Abstract
Hyperbolic discounting induces dynamically inconsistent preferences among consumers. Hyperbolic consumers are predicted to be more likely than non-hyperbolic consumers to have a higher marginal propensity to consume (MPC, hereafter) out of liquid assets and a lower MPC out of illiquid assets, because hyperbolic discounters are more often liquidity-constrained whether they are naive or sophisticated. I investigate the empirical validity of the hypothesis by using a broad panel survey from “Japan Household Survey on Consumer Preferences and Satisfaction” conducted by Osaka University, 2005-2007. I find that hyperbolic households have actually higher MPCs out of current income than non-hyperbolic households, whereas the MPCs out of fixed assets for hyperbolic households are lower than that of non-hyperbolic households. The asset-specific MPC puzzle could thus be resolved in part by hyperbolic discounting.
