Prescription Drug Use under Medicare Part D: A Linear Model of Nonlinear Budget Sets

In a NBER Working Paper, Jason Abaluck (Yale University), Jonathan Gruber (MIT) and LDI’s Ashley Swanson analyze the complicated decision problem faced by Medicare Part D enrollees in their use of prescription drugs. Enrollees must respond to prices that are difficult to find, and that may change as they spend more over time. As in other settings such as income taxation and electricity and cellular telephone markets, these sorts of nonlinear prices make it difficult to estimate behavior, as consumers may not follow rational models of consumption. The authors use Medicare Part D claims data from 2006-2009 to estimate a flexible model of consumption that accounts for these complex prices, that allows enrollees to be imperfectly forward-looking about price changes, and that allows some prices to be more salient than others. They develop methods for estimating consumption behavior using changes in Part D prices between years of enrollment. Their estimates suggest that consumers do respond to price increases by slightly reducing their prescription drug consumption. They also find that consumers are not perfectly forward-looking in choosing their consumption and that more salient plan characteristics affect consumption beyond their effect on out-of-pocket prices. These results suggest that policymakers designing prescription drug plans must account not only for what prices are, but also when consumers encounter them in the year and how they are presented.