Bryan Caplan  

Why Charles Murray Should Take Signaling Seriously

PRINT
Why Do People Keep Calling Tyl... Joe Stiglitz. Word!...
I critique Charles Murray's work on education at the latest Cato Unbound.  My thesis: Murray is great on the facts, but muddled on theory.  He's one of the few scholars to notice the flimsiness of the connection between higher education and job skills.  But he doesn't have a clear story to explain it.  My central complaint:

But when he tries to explain how useless studies translate into big bucks, his story gets fuzzy.  On the one hand, he tells us that "The BA really does confer a wage premium on its average recipient, but there is no good reason that it should."  On the other hand, he insists that "Employers are not stupid."  How can both be true?
My harshest criticism:

Even stranger, Murray often talks as if the entire labor market were centrally planned by university committees.  Perhaps I am being too literal.  But it is one thing for Murray's imaginary education task force to say, "Let's reify the BA."  It is quite another for a task force -- even an imaginary one -- to say, "We will attach an economic reward to it that often has nothing to do with what has been learned."  What's wrong with this picture? Universities don't "attach economic rewards" to their degrees.  That decision is up to millions of competing, consenting employers.  Unless higher education convinces employers that workers with BAs are more productive than workers without, the BA won't have any "economic rewards."
My solution, if you haven't guessed it:

If Murray can't clarify his model, no one is going to take his facts seriously.  Fortunately, I can help.  Here's what Murray should have said: "To a large extent, the BA is what economists call signaling.  Individual students who go to college usually get a good deal; so do individual employers who pay a premium to educated workers.  The problem is that this individually rational behavior is socially wasteful, because education is primarily about showing off, not acquiring job skills."
How's Murray going to respond?  Hopefully he'll takes my criticism in the spirit in which it was intended - a fanboy's attempt to improve the work of the master.


Comments and Sharing





COMMENTS (8 to date)
August writes:

I don't know whether he touches on the regulations surrounding hiring in the U.S., but HR departments tend to use degrees as a proxy for what they aren't allowed to test for- I.Q.

It seems to me that the reason it seems the way Murray depicts the situation is because there is a delay between expectation and reality. When the B.A. meant you had learned something, people took one having a B.A. as an indication that you had learned something. As the B.A. degree became less and less rigorous and the students learned less and less at college, that did not become immediately apparent. Rather, people still expected someone with a B.A. to know something. It took probably a decade or more of graduates consistently demonstrating their ignorance before it became clear that it wasn't just a few that didn't know anything, but that it was everyone. This was partially masked because those with more technical majors could still do much of the work they were educated to do, even if they could not think or write. I think the fact that so many people now have M.A. degrees demonstrates that this signal is beginning to be passed down. The B.A. doesn't mean much anymore, being equivalent to a high school diploma from 100 years ago, so now employers are demanding their employees have M.A. degrees. That, of course, will likely result in the dilution of that degree as well, though.

Mitch Oliver writes:

Its funny: when I was disillusioned with my educational experience and salivating to get into the job market, my father explained the value of a college degree using the signaling model. In short, he said it doesn't matter that you didn't learn anything in school that you couldn't learn better in practice (he claims he learned more during his co-ops than in classes) or at least teach yourself. A college degree sends a signal that you can complete things, that you are somewhat responsible, etc.

While today I would argue that many fields would be better served by trade schools than liberal arts education as it exists today, that piece of paper still sends the same message.

Here's an interesting question to ponder: to what extent do college classes measure perseverance in the way that tests can't measure? Many, if not most, courses are graded primarily based on tests, which you can simply take by skipping classes and self-studying the material. College classes also have homework assignments, projects, and essays, all of which probably measure perseverance better than any test could.

Yet, it may just be a coincidence of our society that all of these are correlated with perseverance. *Theoretically*, a very perseverant student could just self-study APs, and then go to college and go to classes only to take tests and turn in homework. However, most students who do that are mediocre students (there are very strong students who do that - especially at institutions like Caltech [at Caltech's Math 5a, for example, the majority of students skip all their classes - I'm not a Caltech student, but I've read a lot of posts about Caltech on CollegeConfidential], but these students seem to be in the minority of all students who skip classes). If intelligent and well-motivated students who score high on conscientiousness would study more outside of class and get lower GPAs, then perhaps this might pressure employers to take in more students with lower GPAs. But it seems that the predominant majority of conscientious students are pretty religious about going to class (this might be related to personality - as conscientiousness is also correlated with lawfulness/following what's "official" rather than what's "unofficial").

