Purpose: While standard economic theory posits that more options should generally make consumers better off, studies examining choice of everyday consumer goods have called this into question. This has not been examined for more consequential choices. The purpose of this study is to examine how Medicare beneficiaries respond to health plan choice sets of various sizes. Specifically, its goals are to a) examine whether more plan options increase the probability of switching plans, b) understand why beneficiaries respond to choice as they do, and c) explain the quality of Medicare beneficiary decisions.

Methods: We utilize a sample of 1,106 Medicare beneficiaries who reside in high plan penetration areas and are currently or recently enrolled in a Medicare health plan. Data collection is via the internet and participants are drawn from Knowledge Network's nationally representative panel. Participants were given a hypothetical choice experiment that offered the opportunity to switch to a new health plan, whereby the size (2, 4, 8 or 16 options) and makeup of the choice set were randomized. Subjects were shown information about the available options that mimicked the presentation of the Medicare program.

Results: While some choice is clearly beneficial to beneficiaries, consumers do not show a monotonically increasing probability of switching plans as the choice set rises; after 4 plans, there is no statistically significant increase in switching. This appears to be driven by beneficiaries' uncertainty with regard to their preferences at the lower end of the choice set and being cognitively overloaded at the higher end. Additionally, even though participants engaged in the choice, some still violated basic decision rules, calling into question the quality of their decisions. The order in which the plans appear, as well as the introduction of dominated options to the choice set, significantly influences choice; consumers also show evidence of making intransitive decisions. These violations were not limited to the larger choice sets.

Conclusions: Medicare beneficiaries face a large amount of plan choice. This study calls into question whether this choice is always in the best interest of the beneficiary and has implications for consumer involvement more generally. More research is needed to advance the above results, as well as to begin the development of tools or other methods to assist consumers with such choices.