Gerber et al., 2008: Encouraging voting with social pressure increased voter turnout among households in Michigan
180,002 households in Michigan received one of four mailers, or no mailer, 11 days before the August 2006 Michigan primary election. The first mailer, and all others, emphasized only the recipient’s civic duty to vote (“DO YOUR CIVIC DUTY—VOTE!”). The second added that the recipient was being studied (“YOU ARE BEING STUDIED!”—Hawthorne condition). The third informed recipients that who votes is public information, listed the recent voting record of each registered voter in the household, and indicated that the sender would “mail an updated chart” after the election (self condition). The fourth also informed recipients that who votes is public information and listed the recent voting record of each registered voter in the household, and also listed neighbors and their voting records and indicated that the sender would “mail an updated chart” after the primary (neighbors condition). As compared to turnout in the no-mailer control condition (29.7%), turnout was higher in the civic duty condition (31.5%), Hawthrone condition (32.2%), self condition (34.5%), and especially in the neighbors condition (37.8%). The results did not vary much with the degree to which the recipient voted often or rarely.