Arnold Kling  

Alex Tabarrok Explains College Wage Stagnation

PRINT
My Talk at Berkeley and My MSN... Lifted from the Comments...

He writes,


Over the past 25 years the total number of students in college has increased by about 50 percent. But the number of students graduating with degrees in science, technology, engineering and math (the so-called STEM fields) has remained more or less constant. Moreover, many of today's STEM graduates are foreign born and are taking their knowledge and skills back to their native countries.

...In 2009 the U.S. graduated 89,140 students in the visual and performing arts, more than in computer science, math and chemical engineering combined and more than double the number of visual and performing arts graduates in 1985.

Read the whole thing. I think some of my commenters had already guessed that this was the answer.

On a related note, I did not realize that this was an issue:


on September 1 the Obama administration filed a racial-discrimination complaint against Leprino Foods Co., which makes mozzarella cheese, over its use of WorkKeys assessments to screen job applicants at its plant near Fresno, Calif. According to the complaint, only 49 percent of black, Hispanic, and Asian applicants passed the tests, in contrast to 72 percent of white applicants (a Leprino spokesman declined to comment on the allegations).

It hadn't occurred to me that A Means A could run into legal issues with disparate impact laws.


Comments and Sharing





COMMENTS (17 to date)
Alex J. writes:

It never occurred to me that the education establishment would be the bootleggers behind suppressing employment screening tests, though now it makes perfect sense.

matt writes:

This why people like Steve Sailer are important. We need folks willing to rudely point out that intelligence isn't evenly distributed. Until our society comes to terms with this fact we won't overturn the educational establishment and replace it with something more affordable. In the end all objective criteria will have a disparate impact.

The only other option to people accepting the data,is making it illegal to count people by race. This way disparate impact can't be calculated. Racial blindness by the government is a more achievable goal than overturning people's unreasonable blank-slatism.

Steve Skutnik writes:

I think there's yet another dimension to this that is often overlooked when examining this, specifically pertaining to STEM. In particular, how much can you do with a bachelor's in any of the "pure" sciences, aside from teaching high school science? It is a near-requirement for those majoring in "pure" sciences to go on for graduate education - and even typically a master's degree is insufficient (most community colleges, much less universities, make a Ph.D. a condition of employment - now let's talk about the signaling model of education...)

I think this ratcheting commitment for gainful employment in sciences is an often-overlooked component of the problem. How many domestic students can you reasonably scam into majoring in sciences with the knowledge that any reasonable employment for their efforts is not a four-year commitment, but likely a 9-10 year one, on average?

(Obviously, this issue doesn't apply in engineering fields, given the employability with a simple bachelor's degree. But this phenomenon did weigh heavily on my decision to switch fields from physics to engineering after my masters...)

joeftansey writes:

The kinds of jobs available for engineers is changing too. The old generation are essentially glorified mechanics. They are being replaced by rapidly improving software.

So does the new technology create more jobs than it destroys? Whelp, they're moving to software because it gets more work done. So I'd say its tilted towards fewer overall engineering jobs.

Combine this with the fact that all the easy problems have already been solved (hint - you solve the easiest problems first), and the demand for medium-skilled workers declines. Your average engineer with 20 years of experience and a high school dropout have two things in common. They both can't help with offshore drilling, and they can both make cheeseburgers.

In fact, we've already done a lot of the hard stuff. I don't see any reason for easy problems to stick around waiting to be solved by middle-class undergraduates.

Sol writes:

I'm pretty sure there are still jobs for "normal" engineers. At least, my brother-in-law has a BS in engineering and has been happily employed since his graduation five years ago doing work on automotive brake systems. The problems may be easier to handle than they were fifty years ago, but there's still work to be done on them.

Roger Sweeny writes:

It hadn't occurred to me that A Means A could run into legal issues with disparate impact laws.

Oh, my, my, yes. Ever since Griggs v. Duke Power (1971), American employment law has pretty much said 1) you can't require any prospective employee to take a test or report results of a test if that test has a racially disparate impact, but 2) you can require a prospective employee to have a degree no matter how much of a disparate impact that requirement has (whites and Asians are much more likely to have degrees than blacks and hispanics).

"A means A" is illegal if it falls under 1 and legal if it falls under 2. So how would courts and agencies decide?

Lots of things--interest group theory, regulatory capture, sociology, politics--make me think they will go with 1. I suspect Brian agrees with me.

The legal language justifiying the present situation might provide some guidance but in my perhaps overly cynical view, it amounts to saying, "Education is a GOOD THING and educated people [defined as people with degrees] are better people, so of course they should get preference."

Mercer writes:

Someone reading Alex would think art fields were the most popular majors. The link below shows business majors were the most popular fields by a wide margin.

http://nces.ed.gov/programs/digest/d10/figures/fig_15.asp?referrer=figures

Anyone surprised by the employment test lawsuits should read Steve Sailer regularly. You can start by reading his articles on the Ricci case.

Bryan Willman writes:

But wait, there's more!

In the case involving firefighters (Ricci v. DeStefano) the court ruled that the city MUST USE its test results, even though no african-american applicants passed! (That test was very specific to firefighting, and noone credibly claimed it was about anything else.)

So your test can have "disparate" impacts if it is provably specifically related to the tasks at hand, but it must not have if it is not, and most likely you will be sued either way! In fact, you will be sued twice once from each side!

