At the recent APSA meeting, Timothy Vercollitti, and David Andersen presented Protecting the Franchise, or Restricting it? The Effects of Voter Identification Requirements on Turnout. Here is the abstract:

The literature on procedural barriers to voting has focused on registration deadlines, hours of poll operations, and even the physical characteristics of polling places to explain why voters do or do not show up on Election Day. Less is known, however, about the effects of voter identification requirements on turnout. These requirements have taken on heightened importance since the presidential election of 2000, with many states tightening their requirements to combat vote fraud since then. Forty-one states now require proof ranging from voter signatures to photo identification at the polling place. In this paper we examine the effects of these varying requirements on voter turnout. Drawing on previous research, we hypothesize that as the level of proof becomes more costly to the voter, turnout declines. We test our hypotheses using aggregate measures of turnout at the state and county levels in the 2004 presidential election, as well as individual-level data drawn from the Voter Supplement to the November 2004 Current Population Survey. This research has significant normative importance, in that it speaks to the difficulty of balancing the potentially competing aims of election integrity and access to voting.

Thanks to David Kimball for the pointer. UPDATE: This link won’t work for everybody. If you are having rouble, search for one of the author’s names, which should pull up the right paper.