Approximately 30% of voters chose to cast their ballots before Election Day in both 2008 and 2012, according to political scientist Michael P. McDonald
. While there was only a slight uptick between these two elections, that's a big increase from previous election cycles.
As more and more people vote before Election Day, the significance of early voting is moving beyond the realm of lawyers and academics; it also raises some fundamental questions for everyday Americans about how we should conduct elections in this country. It's worth asking what effect early and absentee voting have
. Does pre-election voting increase voter turnout? Does it change the composition of the electorate? And does it affect people's choice of candidate?
Recent news about the FBI's investigation of emails
from Hillary Clinton's top aide and Donald Trump
's dubious tax maneuvers
bring these questions to the fore. Millions of people had already cast their ballots before these stories broke. It's possible that some of them would have voted differently if they'd waited until Election Day, leading Donald Trump to complain about vote-switching and "buyer's remorse."
Yet it's also possible that providing more options for voters will increase the size and breadth of the electorate -- something that voting rights advocates
, who have challenged cutbacks in the period for early voting in court, tout as an unqualified good for democracy.
There is some evidence that restrictions on early voting can have discriminatory effects. In Ohio, for instance, the Republican-led legislature limited the period for early voting and eliminated "Golden Week
," a five-day period for same-day registration and early voting. African-Americans are heavy users of in-person early voting in Ohio, but evidence of intentional discrimination was lacking. There was, however, considerable evidence that African-Americans would be disproportionately affected by Ohio's voting restrictions, enough for a federal district judge to find a violation of the Voting Rights Act. But the 6th Circuit Court of Appeals reversed, finding insufficient evidence that African-Americans would be negatively affected.
There are some serious problems with the 6th Circuit's analysis in the Ohio case, as I've explained here
That said, the evidence on early voting debate is more nuanced than partisans on both sides acknowledge. The jury is still out on the extent to which early voting actually really increases turnout. While early voting certainly changes when people vote, and political scientists Jan Leighley and Jonathan Nagler
find some evidence of small increases in turnout in states that adopted no-excuse absentee voting, it probably doesn't have a major impact on who votes. Most of the social science research suggests that early voters tend to be highly engaged and partisan. They are mostly people who would vote on Election Day if not allowed to vote before it. And because early voters tend to be partisan, it's not likely that late-breaking news would change many of their minds.
Leighley and Nagler also find that an early voting period of as long as 27 days is needed for positive effects in turnout to appear. The evidence is much stronger with respect to Election Day registration, a practice that they and other political scientists have found to have a more significant positive effect on who comes out to vote.
For those who want to see more people exercising their right to vote, there's both good and bad news here. While many voters -- especially people of color -- like the convenience that early voting offers, expanded early voting by itself doesn't seem to have much effect on turnout. Nor is it clear that early voting has much of an effect on who votes or for whom they vote. The bright side is that we probably shouldn't worry too much about whether early voting will affect who wins the presidential election: It almost certainly won't. If we really want to increase the size and breadth of the electorate, we should focus primarily on voter registration, especially reforms that would allow people to register and vote at the same time.