rate suppression; see Harrington (1992) for a discussion of this term. We use the term rate regulation to refer to compression and suppression together. Rate compression is illustrated by California's Proposition 103, which stipulates that, without the additional approval of the insurance commissioner, passenger automobile insurance rates may apply only the following three factors: 1. The insured's driving record 2. The number of miles driven annually 3. The number of years of driving experience Such characteristics as the driver's place of residence, age, sex, and marital status could no longer be used without the approval of the insurance commissioner. These factors were frequently used by insurance companies prior to the passage of Proposition 103.3 Rate suppression is used in Massachusetts, where premiums in many categories are explicitly set below the actuarial cost-a system called tempering. Blackmon and Zeckhauser (1991) report that in Massachusetts, expected costs vary across drivers by a factor of 4.4, but premiums vary by only a factor of 3. Expected costs vary across territories by a factor of 2.7, but premiums vary by only a factor of 2. Insurance companies, of course, have an incentive to reject customers who must be charged suppressed rates. Since auto insurance is mandatory in all states, rejected customers still need insurance, which is generally provided through assigned risk pools. Drivers who are denied auto policies are placed in the assigned risk pool and charged premiums that may be below the actuarial costs. Each auto insurance company in the state is then required to take a share of the assigned risk pool equal to its share of the overall market.4 Standard welfare economics provides no simple explanation for regulatory pressures in this market. There are two factors that may work in the direction of welfare enhancement. First, in a world of imperfect information and costly sorting, rate compression could work to curtail a tendency to form (socially) too many risk categories. Second, by lowering premiums, rate suppression could induce previously uninsured drivers to purchase coverage, thus eliminating externalities associated with uninsured motorists. On the other hand, as with any cross-subsidization scheme, there are obvious welfare losses associated with rate regulation. Drivers who are charged premiums above their true costs will underconsume driving, auto insurance, or both, and vice versa. The net welfare effect of regulation is thus far from clear. Given this, we raise the question, Why has the auto insurance industry emerged as a primary target for increased regulation? The main goal of this paper is to try to answer this question by linking the possible sources and con