A civic education program for high school students is found to have no statistically significant effect on registration or turnout of young voters.

Phillips, John Anthony. 2004. "The Relationship between Secondary Education and Civic Development." CIRCLE Working Paper 14: May 2004.

John Phillips presents the results of two field experiments that examine the effect of extracurricular participation and practical lessons about local politics, "service learning," on the civic knowledge, attitudes and behavior of high school students. This abstract highlights results of Phillips' experimental interventions on high school seniors that estimate the effects of service learning on the voter turnout of newly registered voting age students. Overall, Phillips finds the effects of service learning to be small and elusive. With regard to voting among treated high school seniors, Phillips estimates no statistically significant effect of his civic education treatment on registration or turnout among the treatment group seniors.

Electoral Context: Phillips examines effects of service learning on rates of voter registration and turnout in the November 5, 2002 general Election and the March 5, 2002 primary election, the April 9, 2002 local election, and the June 4, 2002 local runoff election among subjects from his sample of high school seniors in Long Beach, California.

Subject Population: Phillips' sample consists of high school seniors attending school in Long Beach, California.

Randomization Procedure: Phillips randomly assigned individual seniors to treatment and control groups; control groups received civic education lessons and materials while control group seniors were not contacted.

Treatment: Phillip's experiment among high school seniors provided a 1-hour seminar to randomly-selected inner city seniors that demonstrated in detail how they could register to vote, research ballot issues and candidates, cast their ballots, contact local representatives, participate in collective action, and get involved in neighborhood organizations. Besides participating in several role-plays and filling out practice forms, treatment students received a 24-page packet entitled "The Insider's Guide to Getting Informed and Getting Involved in Long Beach."

Findings: Phillips' findings reveal no statistically significant relationship between these lessons and subsequent changes in civic knowledge, attitudes or behavior. With regard to voter turnout, as of February 3, 2003, all 207 seniors in Phillips' study were at least 18 years old. County records showed that 16 out of 112 control seniors (14.3%) and 17 out of 95 treatment seniors (17.9%) had registered to vote by this date. The 3.6 percentage point difference in registration rates is not statistically significant (Chi2, df=1, p=.480). Five of the control group seniors (4.5%) and 2 of the treatment group seniors (2.1%) voted in the November 5, 2002 midterm election (Chi2, df=1, p=.349). No students in the sample voted in the March 5, 2002 primary election, the April 9, 2002 local election, or the June 4, 2002 local runoff election.