Estimating the Effects of the National Voter Registration Act of 1993

Over-time and multivariate cross-sectional analyses of large survey samples are used to estimate the likely effects of the National Voter Registration Act (NVRA) by examining turnout in those states where procedures comparable to any of the act's provisions were in force in 1992. In contrast to previous studies, we find that most state motor voter programs did not resemble the NVRA provision. We analyze one state program that did, and in addition use election-day registration as a surrogate, because it also provides one-trip voting. Our two approaches lead to estimates of turnout increases due to the motor voter provision of 4.7 and 8.7 percentage points, respectively. The lack of state counterparts to public agency registration prevents estimates of this provision's consequences. Eliminating purging for not voting will increase turnout by as much as 2 percentage points. Universal mail registration will have no effect. The turnout effects will be greatest among the two largest groups of current nonvoters: people under the age of 30 and those who moved within two years of election day. Neither group is politically distinctive, except for young people's weaker identification with the major parties and greater affinity for third-party candidates.