Working Paper: NBER ID: w10454
Authors: James J. Choi; David Laibson; Brigitte C. Madrian; Andrew Metrick
Abstract: Economic theory predicts that an unexpected wealth windfall should increase consumption shortly after the windfall is received. We test this prediction using administrative records on over 40,000 401(k) accounts. Contrary to theory, we estimate a negative short-run marginal propensity to consume out of idiosyncratic 401(k) capital gains shocks. These results cannot be interpreted as standard intertemporal substitution, since the idiosyncratic returns that we study do not predict future returns. Instead, our findings imply that many investors are influenced by a positive feedback effect, through which higher recent returns encourage higher short-run saving. Like any other animal, 401(k) participants appear to increase behaviors that have been associated with high rewards in the past.
Keywords: Consumption; Wealth shocks; Behavioral economics; 401(k) plans
JEL Codes: E21; G11
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| 401(k) capital gains shocks (D14) | consumption growth (E20) |
| positive wealth shock (E21) | consumption (E21) |
| lagged consumption response following a positive wealth shock (E21) | consumption (E21) |
| higher recent returns (G17) | saving behavior (D14) |
| reinforcement learning (C73) | saving behavior (D14) |