Bryan Caplan  

Diamonds in the Rough

PRINT
John Goodman Hits a Home Run... Paging the Left...
When people weigh the plausibility of the signaling model of education, they focus on the ease of on-the-job incompetence detection.  If it only takes a couple of months to notice that a person's credentials overstate their performance, how important can signaling be? 

I've responded to this objection before, but recently realized that I've neglected a parallel problem.  Suppose for the sake of argument that employers can quickly figure out whether an employee is a "lemon."  A serious knowledge still remains: Figuring out whether an employee is a "diamond in the rough." 

Example: Suppose you're the manager in a Starbucks.  How long does it take you to figure out if one of your baristas has the right stuff to become a Starbucks executive?  What are the odds that you totally miss a workers' untapped potential?

If you're willing to admit that diamond-in-the-rough detection is unreliable at best, you've opened another door to the signaling model.  If you're a potentially awesome but unappreciated worker, extra education is a great way to get the market's attention.  Employers hiring for better jobs are a lot less likely to throw your application in the circular file if you've got the right diplomas.

But isn't this process productive, broadly defined?  At some level, sure.  But as always in signaling models, there are negative externalities at the margin.  If everyone signals for four extra years, this doesn't improve the quality of our signal; it just waste four years of time and resources.  But that's from a social point of view.  Selfishly speaking, "wastefully" signaling for four extra years can still enrich you by making you look better relative to competing workers.  It's just like standing up at a concert to see better: You make yourself better off by making everyone else worse off - and burn socially valuable resources in the process.

P.S. Here are my Powerpoint slides on signaling from a recent IHS seminar.


Comments and Sharing





COMMENTS (12 to date)
Nathan Smith writes:

1. You lean pretty heavily on that concert analogy, but it strikes me as merely suggestive. I want to see a model. What's the best in the literature, that you know of?

2. I'm not quite convinced that you have good enough reasons to reject the hypothesis that education is signalling, but it's an efficient investment in skill revelation.

3. Do you think there are unexploited profit opportunities here? If I were to found a company to capture the surplus from discovering these diamonds in the rough, how would I go about it?

Dan Carroll writes:

While the signaling model doesn't necessarily equal a higher education bubble, the Internet thows open possibilities of disruption - if the signaling model is untrue.

If the primary value of education is in what is taught, then much of what is taught can be put on the Internet ... plus much more (like everything that has ever been written). In the old days, if you wanted to learn something, you had to pay for access to an expert (aka professor). Now, a handful of branded experts can put their lectures on the Internet, sell them for a fraction of the cost (piracy notwithstanding), and make a lot more money. [Note: a blog is a good way to increase your branded presence, even if you don't get paid for it.]

If the value of education is high, and the value of signaling low, then I would expect enrollment in institutions of higher learning to start dropping precipituously as more lectures are put online.

Unless, of course, the stamp of approval from the admissions board and the prestige of the paper printed by the university are worth more than the combined value of the lectures.

Does your analogy for parenting also hold for teaching? Namely that "students are like flexible plastic – they respond to pressure, but when you release the pressure they tend to pop back to their original shape"?

I'm curious about the claim that education has negative externalities. I get why education might have negative externalities on this dimension. But education might have good effects too. I'm curious what you think about the following highly speculative suggestion.

As you know, a university education might be causally related to both pro-trade and pro-immigration attitudes. The mechanism seems to be that college education socializes people to become more cosmopolitan and tolerant. So, perhaps the expansion of education increases support for free trade and open borders, everything else being equal. You're a defender of free trade and open borders. So, this is a defense of education that even you should endorse.

Nathan Smith writes:

re: "If the value of education is high, and the value of signaling low, then I would expect enrollment in institutions of higher learning to start dropping precipituously as more lectures are put online."

This is certainly logical, but there's got to be something wrong with it, because it's been cheap to reproduce lectures for a long time, and if lectures are really what a college offers, market forces should have ensured that the Great Lecture Series would have displaced the Ivy League long ago. The internet makes it cheapER to distribute, but it was already trivial compared to $20,000+ tuitions. It's very hard to figure out just what the value is that educational institutions are creating to justify their high cost/price. But I don't think iTunes U alone can replace it. iTunes U + Facebook might have a fighting chance...

Floccina writes:
It's very hard to figure out just what the value is that educational institutions are creating to justify their high cost/price.

I think that the value is bragging rights that the parents get.

Troy Camplin writes:

I'm still waiting for someone to answer me why my 10 years of college education, with a B.A., a M.A., and a Ph.D., does not seem to signal to anyone that I should be given a job to do much of anything.

My dad's bragging rights are currently, "My son has a doctorate, but he's unemployed."

Is it maybe that the people in H.R. don't have the first clue as to what my education signals? For a signal to work, the person receiving it has to be competent to translate the signal, right?

Jim Object writes:

Bryan, I have gone on at length about how great you are a few times.

Now you have offered me a powerpoint.

*headdesk*

I'm going to join the North Korean Army or something. I quit.

Troy Camplin writes:

Still waiting...

Alrenous writes:

Potential awesomeness is nearly invisible. Actualized awesomeness is as easy to detect as a lemon - for professionals, perhaps even easier. This means there's a straightforward but currently nonexistent solution for the barista.

If they suspect they could be an awesome executive, they should put their money where their mouth is and go learn the skillset. Then, ask their manager to test them on that skill. As a bonus, passing the test also signals consciousness because it required the employee to have learned all the stuff under their own discipline.

For example, if I were designing a test for knowledge of undergraduate physics, I'd open by asking for an explanation of the Riemann tensor. Even with just with that one question, I get a rather good read. Similarly, by repeating the question with many subjects, I can compare the disparate answers and get a good first impression of the detail level of their thought and of their communication style.

Put another way, programming interviewers don't have to spend two months, let alone four years, to filter out most unqualified applicants.

All of which is to say that university can't possibly be about what it is said to be about. It spectacularly fails to add up.

Misaki writes:

The usefulness of a signal cannot be determined without specific knowledge of its requirements compared to what it's intended to predict and the goals and quality variation of those who choose to signal or not signal.

http://zhongwe2.serverpros.com/cs/

http://pastebin.com/Q86Zhgs9 *note that wage should decrease due to widespread perception of companies being in difficulty, not increase

Misaki writes:

@ Nathan Smith

"3. Do you think there are unexploited profit opportunities here? If I were to found a company to capture the surplus from discovering these diamonds in the rough, how would I go about it?"

Difficult, because the same individuals who decide not to signal in education are also likely to not be interested in signalling in income, which is nearly the only utility derived from having a high-paying job currently.

So the answer to this is something like, "provide a form of compensation that is useful to people who are willing to ignore standard signals to accomplish their goals", as described here: http://pastebin.com/Q86Zhgs9

Comments for this entry have been closed
Return to top