Bryan Caplan  

Channel Charles Murray

Stossel Parody... If This is Polite Society, Wha...
No one in the Cato Unbound exchange on Charles Murray's education book responded to my closing questions.  I asked Murray:

In your view, why precisely does the market financially reward students for taking lots of classes that at best seem distantly related to job performance?  You don't seem ready to sign on to my signaling story.  Do you have an alternative?  As I've said before, I agree with most of your conclusions.  It's your model that leaves me wanting more.
I also asked Pedro Carneiro:

If you mentally review your years as a student, can you honestly say that your typical class raised your marginal productivity?  If you adjust for the fact that you're an academic, doesn't the weakness of the connection between what you learned and what you do trouble you?
Does anyone want to trying channeling Murray and/or Carneiro?  I need closure!

Comments and Sharing

COMMENTS (10 to date)
Curt G writes:

I feel your frustration, as I expected some kind of response. The silence seems odd.

"Bueller ... Bueller ...?"

Daniel Klein writes:

Seems to me that Bryan is missing Murray's thinking about identity/cultural externalities.

It seems to me that Murray's point could be simply represented in a classic Schelling binary-choice tipping model -- the kind in the last chapter of Micromotives and Macrobehavior (1978).

The figure is for only a subset of those young adults that are intellectually able.

The horizontal axis is the percentage getting a BA. Call that variable G (for getting). So G ranges from 0 to 100.

The vertical axis is payoff to the individual. All individuals in this model are identical (though they don't necessarily act the same way).

At the vertical line at G = 0, the don't-get-BA payoff line has intercept at point A, and the get-BA payoff line has intercept at point B. Make A considerably above B.

The don't-get-BA payoff line slopes downward, and hits the vertical line at G = 100 (at the right edge of the diagram) at point Y.

The get-BA payoff line slopes upward and hits the right edge at point X.

Make point X higher than Y.

Make A (which is over on the left) higher than X (which is over on the right).

Call the point of intersection I, which corresponds to some tipping point G = k. At expected G k, everyone will get a BA. Thus, the two stable equilibria are points A and X. (I is an unstable equilibrium.)

Murray's point is is that this population is stuck at X; it would be better for every one of them if they were all at A.

Let's suppose that the model holds constant public policy (taxpayer subsidies, etc.).

However, the slopes of the two payoff lines reflect movements along some sort of cultural/identity function, such that when G = 100, not getting a BA is highly stigmatized, but not so when G = 0. The stigma is proportional, say, to G, hence the curves are continuous and have their pos/neg first derivatives.

Murray's point is: Hey folks, this X is an unfortunate cultural equilibrium, and now that I'm making us all aware of it, I am altering that to flatten the slopes of both curves so that they no longer intersect. Flattening the slopes means less stigma/status-effect as a function of how many are getting BAs.

Alternatively, Murray might somehow be saying: By virtue of this discourse, let's mutually coordinate on some G

By enlightening the public (or at least this subpopulation) and flattening the payoff lines so that they no longer intersect, the unique equilibrium will be A, and everyone represented in this diagram is better off.

Of course, others not so represented will be worse off, particularly all the social democrats who want everyone to be properly acculturated by immersion in and identification with academia.

Even if taxpayer subsidies were eliminated the population could still get caught at something like X.

Bryan Caplan writes:

Reply to Dan:

But if you read my critique of Murray, I'm offering an explanation for *why* competitive employers "stigmatize" non-B.A.s. Murray doesn't really explain how profit-maximization and stigmatization are compatible. Like I said originally, I'm not challenging his conclusion; I just want a better story about how his conclusion is possible.

guthrie writes:

I'll try my hand here, although I'm at a bit of a disadvantage. I'm one of these 'uneducated' professionals who has never completed college (I spent 3 years at a community college from '89 to '91 as a music major… I did not complete the program and never returned to school). I currently work in a bank (well capitalized and quite solvent, Dr. Kling!), in customer assistance.

I’ve followed the conversation and here’s what I think, for what it’s worth:

There’s a big part of me that deeply appreciates Dr. Murray’s suggestion that most career degrees can be shortened and made more practical, or eliminated entirely from some professions. I know first hand how employers use earned degrees as a blanket standard that excludes folks like me. I won’t even waste my time attempting to apply for most of these jobs (unless I really want it, but even then I have little hope and my skepticism is usually borne out). And things that I am good at, interpersonal interaction, self-motivated learning, improvisation, aren’t really taught in high school or college, and can’t be quantified without experience. My income has suffered, I’m sure, because jobs that I might be practically qualified for, the employers wouldn’t consider me, an ‘unpapered’ individual.

I’m leery, however, of his (and your?) insistence that IQ ought to be used as the plumb. Perhaps this is because of insecurity over my own IQ… nothing’s worse then being considered ‘stupid’, right? Perhaps that’s just me… Even so, I am unconvinced that IQ is the only way to figure out if someone is ‘fit’ for a 4 year degree. I have too much experience with too many people throughout my life that seems to me to counter this idea.

