Educational Signaling: A Fad Whose Time Has Come
Noah Smith, writing for Bloomberg view, isn’t happy with the educational signaling “fad.” Here’s my point-by-point reply. Noah’s in blockquotes, I’m not.
Talk to economists, and you’ll find a large number who believe that
college — that defining institution of America’s privileged youth — is
mostly signaling. It makes sense, after all — don’t most people go to
college because they think it will get them a job? And honestly, when
was the last time you actually used any of the things you learned in
college at your job?
It does indeed “make sense.” And according to the only survey of which I know, economists in general are at least open-minded about educational signaling. But Noah strangely neglects to mention that empirical labor economists in general, and education economists in particular, rarely engage the signaling model. Human capital purism really is their dominant paradigm – and they studiously ignore the “when
was the last time you actually used any of the things you learned in
college at your job?” argument.
Possibly the biggest promoter of the signaling theory of education is George Mason University’s Bryan Caplan. Caplan believes so passionately in the model that he’s writing
a book about it, called “The Case Against Education.” He has already
written enough blog posts on the topic to make a small book!
Caplan’s message is bound to appeal to people who dislike the
institution of college, whether because they think it’s too politically
leftist, or they’re worried about high tuition and student loans.
There’s something to this. It’s worth pointing out, though, that many defenders and fellow travelers of the signaling model are left-leaning sociologists.
there are some big holes in the case. Caplan’s GMU colleague, Tyler
Cowen, is rightfully skeptical of claims that college is mostly signaling. Let me add my voice to the skeptical chorus.
First, intelligence isn’t that hard to spot. Performance on any
mental task — a standardized test or even a high school chemistry class
— will give an employer an immediate general idea of ability. No need
to waste four of your prime years on a signal that can be generated in
I’ve said this. I’ve also criticized the many economists who think Supreme Court rulings on intelligence testing are a major cause of the rising educational premium.
So is college a way to signal conscientiousness and willingness to
work? Maybe. But an even better way to signal that would be to actually work at a job
for four years. One would think that if young people needed to do some
hard work to signal their work ethics, some companies would spring up
that gave young people real productive work to do, and provided evidence
of their performance. Instead of paying through the nose to send a
signal of your industriousness, you could get paid. But we don’t see
Like most economists, Noah needs to be more sociological. In a cultural vacuum, working four years might be a great signal of work ethic. But no human being lives in a cultural vacuum. We live in societies thick with norms and expectations. And in our society, people with strong work ethics go to college and people with bad work ethics don’t.
Disagree? Just picture how your parents would react if you told them, “I’m not going to college. I’m just going to get a job.” In our society, your parents definitely wouldn’t respond, “That makes sense, because you’re such a hard worker.” Why not? Because in our society, most hard-workers choose college. If a hard-working kid refuses to copy their behavior, people – including employers – understandably treat him as if he’s lazy. Because lazy is how he looks.
Noah overlooks another key trait that education signals: sheer conformity to social norms. In our society, you’re supposed to go to college, and you’re supposed to finish. If you don’t, the labor market sensibly questions your willingness to be a submissive worker bee.
When you think about it this way, the whole idea of
college-as-signaling becomes a little absurd. People’s careers last for
35 to 45 years. But after you’ve been working for a while, prospective
employers can look at your work history — they don’t need the college
signal anymore. Caplan’s theory therefore is that many young people are
spending four years — and lots of tuition money — on something that
will only affect the very beginning of a career.
As I’ve explained before, getting your foot in the door may seem like a small step, but it’s invaluable nonetheless. Without the right degrees, it’s extremely hard to even start building an impressive work history for employers to judge. Furthermore, if Noah were properly sociological, he’d know that employers are quite slow to fire employees whose formal credentials overstate their job performance – and even slower to publicize their negative evaluations to the broader labor market. Economists may balk at the idea that a mere credential can durably raise your earnings, but everyone long saddled with incompetent co-workers says otherwise.
There are many other reasons to doubt the signaling theory of
college. A more likely explanation for college’s enduring importance is
that it provides a large number of benefits that are very hard to
measure — building social networks, broadening people’s perspective,
giving young people practice learning difficult new mental tasks and so
I’m glad to hear this. Noah inadvertently grants one of my key points: Most of education’s labor market payoff is unrelated to the material your professors explicitly teach you. Once you accept this heresy, you’re stuck with some combination of my multidimensional signaling story, and Noah’s amorphous, evasive “large number of benefits that are very hard to measure” story. If that’s the choice, my story will end up with the lions’ share of the mix. Noah is welcome to the leftovers.
