Association for Public Policy Analysis & Management

Narrow provider networks are attracting widespread attention as a means of potentially controlling healthcare costs. We provide plausibly causal evidence on the effect of a narrow network plan design offered by a large national health insurance carrier in a major metropolitan market. Our econometric design exploits the fact that some firms offer a narrow network plan to their employees and some do not, allowing us to estimate effects that can be interpreted as average treatment on the treated estimates. Our results suggest that narrow network health plans lead to reductions in physician office visits, both for primary care and for specialist care. We also find sizeable decreases in specialist visits as well as laboratory and radiology services and prescription drug use. We found no evidence of increased use of the emergency department in narrow network plans. Spending was consistently lower as measured by both insurance paid and patient paid amounts in narrow network plans.