On how physical proximity affects turnout:
- A 2005 study on Atlanta's 2001 mayoral election showed that people who lived closer to their polling places (especially those without cars) were more likely to show up on Election Day than those who lived further away.
- In addition, those whose polling places were recently moved just a little closer to their homes (due to an increase in number of precincts) were also more likely to turn out, indicating that increased convenience outweighed any confusion over a new polling location.
- A 2003 study in 3 Maryland counties found that for each 1-mile increase in proximity to the polling place, turnout jumped by 0.453 percent, or nearly half a point.
On how the type of polling place facility affects the vote:
- A 2008 Arizona study found that, in the 2000 election, those had voted in schools were more likely to have voted in favor of a school-funding ballot measure.
- A follow-up experiment to the study found that those who were shown pictures of schools prior were more likely to, afterward, indicate support for education funding than those who were shown pictures of a church. Those shown pictures of a church were more likely to indicate support for limits on stem cell research.
- A study published this year found that in 2006 in South Carolina, those who voted in churches were more likely to have voted for a proposed constitutional ban on gay marriage than those who voted elsewhere.
Whether or not early voting means less locational bias - and it's unclear that this is the case -, the serious downfalls of mandatory and exclusive early voting systems (that such systems generally have no great effect on turnout, that they systematically exclude those without consistent permanent addresses, and that they diminish Election Day community excitement and celebration) hardly seem to make early voting an effective type of solution.
Read more about polling place effects, issues with early voting, and voter turnout in general in the rest of the Jacobs' article here.