Managing Behavioral Hazard: Value-Based Insurance Design and Inertia

Thumbnail Image
Degree type
Doctor of Philosophy (PhD)
Graduate group
Health Care Management & Economics
Grant number
Copyright date
Related resources
Hua, Lynn, Mei

Health insurance can be used as a mechanism for more efficient health care decisions. While value-based insurance design (VBID) aligns cost-sharing with clinical value, whether consumers reduce their medical expenditures is unclear. I study the impact of a new value-based insurance design which decreased copays for primary care physician visits, increased copays for specialist visits, and introduced negative cost-sharing with preventive care incentives to reduce the deductible. I find consumers are persistent in their plan choice and there is entry of younger, new employees intothe VBID plan. Old subscribers defaulted into VBID have a greater number of PCP visits while new employees who actively choose VBID have a lower number of specialist visits compared to non-VBID subscribers. To study the demand for this new design and how selection and treatment effects interact with consumers experiencing inertia, I estimate a model of plan choice and level of deductible and investigate responses to counterfactual plan menus which i) reduce the number of plan options, ii) lower the switching cost to zero, and iii) mandate enrollment in the value-based plan. By switching to the value-based plan, enrollees can reduce their premium paid by as much as $4,351 with moderate expected increases in out-of-pocket payments of $85 for subscribers with good health and $245 for those with poor health, on average. These results highlight the importance of active choice coupled with decision aids, targeted information about coverage changes, and strong financial incentives to motivate changes in consumer behavior.

Lucarelli, Claudio
Date of degree
Date Range for Data Collection (Start Date)
Date Range for Data Collection (End Date)
Digital Object Identifier
Series name and number
Volume number
Issue number
Publisher DOI
Journal Issue
Recommended citation