Our results cast doubt on the validity of the norms hypothesis--that physicians treat patients in accordance with the average or modal insurance coverage in an area. The one study in the literature directly supporting this hypothesis is misspecified; when more correctly specified, it provides no support for the hypothesis. Two versions of the hypothesis are distinguished. the stronger maintains that physicians do not take account of within-area variation in insurance coverage. This version is firmly rejected by the data. A weaker version says that the physician does take account of within-area variation but that between-area differences in average coverage have an independent effect on demand. Little or no support is found for this version, but our data do not permit very powerful tests. (1978)
Journal of Human Resources
1978
Joseph P. Newhouse and M. Susan Marquis
http://www.ncbi.nlm.nih.gov/pubmed/722070