have good, inflated grades.
Even if you are persuaded by Caplan’s arguments that schooling is mostly signaling and doesn’t really contribute much to the skills and talents of most students, you may be less persuaded by his recommendation for dealing with the high costs of signaling wars.
His solution: eliminate taxpayer support of education. As a life-long educator who has enjoyed feeding at the trough of the public fisc, I cringe at this recommendation. I am especially reluctant to endorse it for grades K-12. At the university level, though, it makes some sense to at least consider reducing the sizes of the government subsidies. I realize my experiences may not be generalizable, but they probably are: there are far too many students in university who will learn very little and use next to nothing of what they were taught. They are there, in part, because governments subsidize their participation in the credentialization wars.
Caplan’s book is an easy read. He writes with serious humour, if that makes sense. I have smiley faces in the margins all through the book. My major complaint is that the publisher chose to use endnotes instead footnotes, meaning I was constantly flipping back and forth to check the footnotes, many of which are well-worth the effort.
I certainly urge those who might be interested to read the introductory material in Chapter One. After that, Chapters four and five become tedious --- necessary to bolster his case, but tedious, as he develops and carefully references his case numerically. Also his conversations near the end of the book are interesting. But for those who want the Readers’ Digest version of the book, I highly recommend Caplan’s column in the Los Angeles Times, which presents his case extremely well.
[Addendum: if high school dropouts quickly learn that there is great advantage to having a diploma, I strongly suspect that many of them outright lie and say they have a diploma, fully expecting that potential employers won't bother to check. If so, then the signal for having a high school diploma has likely been seriously distorted and weakened.]
Note: a condensed, edited version of this review will appear in the April edition of the newsletter of the Society for Academic Freedom and Scholarship [SAFS], a Canadian organization created to foster and defend academic freedom.