Ramirez 2005 - Multi-method Mobilization Among Latino Voters
Of three mobilization methods: robo calls, direct mail and phone calls, only live phone calls produce statistically significant mobilization effects among Latino voters.
Ramirez, Ricardo. 2005. "Giving Voice to Latino Voters: A Field Experiment on the Effectiveness of a National Nonpartisan Mobilization Effort." The Science of Voter Mobilization. Special Editors Donald P. Green and Alan S. Gerber. The Annals of the American Academy of Political and Social Science. vol. 601: 66-84.
Ricardo Ramirez presents findings of a randomized field experiment of 465,134 registered Latino voters. The National Association of Latino Elected and Appointed Official's(NALEO's)Voces del Pueblo voter mobilization effort in 2002 explored three alternative modes of communicating with voters: direct mail, robotic phone calls, and live phone calls from volunteers. Of the three, only live phone calls produced a statistically significant increase in voter turnout. The ineffectiveness of direct mail and robotic calls is consistent with results from other experimental campaigns. What remains unclear is the extent to which direct mail and robotic calls targeting low-propensity Latino voters would be more effective in presidential elections. For the present, it appears that, of the methods examined in this study, the most effective way to mobilize low-propensity Latino voters is through phone banks staffed by volunteers.
Electoral Context: This experiment was conducted during the November 2002 national election.
Subject Population: The 2002 Voces del Pueblo campaign sought to increase turnout among low propensity Latino voters. The voters targeted reside in precincts where Latinos make up at least 70 percent of the voters and where turnout in the 2000 presidential election was less than 50 percent. The registered voters targeted in this campaign had a voter turnout rate of 38 percent in the 2000 election, which was approximately half the voter turnout rate among registered voters nationwide. The GOTV effort was conducted in six sites: Los Angeles County, Orange County (California), Harris County (Texas), Denver Metropolitan Area (Colorado), New York City, and the state of New Mexico. The sites were chosen based on three criteria: infrastructure, resources, and Latino population.
Randomization Procedure: The initial database that was purchased from commercial vendors consisted of 465,134 registered Latino voters in low-propensity precincts. Of these, 405,058 individuals were randomly assigned to receive a voter mobilization contact through live calls, direct mail, and robo calls. The remaining 60,076 names were assigned to the control group. NALEO also produced radio and television public service announcements that were aired in its target communities.
Live calls. The live calling campaign was staffed by fifty-one bilingual paid volunteers who conducted calls in Los Angeles and Orange Counties. Callers in both counties logged 781 hours between October 24 and November 4, 2002 and attempted to contact 52,315 Latinos in Los Angeles and Orange Counties (35,853 and 16,462, respectively) who lived in the targeted precincts. Short training sessions were conducted prior to the phone, and phone canvassers were given a "scripted message" to follow but were encouraged to conduct calls in a more conversational style. In practice, the conversational approach was the norm among phone canvassers. Canvassers asked subjects if they intended to vote on Election Day and recorded the answers. The contact rate for the live phone calls was approximately twenty five percent.
Robo calls: In the California, Houston, and New Mexico sites, voters received two calls of a script read by the very prominent Univision anchorwoman Maria Elena Salinas. In Colorado, NALEO used Denver council member Debbie Ortega, who read a similar script in English. The decision to use an English script reflects the low rates of bilingualism or Spanish use in Denver. Whether the calls were in Spanish or English, the possibility remains that some recipients were not conversant in the language in which the robo calls were presented.
Direct mail: During the final weeks of the campaign, NALEO sent four pieces of direct mail to Los Angeles residents, two pieces to Orange County residents, and three pieces to those in other sites. It is not possible to determine the contact rate for direct mail.
Findings: Ramirez finds that the intent-to-treat effect of the direct mail was modest with only one out of seven areas exhibiting a statistically significant increase in turnout from direct mail alone. Robo calls had very little effect with an estimated intent-to-treat effect of 3.5 percentage points (standard error: 1.96 percentage points). Live phone calls produced the greatest increases in voter turnout. Ramirez' regression results estimate a treatment on treated effect of actually receiving a live call from the local phone banks raised turnout by 4.6 percentage points (standard error: 1.84 percentage points).