Working Paper: NBER ID: w27078
Authors: Tatyana Deryugina; Benjamin M. Marx
Abstract: Do new societal needs increase charitable giving or simply reallocate a fixed supply of donations? We study this question using IRS datasets and the natural experiment of deadly tornadoes. Among ZIP Codes located more than 20 miles away from a tornado's path, donations by households increase by over $1 million per tornado fatality. We find no negative effects on charities located in these ZIP Codes, with a bootstrapped confidence interval that rejects substitution rates above 16 percent. The results imply that giving to one cause need not come at the expense of another.
Keywords: charitable donations; tornadoes; natural experiment; displacement effects
JEL Codes: D64; L31; Q54
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Tornado fatalities (Q54) | Total donations (D64) |
| Tornado fatalities (Q54) | Donations to local charities within 20 miles (D64) |
| Tornado fatalities (Q54) | Donations to charities located more than 20 miles away (D64) |
| Tornado fatalities (Q54) | Substitution rates (C29) |