by Sheng Guo, Florida International University, Jonathan Skinner, Dartmouth College, and Stephen P. Zeldes, Columbia University
Whether Americans are prepared for their retirement is a hotly debated question. Studies that assess households’ financial savings find that their savings are often abysmally low, while research examining retiring households’ overall consumption (including home production) finds little or no material decline. We attempt to test and reconcile these two opposing views using the Health and Retirement Study data to study the dynamics of pre- and post-retirement consumption across U.S. households. In preliminary work, we find among those least well-prepared for retirement, there is both a small abrupt change in consumption in retirement, but more importantly a steeper gradual decline in consumption over the next decade that cannot be explained by conventional factors. In this proposal, we seek a more general empirical model of consumption (including home production) at retirement, and to explain the puzzling patterns we find using the Gabaix (2016) “inattention” approach of sparse dynamic programming. We plan to establish both empirical patterns of retirement consumption, and to solve and calibrate the inattention model to test whether such a model can match consumption and saving patterns we observe in the data.