Are Political and Charitable Giving Substitutes? Evidence from the United States

Using microdata from the American Red Cross (ARC) and the Federal Election Commission (FEC) in two natural experiments, we provide evidence that political giving and charitable giving are substitutes. In the first natural experiment, we estimate the effects of a positive shock to charitable donations to the ARC: foreign natural disaster events. We find that although charitable donations to ARC increase by 34.9% in the six weeks following a disaster, political donations decline by 18.8% in the same period. Put differently, each 1% increase in the charitable giving to ARC is accompanied by a 0.53% drop in political donations. At the average county-week-level donations, the implied effect of a $1 increase in charitable giving is a $0.42 decline in political donations. In the second natural experiment, we estimate the effects of a positive shock to political giving: advertisements for political campaigns. Exploiting geographic discontinuities in advertising markets, we find that political advertisements increase political giving, whereas they decrease charitable donations to ARC. Our estimates imply that each 1% increase in the political giving is accompanied by a 0.59% drop in charitable donations to ARC. At the average countyweek-level donations, the implied effect of a $1 increase in political giving is a $0.33 decline in charitable donations. The crowding -out elasticities suggest that political giving and charitable giving are relatively close substitutes. We provide a number of robustness checks, and we discuss potential causal mechanisms.