Final challenge for Noah: If education’s rewards stem from this “large number of benefits that are very hard to measure,” why on earth would the payoff for graduation vastly exceed the payoff for a typical year of education? My explanation, of course, is that given the vast social pressure to cross educational milestones, failure to graduate sends a very negative signal to the labor market, leading to discontinuous rewards. What’s Noah’s alternative? Do schools really delay “building social networks, broadening people’s perspective,
giving young people practice learning difficult new mental tasks and so
forth” to senior year?
Apr 28 2015 at 1:58am
In countries with more meritocratic elite university admissions processes, e.g. India, China, Japan, Korea or France, a big part of the value of the degree stems from the fact that it certifies you came Nth out of the entire eligible population in some tough competitive entrance exam, not a pathetic multiple-choice joke like the SATs or GMATs.
That’s also true of a few US universities like Caltech where admissions are not distorted by legacies, donors’ offspring or racial quotas. The quality of the education is almost secondary to the informational value of the selection filter. The flip side is that students who worked really hard to get in can sometimes coast after that.
Apr 28 2015 at 3:21am
Bryan, I agree with you that in today’s world there are good reasons why leaving college is a reliable signal of laziness/nonconformity. However, this doesn’t explain why this is a *stable* equilibrium, and so doesn’t seem to rebut a (steel-manned version of) Noah’s objection: Why don’t alternate, less-wasteful methods of a signaling conscientiousness and conformity start appearing? For people who are just below the level where they could graduate from college, why haven’t employers offered internships that test employees similarly (but at a lower level) as college, but also produce valuable output? Instead, community colleges have exploded.
Your points like the sheepskin effect and the non-applicability of college material are quite striking, but any theory that simply posits a signaling equilibria without showing why it’s stable attractor is incomplete.
Apr 28 2015 at 3:28am
To more rigorously expand on what Jess Riedel said:
One thing that has always bugged me about the signalling model is that it fails to explain the lack of arbitrage occurring. (Of course it could be occurring and I haven’t searched hard enough to find it in which case, someone please link!)
Presumably if education were largely signalling students and companies are overpaying for human capital. The former claim is obvious, but the latter stems from the student, immediately after obtaining a degree, commands a much higher demand for his services than someone from, e.g. an education center set up by a company. Thus the company must pay this graduate a premium simply because they use the same signalling requirements as everyone else.
It follows that there exists an arbitrage opportunity for companies with very specific human capital requirements. Set up education centers or contract with firms who will provide education, students then pay a lower tuition, learn for X years, then get paid less once they “graduate”.
Alternatively just have prospective employees take low-cost-to-administer tests which cover everything from cultural background to work ethic to industry specific human capital. Really any signal that differs from the signal used by everyone else so one can obtain a quality worker at a lower price. For-profit college degrees are another viable signal.
There would be little reason to go to “traditional” college other than as consumption as soon as this arbitrage became commonplace. The converse is probably true as well: if the necessary infrastructure for arbitrage does exist and individuals do not take advantage of it, then this indicates that the non-human capital inducing parts of college are consumption. This makes the signalling argument moot.
Apr 28 2015 at 3:38am
My first job out of college was working as an investment banker. In my five year stint, I never met a single person who did not have a college degree, which is of course compatible with either explanation. I did, however, meet people who did not have degrees relating to finance or our industry. And since all dealflow came from partner relationships (that’s what they’re there for, after all) I don’t see how broad social networks could have come into play at the analyst level. At some point you have an “education of the gaps” thing going on. If you can point to 10 things one supposedly learns or acquires in college, and every measurable or determinable aspect turns out to not be a bona fide prerequesite, it makes me skeptical of your overall theory of educational value.
On top of that is the fact that many people in the industry specifically believe the signalling model of education, and hire based on that fact. It would be very odd if people thought they were hiring from ivy leagues because of the signal it sends to potential clients, but it turns out the reason they were actually hiring them is because of the indelible mark the school left on their character.
Apr 28 2015 at 3:51am
I also thought about the arbitrage point. I think though that you do not even need tests to do it. There are very strong predictors of final college performance after 1.5 years (top GPA, participate in all investment banking things, doing spring internships etc).
