Education, or, What do people do?

David Glenn in the Chronicle asks 'what explains the growing gap in wages?' The arguments covered run that the number of people going to college hasn't kept pace with the demand for skill by employers, and that the slowing of educational attainment has caused wage inequality to increase. I'm biased towards any argument for universal education, even up to the college level (whether this would really increase the level of 'skills' is another matter, but I think that's secondary), but I'm also a little skeptical that massively expanding higher education would bring down wage inequality in the way Glenn and his cited studies speculate.

First we have to worry about the usual signaling story - is college just an indicator that someone's smart rather than something that makes them smarter? - but more than that, what would people do? What do people do? The 2000 US census has this (pdf link):

Management, professional and related: 33.6%
Service: 14.9%
Sales and office: 26.7%
Farming, fishing and forestry: 0.7%
Construction, extraction and maintenance: 9.4%
Production, transportation and material moving: 14.6%

'Production occupations' on their own make up 8.5%. Now, a potted history of modern development might go something like this:

1) agriculture; we don't all have to be farmers anymore, let's start making stuff: industrial revolution
2) cheap global trade in stuff; we don't all have to make stuff anymore, let's do professional services
3) now what?

If you are even willing to entertain that kind of a story, you have to wonder what it would really mean in the labor market to hypothetically educate everyone to college level (again, I like it for its own sake; not trying to argue against education). Buying the skills angle from the Chronicle article probably means believing that if we educate everyone they can all get the kind of Wall Street/banking/consulting/legal jobs that are typical of the direction OECD economies are heading in, right? Perhaps some of the scientists can do tech research, but for a bunch of college graduates you're looking at a lot of things that don't exactly scream 'tangible production'.

That's not quite the same as the argument the Chronicle's citations are opposing, which is to say that wage inequality has gone up as demand for skilled labor has fallen; it's more to say that the whole production side of the US economy is just a whole lot different than it was when wages were more equal. We've got a chicken/egg problem: if we educated everyone, would
a) everyone, and thus the country, be more productive, and if so in what industry;
b) the wage gap close by bringing the bottom up and the top down;
c) a lot of college graduates work in unexpected places;
d) the breakdown of occupation change in some way, and if so in what way?
The last one would be very interesting, but which way would it go? Would we simply have yet more bankers and consultants, or would there be a more fundamental shift? Is it possible to get by on a smaller service sector, or would we end up with the same service sector populated by college grads?