How much of the college education premium is payoff to useful human capital?

[KrazyHorse]

Of the remainder, how much is due to simple selection bias and how much is due to the signaling effect of a degree?

These are key questions because, while it's apparent that a college education does pay off (naively, assuming no selection effects) it is not at all apparent that higher education levels as currently seen in the First World are actually socially useful.

My guess is as follows: of the naive aggregated college education premium, 10% is due to useful human capital accumulation, 60% is due to selection bias and 30% is due to signaling.

Other guesses? How would you disaggregate these effects? Is anybody aware of good empirical work on this issue? Have I missed any important contributory effects (i.e. is there a whole other category I could include)?

[Dracon II]

Here's a statistic for you... the unemployment rate.

[Drake Tungsten]

What metric would you use to measure usefulness?

[KrazyHorse]

Dracon, go somewhere else. This thread is for people who want to seriously consider this specific issue. Not for morons who get sucked into Drake's trolls.

[KrazyHorse]

Drake:

Ideally, the metric would be the marginal change in productivity of the whole society (marginal with respect to a single additional college grad, integrated over the lifetime of the grad, NPVed etc)

One key point is that you can't simply disaggregate the selection effect, as there is a signaling component; when I get a better job due to my college degree (or due to my GPA or the brand name) some of this premium is being paid by the person I displace.

It must also be noted that signaling is not purely wasteful, as it does let businesses more efficiently place better candidates. Without the signaling of college they would need to go through another discriminant, e.g. job performance during a review period. This would itself involve costs...

[Dracon II]

According to my screen Drake posted one minute after me... so evidently the vortex which his troll created (into which I was evidently sucked) must have somehow warped time.

[Dracon II]

OR... I'm perfectly capable of being a trolling moron of my own volition and initiative


But seriously... I was making a serious (if somewhat cynical) point. Though I am drunk.

[KrazyHorse]

According to my screen Drake posted one minute after me... so evidently the vortex which his troll created (into which I was evidently sucked) must have somehow warped time.

You're a ****ing moron if you don't know what I'm referring to.

[Dracon II]

Well, I guess I'm a moron, 'cause I don't see any posts/trolls before mine.

[KrazyHorse]

I'm a moron

Yes.

[Dracon II]

I must REALLY be a moron, because I had NO idea you were going to twist my poor widdle words like that!

[Jon Miller]

I would like to say that a college education is important in creating an educated populace. A populace that can appreciate art, understand a basic level of science, and determine things rationally.

I have doubts, but this might even be the case.

https://www.collegeboard.com/prod_downloads/press/.../EducationPays2004.pdf

JM
(I know that this does not deconvolute things at all, really.)

[flash9286]

It must also be noted that signaling is not purely wasteful, as it does let businesses more efficiently place better candidates. Without the signaling of college they would need to go through another discriminant, e.g. job performance during a review period. This would itself involve costs...

I think this is key. Are there more efficient ways that society can let employers to discriminate. A college degree from a good school signals that you are reasonably smart and have a good work ethic. For most jobs it seems that simple IQ test plus a probationary period (to test work ethic) would be a less costly way to discriminate on a social level than the college system. Of course, it costs employers nothing to use college as a signal since they aren't paying for it.

On a related note college is more useful in some areas than others. For example, I'm guessing an aerospace engineering degree would be useful if you got a job at Lockheed working on planes, but would it be more cost effective to have on the job training plus other training the employer wants to give plus experience plus a long term K with a non-compete.

[Kuciwalker]

Of course, it costs employers nothing to use college as a signal since they aren't paying for it.

It costs employers a huge amount of money; however, they didn't get to decide to spend it.

[flash9286]

How so? Are you talking about through increased taxes.