## February 03, 2005

### Tort and Insurance; Skating Example

We had a very good discussion at the law-and-econ lunch today of tort liability. We went over old ground, but I think I understand it better now.

Suppose we have a skating rink that has a cost of \$30 per person who skates and competes with other rinks that have identical costs. In addition, there is an average probability of 1/1000 that a skater will be injured and suffer hurt worth \$25,000. Here are two possible legal rules:...

...

1. NO LIABILITY. If the person gets hurt, the rink pays nothing. The result will be a ticket price of \$30.

2. LIABILITY. If the person gets hurt, the rink pays the \$25,000. The result will be a ticket price of \$55 (=30+ 25000/1000). In effect, each skater is required to purchase a bundle of two goods from the rink: skating, and insurance.

In either case, the rink continues in operation and people continue to go to it. So if nothing else is brought in, either system is efficient.

One thing that matters is negligence by the ice rink. If it is negligent, then that makes LIABILITY more efficient. Negligence by the skater also matters, and makes NO LIABILITY more efficient.

PROBLEM 1. If a skater already has his own insurance, then under current U.S. law, he collects double under LIABILITY. He will get \$25,000 from the insurance company and another \$25,000 from the skating rink. As a result, he will have too great a tendency to skate even when he knows he is more accident prone than average, and will not be as careful as he should be on the ice. Under NO LIABILITY, on the other hand, he is still insured, and the moral hazard problem is less, though still present (and, actually, coinsurance and deductibles will help out).

PROBLEM 2. Under LIABILITY, 1/3 of the \$25,000 payment goes to pay the skater's lawyer, and a similar sum must be paid to the rink's lawyer. Thus, this form of insurance-via-litigation is much more costly than ordinary insurance.

PROBLEM 3. As with any insurance, the insurance-by-litigation gives rise to adverse selection and moral hazard. These are the two things I mentioned in (1) (perhaps the same problem, but it arises even if the skater does not have separate medical insurance). High-risk skaters will skate more, and all skaters will take less care. Moreover, if high-risk skaters are more likely to accept the \$55 package, low-risk skaters will drop out, rinks will be forced to increase the price to above \$55, more low-risk skaters will drop out, and there will be a lot of inefficiency.

As an example of this, there might be a skater with a very low probability of accident, so the insurance is only worth \$1 to him, not \$25. If his benefit from the skating is \$40, then under NO LIABILITY he will skate, at a price of \$30 and a net benefit of \$40-1=39, which is efficient-- there is \$9 in gains from trade. Under LIABILITY he will not skate, because the price is \$55 and the net benefit is still \$40+1 = 41. The gains from trade are lost.

Posted by erasmuse at February 3, 2005 12:39 PM