Ballot Secrecy Concerns and Voter Mobilization: New Experimental Evidence About Message Source, Context, and the Duration of Mobilization Effects

Experiment 1: VPC obtained a list of eligible Wisconsin registrants from an outside private vendor; the voter records used to generate the sample for this experiment came from a voter file produced by the Wisconsin Secretary of State in April of 2012. the organization selected the study population from this list of registrants. Records were first removed if they had an invalid mailing address or if they were included in another experiment run by the organization. From the remaining records, registrants were retained if they (a) were unmarried women or members of a minority group10 (i.e., non-Caucasians), (b) were listed as active registrants, (c) had registered in December of 2006 or later, and (d) had no record of having voted in any general primary, general election, or special election. We limited our analysis to the 10,200 individuals who had been registered before the November 2008 election but had not participated in that high-salience race.Experiment 2: See http://hdl.handle.net/10079/ffbg7p5