Of course if all firms hired based on those observations, people could cheat and work hard for a year like Bryan posits. But if I’ve thought it through correctly the equilibrium requires that there should not be *any* good proxy at year 1 in college for performance in year 4, as hiring on the proxy will dominate hiring after 4 years (as you can offer a lower salary to compensate for tuition and foregone earnings).
Bryan simply says that if you wait for people to apply then non-college gradautes might be lemons. But if the strategy is to seek out the most promising people in year 1 and offer them a job, you need a much more complex explanation of them turning it down unless they’re not conformist enough. This explanation would also require statistics on how many in year 1 with top grades, loads of extracurriculars etc actually are lemons – and my suspicion is that they are rare.
Apr 28 2015 at 5:23am
“Because in our society, most hard-workers choose college. If a hard-working kid refuses to copy their behavior, people – including employers – understandably treat him as if he’s lazy. Because lazy is how he looks.”
I don’t know what to make of this paragraph. A more poorly written, unqualified statement rarely makes any blog, much less Econlog. Noah should counter-file.
Hard workers choose college? Does it hold that those who go on to achieve advanced degrees are the hardest working of all?
My experience in the work force has been the opposite. Those with advanced degrees rely on their credentials to do less work, presumably on the notion that; “I’ve already done the work, I’ve a gone the extra mile, I have nothing further to prove.” Whereas, the worker whose labor and performance is his only credential has to prove himself every day. That was how the labor market worked when I first entered it in the early 1980’s. College degrees were less common. You could still rise through the ranks, and the measure of performance was reliable output and dedication.
Hard workers choose to work hard. A good work ethic has nothing to do with schooling.
Apr 28 2015 at 6:20am
I think you’re focusing on the degree too much. The goal of an employer is to find the optimal trade off between the best-fit worker for the job and his salary. None of that requires, a priori, completing college, regardless of the job. So a signal that predicts the completion of college is a signal proxy for another signal that signifies work-ethic, etc.
The first 1.5 years is basically all signalling anyway, since the first 1.5 years is when many non-TEM students get their GEs and intro courses out of the way as they find out what interests them.
Now that I think about it though a much bigger problem arises if signalling does not exist: If signalling does not exist then nearly everything someone learns or does in college is necessary for the job. How does this model explain the heterogeneity of educational backgrounds within certain professions?
Apr 28 2015 at 9:13am
I think this is where Caplan is at his strongest, combining empirical data with common sense.
In response to some of the other commenters, I think you probably should be looking for cases of arbitrage. Peter Thiel seems to have a good idea going, and I’m sure there are other examples. One thing about arbitrage is that it takes pretty smart people to find it. The vast majority of people, including employers, will go with the flow.
Apr 28 2015 at 10:46am
I work in Canada as an Air Traffic Controller. I dropped out of college after my first year because I was taking English and while I greatly enjoyed the subject, it was clear that it was becoming overrun by very boring identity politics, feminism etc. It was also clear that I could spend years studying it and likely still be working in a coffee shop. ATC was brought to my attention after my brother applied and succeeded at it. There was no degree necessary just write an exam (at a cost of $200) that tests your spatial reasoning and a few other traits and if you scored high enough you would receive an interview. After interview they required two fairly extensive references from former employers and a security background check. Once approved for a “training opportunity” you would be asked to report to the Training Center and commence studying. 8 months of education and simulator training in a class of 16 of which 8 of us graduated (receive less than 80% on any exam and you’d receive one chance to redo. Fail again and you’re out). That was 8 years ago and I’m still doing it, making about 90K a year. Looks like the FAA in the U.S has a similar approach:
Apr 28 2015 at 11:05am
Or, alternatively, the signaling model of education matches up well with their personal experience.
Apr 28 2015 at 11:44am
“It’s better to be a bum with a college degree then a bum without one.”
My self-described communist aunt discussing my future 25 years ago.
“All college does is prove you can be trained.”
My first employer when I was 15.
Apr 28 2015 at 1:58pm
Noah’s comment, to be as uncharitable as he is, is absurd.
The first white collar job is essential. Without a college degree you may NEVER get a shot at one. And college is a pleasant way to break into one.
Looking at my work as dishwasher would not tell employers much about my ability to write computer programs, though it could tell how hard I work.
