It is common to hear opponents of single-payer health reform point out that if we bought health insurance the way we bought car insurance, we'd all be broke. Ezra Klein says this is a stupid analogy:


1) You Can Drive Without Swiping Your Auto Insurance: This is the most obvious, and most important, difference. If your car required proof of insurance to start, more people would be insured. But it doesn't. Cops don't even ask for insurance when they pull you over. The only time insurance is relevant is when you're in an accident.

Health insurance mandates, by contrast, make proof of insurance a prerequisite to using the system. To see a doctor, you do have to swipe your insurance card. To enter a hospital, you do need to show proof of insurance. In the Edwards plan, you'll need to enter your policy number on your tax form. If you can't, you'll be enrolled in the basic plan. This alone will reduce noncompliance dramatically, because compliance will be necessary to use the health system. Auto insurance, by contrast, is not necessary to use a car.

2) Car accidents are less likely than flus: Everybody reading this has been to the doctor. Most reading this have never been in a car accident. People are not unaware of the relative rarity of those events. It's easy to skip car insurance under the theory that you will not need it. It will be harder to skip health insurance under the theory that you will not need it. (This is Obama's argument. The danger is that people will skip it until they do need it, as explained here.)

3) We do not subsidize auto insurance: At least so far as I know. By contrast, all of the major Democrats, including Obama, subsidize health insurance to 300 percent or 400 percent of the poverty line. If lower income folks were given help on auto insurance, more would purchase it. As it is, they will be given help on health insurance,and so more will purchase it. This isn't rocket science. It's also one of the useful things about a mandate: The very fact of its universality means a hue and a cry will go up if the government doesn't ensure affordability. Thus, the government has to ensure affordability, or publicly dismantle its mandate. My hunch is it will do the former.



This seems very muddy to me; Ezra is flipping back and forth between talking about using the car, and using insurance. If you restrict the comparison to the use of insurance, most of his objections are moot. Sure, you don't have to show your insurance card to get your car repaired, but they won't let you drive it out of the garage again unless someone has paid; and similarly, they will treat you without insurance, it's just that you'll get an ugly, monster bill.

And while car accidents are less likely than flus, most people should be doing routine maintenance on their cars at least as often as they go to the doctor, which is precisely why free-market types like the analogy: it clarifies the distinction between other types of insurance that are used to cushion against catastrophic events, and the weird way that Americans use health insurance to essentially prepay for largely predictible expenses.

The real reason to distinguish between car and health insurances to my mind is that one of them allows you to do something we think of as optional (drive), and one of them allows you to do something we think of as a right (live).

Nonetheless, I think it's very useful to compare the markets, and policy responses to failure therein. Do mandates work to ensure universality, or something close enough thereto? In the market for car insurance, they do. Most people get insurance because it's illegal to drive without the stuff, even if they're reasonably sure they'll never be in an accident. Yes, some small segment of the population is uninsured. But they're mostly poor or irresponsible people who wouldn't have contributed much money to the system anyway.

That's why, even assuming that the market failures such as adverse selection are actually serious problems, there's no reason to think that we need every single person in the country to be insured in order to fix them. The adverse selection problem will disappear before the uninsured do; markets can tolerate some level of non-compliance.

Remember that almost all of the people you're worried about not complying will almost never show up in a hospital; anyone who's actually sick and uninsured will presumably jump on community rating. So this isn't a problem about system access; few of them will try to swipe that card. It's a finance problem: what we're really worrying about here is that we can't take money from those young, healthy, uninsured people in order to subsidize the older, sicker people who buy insurance. But as I pointed out above, the people who will fail to buy health insurance under the new system are almost certainly the kind of people who have very little ready cash to contribute to the system. So it seems unlikely that the overall effect of their absence would be large.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.