MODELING THE EFFECT OF PHOTO IDENTIFICATION REQUIREMENTS ON VOTER TURNOUT

Abstract

Voter identification laws have become a highly contentious topic in American politics as society debates how to balance safeguarding the integrity of our elections with citizen participation. This thesis evaluates the claim made by voter identification requirement critics that increasingly stringent requirements lower voter turnout, particularly among certain demographic groups that may have a more difficult time meeting voter identification requirements. Specifically, this paper analyzes the effects that various levels of voter identification requirements may have on voter turnout, hypothesizing that more stringent voter identification requirements, such as requiring photo identification at the polls, are associated with lower voter turnout. This research uses individual level Current Population Survey (CPS) data for the 2004 through 2010 congressional and presidential elections, as well as compiled state election information. As cultural and other state level differences that may potentially impact voter turnout exist between states, such as the availability of mail-in voting, early voting, or same day registration, the analysis specifies a logistic state and year fixed-effects model of voter turnout with additional controls for individual characteristics. This study's primary finding is that photo identification requirements disproportionately impact young voters between the ages of 18 and 24, decreasing the probability that they turn out to vote.