Bryan Caplan

What Does Education Signal? Answering Arnold's Challenge

PRINT
Samwick, Thoma, and Pearlstein... Name My Book, Part II: Vote on...

Arnold raises a common objection to the signaling model of education:

My problem with the signalling model of education is that it suggests that there is a huge unexploited profit opportunity for employers and employees who can come up with alternative signals. And yet nobody tries to set up a system for identifying and hiring smart high school graduates.

This is a powerful objection against the view that education is purely a signal of intelligence. But there is a lot more to being a good worker than being smart. It is also important for employees to be conscientious and conformist. And while we can accurately assess someone's intelligence with a short IQ test, it's a lot harder to find out how conscientious and conformist someone is. Only Jack Black or Homer Simpson would admit in an interview that he's lazy or weird.

Oh no, in interviews, the only character flaw that anyone owns up to is being a "workaholic" or a "perfectionist"!

But why does school have to go on for years? Simple: Even a lazy weirdo can pretend to be hard-working and conformist for a few months. Now suppose an employer wants people at the 90th percentile of conscientiousness and conformity. He's got to set the educational bar high enough that 89% of people give up despite the rewards. Especially in an environment where government heavily subsidizes education, that could easily mean you have to get years and years of school to distinguish yourself from the pack.

An interesting implication is that high-IQ people who don't go to college are actually signaling that they are unusually lazy and/or weird. It's easy for a high-IQ person to breeze through school. If one fails to do so, a sensible employer will naturally ask "What's wrong with him?" It's not surprising, then, that few employers will give an eighteen-year-old genius a responsible job. The very fact that you refuse to go to college suggests that you aren't going to be as good at your job as your test scores would normally indicate.

I hasten to add that I don't mean to disparage my self-taught readers. I'd probably really enjoy meeting you. But I'd be lying if I said I'd be eager to hire you. No offense intended - if I were running a business, I wouldn't hire myself either!



TRACKBACKS (3 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/454
The author at The Burden of Proof in a related article titled The Entrepreneur & Education writes:
    Just a quick note to direct your attention to 2 great discussion taking place at EconLog: Galbraith vs. the Internet, about how Galbraith was wrong and free marketeers were ... [Tracked on February 9, 2006 8:32 AM]
COMMENTS (11 to date)
Al T writes:

Then why do colleges have so many lazy weirdos that manage to get degrees. I hate to think that the rest of humanity is either lazier or weirder than the people I went to school with.

Also, what would the signaling theory make of the adage that my CompSci professors taught me. "Those who get the A's in class will probably work for those that get the C's." The idea (well one of them anyway) is that those who get C's spend their time doing things like working on their social skills, whereas those who get the A's miss out on that aspect of college life.

JKB writes:

Well, there actually are those who have developed an alternative signal. They're called start ups. Paul Graham offered this premise in his "Hiring is Obsolete" essay. Education is the signal for the risk adverse. But you are correct in that education is a signal of conformity. Those who conform the best, make the best grades. For an old, established enterprise, the conformity might be the attribute with the highest value.

One additional aspect of education is that it expands your horizons. Even in a state university, you are exposed to a wider range of subjects, personalities and ideas than if you are in high school, which self selects for those of similar background. More of exposure is available to the individual via the internet these days but the university environment does concentrate the differences and in many ways forces the student to at least recognize they exist.

In good times, education is the signal for employment since there is such a large pool of graduates. In tough times, education isn't enough, most employers require work experience as well. Of course, running a successful startup, trumps both of these signals by demonstrating proven accomplishment.

Ben writes:

I don't understand why this is such a powerful objection. Wouldn't attempts to set up alternate signaling systems have the same types of anti-discrimination legal barriers that IQ testing has?
http://www.gmu.edu/departments/economics/bcaplan/e321/lab7.htm
Section is XI. F is the most relevant,

PS You had google's first entry for
anti discrimination law IQ tests hiring

Kerry writes:

As usual, your analysis of education ignores the fact that many technical fields require such a breadth and depth of knowledge and/or experience that even an intelligent person cannot become competent in a short amount of time. I am a scientist, and I think the current length of modern American science education(K-12, 4 year undergrad, 4+ years grad) is near the minimum required to reach the state where one is able to do the real work of science: Making, testing, and communicating original hypotheses.