BTW, my wife earned an English degree with a creative writing emphasis, from a decent program (UW-Madison), and can’t find work. Go figure. One way or another, there are probably other qualifiers that many employers are looking for aside from degrees, but in many cases it will at least get your resume a second look…

As far as the ‘story’ is concerned, my guess is that employers see a degree itself as a ‘product’, as an example of what one can produce given the time, tenacity, and funds. Wages are duly linked to how much they value this particular widget, evaluating (but not too deeply) what was studied and where, as it relates to other widget’s they can glean from a resume or interview. Perhaps think of a Christmas tree with various ornaments, one of which is a degree. How pretty it is or how it showcases the rest of the tree depends on the viewer, no?

Perhaps some entrepreneurs, themselves ‘unpapered’, surround themselves with the pedigreed to display credibility to potential investors. The value comes from the perception of smarts that mitigates whatever lack of practicable experience each individual may have. I’m sure these ideas are inadequate in fully reconciling the paradox, but I think it comes close. What say you?

As for present conditions for a ‘tipping effect’, I’m thinking that most employers will act conservatively, and therefore any change will be slow (as per usual, no?). No one wants to be first. Second is best, or perhaps third, but never first or forth.

It would be interesting to hear from an actual entrepreneur or HR person on this.

Allow me now to attempt your other question… please bear with me!

If one is on an assembly line (and let’s not kid ourselves, ‘data entry’ is as much an assembly line as sticking an O ring on a cylinder over an over again), then no, a wide knowledge base isn’t absolutely necessary, and neither employers nor potential employees ought not feel the need to use a BA/BS as the arbiter of ‘what’s required’.

However, in most businesses, things are constantly moving. ‘If you’re not moving forward, you’re moving backward’, and ‘If you’re not growing, you’re dying’, are the clichéd mantras that, even so, are true. I think the question is this: who knows where an answer for a particular problem will come from? For example, the guy who invented computer punch cards got the idea from how Persian rugs are manufactured! One may not remember everything in a cognitive sense at any given moment, but the one moment when an issue arises that may be solved by remembering how wool was overproduced in Europe in the Middle Ages, or the difference between ego and id, it might be handy to have taken some of those classes!

We humans have brains that make leaps in logic and irrational connections like that all the time and, counterintuitively, they can actually make sense and add to our value. The problem again is that no one can know when a bit of knowledge will be useful in any given situation. Better to have a shotgun when you don’t know where to aim. Or, better perhaps, it’s best to have a full rucksack when on a wilderness trek. The one thing you didn’t pack is the one thing you’ll need in an emergency.

And isn’t one lament that I’ve read on the site here that many good economists aren’t very good storytellers? (Goodness knows Mr. Roberts could benefit from some training! His island analogy is good as an analogy but not very good storytelling, sorry to say!). I believe that any profession would benefit from having storytellers. Storytelling is a learned skill that comes from listening (reading), repeating, innovating, failing, and repeating again. Literature, Theater, and other ‘arts’ classes aren’t really required to do what you do, but could help you overcome this ‘problem’.

You may tell ‘stories’ in your classroom to a captive audience, but if you and your cohorts were to put together an ‘Econ Road-Show’, and made your class a ‘performance’, you just might get a better idea how to hold an audience’s attention… I’m sure, however, that’s not really an issue for you, right Bryan?!

Just as an aside, being an improvisational actor, I have found a significant benefit to having a rather broad learning base. If an audience member suggests ‘Helmut Kohl’ or ‘Manzikert’ it’s helpful to know at least a little of what they may be talking about in order for the audience to have a good time. If given the shot, I’d gladly work for a 4 year degree!

I thank you for your patience with me! Now, I admit (and it’s obvious I’m sure) that I have a less-than-firm grasp of ‘signaling theory’, ‘externalities’, and probably a host of other econ theories and jargon required to adequately ‘channel’ either Murray or Carneiro, but hopefully I’ve held my own here... Please be kind to this unpedigreed whelp!

Daniel Klein writes:


You write at Cato Unbound:
- - -
If the BA is a signal of intelligence and perseverance, why haven’t certification tests caught on? The obvious explanation is that the first people in line to take the test would have high intelligence but low perseverance. After all, if they are smart enough to do well on the test, why are they so eager to avoid college? Are they lazy or something?[3] Consider further: If elite college is mostly a signal of terrific high school performance, why don’t employers poach the nation’s top high school grads before they set foot at Harvard? Perhaps employers would wind up hiring the top slackers of the Ivy League. The best students wouldn’t be looking for an easy way out.

