Committee Schedule

Testimony

Testimony of Mr. Stephen Ansolabehere

Professor, Department of Government
Harvard University

Hearing: Voter Registration: Assessing Current Problems
Wednesday, March 11, 2009

[view file]

Executive Summary

In 2001, the Caltech/MIT Voting Technology Project estimated that 4 – 6 million votes were lost owing to administrative problems. These people tried to vote but could not because of difficulties with polling place operations, voting equipment, absentee ballots, or registration. The largest sources of the problem lie evidently in the voter registration system, which accounted for roughly half of all lost votes.

Eight years later, we have witnessed significant improvement in election administration. Updating of voting equipment has cut the rate of residual votes in half. States have put in place provisional balloting and other checks to prevent lost votes. Voter registration, however, remains problematic. State efforts to improve registration in compliance with HAVA are underway, but limited resources might hamper the ability of states and counties to implement fully computerized registration systems.

This testimony examines the experiences of the electorate in 2008 with the voting systems, especially registration and other parts of the authentication process. In 2008, there were approximately 230 million people of voting age, 212 million eligible to vote, 168 million registered, and 133 million who actually voted. These facts demonstrate the pivotal importance of voter registration. Approximately 79 million eligible voters did not participate. Of these, 44 million were not registered, and 35 million registered citizens did not vote. According to data from the Cooperative Congressional Election Study, 4 to 5 million people report administrative procedures as the reason for not registering, and approximately 4 million of registered voters did not vote because of administrative problems, approximately the same magnitude as we saw in 2000. Improving registration and authentication systems ought to remain a high priority.