Marginal Propensity to Consume in Recessions: a Meta-analysis

This paper assembles a dataset of 1244 estimates of marginal propensities to consume (MPC) out of stimulus checks and other small transitory or predictable payments—as reported by 40 studies. I use meta-regressions to uncover the sources of systematic variation in estimates and provide fitted MPC for a number of policy-relevant scenarios. An increase in unemployment by one percentage point is associated with an MPC estimate that is higher by 4-5 percentage points. MPC estimates systematically vary depending on payment characteristics: they decrease with the size of the payments; MPCs out of stimulus checks are higher than those out of some other payments that are recurring. MPCs are lower for households holding ample liquidity. These results imply that the effects of stimulus payments and other policy interventions crucially depend on the circumstances and the manner in which funds are disbursed, and highlight the importance of considering state-dependent multipliers, liquidity constraints, near rationality and mental accounting.