In addition, grades are important for weeding out not only nonconformists, but also people who don't understand the material.

Anthony writes:

One further thing to add to the mix, you really should distinguish between Education and Training.

I agree to an extent with the signalling theory of education - however some further education is not education. In my view going to medical school is akin to learning to be a plumber - if a goes wrong you can do b, c, or d etc. Yes there may be more to learn, but they are not that dissimilar

With the last poster mentioning science as one example of this (I'd say a bit of both - UK graduates reach the same level in three years), computer science is another (some degree of training in coding, but some education ie. Maths, which isn't strictly necessary to complete the task at hand). Economics however pretty much goes fully into the Education category, along with History of Art, and a few more (though all have some element of training in learning new skills).

In conclusion I'd say that you DO learn things at university, but it's a small percentage of the total amount, varying by subject. For some people this means that without their degrees, they couldn't do their job, for many more, it's a signal that you are capable.

If you're right then there's a great opportunity to make money with a business model that's less reliant on the conformism of workers or their hard-working-ness.

It's always a good idea to look at employing the "unemployable"... for example, if I would need an econonomics PhD and I would live in a racist society I would imagine I could get a black graduate, cheaper and easier than a white one, ceteris paribus. (Sorry if this example is not P.C. enough)

T.R. Elliott writes:

This is all purely armchair speculation akin to debating how many angels can dance on the head of a pin. For every argument, a reasonable counter-argument can be formed.

Caplan needs education to be a signaling device to support his broader theories on IQ and the role of education in society. Therefore he will always find a reason to counter argue any undermining of his broader thesis.

We call that ideology.

In the mean time, the real world continues on.

Dan Hill writes:

If one of the signals is about application and conformism then employers need to be aware that they are selecting from the middle of the bell curve. Sure, they're avoiding the risk of selecting from the bottom tail but they are also losing the opportunity to recruit superior performers. Whether this is a problem or not depends on what you are trying to achieve, but like any investment strategy, you don't get above average performance by avoiding risk.

Signalling is of course about trying to reduce the cost of selection which is especially high for employers who lack the knowledge themselves to actually identify what really matters (in my field - IT - the less knowledgeable the employer the more the requirements focus on arbitrary and incredibly detailed technical knowledge even when it's not really what matters to success). Dare I say, the more pointy haired the boss the more important signalling is.

Matt writes:

What about testing for differences in time preference and thresholds for pain to determine someone's potential worth as an employee? I suspect that some people (you know, some of the high-IQ, lazy wierdo types who need to go to college to show they have characteristics worthy of an employer's attention) figure if they tough it out for only eight years in college, they can then live the next 50+ years of their lives being as lazy, wierd and as smart as they already were. Like a college professor?

Bob Knaus writes:

As a high-school dropout management consultant, I say "No offense taken!" :-)

I was 19 when I quit working for my dad as a sharecropper on the family farm. I was 34 when I applied for my first job. At the time, I was a pony-tailed "network guy" who had no degrees but lots of relevant experience. The consulting firm called me back the next day with a job offer.

I've hired several people over the years who don't "fit the mold". The ones who have succeeded have told me they were grateful for the opportunity. One saying I learned as a consultant was "There are no bad hiring decisions, only bad firing decisions."

I can't think of a better argument in favor of labor market flexibility.

Bill writes:

UK graduates reach the same level in three years

That's because UK grads are more narrowly educated. The European schools generally do not require the general education that US schools do. In the US, we are not only trying to train workers, we are also trying to train citizens.

I wonder if Bryan's book will contain a discussion on specialization's effects on voters.

Comments for this entry have been closed
Return to top