How to Get Out the Latino Vote: Evidence from Field Experiments

During midterm elections, Latino voter turnout is known to lag below that of other racial groups in the United States, despite a growing population. In 2010 the census estimates Latino turnout was 61% among registered voters, compared to 70% turnout for the population as a whole. In contrast, Latino turnout had surged to a high of 84% just two years earlier in 2008. Primarily the result of lack of outreach efforts, fewer Latinos go to the polls in midterm years than presidential. However this can be fixed. Research demonstrates that it is possible to mobilize and turnout Latino voters, with the proper investment, targeting and messages. Get-out-the-vote (GOTV) field experiments provide valuable guidance on how to turn potential votes into actual voters, which could be critical in 2014.  Below, I highlight one example in which Latino turnout increased by 43 points among those successfully targeted.

GOTV field experiments use randomized assignment to treatment to isolate the effect of being invited to participate in an election. Prior to contacting voters, those in the target pool are randomized into treatment and control groups. Those in the treatment group are then targeted for mobilization, while those in the control group are not contacted or are targeted for contact with a non-electoral message (e.g. please recycle). After the election, comparison of turnout rates between the two groups that takes into account the failure-to-treat problem (wherein some individuals assigned to the treatment group are not contacted) can generate robust estimates of campaign effects.

In 2001, in a small town in central California, I conducted a door-to-door canvassing experiment to increase Latino voter turnout in a non-competitive school board election (Michelson 2005). Other early experiments, in Fresno, Bridgeport, and Phoenix found that Latino door-to-door canvassers were more effective than non-Latinos at making contact with targeted Latino voters and that Latinos were strongly motivated to vote when appeals to participate were linked to a specific community concern (e.g. keeping a local hospital open or instituting a higher local minimum wage). Campaigns in Bridgeport and Phoenix increased turnout among contacted voters by double digits (see Table 1).


Since 2001, I have conducted hundreds of experiments aimed at learning how best to get out the Latino vote, often in partnership with non-partisan community-based organizations. In our book Mobilizing Inclusion: Transforming the Electorate Through Get-Out-the-Vote Campaigns, Prof. Lisa García Bedolla and I describe 268 get out the vote experiments conducted across six electoral cycles from 2006 to 2008; many of these experiments specifically aimed to increase low-propensity Latino voters to go to the polls – the very Latinos on the bubble in 2014.

Repeatedly, we found that well-conducted mobilization efforts using door-to-door canvassing or live telephone calls could successfully mobilize Latino voters. One particularly powerful experiment for the November 2006 midterm election was a two-call phone bank conducted by the Southwest Voter Registration Education Project (SVREP). Individuals in the treatment group voted at a rate of 36.6%, compared to 34.3% in the control group. This 2.3 %-point intent-to-treat effect translates into an increased turnout rate among those successfully contacted (taking into account the failure-to-treat problem) of 9.3 %-points. Controlling for voter history increases the estimate for the phone calls to 10.3 %-points (Table 2).


The key to making these blandishments to vote effective is for canvassers to make personal contact with targeted voters—to make them feel personally invited into the polity. Campaigns that take advantage of social networks and trusted local organizations are more likely to successfully move Latinos to the polls. This is well illustrated by results from another experiment from our book, conducted in June 2006 in cooperation with the Center for Community Action and Environmental Justice (CCAEJ). CCAEJ had been active in the Mira Loma and Glen Avon neighborhoods in the northwest corner of Riverside County, California, for decades before deciding to try to increase voter turnout. However, they were well known for their work and had a strong team of experienced canvassers. The original mobilization plan called for door-to-door visits with targeted voters, followed by the placement of doorhangers at the homes of successfully contacted individuals on Election Day. While the CCAEJ volunteers were distributing the doorhangers, however, they often made a second face-to-face contact with the target voter.

Overall, the campaign was enormously successful: when a CCAEJ canvasser actually contacted a person on the target list, they increased her probability of voting from 11.1 percent (the turnout rate in the control group) to 54.2 percent (the turnout rate in the treatment group) — the largest estimated treatment effect to emerge from a voter mobilization field experiment. Other door-to-door experiments have not replicated that enormous effect, but the lessons learned from the CCAEJ effort did persist in future efforts. Two contacts are better than one, and local volunteers contacting their neighbors are more effective than strangers.

There is ongoing debate over the specific content of the GOTV message. While some research has found generic “civic duty” messaging can work among Latinos, other research suggests that targeted ethnic appeals are more effective. Recent bilingual phone bank experiments conducted in 2010 in California and 2012 in Texas, in cooperation with Prof. Ali Valenzuela, suggest that appeals to ethnic solidarity can be more effective for Latinos who are less incorporated and who have stronger ties to their Latino identity. Other experiments have demonstrated the importance of using bilingual canvassers or mailers when reaching out to Latino voters.

Latino voter mobilization has the potential to change the outcome of a number of 2014 midterm election contests. Those who dismiss outreach to Latino voters as a waste of resources because they are unlikely to participate are merely replicating a familiar self-fulfilling prophecy. Latino voters, just like other voters, are fairly easily motivated to participate. They just need to be asked.

Melissa R. Michelson, Ph.D. is Professor of Political Science at Menlo College. She has conducted over 300 get-out-the-vote field experiments and written dozens of articles and book chapters on Latino politics, as well as the award-winning book Mobilizing Inclusion (Yale University Press, 2012). Her newest book, Living the Dream, will be released by Paradigm Press in September.  She can be reached at


, ,