社会违约与计划选择：配偶追随案例

Policy Research Working Paper Social Defaults and Plan Choice The Case of Spousal Following Tal GrossTim LaytonDaniel PrinzJulia Yates Policy Research Working Paper11207 Abstract This paper studies how couples in the Medicare Part D pro-gram choose an insurance plan. Over 70 percent of enrolleeschoose the same plan as their spouse. Even among thosewith differing health care needs, well over half do so. Dis-crete- choice models suggest that beneficiaries place a value as their spouse. Using a regression-discontinuity design,the paper shows that younger spouses disproportionatelyfollow their older spouse’s plan choice. Joint plan choicecontributes modestly to overall overspending, but increases Social Defaults and Plan Choice:The Case of Spousal Following Tal Gross† JEL codesI13, H31 ∗We are grateful to Keith Ericson, Sandra Gain, Jim Rebitzer, ´Agnes Szab´o-Morvai, Roula Yazigi,and seminar participants at Lehigh University, Purdue University, Stanford University, the Univer-sity of Georgia, Georgia State University, the University of Ottawa, the University of Pennsylvania,the University of Virginia, the University of San Diego, the University of Duisburg-Essen, and theCERS Institute of Economics for useful feedback. Amanda Stiebris provided superb research assis-tance. Research reported in this publication was supported by the National Institute on Aging ofthe National Institutes of Health under Award Number P30AG012810 through the National Bureau †Boston University and NBER, talgross@bu.edu‡University of Virginia and NBER, timothyjlayton@virginia.edu§World Bank, dprinz@worldbank.org¶University of Michigan, jryates@umich.edu Default options guide many important economic decisions.Many consumers, forinstance, choose a retirement-savings plan passively by opting into whatever plan isthe default option (Madrian and Shea, 2001; Beshears et al., 2008). Similarly, manypeople choose whether or not to become organ donors based on whether or not they Other decisions require an active choice. Americans are not automatically assignedto a car insurer. They must choose one if they want to drive legally. In many contexts,the choice of a health insurance plan is similar:consumers are not automatically Such active choices may facilitate better decisions (Carroll et al., 2009; Bernheimet al., 2015; Goldin and Reck, 2022).Without automatic enrollment in a defaultoption, perhaps consumers are forced to weigh all options equally and so are more Then again, the absence of a default option does not imply that consumers makesuch decisions in a vacuum.They may use heuristics to make these decisions.Inparticular, consumers observe the choices of their peers, and their peers’ choices may act as a “social default.”1Instead of a social planner singling out a particular option For complex products, social defaults may be problematic.It is one thing ifconsumers choose the same brand of packaged food as their neighbor (Bronnenberget al., 2012) – it is quite another if they choose the same insurance plan. Insuranceplans are often horizontally differentiated, and so the optimal plan for one person In this paper, we study an active choice over a complex product in which mostconsumers are not initially offered a default option. Medicare recipients who join the Medicare Part D program must choose a private drug plan.2New Medicare recipients them.3 We study whether the choices of their peers affect their own choice of plan. To do so, we focus on a person’s closest peer, their spouse. Our empirical analysis proceeds in three steps.First, we present a descriptiveanalysis to measure the share of couples on the same plan. Over 70 percent of couplesin the Medicare Part D program are enrolled in the same plan. One driver of thatbehavior may be homophily:couples tend to have similar prescription-drug needsand preferences. And yet, even among couples with very different healthcare needs, Second, we estimate discrete-choice models. The models allow beneficiaries’ will-ingness-to-pay for each plan to vary according to their own expected out-of-pocketcosts in that plan. We include in the model an indicator for whether the beneficiary’sspouse is enrolled in a given plan, allowing us to estimate how much the beneficiaryis willing to pay to be on the same plan as their spouse. Those models suggest that younger spouses value being on the same plan as their spouse at about 76 percent ofthe value they place on remaining in the same plan from one year to the next (that The third step of our analysis takes a quasi-experimental approach involving a pairof natural experiments. Humana, a large insurer, twice introduced new plans to themarket that immediately became very popular. Via a regression-discontinuity design,we find that those who joined Medicare Part D just before the new Humana planswere introduced were less likely to join those plans as compared to those who joined We then study younger spouses who enter Medicare once the new Hu