Here's Why Health Insurance Is So Weird

Every so often, I write about how Obamacare transfers wealth from the healthy to the sick, and while that's a just thing, we can't necessarily expect the healthy to be happy about it.

When I write this, I tend to get responses from liberals arguing that transfers from the healthy to the sick aren't a policy choice; they're just an obvious characteristic of insurance:

This objection is wrong, and it reflects a misconception that most people have about insurance. This misconception obscures a key policy difference between conservatives and liberals that drives the debate over health care.

All insurance is a transfer to people who experience losses from people who do not. Health insurance is special in that it is a transfer to people who have high expected losses from people who do not. Liberals think this special feature is desirable; conservatives are uneasy about it.

Let's look at a simple property insurance example. Imagine a state where 50% of homes are located on the ocean and 50% are not. The oceanfront homes have a 10% chance of storm damage in a given year and the other homes have a 1% chance. Storm damage always results in a loss of $50,000 and the insurer has a profit margin of 10%.

By my calculations, oceanfront homeowners in this state will pay property insurance premiums of $5,500 and non-oceanfront homeowners will pay $550.

This insurance market will, in any given year, transfer money to the homeowners hit by storms from the homeowners who escape storms. But it will not, in the long run, transfer wealth from low-risk homeowners to high-risk homeowners. Over time, homeowners on the ocean can be expected to pay 91% of all premiums and make 91% of all claims.

Health insurance does not work like this. A variety of subsidies, regulations and norms conspire to ensure that people with high expected health costs - the equivalent of the oceanfront homeowners - pay premiums that are less than their expected claims. Their claims costs are cross subsidized by healthy people - the equivalent of inland homeowners - who pay far more in premiums than they can be expected to make in claims.

There's a challenge with any insurance market like this: How do you get the healthy people (or the inland homeowners) to buy insurance that's a bad deal for them? In the case of health insurance, we use a pile of subsidies, regulations and norms to goad the low risks into overpaying so they can subsidize the high risks.

We give out generous tax subsidies to buy health insurance, require employers to offer group health insurance on roughly the same terms across their employee populations, provide direct government health insurance to the old and the poor, and (under Obamacare) add new subsidies and penalties to push people into insurance. This stuff is all enormously expensive, but it gets most everybody covered.

As insurance markets go, health insurance is totally weird. But it's not totally random. Health insurance has to be unlike other insurance for a few reasons:

• People choose to live on the ocean; they don't choose to have a chronic health condition. So, it's a lot harder to say it's fair to charge higher premiums to people with higher expected costs. (Though, as an aside, this distinction doesn't stop lawmakers from hurricane-prone coastal areas from arguing that property insurance should be subsidized for their constituents.)
• In the case of auto insurance, higher premiums for dangerous drivers are, among other things, a mechanism for encouraging people to drive safely. It's not obvious how much we want to try to induce better health by charging higher premiums to the sick, though employers are doing a little bit of this by charging higher rates to smokers or overweight people who do not engage in fitness programs.
• People can face unexpected changes in health status that make annual insurance contracts less workable for health insurance than other kinds of insurance. Your home is not likely to be suddenly diagnosed with hurricane or earthquake risk. But if you're diagnosed with diabetes or HIV/AIDS or cancer, your expected lifetime path of medical claims will suddenly rise, and so will the premiums insurers would want from you. Current laws try to work around this with guaranteed renewability: Your insurer has to keep renewing your health insurance without a special premium increase even if your health status worsens. But this doesn't help you if you want to change insurers or have to move to another state. Addressing this problem requires some sort of redistribution toward the newly sick.

So, there are good reasons for health insurance to be weird, and to include transfers toward people with high expected losses. But conservatives are uneasy about using health insurance to shift the cost of expected losses for some valid reasons.

• There's a reason we don't prepay for most goods and services through insurance structures: It encourages people to overconsume, since they don't pay (or don't pay much) at the margin for what they do consume.
• While health status isn't chosen, it's also not entirely a matter of luck. Should we worry that shielding people from health cost increases will encourage unhealthy behaviors like smoking and obesity?
• How much do we really want to use public policy to shield people from losses due to factors beyond their control? For example, the ACA is designed to prevent premium variation between men and women, even though healthy women tend to have greater health care costs than healthy men. But often, similar variations go unaddressed or only partially addressed by public policy. Childbirth is more disruptive to women's careers than men's, and how much the government should do to compensate women for this is a topic of major policy contention in lots of countries. Short people make less money than tall people, but we don't have a program of fiscal transfers to the short. I think premium equalization for men and women is good policy, but it's not as obvious as liberals often make it out to be.