In the United States, as in many countries, local governments often provide fire protection. In general — there are exceptions, but they’re still rare enough to make news — this is a free service, available to everyone who lives in whatever jurisdiction provides it. No one has to sign up or pay for coverage. To most people, I suppose, this is a normal and reasonable thing to do.

One effect of fire protection is to stop peoples’ homes from burning down. As it happens, rich people are more likely to own homes than poor people. And when people with lower incomes do own houses, they are generally less expensive. So, the distributional effect of preventing houses from burning is clearly regressive.

Why should everyone have to pay to keep millionaires’ mansions from burning? Modern apartment buildings probably aren’t even at that great risk of fire, what with sprinkler systems and so on. It’s the big houses up in the hills that are in the greatest danger.

So now comes a new mayor — let’s call him Mayor Pete — who proposes to abolish the municipal fire department and replace it with private fire services that people can contract with. Maybe he’ll take a page from ACA and have gold, silver, and bronze levels of fire protection, sold on exchanges. It’s smart to build on what works, after all.

Naturally there will be means-tested vouchers for poor people to pay for fire protection. Or we can, say, cap the cost of fire protection at some percent of household income, with the difference made up by a subsidy. Just be sure you can fully document your income and assets each year, and don’t forget to fill out the forms. Of course, not everyone needs fire protection — the homeless are free to opt out, and renters can decide for themselves if they prefer a building with fire coverage or cheaper rent.

Obviously, I am making an analogy with free college. And obviously, people who don’t support free college (and probably many who do) are going to reject this analogy. Here are some possible counterarguments:

• Everyone wants to not die in a fire. Not everyone wants to go to college.
• If someone falls through the cracks and doesn’t get fire coverage, the effects can be catastrophic — loss of home and possessions, serious injury, death. If some people end up unable to attend college, that is certainly unfortunate but not a disaster in the same way.
• It’s much more efficient to have a single fire service serving a whole area than to have lots of different contractors providing different levels of coverage in overlapping areas. There would be wasteful duplication of facilities, equipment, and personnel, and in an emergency, confusion about who was responsible for what.
• If one house is allowed to burn, that creates major risks for the houses nearby. Because fire spreads, fire protection isn’t something you can really opt into or out of on an individual level.

These are not unreasonable objections. On the other hand, we can debate how different fire service and college education really are on these dimensions.

Mike Konczal or I might say that they are not really so different — that many of the same practical considerations that favor a single free, universal system of fire protection also apply to college. We might say that higher education is not a luxury in the contemporary United States, and that if measures to keep the rich from getting a free education at a public college end up also excluding some non-rich people — as they inevitably will — that is a major cost. We might say that the machinery of assessing eligibility for various subsidies, vouchers, etc., collecting fees, and excluding or penalizing those who haven’t paid, is immensely wasteful. We might say that the benefits of higher education are social and public, and that these broader benefits are undermined when education is treated as a private good.

Noah Smith or the real Mayor Pete might say on the contrary that there are big differences — lots of people don’t go to college and that’s fine; means-testing is accurate and reasonably efficient, at least compared with running duplicate fire departments; and claims about the importance of higher education to a fulfilling life or a robust democracy are mostly just the self-flattering fantasies of college professors.

Well, we disagree. But however you apply them in this particular case, these all seem like relevant arguments in thinking about how desirable it is to make a public service free and universal.

What they are not, is arguments about distributional impact. There’s no controversy over the distribution of student debt or tuition spending — they rise with income, but fall as a proportion of income. We can debate over whether that makes forgiving student debt progressive or regressive, but I don’t think that’s what’s motivating either side here. Disputes over whether something should be free and universal hinge rather on whether we see it as a fundamental right or a luxury; whether we see the risks of under- and over-provision as symmetrical; whether technical considerations favor provision through a single uniform system; and whether the service is a public good in the traditional sense, and whether it has significant externalities. If we were actually debating the elimination of universal fire protection, these would be the kinds of arguments people would make. Not ones about the direct distributional impact.

The distribution of college spending is quite a bit flatter than the distribution of home equity. So if you don’t oppose free universal fire protection on the grounds that it favors the rich, then I’m pretty sure you don’t actually oppose free public college on those grounds either. Mayor Pete certainly does not have any general objection to public spending from which the rich derive more direct benefit than the poor. Indeed, since public goods are mostly complementary to private goods — roads and cars; airports and airlines; meat inspectors and meat; police and private property — this is probably true of the great majority of public spending, at least if you look at it in the same narrow financial terms that people are looking at college debt forgiveness.

So I don’t think distributional concerns are the real reason that people oppose free universal public college. Presumably the real reasons are some mix of “I think it is very important that everyone is protected from fires, but I don’t think it’s that important that everyone can go to college,” and “Charging people individually for fire protection is impractical, but charging people to go to college seems to work OK.” Which might be reasonable positions! But let’s debate those.

I’d love to have that debate. But I must add that the fact that people who oppose free college keep bringing up the distributional impact, suggests that they may not be confident of winning on other grounds. It suggests that they, at least, don’t believe that most Americans see college education as simply a private good.