Greater polling-place distance lowers in-person voting

What the study found: Greater distance to a polling place was associated with a lower likelihood of voting in person. The abstract says that a one-mile increase in distance reduced in-person voting by 1 to 3 percentage points, with larger effects among people closest to the polling place.
Why the authors say this matters: The authors conclude that polling-place location affects voting costs and turnout. They also say their counterfactual exercises identify polling places that would maximize turnout, and that some precincts have large potential gains even when only existing buildings are considered.
What the researchers tested: The study used a geographic regression discontinuity design, a method that compares people near a boundary to estimate causal effects, using data from Pennsylvania and Georgia. It examined how distance to one’s polling place relates to the likelihood of voting and whether voters substitute to mail-in voting when distance increases.
What worked and what didn't: A one-mile increase in distance was associated with a 1 to 3 percentage point drop in in-person voting. The effects were two to three times higher among those closest to the polling place. When mail-in voting was available, some voters shifted to that option as distance increased.
What to keep in mind: The abstract does not describe detailed limitations beyond the states studied and the use of counterfactual exercises. The findings are based on Pennsylvania and Georgia data, so the scope is limited to those settings in the provided summary.

Key points

• A one-mile increase in distance to a polling place reduced in-person voting by 1 to 3 percentage points.
• The effect was two to three times larger for people closest to the polling place.
• When mail-in voting was available, some voters switched to that option as distance increased.
• The authors used data from Pennsylvania and Georgia and a geographic regression discontinuity design.
• Counterfactual exercises identified polling places that could maximize turnout, including some using existing buildings.