Empirical MPCs and the Dynamics of Fiscal Stimulus

Debt-financed fiscal stimulus is a primary policy tool in times of economic crisis, yet its distributional impact remains understudied. Auclert et al. (2023) create a heterogeneous-agent New Keynesian model to show that debt-financed transfers may generate a “trickling-up” effect as high-MPC households spend down their savings, funnelling them to lower-MPC households. In my thesis, I apply a k-means clustering algorithm on data from the Italian Survey on Household Income and Wealth to recover a joint distribution of income, liquid wealth, and MPCs to calibrate the model — the first such empirical recovery — and find no evidence of a household type with MPC=0. I show analytically and simulate numerically that when all household types have positive MPCs, the trickling-up model converges to a permanently elevated output level, and that no finite monetary policy response can fully eliminate this excess demand. An open economy extension resolves this: for any positive degree of openness, the foreign sector acts as the demand sink absent from the domestic economy, restoring demand to its pre-transfer level exactly.