SciLine interviewed Jan Leighley, professor of government in the School of Public Affairs at American University in Washington, D.C., on Nov. 4, 2022. Leighley discussed how early voting affects turnout, how turnout differs for midterm and presidential elections, how pollsters predict turnout and how to understand the persistent gap between people’s intention to vote and actual turnout.
Below are some highlights from the discussion. Answers have been edited for brevity and clarity.
How do we know after an election how many people ultimately voted?
Jan Leighley: It’s usually January by the time we really have a good sense of how many people voted, and there can be corrections after that. Now those are usually small corrections. But it takes a lot of people a lot of time even to ensure that number is correct.
How do newer forms of voting, such as early voting or mail-in voting, affect turnout?
Jan Leighley: The assumption has always been if we make it easier to vote using these methods, more people will vote. (But) most research suggests that the people who take advantage of these newer ways of voting would have voted anyway on Election Day if they weren’t able to vote in those other ways.
The original intention of those reforms was to get new people into the voting pool. We have a little bit of evidence that – under some circumstances – we can indeed increase turnout in the range of 2 percentage points or so.
But it depends on what parties do. It depends on the competitiveness of elections and other factors unique to specific elections.
How does turnout differ between midterm and presidential elections?
Jan Leighley: Presidential turnout is substantially higher than midterm elections, when members of Congress and a set of senators and governors might be up for re-election or election.