By Noah Smith

Bloomberg News

One of presidential candidate Sen. Bernie Sanders’ tentpole ideas is the promise of a free college education for all Americans. Sen. Elizabeth Warren, who is also running in 2020, also supports the idea. To a millennial generation drowning in student debt, whose job prospects were severely curtailed by the Great Recession, free college is a tempting prospect. Supporters believe it could also help rectify the immense opportunity gap between the children of the rich and the children of the poor:

But “free college” could mean any number of things, not all of them beneficial.

If you want to make something cheaper, there are two ways to do it: you either help pay for it, or you limit the price that sellers can charge. If “free college” were simply a bunch of government subsidies for tuition, colleges would simply raise prices. We know this is true because something similar has happened before: Widely available subsidized student loans helped push up tuition much more than the cost of other goods. So in practice, “free college” is going to mean price controls.

The typical danger of price controls — that they create shortages — isn’t a problem here, since college spots are already rationed by the application process. But price controls might reduce quality. The U.S. has the most expensive tuition in the world, but it also has most of the best universities in the world. Forcing top public schools like the University of Texas or the University of Michigan to cut their budgets might weaken their ability to hire top professors and provide supportive living environments for students, thus reducing education quality.

It also might weaken U.S. technological dominance and the health of local economies. Many people think of colleges purely in terms of education, but this ignores their function as engines of research. Universities have produced, and continue to produce, many of the breakthroughs that make the U.S. a technological leader. Research, not education, is also the main way that universities benefit the economies of surrounding regions and towns; students tend to leave campus after they graduate, but research stays local, and helps to draw in private investment. Government provides some of the money for university research, but undergraduate tuition also helps fund it.

A final danger of tuition caps is that they could end up benefiting the rich a lot more than the poor and middle class. Currently, many colleges offer need-based financial aid, and the government offers additional assistance such as Pell Grants, while the rich and the upper-middle class are generally forced to pay the full sticker price. Tuition caps would therefore lower prices for the rich while doing little for the poor. But that doesn’t mean tuition caps would make colleges more interested in admitting poor kids; because tuition caps would force colleges to rely more on alumni donations for funding, they might feel pressure to seek out rich kids who would be more likely to contribute money after graduation.

So simply capping tuition is a bad idea, as is simply subsidizing tuition. But what about a combination of both? Sanders’ bill would provide federal funding to states in exchange for them eliminating tuition. This mixed approach is better, but it still runs into some of the same problems. Since the federal government would decide how much funding a university should have, it would stop any individual university from being able to increase its spending on upgrading its educational experience or their research activities (unless it could solicit more private donations). That would make it harder for universities to improve. Additionally, if federal funding wasn’t regularly increased to keep up with inflation, rising research costs, or higher professor salaries, the quality of U.S. institutions could steadily degrade. And the problem of helping the rich rather than the poor would remain.

But this doesn’t mean that all plans to make college cheaper are hopelessly flawed; it simply means they need to be better targeted.

Grants and other non-loan aid for low-income students should be increased. These may push up tuition a bit, but they’ll also shift the cost of college onto wealthier families. Many states are already doing this, and some Democrats in Congress have introduced a bill to do it at the federal level as well.

Second, community college can be made free, by replacing tuition with government funding in a Sanders-esque manner. Since community colleges don’t do research, and since they tend to serve lower-income students, there is little downside to making these free. California Gov. Gavin Newsom is already trying to do this.

Third, states and the federal government can increase funding for public universities on the condition that they limit rising administrative costs, which are a substantial source of rising tuition. And finally, federal funding for research can and should increase significantly, allowing colleges to rely less on tuition to pay for labs and researcher salaries. Recently, research funding has been heading in the wrong direction:

So although centrist Democrats like Sen. Amy Klobuchar are right to say that free college is probably a bridge too far, cheaper college and free college for low-income kids is an achievable and desirable goal. This is a case when big promises should give way to smart policy.

Noah Smith is a Bloomberg Opinion columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.