The special election in California’s 37th congressional district is a race with enough Latino voters to suggest that the political professionals there would maximize their chances of delivering a win for Senator Jenny Oropeza or Assemblywoman Laura Richardson if they gave Latinos more than one banner point in their polling crosstabs. An article in Congressional Quarterly points out that more than 40 percent of the residents in the district are Latino, though the percentage of voters is significantly less. By examining the results of a recent California primary election with a significant number of Latino voters, we can see the importance of drilling a little deeper than just ethnic self-identification when targeting Latinos.

In 2006, I was doing the research of a independent expenditure that was working in Democratic state legislative primaries throughout California. One primary was for an open seat in a Southern California district with more than 40 percent Latino voters. The top two candidates in this multi-candidate primary were both well-funded Mexican-American local officeholders. The pollster for one candidate (”Candidate A”) conducted a poll in February showing that candidate with a 25-point lead on Candidate B (50 percent to 25 percent). “With so few undecided voters,” the pollster wrote, “it will be nearly impossible for the other candidates to make up significant ground on [Candidate A] and narrow the margin.”

Candidate B ended up winning in June by 20 points, 52 percent to 32 percent.

One of the reasons for the dramatic turnaround was that Candidate B went after the Latino vote aggressively - including Spanish-language radio, mail and field. Candidate A did not do any Spanish-language media. As the above graph shows, Candidate B’s efforts paid off. In the first poll I conducted in the district, Candidate A was ahead with all Latinos, 38 percent to 34 percent. By my second poll, Candidate B had moved ahead with Latinos, 51 to 32 percent. Candidate A’s vote with Latinos who said they either spoke mostly English or both English and Spanish at home remained constant at 35 percent, which makes sense since Candidate A’s media was all in English. With Latinos who said they mostly spoke Spanish at home - about a quarter of district’s entire electorate - Candidate A dropped from 42 to 29 percent. Meanwhile, Candidate B’s vote grew with both groups.

The moral of this story is that Candidate A suffered from painting Latinos with too broad a brush. After the November 2006 general election, I conducted a poll of Latino voters who had just voted. One of my findings from that poll was that 33 percent of the Latinos who voted in November were immigrants. Fifty-two percent of immigrant voters said they were likely to speak Spanish at home and 26 percent said they got most of their news about the recent election from Spanish-language television. Only 17 percent of the sons and daughters of immigrants said they spoke mostly Spanish at home, dropping to 3 percent of second generation Latino voters. Just 4 percent of Latino voters who were born in the United States - 4 percent - said they got most of news about the recent election from Spanish-language television. In other words, Latino immigrant voters in California should get a significant portion of their communication in Spanish; other Latinos are better reached in English. Yet too many campaigns do not ask the questions on their polls that will help them gain this sort of understanding of the Latino voters in their election. Are they immigrants? Are they Mexican? Are they evangelicals? All these traits have a dramatic impact on Latinos’ voting behavior.

And, as Candidate A learned, they can tell you a lot about how to reach your voters.

This entry was posted on Saturday, June 2nd, 2007 at 5:23 pm and is filed under Blog, Latino Vote. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.