Is a Dollar a Dollar? How Transfer Design Shapes Household Spending

Résumé fourni par la source

Government transfers vary along two design dimensions that standard models predict should not matter: whether benefits are paid in cash or kind, and whether they arrive as one-time or recurring payments. We test to what extent these design choices a!ect household spending using sharp changes in four pandemic-era transfers to lowincome families with children: one-time food vouchers, monthly food benefits, one-time cash payments, and monthly cash payments. Linking variation in benefit timing and amount to high-frequency household scanner data, we estimate spending responses in a harmonized di!erence-in-di!erences and event-study framework. We find that households spend more out of in-kind benefits than out of cash transfers of similar value, and more out of recurring payments than out of one-time transfers in the short-run. Our estimates produce the first internally comparable set of shortrun food-store MPCs across di!erent transfer types ranging from 0.18 (one-time) and 0.38 (monthly) for in-kind benefits, versus 0.06 (one-time) and 0.20 (monthly) for cash. The patterns suggest that the form and timing of transfers systematically shape household spending in ways that are consistent with mental accounting, labeling, and forward-looking responses. More broadly, our results imply that transfer design-not just generosity and targeting-is a first-order policy lever for both social insurance and short-run demand stabilization.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.