In a recent poll for NBC and the Wall Street Journal, the Democratic firm of Peter D. Hart Research Associates and their Republican counterpart, Public Opinion Strategies, have the good sense to ask a question about generation:

Were you or was one of your parents born outside the United States?

This is very much the sort of question I wish Democratic campaign pollsters would ask all the time when they are polling in jurisdictions with a significant Latino population. In addition, unless they have good reason to believe that the Latino population in that jurisdiction all come from a particular country, I also wish they would ask country of origin. And then cut straight to the chase by asking what is the main language spoken in the respondent’s home.

Those three questions, when combined with the usual questions in any good poll, add so much to a campaign’s understanding about how to speak to Latino voters: from messages and images to language. Sixth generation Mexican Americans are very different from recent immigrants. Naturalized Cubanos are very different than naturalized Central Americans. Treating them all the same - not asking the questions above - has consequences.

Assembly districts in California are huge - almost as many voters as a congressional district. Many of the campaigns end up being multi-million dollar affairs, especially once the money spent by independent expenditures is factored in. I was working for one such entity during the recently concluded Democratic primaries in California. A couple of our districts had sizable Latino populations.

One, in Southern California, had two Latinos and a Latina running against each other. In February, according to one of the candidate’s polls, the Latina had a 30 point lead. When I did my own poll of the district, I learned that more than 40 percent of the primary electorate was Latino, that 24 percent of the electorate spoke mainly Spanish at home and that 72 percent of the Latino voters were either born in Mexico themselves or had a parent who was born there. In other words, the data showed not only that Latino voters were critically important to winning the election, but also that there was a tremendous opportunity to reach out to voters in Spanish and evoke a proud immigrant story.

Independent expenditure campaigns are not allowed to coordinate with the campaigns of the candidates who they are supporting. As a result, my independent expenditure did everything it could to target Latinos but could not make up for the finding that the candidate we were supporting, the Latina, was conspicuously going after whites while hemorrhaging with Latinos. (A story for another day is that the hemorrhaging seemed to go undetected for too long by the other entities doing polling in the district.) Her drop in support with Latinos took place despite the fact that she had as compelling a story to tell to Latinos as either of her opponents - in fact, she was actually born in Mexico, just like so many of the voters she was going to need to win. But it was her well-funded opponent, also born in Mexico, who bombarded the district with Spanish-language field and mail. This played no small part in his ability to turn a 30-point deficit into a 20-point win.

I don’t think this would have happened if the Latina candidate had asked about generation, country of origin and language of choice in her baseline poll. Latino voters nationwide tend not to be Spanish-dominant speakers, but Latino voters in this particular district are. And the Latina didn’t speak Spanish to them.

We can’t just assume that Spanish is or isn’t the way to win the Latino voter, whether in a mayoral race, a congressional district, a gubernatorial election or nationwide. We need to understand who are the Latinos we are trying to deliver. We need data.

This entry was posted on Tuesday, June 13th, 2006 at 5:51 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.