College also filters for IQ. I have heard that the average college grad has an IQ above 115 where 100 is the average for the population.
College also signals interest in a subject. To non-academically inclined people (I am thinking of my son he plumber) much white collar work is extremely boring so it is good to start with someone interested in the subject.
Apr 28 2015 at 3:11pm
According to Caplan, graduating from college signals one’s intelligence, energy, general knowledge, conscientiousness and, above all, conformism. And why is graduating from college chosen as the means of signaling, in preference to other, less costly means? Caplan’s answer: because that’s *our social norm*.
So far, so good. But one can always ask another ‘why’-question, such as “Why is that our social norm?”; and if Caplan provided an answer, delving into U.S. social history, his case would be even more convincing than it already is.
But, of course, at some point we’ll have to leave off asking more ‘why’-questions.
Apr 28 2015 at 4:08pm
I would have to agree that educational signalling is becoming an important hiring metric in some fields, paradoxically because of the increasingly poor work ethic displayed by corporate human resources departments and the division management served by those H.R. personell.
In the same fashion that “free” trade agreements substitute for dynamic and effective business development practices, signalling paradigms allow the intellectually bankrupt to allow statistical modelling to take the place of competence in hiring practices.
Rather than sift through the application information to find the best fit for the work, Management prefers to use irrelevant statistical data to place workers in positions for which they may be entirely incompetent. It’s shameful, but necessary in workplace environments that are increasingly populated with decisionmakers whose primary motivation is personal greed, rather than the overall wealth and sustainability of the entire organization.
Ordinarily, enlightened self-interest should find enough points of congruence between individual greed and organizational advancement, but these are not ordinary times. When employees are increasingly thought of, not as assets, but as human “toilet paper” to be treated in the way the reference indicates, even competent managers are adrift in a sea of uncertainty regarding the appropriate points of congruence of personal and organizational interests.
Apr 28 2015 at 9:42pm
I hate this argument and it seems to come up all the time. Wouldn’t any reasonable human being agree it is a combination of the two.
Mark V Anderson
Apr 28 2015 at 10:09pm
Decades ago I hired several accounting clerks. I made up a math test to help me judge the applicants. Over time I realized that this test was more important than any other factor in the interview. I think that firms would indeed be do better to use more tests for hiring.
I did use schooling also as a filter of the resumes to determine who to hire. I didn’t do this because these schools taught them anything useful, but because I figured those who received accounting certificates at least had some interest in the subject. That is, signaling.
About 15 years ago, I needed to hire a staff accountant, so I put together an Excel test to give my applicants. Unfortunately the HR staff had since grown, and they forbid me to give applicants this test. IF a clever lawyer could prove that certain protective classes did worse on my test than others, we could lose a case of discrimination. Duke Power was a case in the early ’70’s where the Supremes ruled that giving an intelligence test could be considered discriminatory if it could not be shown as directly related to the job. And cases after that reinforced the rule that tests with disparate impact could be considered proof of discrimination. So, no arbitrage of employees using tests instead of education is legal. Bryan seems to minimize this issue in his posting, but I think it is pretty major.
I think it is also true that it can be very difficult to buck the trend to try to benefit from the over-education of out young. If one is looking for ambitious, intelligent, and conforming workers, one will find 99% of them go to college. It may theoretically be a great benefit to firms to identify that 1% that didn’t go to college, because they will be be cheaper and more desperate for a good job, but it would also be expensive to identify them, and perhaps impossible to find them (especially without the benefit of a test, which would likely be illegal). I think there is some of this arbitrage happening, especially in smaller firms, but it is a very slow process.
Apr 29 2015 at 10:01am
Colleges are not just about education and signalling, they also perform a “sociological” role by being places where young people can mature before entering the working world and can also find spouses.
Colleges where these things matter less tend to offer curricula focused more directly on job skills: business, education, IT, etc., which is what you would predict.
The people engaged in this debate tend not to go to those kind of universities and I think they would be surprised how career-focused many majors and universities are.
Apr 29 2015 at 3:33pm
In particular, it’s interesting because Noah is a bit of a Japanophile who spent some time there, and the signaling model seems to be fairly popular there. It’s widely accepted that, say, all the real work is getting into Tōdai by passing the entrance exam, and that college itself is a time to party and network while also showing your willingness to conform.
Comments are closed.