[gov3009-l] Mullainathan on 'How We Choose'

Dear Applied Statistics Community, Please join us for the final applied statistics workshop of the semester when Sendhil Mullainathan, Professor of Economics Harvard University, will present 'How We Choose: Medicare Drug Plan Selection', work that is joint with Jeff Kling, Eldar Shafir, Lee Vermeulen, and Marian Wrobel. Sendhil provided the following abstract: Choices increasingly abound for various government supported services, ranging from charter schools to health plans. 24 million elderly Americans have enrolled in Medicare Part D prescription drug coverage during the past two years, and may choose among at least 40 plans. In this paper we examine the informational context in which choices are made and conduct an experiment of information provision, focusing on the decision about whether to switch plans during the open enrollment period in 2006, one year after the program began. We find that most participants obtain their information from mailings from plans and from Medicare. This information is not personalized, although the costs and benefits for a given plan vary greatly depending on specific prescriptions are used. Knowledge of how plans work is low. Personalized information is available by calling Medicare, but most participants do not seek information. Our randomized experiment provided an intervention of personalized information (highlighting the predicted out-of-pocket cost of the current plan and the least expensive plan, and also listing costs of all plans -- based on information about prescription use) in comparison to a group that was provided information about accessing the Medicare website. The intervention group plan-switching rate was 28 percent, while the comparison group rate was 17 percent. The potential cost savings for those affected by the intervention was at least $230 on average. The impacts on switching and potential savings were larger for those with greater absolute and relative potential savings, and for those in small market share plans. The impacts on switching were larger for those initially in low premium plans. We conclude that additional efforts to distribute simple, personalized drug plan information would lead to significant reductions in Medicare beneficiaries' out-of-pocket costs and that the costs of such a program would likely be offset by reduced Medicare expenditures on subsidies to drug plans. I've also attached a draft of the paper. As a reminder, the workshop meets at 12 noon and we provide a light lunch. We are located in room N-354, CGIS Knafel, 1737 Cambridge St. Please contact me with any questions, or suggestions for next semester's schedule Best, Justin