Medicare's prescription drug benefit (Part D) has been its largest expansion of benefits since 1965. Since the implementation of Part D, many regulatory proposals have been advanced to improve this government-created market. Among the most debated are proposals to limit the number of options, in response to concerns that there are "too many" plans. In this paper we study the welfare impacts of limiting the number of Part D plans. To do this, we first provide evidence that consumers view Medicare Part D plans as differentiated products. In doing so, we determine how much Medicare beneficiaries value the plans' various features -- an important measurement not only for our analysis, but also because these features are heavily dictated by policy. Second, using our demand- and supply-side estimates, we conduct several policy experiments to understand the implications of reducing the number of plans. Specifically, we assess the effects on equilibrium premia and welfare from removing plans that cover "the gap," reducing the maximum number of plans each firm can offer per region, and, for validation purposes, the impact of a recent major merger. Our counterfactuals regarding removal of plans provide an important assessment of the losses to consumers (and producers) resulting from government limitations on choice. These costs must be weighed against the widely discussed expected gains from limiting options (due to expected reductions in consumer search costs) when considering new restrictions on the number of plans that can be offered. We find that the search costs should be at least two thirds of the average monthly premium in order to justify a regulation that allows only two plans per firm, and that this number would be substantially lower if the limitation in the number of plans is coupled with a decrease in product differentiation (e.g., by removing plans that cover "the gap").
You may purchase this paper on-line in .pdf format
from SSRN.com ($5) for electronic delivery.
Machine-readable bibliographic record -
MARC,
RIS,
BibTeX
