In 2001, the House Commerce Committee examined the role of the media in the 2000 election debacle. The hearing questioned whether the exit polling by the networks and the practice of the networks “calling” states for one party or the other before all of the votes were counted was problematic. At the outset of the hearing, the Commerce Committee showed a video montage of the 2000 election night media coverage. It showed the network anchors promising to be very careful in how they made decisions to call states in the presidential race and then showed the tragic comedy, as the networks tried to determine who had won the State of Florida.
 
Looking back on that montage, the interesting thing about it are the words that were not used by the network anchors. Punch cards. DREs. Optical scan ballots. Canvassing board. Audit trails. Recount. Voter confidence. These words and phrases have only come into our common lexicon since mid-November 2000, when the public first saw the way in which the mechanics of elections operate.
 
Since then, the study of voting technology and election reform more broadly have become growing research fields. Beginning in 2004, various collaborations between colleagues at the University of Utah, the California Institute of Technology, and Brigham Young University have studied voter confidence in the electoral process. These studies have led us to understand basic issues related to the factors that make voters confident that their ballots are counted accurately. The theoretical basis for understanding voter confidence builds on literatures from trust in government, consumer satisfaction, and the history of voting in America.
 
In national surveys of public confidence conducted from 2004 to 2006 by Thad Hall, assistant professor in political science and research fellow at the Institute of Public and International Affairs at the University of Utah, and two colleagues at Caltech—Michael Alvarez and Morgan Llewellyn—have found that there are several key factors that affect confidence. First, not surprisingly, Democrats have been less confident that ballots will be counted accurately compared to Republicans. Given the fact that Democrats lost the 2000 and 2004 presidential elections, this lack of confidence likely reflects the impact of these losses. African Americans are also less confident than are White voters. More recently, they surveyed individuals both before and after the 2006 elections, an election when Democrats took control of Congress. Interestingly, the confidence of Democrats rose dramatically after the election, suggesting that the outcome of an election can affect confidence.
 
When examining the interaction of voting technology and confidence, the data show that voters are least confident voting absentee. This makes sense, given that absentee voters do not know what happens to their ballots after they put the ballot in the mail. A recent study by Michael Alvarez, Betsy Sinclair, and Thad Hall, using data from Los Angeles County and its 5 million registered voters, found that a small but significant percentage of absentee ballots are not counted because they either arrive too late to be included in the tabulation or the voter makes a mistake completing the ballot that invalidates it (such as not signing the ballot).
 
Alvarez, Hall, and Llewellyn have also found that there is an interaction between a voter’s views about electronic voting and their confidence in the technology. Voters who like electronic voting are just as confident as voters who cast optical scan ballots that their ballot will be counted accurately. However, if a voter is not confident in electronic voting, their confidence that their ballot will be counted accurately declines precipitously. These authors have also found that the paper audit trails that have been added to electronic voting machines increase voter confidence. These findings dovetail nicely with the findings of studies conducted by a consortium of researchers at the University of Maryland, the University of Michigan, and the University of Rochester that has examined the usability of voting systems.
 
The other key aspect of the voter confidence puzzle is the factors on election day that affect voter confidence and the confidence of poll workers that the election process produces fair outcomes in which the ballots are counted accurately. Here, there has been a collaboration between the University of Utah, Brigham Young University, and the University of New Mexico to study both voter and poll worker confidence in the electoral process. Studies from this collaboration have found that the interaction that voters have with poll workers has a great impact on the voter’s confidence that their ballot will be counted accurately. Voters are sensitive to the competence of their poll workers and when they have an encounter that they rate as excellent, they are more confident than if the experience was just average. Voters are also sensitive to the encounter that they have with their voting technology. If they have a problem with their paper ballot or with the electronic voting machine, but actions markedly reduce the confidence of the voter.
 
The voter and poll worker interaction, as well as the voter-technology interaction, are affected by the training of the poll workers. The better training that the poll workers have, the more confident they are that the ballots are counted accurately. Not surprisingly, training also affects the number of problems that occur at the polls; better training is associated with fewer problems. In comparing Utah with Cuyahoga County (Cleveland) Ohio, studies find that Utah fared much better, in part because the Utah poll workers felt more confident about their training.
 
As Americans prepare for the 2008 presidential elections, the issue of voting technology and poll workers will be in the media spotlight. Educating the public and the media about how elections work will, hopefully, occur before the election and before there is some election debacle in a state or local jurisdiction. In addition, local election officials should draw upon the lessons in states like Utah, which has moved to small, hands-on training, as a model for how to educate poll workers about election processes and procedures. Voters in Utah also benefit from the increased levels of voter confidence that come from voting on an electronic voting machine that includes a paper audit trail.
 
 
 
Working papers and publications on this topic can be found at http://www.vote.caltech.edu and at http://www.ipia.utah.edu/workingpapers.html.
 
Thad Hall is an assistant professor of political science and a research fellow in the Institute of Public and International Affairs at the University of Utah. He is the author of the book Point, Click, and Vote: the Future of Internet Voting (Brookings Institution Press), Authorizing Policy (Ohio State University Press) and the forthcoming books Electronic Elections (Princeton University Press) and an edited volume, Election Fraud: Detecting and Deterring Electoral Manipulation, to be published by the Brookings Institution Press.