By VICTORIA M. DEFRANCESCO SOTO and JENNIFER L. MEROLLA
Gone are the days of Gerald Ford biting into a tamale, husk and all, or of George H.W. Bush referring to his Latino grandchildren as, “the little brown ones.” Campaigns today are more adept at targeting Latinos. And these micro-targeted efforts have paid off in gaining the favor of Latino voters. But as our research indicates, the downside to the increasing accuracy of hitting the Latino bull’s eye is the resultant ricochet — the potential negative effect on non-Latino voters.
In a research study conducted during the 2008 presidential election, we presented a sample of registered voters a series of general and Latino-targeted Obama and McCain ads. There were three types of Latino ads. In the first type (we called this the uberLatino ad), Spanish is spoken and there is a Latino endorser. In the context of this election, think for example of a Romney ad in which Sen. Marco Rubio speaks in Spanish. The remaining two types of Latino ads were mixed — they featured either a Latino endorser speaking English or an Anglo endorser speaking Spanish. In recent ads, this would take the form of, say, salsa singer Marc Anthony’s English language ad endorsing Obama or Craig Romney’s Spanish-language ad where he speaks about his father.
At the same time, we found no effect of the general ads or the Latino endorser speaking English ad on increasing (or decreasing) the likelihood of voting for a given candidate. And these effects held, regardless of the respondent’s dominant language. Put simply, you get more bang for your buck with a Latino targeted ad by employing Spanish.
When we turned to the effects of these same ads on non-Latino whites, the two mixed Latino ads seemed to backfire completely. The English ad featuring a Latino endorser and the Anglo endorser speaking Spanish decreased the likelihood of intending to vote for either Obama or McCain by at least 5 percentage points. Somewhat surprisingly, the overtly ethnically targeted ad, the Latino endorser speaking Spanish, did not have any effect. It appears that non-Latino whites may have tuned such ads out, likely because they clearly targeted Latinos.
Among black voters, we found a much more marked negative effect of Latino targeted ads. Support for Obama among black respondents who had not been exposed to an ad was 98 percent, but among those exposed to a Latino-targeted ad, support for Obama dropped 20 percentage points compared to those exposed to a non-targeted message.
In the 2008 presidential campaign, as in the current one, the Republican candidate has spent less time courting Latinos than George W. Bush did. Nevertheless, Romney, recognizes that he has to solidify a base of Latino Republicans. He also needs to persuade Latino independents or those disenchanted with Obama to give the Republican candidate a try, or at least not to vote for Obama. Given the presence of Latinos in a number of swing states, effectively reaching this electorate is the difference between 0 or 29 electoral votes in a key state like Florida.
In these last weeks, both the Republican and Democratic parties will unleash a barrage of Latino targeted ads. And while in war there is always the possibility of friendly fire casualties, the campaigns would be well-advised to take great care in how they target the Latino electorate so as not to disperse allies. It seems that the best way to hit the bull’s eye among Latinos without getting a ricochet among whites is to air Latino targeted ads in Spanish. This may be an effective strategy for Romney. However it will not be as effective for Obama because all the Latino targeted ads in the past presidential cycle decreased support among African-Americans, a key Democratic constituency.
This op-ed originally appeared on Politico. Dr. Victoria M. DeFrancesco Soto is an adjunct professor at the University of Texas Lyndon B. Johnson School of Public Affairs and a fellow at the Center for Politics and Governance; Dr. Jennifer L. Merolla is an assistant professor in the Department of Politics and Policy at Claremont Graduate University. This research project was conducted in collaboration with Matt Barreto at the University of Washington and Ricardo Ramirez at Notre Dame University.