Joining the Spence Club

Bryan Caplan wants to start the Spence club, for economists who believe in the signaling model of education. Sign me up. Here's some  evidence from the UK that corroborates signaling theory.
1. Men who went to private school earn 4.4% more (pdf) a year, on average, than oiks, even controlling for qualifications, university and subject studied. This is difficult to reconcile with human capital theory,  as the result controls for qualifications and experience. But it's consistent with signaling - coming from a posh school signals that one has good attitudes that employers value.
2. Men (but not women) with arts degrees earn less (pdf) than if they had not gone to university at all. Perhaps this is because an arts degree signals that a man is airy-fairy. But as studying arts is a suitable subject for girls, no such adverse signal is sent by female arts graduates.
3. Returns to many vocational qualifications - but not degrees - are zero in the public sector, and returns to NVQs are generally negative.  This could be because although such qualifications help build human capital, they also signal that their holders were not smart enough to pursue general education.
4. University drop-outs earn less than those who never went to university at all. This is inconsistent with human capital theory, which predicts that two years of university education should be roughly as good as two-thirds of a degree. But it's consistent with signaling theory, which says that dropouts signal a lack of application and self-discipline, and that the value of a degree consists in the signal sent by the piece of paper.

Comments

[dearieme]

Technical question. "Men who went to private school ... even controlling for qualifications, university and subject studied." If there is a pronounced correlation between going to private school and (a) university attended, and (b) subject studied, how easy is it to be confident that "controlling" is a job well done?

[john b]

I'm pretty convinced that the signalling effects are real. However, it's also worth noting that individuals' lifestyle choices are likely to reflect their personal values and preferences:

2) a bloke who chooses to study English is likely (not certain, but likely) to be less motivated by the prospect of high earnings than one who chooses to study economics.

4) someone who drops out of university is likely (not certain, but likely) to be less motivated by 'conventional' measures of success than someone who completes their course.

i.e. if you don't view making lots of money as a key goal in life, you're much more likely to choose a degree course that will mean you don't make lots of money.

3) seems like the best example of 'pure' signalling.

[chris]

Dearieme - it all depends on the size of the correlation. If it were one (say, Oxbridge comprised nothing but public schoolboys) we couldn't distinguish between a university and school effect. Luckily, correlations aren't so high, except in a few instances (classics?) that don't detract from the main point.
John B - you're entirely right. But these explanations aren't mutually exclusive. The dropout might well be less motivated - but he's certainly signaling this much to employers.

[dearieme]

Ta. Here's a story about signalling. (But it's about Australia and I was told it by a Kiwi, so I don't vouch for it.) Chinese youngsters in Oz apply to medical school because it's hardest to get in to. They graduate, so proving a point, but don't then practise medicine but instead go to work for daddy's business. Thus there's a shortage of medics in Australia. Good story, eh? Does it contain any truth, though?

[Meh]

I think a better question to ask ourselves would be why it is given all this "efficient signaling" that UK business is rather well known for being bad at picking out talent, particularly managerial talent. Most studies show that UK businesses taken over by US firms show increased productivity and other signs of "better management."

[Jim]

Isn't (1) consistent with those who went to a posh school simply knowing more people who are likely to be in a position to set them up with a good job? Suppose you could argue that this is just another form of signalling ("He's one of us, look where he went to school"), but it seems to me it's more about different search costs due to different social networks.

[ad]

'I think a better question to ask ourselves would be why it is given all this "efficient signaling" '

Who said anything about efficient signaling? A signal that takes a minimum of three years to send is surely extremely inefficient.

We need cheaper signals.

[youngman]

I agree with your overall point, but disgree with 3 of your examples.

My gut feel generalisations are

1. Private school helps to make its students more rounded.

2. Males doing arts degrees don't just seem more airy fairy they are! Not surprisingly airy fairyness does not help one to make money.

4. Uni drop outs often have issues (committment etc) , not surprisingly these efect earnings.

Still your on the money. Uni mostly just signals your intelligent (ish) and can committ to something.

[Luis Enrique]

But there are also plenty of facts that are inconsistent with a pure signalling model of education too (I don't have the papers to hand, but I'm sure you could dig them out with little trouble). So education involves an element of signalling and an element of skill acquisition. Why the need to divide into camps?

[Steve]

This Grauniad article reckons that History graduates are over-represented on the boards of the top 100 companies:

http://education.guardian.co.uk/higher/news/story/0,,1531834,00.html

I'd guess that most of those would be men.

[Katherine]

"an arts degree signals that a man is airy-fairy. But as studying arts is a suitable subject for girls, no such adverse signal is sent by female arts graduates.". Perhaps this signals that employers are tied into gender stereotypes that have little or no basis in reality. There is nothing rational about thinking "man doing girly subject = bad thing".

[dearieme]

Spot on. And there is nothing rational about thinking "girl doing boyish subject = good thing" either.

[john b]

Steve - but that'll be a case of "returns to *Oxbridge* degrees", which will have a very different earnings profile from the population overall.

[Steve]

John, do you know that they are mostly Oxbridge history graduates or is that just an educated guess?

[Uk.cv.com]

Great Post, you’ve done a very nice article, thanks a lot.