by Jessica Schultz
This paper explores how people track inflation over their lifetimes while facing tradeoffs between attention and certainty. It first employs a flexible modification of the Recursive Least Squares Learning approach from Malmendier and Nagel (MN) (2016) to find that households place weight on each inflation observation in a hump-shaped pattern over age when using past observations to set expectations about the future. This finding departs from MN, which models a strictly increasing weighting scheme with age. This paper then uses these findings to motivate a theory of Rational Inattention (RI) in inflation: as households age and accumulate wealth, their knowledge of the inflation rate becomes more important in their financial decisions–so they pay more attention to inflation. Consequently, as they decumulate wealth during their retirement, they have less reason to track inflation as accurately.
This paper subsequently formalizes this theory in a two-period RI model in which inflation-driven uncertainty in the interest rate between a working period and a retirement period can be reduced at a cost; this reduction in uncertainty occurs through observing an endogenously chosen signal that is correlated with the interest rate. It finds that as wealth increases before retirement, the optimal choice of signal precision increases as well. These findings help explain the hump-shaped weighting scheme for inflation observations in the empirical section, assuming changes in these weights over age are related in part to changes in household wealth. Ultimately, these findings suggest that monetary policy that focuses on long-term inflation stability or accounts for this heterogeneity may be most effective in anchoring consumer inflation expectations and increasing consumer welfare.
Advisors: Professor Francesco Bianchi, Professor Michelle Connolly | JEL Codes: E2, E21, E31
