Discrimination in insurance underwriting is regulated at the state level. Surprisingly, there is a great deal of variation across jurisdictions in how and the extent to which risk classification by insurers is limited. Some states expressly permit insurers to consider certain characteristics, while other characteristics are forbidden or limited in various ways. What explains this variation across states? Drawing on a unique, hand-collected data-set consisting of the laws regulating insurer risk classification in all 51 U.S. jurisdictions, this Article argues that much of the variation in state-level regulation of risk classification can in fact be explained by focusing exclusively on three factors: the predictive capacity of the characteristic in question, the extent of the adverse selection problem created if the characteristic is restricted, and the extent to which the characteristic is considered a socially suspect classification.


Insurance Law

Date of this Version


Included in

Insurance Law Commons