Lindsay Nielson and Dr. J. Quin Monson, Political Science
Early voting is a rapidly-growing development in American elections. This trend includes both absentee voting, in which a voter submits an application to have a paper ballot mailed to his or her home, or in-person early voting, in which a voter may cast a ballot at an official polling place before the designated election day. Absentee voting was first used during the Civil War to allow military personnel stationed away from their home polling locations to cast their ballots, and most states only made this option available to the general voting populace in the last half of the twentieth century. Early voting, on the other hand, began in Texas in the 1980s. Although absentee voting is still more widely used, early voting has been growing at a rate eight times faster than that of absentee voting.
In recent years many states have liberalized their early voting laws to make them less restrictive. Thirty states allowed no-excuse early voting in the 2004 elections, up from 26 states in 2002. The increasing ease of voting early raises the questions of whether structural differences lead to differences in who is more likely to take advantage of the early voting system, and under what circumstances people can be enticed to vote early. The primary effect of changes in early voting laws is that it is now easier and more convenient to vote, as participants are no longer required to show up at a certain time and place to cast a ballot. However, an easier voting process alone does not raise turnout—voter awareness is crucial. Some voters are only active participants to the degree that they are mobilized by political parties, interest groups, candidates, and campaigns. To this effect, I hypothesize that campaign contacts (specifically those which ask voters to vote early) and the way that elections are structured affect an individual’s likelihood to vote early.
My analysis used data from the 2004 Campaign Communications Study conducted by the Center for the Study of Elections and Democracy at Brigham Young University. A national random sample of 2108 voters was asked to keep detailed logs of all of their campaign contacts in person, by telephone, and by mail. They were also asked a series of standard political survey questions, including whether or not they voted early or absentee. To examine trends in early voting, I estimated a multivariate logic regression model of whether or not the voters in the Campaign Communications Survey participated on Election Day or voted early or absentee. My dependent variable is whether a subject voted on election day or cast an early ballot. My independent variable of interest is whether the subject was contacted by a campaign to vote early. I also controlled for other factors that influence turnout—state laws allowing for early voting; restrictions that state laws place on early voting; whether the voter is a strong partisan; whether the voter lived in a battleground state; sociodemographic factors such as gender, age, education level, income, and race; attitudinal factors, particularly political knowledge and trust in government; the total number of different types of contacts the respondent received; and prior voting history, defined as having voted in the 2000 presidential election.
My findings showed that early voters were more likely to be older and to have more education than election day voters. Curiously, early voters in 2004 were also significantly more ideologically conservative and were more likely to identify as Republicans, and particularly as strong Republicans. Early and election day voters were similar in their levels of political interest and trust in outside groups. A higher proportion of early voters than election day voters received phone and in-person contacts. However, my regression model found that mail and in-person contacts were not statistically significant factors; phone contacts, while being significant, had a negative effect on the probability of voting early. This could indicate that voters are likely to suffer from campaign fatigue, and that campaign contacts may not be as effective as campaign organizers hope. Many of the control variables I included were not statistically significant, including gender, age, race, income, education, and prior voting history.
Structural and campaign factors also had strong influences on the likelihood of voting early. State laws allowing for no-excuse early voting resulted in a strong positive increase in the likelihood of voting early; voters who identified as strong partisans, who lived in a battleground state, and had higher levels of political interest and trusts in government, were also more likely to vote early. There are several explanations for these factors. Strong partisans may be more likely to vote because they comprise a political party’s base of core voters who generally vote regardless of other factors. The strong influence of partisanship on the likelihood of early voting also indicates that early voters may not be representative of the general public at large. Battleground states also see heavier campaigning, which could lead to higher voter awareness in general. Whether or not the voter lived in a battleground state was also an important determinant of whether or not he or she voted early. However, this effect could be due to increased campaigning in battleground states, and thus higher voter awareness.
Because early voting is still a new development in elections and voter turnout research, the exact ways in which it will affect campaigns and elections are still uncertain. However, there is no question that the implications of who votes early and how they are motivated to vote early will change campaign dynamics and require candidates, political parties, and interest groups to alter the ways that they mobilize voters. My findings illuminate some of the reasons behind early voting and describe some of the effects of early voting on campaign dynamics. However, it is also important to remember that no matter how easy it is for voters to participate in elections, most voters will only participate to the extent that they are politically engaged. Thus, any reforms in elections laws must also be accompanied by attempts to increase the public’s level of political engagement.