Wednesday, October 22, 2003

An Insurance Primer

When talking about the issue of health insurance, I'm startled about the degree of ignorance about just what insurance is for. Leaving aside the issue of how insurance should be provided, I've noticed that there are plenty of people who seem to object to the very concept of insurance. It's as if they see insurance, even private insurance, as an opportunity for people to get something they didn't pay for.

I. Insurance in General

Why do we buy insurance (Life, Auto, Homeowner's, Health, etc...)? The reason we buy insurance is that we don't like risk and uncertainty. Individuals are risk averse - that is, all things being equal, we prefer certain outcomes to uncertain ones. Suppose you had two choices -- either you could choose a job which gave you an income next year of $100,000, or one in which you had a 50% chance of receiving an income of $50,000 and a 50% chance of receiving an income of $150,000. Risk aversion leads to individuals choosing the former. The expected, or mean, income in both cases is the same, but the second choice increases uncertainty.

We buy insurance to remove risk from our lives. We care about it - but insurance firms with a lot of customers don't. If we think about, say, theft insurance - if a company has a portfolio of 10,000 customers, each of whom has a 20% chance of having $2,000 stolen from them - then an actuarially fair price for theft insurance would be .2*2,000 or $400. With enough customers, the insurance company faces no risk - about 2000 customers get robbed in any given year, give or take. By having a pool of customers, the insurance company has no risk (well, perhaps some -- see tomorrow's entry in this fascinating series), and the insurance buyers have paid money to remove the risk from their lives. Each year they will "lose" $400 to the insurance company, and if they get robbed the company will cover their losses. Obviously, in the real world actuarial tables are more complicated than this, but the basic rule applies.

II. Why Insurance is Different

The market for insurance can, roughly speaking, satisfy some of the requirements for being a competitive market, but nonetheless the insurance market will always be different from other markets for two key reasons - moral hazard and adverse selection.

Moral hazard is the more obvious one. Insurance doesn't just remove the risk of things we have no control over - it also removes the risk of things we do have some control over. If we don't suffer the consequences of riskier behavior, we'll be more likely to engage in it. The standard example is that collision damage loss waiver on that shiny new rental car. Check that box, and beat the hell out of the thing. Who cares? It's a rental and it's insured.

Adverse selection brings out another key difference of insurance - the fact that insurance markets operate under imperfect information. Generally, the buyer of insurance has a better idea how risky his/her behavior is, and therefore how costly they will be on average, than does the seller. Let's assume there are two types of individuals - high risk and low risk. High risk individuals have a much greater probability of experiencing a costly event than do low risk. People know their own types, but the insurers do not. So, an actuarially fair (and competitive) insurance price would be based on the average level of risk. The problem is, this price is a *very* good deal for high risk types, but generally a *bad* deal for low risk types. In other words, only high risk individuals buy insurance and low risk individuals don't.... but, since only high risk individuals buy insurance, the "actuarially fair" price will reflect the fact that the risk pool only includes these individuals.

Continuing tomorrow with why health insurance is different...