You're assuming that investors will look at only the risk/return in evaluating who to extend loans to. The credit market doesn't work like that. If you have a lower middle class kid whose parents have missed a few payments on their credit card, he's not getting a good enough credit rating to get loans for the same exact law school that someone from an upper middle class family whose parents have sterling credit will get.
I started to type this same thought, but you said it better.
It's no different than a bank loan to start a business: if you don't have collateral, no bank gives a fuck how good your business plan is, or how much money you could make.
Well, actually, missing a few payments on a credit card and/or not having collateral would indicate risk and be a perfectly appropriate component of a risk/return calculus. But nonetheless, your general point is well taken.
I'm just not sure it's a bad thing. Meaning -- starting a business, much like going to law school, is a risky proposition (which is also why the government shouldn't be in the business of giving loans to private sector businesses -- see, inter alia, Solyndra). Allowing the market to reflect that reality would help potential students make rational decisions about their own risk/cost equation. A bunch of people who shouldn't go to law school, wouldn't be able to go to law school (much as in how many people who shouldn't be able to buy houses, cannot buy houses). Is there an empirical good in having more lawyers? Does the federal government have an interest in ensuring equitable access to professional school to all members of society?
The world's not perfect -- I guess I'd prefer less lawyers from lower-middle class backgrounds than a ton of law school grads who had no business going to law school in the first place. At least when the government tried to encourage home ownership there was an underlying asset of some value (like, you can live in it). A JD without a job is worthless. Either way, there will be suffering and "unfair" outcomes. I'm not sold that one is better than the other.