Sunday, September 16, 2007

An Experiment in Adverse Selection

An article in Friday's Wall Street Journal, written by Chad Terhune, describes new health insurance policies being offered by some insurance companies. I can't give a reference to the WSJ piece, but another paper reports on it here. The ones offered by American Community Mutual Insurance are particularly interesting. They are targeted at the "healthy young" and offer low annual premiums, around $1000 per year. That doesn't get you much, however, as there is a rather small maximum benefit and a large deductible. What it does get you is the option to buy a much larger benefits cap -- up to $5 million -- if you do get sick and want to initiate the higher coverage. "Coverage on Demand," the company says. Of course, that additional coverage is expensive. It HAS to be, because only the ones who buy it will need it! So given the adverse selection that has to happen, is there a price that will allow the company to break even at least, and that will attract some customers into the program? Or at any price, will the only customers who sign up be so costly that the company will have to lose money? The higher the price of the "coverage on demand," the sicker and more costly will be the individuals who sign on. Or, maybe, the individuals who get sick will have enough uncertainty and be so risk-averse that they will buy expensive coverage even when their medical bills will not be all that large.

There is a large market of rational individuals who find medical insurance too expensive and therefore go uncovered, so I can understand the experiments at attracting them. And there are a lot of new plans coming out, especially as some states like Massachusetts require insurance. The fine print is going to lead to a lot of litigation, I predict: what maximum benefits really are, whether they were disclosed, non- covered conditions, etc.

Great case to cover in a class on the classic information economics problems of adverse selection and moral hazard.


Anonymous said...

Can you comment on whether adverse selection operates at all in a universal system? If no one "elects" into the pool (they are all automatically selected) does that mean adverse selection can't happen. It still seems like there would be a lot of lurking "lemons" in a universal pool. Or is that just info asymmetry due to lack of transparency on health issues? Please, set me straight.

Robert G. Hansen said...

Well, if it is truly universal, there should be no adverse selection. That is generally the advantage to pools in insurance -- by employer, or all old folks, etc. But if there is any selection in or out, there is the potential for adverse selection. In many cases of near-universal coverage this will be a small problem and the costs of it will be overwhelmed by the amount folks are willing to pay to avoid risk. I think that is what these "insurance on demand" policies are banking on.