One possible way to mitigate this problem is a Pigouvian tax on signalling (perhaps compensated for by a reduction in income tax).
Signalling is a lot like standing up at a baseball game. It gives you a great view, at the cost of having your legs get tired. Once everyone else starts doing it, your view is no better than before, and everyone is a little more tired than before. But no one can sit down, because they don't get to see the game.
A tax on signalling might reduce the benefits and thereby reduce the amount of resources wasted. Additionally, the people who don't engage in signalling would enjoy reduced taxes.
To begin, I'd propose an education tax, since education seems to be our nation's most expensive form of signalling. You want an education? Fine, but be prepared to pay an extra $5-10k/college year to compensate the people who don't get a job as a result of the degree inflation you cause.
(Obviously, ending subsidies for higher education would also be a good start.)
The problem with a fixed cost is for plenty of poeple an extra 10k/year is meaningless but for others it's a huge deal. A possibly better approach would be a 1% increase in your income tax (for life) if you get a Masters/MBA/etc. The net effect is similar (3 mil lifetime earnings = 30k in extra tax). But it scales with income so it attacks the value of signaling.
The harm to society is also fixed, so charging a % of profits is a strange way to tax. Think of degree inflation as being a form of pollution.
Now think directly about pollution taxes. Emitting 1kg of pollution to create an iPod is not a bad thing - iPods are more valuable than having 1kg less pollution in the air. Emitting 1kg of pollution to make a sandwich is bad, however, since the value of a sandwich is much lower than 1kg less pollution. The end goal is to pollute only while making iPods, not sandwiches.
Charging as a % of profits or revenues would not create this incentive.
Similarly, the goal of an education tax is to create a disincentive for acquiring a degree for people who won't gain a lot from the degree.
"The harm to society is also fixed". No it scales with the amount of productivity lost during that time period. A football player that get's a masters before going to the NFL has a greater cost and a lower benefit than a teacher that get's a masters before teaching high-school English.
Pollution has a cost and production has a benefit. If $BENEFIT - $COST > 0, we want more pollution.
The problem is that since pollution is an external cost, the producer/consumer has no incentive to accurately make this calculation. As long as $BENEFIT > 0, the producer/consumer will pollute, and society will pay $COST.
The point of the Pigouvian tax is to give the producer/consumer an incentive to make this calculation correctly, and to compensate the rest of society for the costs they impose.
In order to obtain a job, you need to satisfy an employer's stringent and sometimes arbitrary requirements. In order to do so, you must spend extra years of your life in school.
That's a naive analysis - given a pool of unknown candidates (e.g a stack of CVs from people unknown to the hiring manager) he or she will obviously look for "the best" candidates, and lacking any other indication (e.g. from fresh grads with no work experience) will go for the best qualified. It's not the case that employers are driving up the academic requirements - they are merely responding logically to a more academically qualified pool of candidates.
On the one hand, the spirit of inquiry and promotion of learning that higher education (should be) driven by is an excellent thing for both "productivity" and society as a whole.
On the other hand, the idea that colleges exist merely as a credentialing system, and the intellectual complacency that pervades much of academia since it can extract such high rents as a gatekeeper, result in a vast misallocation of resources.
Untangling the two (education vs. credentialing) will take time, but with the enormous debt-loads of graduating students these days, I find it hard to believe that today's college students will accept sending their own children through an educational/credentialing system that resembles today's status quo.
China's "education inflation" problem is not like ours. The problem in China: it's an underdeveloped economy. This may change in 20 years-- I think everyone in the world hopes it does-- but right now, there are so few middle-class jobs available in countries like China, Vietnam, and the Philippines. The competition is incredibly tight and (as anywhere else) mostly on socioeconomic status and connections, not education.
Our problem is different: technology and the national job market have created The Résumé Blizzard. The cost of sending a CV to an employer has dropped from about $1 (today's dollars) in postage, nice paper, and printing, to $ε in electron-pushing. Most HN posters are relatively successful in their careers and probably send out 3-5 CVs per job search, but there are unqualified lottery-players who send out hundreds or thousands. Front doors don't work anymore-- "closed till the Blizzard ends"-- and colleges and universities now provide a very expensive sorting mechanism for entry-level jobs. The truth is that job-matching is a difficult unsolved problem that no one has solved very well, and the universities and business schools are, at the entry level, the best solution.
What I don't like about this idea is that, seeing as companies have scads of money and job-seekers generally don't have any, it seems unfair to put fees on job applications. Unless they were limited to a certain level (say, $5.00) I wouldn't support it.
What I would do is create a job-search site where each application costs 1 credit, and encourage employers to spend a lot of time using my well-filtered site. Each member gets 12 credits for free every year, and more credits can be purchased via a premium membership. As the site branched out into a career site, with discussion boards and networking, credits could also be given as a bonus for positive contributions. Limiting the application volume solves the spam problem, though.
Signalling is a lot like standing up at a baseball game. It gives you a great view, at the cost of having your legs get tired. Once everyone else starts doing it, your view is no better than before, and everyone is a little more tired than before. But no one can sit down, because they don't get to see the game.
A tax on signalling might reduce the benefits and thereby reduce the amount of resources wasted. Additionally, the people who don't engage in signalling would enjoy reduced taxes.
To begin, I'd propose an education tax, since education seems to be our nation's most expensive form of signalling. You want an education? Fine, but be prepared to pay an extra $5-10k/college year to compensate the people who don't get a job as a result of the degree inflation you cause.
(Obviously, ending subsidies for higher education would also be a good start.)