Working Paper: NBER ID: w30972
Authors: Gordon B. Dahl; Joseph Engelberg; Runjing Lu; William Mullins
Abstract: We estimate 3% of the U.S. voter population is registered to vote in two states. Which state these double-registrants choose to vote in reflects incentives and costs, being more prevalent in swing states (higher incentive) and states which automatically send out mail-in ballots (lower cost). We call this behavior cross-state strategic voting (CSSV) and estimate there were 317,000 such votes in the 2020 presidential election. Because both Democrats and Republicans engaged in CSSV, the net effect was small, although it could matter in closer elections (e.g., Florida in 2000) or if one party increased CSSV relative to the other.
Keywords: No keywords provided
JEL Codes: D72
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| state swing status (P27) | likelihood of voting in first state (K16) |
| automatic mail-in ballots (K16) | likelihood of voting in first state (K16) |
| voting costs (K16) | overall likelihood of voting (K16) |
| voting incentives (D72) | choice of voting state (K16) |