Jason Malloy writes:

Dr. Caplan's response is good. Dr. Murray has directed his argument at young people and universities, but employers are the one party overwhelmingly responsible for the system. Young people and universities are just making the most profitable decision for themselves. We know this is true. On the other hand, employers are the ones supposedly getting screwed by overpaying know-nothings. So these are the people you need to convince — and they probably want proof.

And this is where I disagree with both Murray and Caplan. I don't know the answer to whether the BA actually is screwing or helping employers. I strongly disagree with Bryan:

"[employers] have a strong incentive to see through academic hype. When firms overpay the overeducated — or needlessly “stigmatize” applicants without a BA — the market charges them for their mistake... As far as employers are concerned, the BA works. When they pay college grads more, they get their money’s worth. You can try jawboning Microsoft into switching to certification tests. But can we really believe that Murray has seen a profit opportunity that Bill Gates hasn’t?"

Gates is notorious for relying on IQ tests, so this was a bad example. Also I don't know how important college is in MS hiring decisions, but I think it hasn't been denied yet that technical degrees like engineering actually do usefully signal job-applicable knowledge, unlike, e.g., an English degree.

There is no data on whether employers who select by job knowledge or IQ tests make more profit than those that select by degree, so there is no way for employers to know this. In theory, I realize the market should invisibly sort the companies with unprofitable hiring practices out of competition. It's a nice theory, but I don't buy it. At all. Especially because most companies appear to be hiring in a similar way.

Dr. Caplan has a lot of faith that companies must be acting intelligently, but I think we already know this is false. Employers fail to even distinguish and differentially pay and promote high talent employees from low talent employees with the same degree! (see linked post above) So even if BA earners were more diligent, diligence would have to outweigh the extra productivity information lost from not giving IQ tests. And, obviously, I don't see how requiring both could hurt.

It's a very real possibility that the whole inflated BA system is based on bad, inefficient. and costly selection decisions by employers. If employers are acting a certain way before there is data to justify it, then I think there is a good possibility that they aren't acting on some indirectly obtained knowledge, but on habit and tradition. This is why good data really can improve markets.

Of course the real reason for the incredibly expensive and inefficient BA signaling system might be anti-discrimination law rather than employer stupidity.

Dr. T writes:

Signaling works even though our current education system sucks. I'd rather dump the signaling and reform the education system. Charles Murray wrote good suggestions about improving education and changing from a second-hand signaling system to a first-hand measured quality education system.

I had to pass five different board examinations to become a clinical pathologist. There are rigorous exams for medical technologists, clinical perfusionists, and other health care professionals. Lawyers and accountants have qualification exams. Some engineering fields do, too. There is no reason this practice cannot be expanded. Employers could look at exam scores instead of vague signals. Students could design their own curriculum instead of being locked into a rigid set of BS or BA requirements.

-----
August: College degrees are not proxies for IQ tests. In today's world, college degrees measure persistance and conformity rather than intellect.

In response to Jason's comment, it's possible that PR might be partially responsible for decisions by employers. If an employer explicitly says that college degrees are not required for employment, well, what does that say about the employer? Keep in mind that most people are more or less socially brainwashed to believe that education and job performance are closely linked with each other, so it's quite possible that the employer doesn't want people to think that the employer employs low-calibre employees (especially since smart people LIKE working with other smart people)

On a side note, here's an interesting thread I posted on something similar a while ago (although I got flamed for it):

http://talk.collegeconfidential.com/college-confidential-cafe/418274-people-should-able-earn-credit-toward-college-degrees-taking-standard-tests.html

Jason Malloy writes:

I agree with little in the Kevin Carey response. Carey repeats the Caplan criticism that the market has optimized hiring, firing, pay, and promotion, which is both theoretically and empirically dubious:

"The fact that the world is not already the way Murray wishes it to be assumes a catastrophic ongoing failure of intelligence and rational self-interest on the part of the business community."

Comments for this entry have been closed
Return to top