This is a follow up to James’ post on whether or not state and federal employees should pay the same premiums for health care. In a simple insurance market people would purchase insurance and their premiums would be based on how likely they are to get sick. There would be no pooling outcomes. To see this suppose there is a pool of individuals, some high risk and some low risk for needing health care. In this pool everybody pays the same premium for their insurance. However, a competing insurance company could lure away the low risk individuals with lower premiums (and higher deductibles) leaving only the high risk people in the initial pool. With the depature of the low risk people from the pool, the premiums would increase in the pool. This is why you’ll hear economists sometimes say that a pooling equilibrium is always broken by a seperating equilibrium. It is the “cream skimming” strategy above.
Now to compound the problem a bit. If we initially have a pooling equilibrium those who are high risk are to some extent benefitting at the expense of those who are low risk. Thus, changing from a pooling to a seperating equilibrium is going to create winners and losers. On top of this there is also the problems of moral hazard and adverse selection. With moral hazard the presence of insurance will alter people’s behavior. For example, if I have car insurance I might drive a little less carefully than if I didn’t have insurance. Sure it would not be good to have a car accident, but I might figure the benefits of speeding up a bit to get somewhere a bit faster might outweigh the possible downside of having an accident. Toss in the fact that people tend to underestimate the probabilities of bad events and overestimate the probabilities of good outcomes (i.e. I think speeding up is really going to help me get someplace faster, and there isn’t (much of) an increase in the probability of an accident if I speed up), and this result should be fairly obvious. Or think of it this way:
- Have you ever over slept and then hopped in the car driving much more frantically to get to work on time for something important?
- Have you ever done this and thought you could have an accident then said to yourself, “Well, I’m insured”?
Basically, when you change the outcomes, you change the behavior as well. This doesn’t always sink in with many people. They think things like, well car accidents are bad…so therefore people are going to drive carefully. But insurance reduces the loss in the bad state (a car accident, getting sick or hurt, etc.). That is a car accident, while still bad is now less bad. The same effect works with health insurance as well. Sure breaking a leg while skiing sucks. It sucks even more if you don’t have insurance. Who do you think is least likely to go skiing? The guy with insurance or the guy without? So the problem with moral hazard is that the mere presence of insurance will change people’s behavior. With health insurance they will engage in riskier activities than they probably would without health care.
Adverse selection is where one type of individual masquerades as another type. With insurance a high risk person might want to try and pass themselves off as a low risk person. Returning to the seperating equilibrium, since the insurance company often cannot observe what type of risk a person is, then the high risk individual will be better off if he can pass himself off as low risk. One response to this is to offer incomplete insurance for low risk individuals that high risk individuals find undesirable.
All of these things complicate the situation noted in James’ post. Currently high risk individuals benefit from the current insitutional arrangement and low risk individuals. However, switching to a new institutional arrangement that requires individuals to pay higher premiums if they smoker or engage in other risky activities could induce some people to opt out of insurance completely. This is especially true when one considers that irrespective of ability to pay, one’s medical costs will be covered (by the rest of us). The same is true for those who are simply high risk (i.e. people who were unlucky to be born with higher risks of say heart disease due to genetics). These people also might opt out of employer provided insurance if their premiums go up. So now with all this we can look at this part of Radley “Randy” Balko’s article,
The question becomes, who should pay the health care costs of a state worker who chooses to smoke, then gets sick as the result of that decision: the person who chooses to smoke, or Georgia taxpayers? I have a hunch what most Georgia taxpayers would prefer.
Some may worry that we’re on a “slippery slope,” here that health insurance companies may soon factor in traits and habits such as obesity, regularity of exercise, or alcohol consumption, too.
But why shouldn’t they? Certainly we should be free to make our own decisions about what habits and vices we choose to indulge. But we should also be prepared to accept the consequences of those decisions. Eat, smoke and drink what you want, but don’t expect people who made better choices or taxpayers to bail you out.
In the ideal world of theory Radley is right. If you decided to eat unhealthy food, drink alcohol (to excess) and smoked, then you’d be a higher risk individual and you’d bear the costs of these decisions either via higher premiums or by bearing the full loss yourself if you opted not to purchase insurance. The problem is we don’t live in that world. When the guy who pigs out on pork rinds, drinks a case of beer and smokes a pack of cigarettes a day keels over of some health ailement and doesn’t have insurance he will still get medical attention. He will be wisked away to a hospital (most likely in an ambulance) and he will get doctors and nurses trying to resolve the medical problem, plus all the drugs and medical procedures. Who will pick up the tab if he cannot or decides not to pay? The rest of us. We’ll either pay via higher taxes, higher health care costs, or both. Further, even without this problem the market for insurance is likely going to result in less than full coverge for at least some people, or what is more generally known as a Second Best outcome (First Best is where nobody can be made better off without making somebody else worse off). This is really a fancy way of saying, that we could, at least in theory, do better.