An unfortunate implication of the signaling model is that cutting the BA down to size will be a lot harder than Murray thinks. 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?
- - -

It seems to me that this does not speak to the notion of an intellectually able subpopulation being caught in an inferior equilibrium like X (as explained in the diagram I described). In that diagram, the costs/benefits of signaling are incorporated. I take Murray to be suggesting that "we" (or, at any rate, a subpopulation) are stuck in an inferior equilibrium. He writes his book to raise awareness of the problem and the viability of a better outcome. Maybe we can change the cultural parameters behind the shapes of the payoff curves. What you write does not address this point, it seems to me.


Max Liberty writes:

Not having read all the conversation, I think you might be missing the effect of (a) the mismatch of interests between employers and their agents in charge of hiring and (b) conscious clustering of behavior (or the herd effect).

HR departments make a BA into a hiring requirement for a variety of reasons, not all of which they would admit:
- They do it because that's the way it's "always" been done.
- They justify it to themselves: "If I need my BA to do my job, surely that position needs one, too."
- Culturally, HR departments do not like making decisions based on things that are not documented. That means that the intangibles that can make a great employee take a back seat to having a diploma.
- This last point is partly reinforced by the need to be able to show, in court, that you made a hiring decision based on classifications that are neutral as to any protected clas. Generally, educational requirements have been upheld as protecting legitimate business interests.


howard432 writes:

The only class I ever took that directly related to anything I did in the "real world" was high school typing.

guthrie writes:

Well said Max! I think your point is better than mine in explaining Murray's paradox.

In Thermodynamics an object stays at rest unless acted upon by an outside force, and I think that works here.

The BA is the cultural standard for the HR set. Until and unless there is overwhelming 'force' in the form of evidence that 'proves' an exam of some kind can better indicate talent for a particular job then a BA, the combination of an HR person's incentives (the CYA effect), along with the herd effect will be very difficult to overcome.

The incentive to see through educational hype may be strong but not strong enough. This may also explain why the BA can be perceived as a 'force of nature'. Again, well put Max!

And Bryan! I'm taking a shine to the 'Econ-Road-Show' idea! Think of it... in a thundering, gravelly, reverbed basso-profondo:

"sunday, Sunday, SUNDAY!!! Come down and experience the GMU ECON-ROAD-SHOW!!


Come see the towering giants of economic theory thrash out fantastic fables of high finance!!

*thumping house music*

You’ll hear from Tyler, Cowan, Russell, Caplan, Kling, Henderson and the immense JAMES BUCHANAN!!!

*ligtning clap*

It’s a veritable ECONOTHONARAMAPALOOSA! Don’t miss it!!"

C'mon Bryan! Throw me a bone!

Brian Macker writes:

"... can you honestly say that your typical class raised your marginal productivity?"

Yes, I can. Of course I do real work in a field the requires thinking, computer science.

My typical classes were english, math, and science. My untypical ones were history, art, and phys-ed.

Even if I did fake work like being a genders studies professor I imagine the BS classes I did take would be related to my job.

I certainly understand and believe in the signaling aspect for other jobs, but some jobs are specialized and require special training.

Daniel writes:

One reason employers have increasing made "longer education" a job requirement is government licensing law. In order to become licensed as a certified public accountant for example, you have to complete 150 college credit hours (in the vast majority of states). A great CPA exam score and bachelor's degree in accounting isn't enough. Funeral directors are also licensed by state governments and these licensing boards often require college degrees from accredited institutions. Passing certification exams and gaining hands-on experience in the field aren't enough in state regulator's eyes. There are many other occupations like this. Go visit your state's licensing division website to get an sampling of this.

Murray must be aware of the effect of licensing law so I'm baffled why he doesn't mention this in his analysis.

Even so, there are many jobs where no license is required. In theory employers who want bright employees could use intelligence tests but as Murray documented in "The Bell Curve", employers can get in big trouble in court and with regulators for using any written test that isn't narrowly tailored to highly specific job functions (see the Supreme Court case Griggs v. Duke Power Co.). Creating job-specific tests that have predictive value can be difficult though. Therefore many employers don't use any tests and instead look to the college degree as a signal of intelingence. Your signaling model seems to me to be a good explanation of this phenomenon.

My best guess is that Murray secretly finds your signaling model persuasive but isn't willing to say so. This may be because he is afraid that embracing the signaling model will lead people to see general IQ tests (rather than subject specific CPA Exam style tests) as the substitute for the college degree. I base this on an article he wrote a few years ago that called for the abolishment of the SAT I Exam. One of his arguements was that people remember cognitive exam scores (like the SAT I or IQ tests) but tend to forget subject specific test scores (like the AP Exams, SAT II subject tests). Therefore SAT I scores tend to be a class and status marker in a way that subject specific tests are not. Murray seems to have a strong desire to avoid anyone catching "stigma" so he would rather not embrace a signaling model.

Comments for this entry have been closed
Return to top