Exploiting three earthquakes in Italy as quasi-experiments, we analyse the response of homeowners’ consumption to transfers targeted to finance housing repair and reconstruction. To the extent that funds are made available up-front, these transfers are akin to loans, mainly affecting the liquidity of households’ wealth. We show that these transfers have little effect over a multi-year horizon—they are not a windfall. Yet, access to reconstruction transfers has a strong and significant effect on non- durable consumption on impact, especially for households with a low level of liquid wealth and bank debt. In contrast, we find no significant consumption change in response to the in-kind equivalent of cash transfers. Our study contributes to the recent literature on the dynamics of the consumption demand by the wealthy hand-to-mouth, providing micro-evidence in line with the main predictions of the theory.