Oh, and if you hire people at random, you are likely to be sued by some employees when other people you should not have hired do criminal or dangerous things.

So an employer can be sued for testing, sued for ignoring tests, sued for not hiring people, and sometimes sued for who they hired.
(And we wonder why unemployment is high???)

Of course, this is an example of a legit problem in the law - lots of folks surely used tests or bogus credential claims of one kind or another to carry out discrimination. And it's very hard for a blunt instrument like law to cleanly distinguish.

What's more, society *says* it wants to ban discrimination, but of course, to some degree, is only pretending. (Hence we report a relatively narrow unemployment number that doesn't reflect more disadvantaged communities, we ban IQ tests and the like in hiring, but allow the employeer to require degrees that may be irrelevent, and so on.)

Bryan Willman writes:

And how much is Filtering vs Signalling?

By signalling I mean "everybody who shows up in a blue blazer and a white sailor hat is a yacht person, let them in" - and if there are many many of them, we celebrate the growth of the club.

By filtering I mean "we only want the top 8%. We're not allowed to test and rank for the top 8%, so we'll look for proxies for that."

To the extent that college degrees were about filtering rather than signaling, any growth in the percentage of the population that holds degrees will reduce the value of a degree, even if the degree increases human capital by a substantial amount.

Glen S. McGhee, FHEAP writes:

I agree with Sweeney regarding Griggs. Let's not overlook its contribution to underlying credential inflation.

Here's Prof Gallaway's article on Griggs' negative impact on higher education:
http://www.popecenter.org/issues/article.html?id=1749

Bryan, if you compare the decoupled and structurally broken US links between education and the workplace with Germany and Scandinavia, you will see that distinguishing between filtering and signaling is beside the point.
Remember how Michael Spence warned how overinvestment in education was always a danger?

roystgnr writes:
I did not realize that this was an issue

It's a sad, eye-opening day for everyone. I didn't realize that commenting here was so fruitless.

Ameet writes:

A Means A could probably succeed in a country like Chile, where parents pay for the majority of education costs (I think the state only pays 16%, according to last week's Economist).

So could a school based on students learning at their own pace, and being at the forefront of educational innovation, with the Khan Academy, lectures recorded by top notch professors (and grading and skype office hours with their grad students), and so on, because the consumers of the final product (an educated student) are actually paying for the education.

I sincerely hope some innovator/entrepreneur looks at Chile, sees the ripe opportunity, and goes for it. It would show up our educational system.

Mark Brophy writes:

I find it very disturbing that the government doesn't publish the complaint on their site. It merely publishes a press release:
http://www.dol.gov/opa/media/press/ofccp/OFCCP20111291.htm

Steve Sailer writes:

Dear Arnold:

As Roger Sweeny says, the courts tend to treat higher education as inherently virtuous. (Similarly, the courts tend to be scared of the military.) So, you would want to position your enterprise as part of the higher education system as opposed to the profit-making system, which (apparently) requires constant oversight to prevent profit-seekers from shooting themselves in the feet by irrationally not hiring employees who would make them the most profits.

College Board / ETS, for example, do a good job of positioning themselves as part of the Good People. Hence, the fact that their tests _always_ have disparate impact is not a big deal to the courts because they are non-profits and higher ed insiders.

Evan writes:

@matt

This why people like Steve Sailer are important. We need folks willing to rudely point out that intelligence isn't evenly distributed. Until our society comes to terms with this fact we won't overturn the educational establishment and replace it with something more affordable. In the end all objective criteria will have a disparate impact.
I agree with half of what you said. We need people to point out that intelligence isn't evenly distributed. But we don't need people to do it rudely. Like it or not the average person evaluates an argument based on presentation, not just facts. Rudely pointing out that intelligence isn't evenly distributed will probably decrease general support for that belief.

One of the reasons I'm not a fan of Sailer's work is that I don't think it would convince someone unless they were already leaning towards that belief. It seems mainly to affirm the beliefs of people who already believe in uneven intelligence distribution than to convince someone who doesn't hold that belief.

If you want a good example of a polite, non-confrontational argument for uneven intelligence distribution, I'd recommend 50 Great Myths of Popular Psychology . In myth 15 they do an excellent job of arguing against the idea that tests are biased against certain groups. But they do it in a very polite way. None of the negative reviews on Amazon even mention myth #15, in spite of it's "hot button" nature.

david nh writes:

" So an employer can be sued for testing, sued for ignoring tests, sued for not hiring people, and sometimes sued for who they hired.
(And we wonder why unemployment is high???)"

Plus, free market alternatives that better suit the needs of employers and employees without burdening students with massive amounts of debt before they even start working are discouraged.

Everywhere one looks the means for the US to adapt and heal itself are blocked. America really is kind of screwed isn't it?

Glen S. McGhee, FHEAP writes:

"Like it or not the average person evaluates an argument based on presentation, not just facts."

But what is a "fact"?

Boy, Evan, do I have a book for you!
Berger and Luckmann, The Social Construction of Reality (1966).

And it you don't like that, an earlier book by Ludwik Fleck, Genesis and Development of a Scientific Fact (1935) presents essentially the same idea. Thomas Kuhn was heavily influenced by Fleck, as was Mary Douglas, and many others.

Broad epistemological assertions are, as they have always been in the history of philosophy, problematic.

Comments for this entry have been closed
Return to top