Buying Data from Consumers: The Impact of Monitoring in U.S. Auto Insurance (Job Market Paper)

jmp_jin.pdf5.39 MB

Abstract:

We study the impact of a voluntary monitoring program by a major U.S. auto insurer, in which drivers are tracked for a short period of time in exchange for potential discounts on future premiums. We acquire a detailed proprietary dataset from the insurer and match it with competitor price menus. Our analysis has two steps. First, we quantify the degree to which monitoring incentivizes safer driving and allows more accurate risk-based pricing. Second, we model the demand and supply forces that determine the amount of information revealed in equilibrium. Structural demand parameters are estimated to capture correlations among consumers' monitoring and insurance choices as well as the cost to insure them. A dynamic pricing model makes the firm's information on driver risk endogenous to prices. We can then jointly characterize information and market structures in counterfactual equilibria. Overall, we find large profit and welfare gains from introducing monitoring. Safer drivers self-select into monitoring, with those who opt-in becoming 30% safer when monitored. Accounting for the resource costs of monitoring and price competition, a data-sharing mandate would have reduced short-term profit and welfare.

Last updated on 07/25/2019