Barry Burden highlights the main points of his article, “Election Laws, Mobilization, and Turnout: The Unanticipated Consequences of Election Reform.” This paper, co-authored with David T. Canon, Kenneth R. Mayer, and Donald P. Moynihan, appears in the January 2014 issue of the AJPS.

Over the past 20 years, the percentage of votes cast before Election Day has more than quadrupled. Many voters, policy makers, and advocates support early voting because they believe that the added convenience increases voter turnout. Our article challenges this popular wisdom. We argue that turnout actually declines when early voting is an option.

To understand this counterintuitive relationship, our theory separates the “direct” and “indirect” effects of election laws. Early voting is a convenience that lowers the direct costs of voting, and this alone should increase turnout. But early voting simultaneously raises the “indirect” costs of voting that a traditional election day helps to lessen. This is because activities focused around an old-fashioned election day helped to mobilize many voters by underwriting the costs of participating.

When all election activity is concentrated on a single day, potential voters are given lots of resources for voting. The media cover the election heavily, their coworkers will be wearing “I voted” stickers, friends and neighbors are expected to be at the polls, campaigns will be calling with reminders, and churches will offer free rides to polling places. All of this information and social pressure helps to get out the vote on a traditional election day but is diluted during a lengthy early voting period.

Early voting is a convenience to dedicated voters but it does not solve the problem of getting additional people to cast ballots. It does nothing to help get more people registered; a person who isn’t registered can’t take advantage of early voting. In other words, early voting is a treat for regular voters but not a great way to mobilize new voters.

We show that the availability of early voting on its own in a state actually decreases voter turnout by several percentage points. Using a variety of data and models, we find a consistently negative relationship between a state offering no-excuse voting before Election Day and overall turnout. Even more surprising, in both the 2004 and 2008 elections, turnout was slightly lower for each additional day of early voting that is offered.

There is a way to combat this negative effect. Turnout generally rises when states combine early voting with a form of “one-stop shopping” that allow people to both register and vote at one location. When same day or Election Day registration is offered alongside early voting, it overcomes the negative effect of early voting on its own and may result in a net increase in turnout. Early voting might well have a number of benefits including lower administrative costs but in isolation it is not a prescription for increasing voter participation.