For a few years now, I’ve been saying that no Republican Candidate of Color has won a race among voters of color. Susanna Martinez, Marco Rubio, and their ilk have not been able to win a majority of people of color in their states. Instead, they’ve won by garnering enough support from conservative Whites to overcome what are usually very low levels of support among the people of color in their states. I often include slides on this important fact in my stock PowerPoint presentation on Latino politics. Well, as much as it pains me to say it, this is no longer an accurate statement, according to recent polling we conducted.
Earlier this month, we polled over three thousand Mexican American registered voters in Texas in five of the counties with the largest number of Latino voters: Bexar, Dallas, El Paso, Harris, and Tarrant. What we learned was eye-opening. As you may recall, Texas and a few other “unimportant” states were left out of the National Election Pool exit polls in November, 2012. Thus, we had no public high quality indicators of what level of support that Texas Mexican Americans gave to Barack Obama or the other candidates on the ballot that cycle. Our poll findings indicate that 70% of Mexican Americans supported President Obama’s re-election, which was a full 7 points higher than the level of support he’d received in 2008. This strong support for the Democratic Presidential candidate makes sense, given the strong alignment between Mexican Americans in Texas and the Democratic Party. In our poll, 59% of Mexican American registered voters felt that the Democrats did the best job of representing their views.
But the poll also showed that, for the first time I’m aware of, Latino voters preferred a Republican Latino candidate over a Non-Latino Democrat in a major statewide race. (Feel free to email and provide an example of one prior to 2012 if you can find one.) Of those Texas Mexican Americans who voted for a Senate candidate, 55% chose Cruz and only 45% chose Sadler, the Democrat. It is important to bear in mind that 42% of the Mexican Americans polled chose neither of the candidates, so the levels of support enjoyed by each candidate among the registered Mexican American voters is even lower than what these numbers may imply. That is because some people did not vote at all, and some voted on other races but skipped voting on the Senate race, and some people may have chosen a write-in candidate or voted for an obscure third party candidate.
To better understand what these voters were thinking, we examined how people who supported President Obama overlap with people who supported Democrat Sadler. Clearly, nobody who supports that Muslim socialist Barack Obama could possibly also support Tea Party darling Ted Cruz. Wrong! Over a third of the Mexican Americans who voted for President Obama also voted for Ted Cruz. Interestingly, only 7% of Mitt Romney’s supporters crossed the party line and voted for Sadler, which makes you wonder about the stability of the Democratic voter base. But that’s another blog post.
Given how puzzling and troubling the phenomenon of Obama and Cruz supporter overlap is, we then examined whether the Latino Cruz supporters knew they were voting for a Republican. Perhaps they had been confused and inadvertently cast votes for a person whose policy positions are about as damaging to the vast majority of the Tejano electorate as they could be. It turns out that over a quarter of the Mexican Americans who voted for Ted Cruz think he is a Democrat (26%). (I know. Catch your breath. I’ve told you before that we have a great deal of work to do in Texas.) Only 69% of his Mexican American supporters accurately identify him as a Republican.
In the coming weeks, we’ll be analyzing this data in many more ways to try and unpack these troubling findings more thoroughly. Here is a link to one of Texas’ most prominent political blogs, Off the Kuff, that offers an explanation for this puzzle. In the meantime, I’ve got some slides that I need to delete from my standard PowerPoint presentation on Latino politics.