Consumers' enrollment decisions in Medicare Part D can be explained by Abaluck and Gruber’s (2011) model of utility maximization with psychological biases or by a neoclassical version of their model that precludes such biases. We evaluate these competing hypotheses by applying nonparametric tests of utility maximization and model validation tests to administrative data. We find that 79 percent of enrollment decisions from 2006 to 2010 satisfied basic axioms of consumer theory under the assumption of full information. The validation tests provide evidence against widespread psychological biases. In particular, we find that precluding psychological biases improves the structural model's out-of-sample predictions for consumer behavior.

Under Medicare Part D, senior citizens choose prescription drug insurance offered by numerous private insurers. We examine nonpoor enrollees' actions in 2006 and 2007 using panel data. Our sample reduced overspending by $298 on average, with gains by 81 percent of them. The greatest improvements were by those who overspent most in 2006 and by those who switched plans. Decisions to switch depended on individuals' overspending in 2006 and on individual-specific effects of changes in their current plans. The oldest consumers and those initiating medications for Alzheimer's disease improved by more than average, suggesting that real-world institutions help overcome cognitive limitations.