Participation in First-Time Voter Programs, in which high school students attend casual, informative sessions about voting and political participation, is found to have strong mobilization effects on new voters.

Addonizio, Elizabeth. 2006. "A Social Approach to Voter Mobilization and Election Day." Unpublished Manuscript. Institution for Social and Policy Studies, Yale University.

Elizabeth Addonizio presents results of an experiment designed to test the effects of early voter education and mobilization programs among first-time voters attending high schools in Connecticut, Kentucky and New Jersey. Students assigned to treatment groups attended a First-Time Voter Program. The program consisted of casual and social informational sessions which highlighted issues of voting that pertain to young people and educated students on how to register and vote. Findings indicate that this voter education approach is very effective in mobilizing new voters.

Electoral Context: The experiments reported here were carried out prior to elections in November 2002 and 2003 and spring elections in May 2003 and 2004. The experiments were conducted in highly, moderately and weakly competitive elections.

Subject Population: The subject population consists of high school seniors in participating schools in New Jersey, Connecticut and Kentucky. In some cases, the sample included all high school seniors at a school, and at others it included certain classes. The populations of the schools represented a range of socioeconomic levels and demographic characteristics.

Randomization Procedure: Individual students at each school were randomly assigned to treatment and control groups.

Treatment: Members of the treatment group attended the First-Time Voter Program while members of the control group did not attend the program. The First-Time Voter Program is a social, informal program about voting. A young person leads each session in a casual and social manner. Sessions ran from thirty to forty minutes and included between twenty and thirty students. Through questions and interactive discussions, each session addressed issues of relevance for students including financial aid for college, the military draft, sales tax, etc. In a second part of the program, the presenter explained voter registration requirements and procedure and passed out voter registration cards. The presenter also explained the importance of becoming informed about candidates and issues and suggested places to find related information. In the third part of the session, the presenter used a voting machine to explain the process of voting and invited students to practice.

Findings: Bivariate probit model first regresses voter turnout on whether the student attended the program and, in a second equation, regress the decision to attend the presentation on the dependent variable of whether the student was assigned to the treatment group. The model also controls for competitiveness of the election and socioeconomic status of the town in which the school is located. Three schools were excluded from the analysis because none of the students that participated in the program turned out to vote in the subsequent election, and three schools were excluded from the analysis because none of the students that participated were eligible to vote in the subsequent election. A probit estimation of turnout on whether the student was assigned to the treatment or control group implies an average intent-to-treat effect of 19 percentage points, which is statistically significant (p-value: 0.004, two-tailed test), and indicates that being assigned to participate is associated with a 19 percentage point increase in the likelihood of voting. Bivariate probit analysis of the effects of participation in the program on turnout implies that the voter turnout level for students who attended the program was 24 percentage points higher than for students in the control group (p-value: 0.009, two-tailed test).