How effective are universal payments for raising consumption? Evidence from a natural experiment.

We investigate the impact of the universal stimulus payments (100-350 thousand KRW per person) distributed by the largest Korean province of Gyeonggi during the COVID-19 pandemic on household consumption using large-scale credit and debit card data from Korea Credit Bureau. As the neighboring Incheon metropolitan city did not distribute stimulus payments, we employ a difference-in-difference approach and find that the stimulus payments increased monthly consumption per person by approximately 30 thousand KRW within the first 20 days. The overall marginal propensity to consume (MPC) of the payments was approximately 0.40 for single families. The MPC decreased from 0.58 to 0.36 as the transfer size increased from 100-150 to 300-350 thousand KRW. We also found that the effects of universal payments were very heterogeneous across different groups of people. The MPC for liquidity-constrained households, which account for 8% of all households, was close to one, but the MPCs of the other household groups were not significantly different from zero. The unconditional quantile treatment effect estimates reveal that there was a positive and significant increase in monthly consumption only in the lower part of the distribution below the median. Our results show that a more targeted approach may more efficiently achieve the policy goal of boosting aggregate demand.