Listen to the press call here.
Washington, D.C — On Thursday, November 1, 2012, the Center for American Progress Action Fund and Latino Decisions hosted a press call to discuss the causes and consequences of many pre-election national and state polls’ misrepresentation of the Latino Vote. CAP Action polling expert Ruy Teixeira and Latino Decision’s Matt Barreto discussed how for years, polls have failed to accurately account for Latino voters in their samples across the country for three reasons: 1) their sample sizes of Latinos are often far too small; 2) their Latino samples are often not representative of the Latino population; and 3) Latinos are often not being interviewed in Spanish at the correct proportions.
Matt Barreto put it this way, “When it comes to polling data on Latinos, it varies widely, but unfortunately it also varies wildly. We see estimates from a 50-50 tie to an 80-20 split… And that’s a real disservice. Any time that you see the numbers fluctuate like that at least one of the two polls is not correct. Not enough people are asking hard questions [about how they poll]…The poll itself could be wrong and not reflective of the Latino electorate as a whole.”
This miscalculation and misrepresentation of Latino voters continues today as Latino Decisions pointed out in a recent blog. Take the recent Monmouth University poll, in which only 48 percent of Latino voters supported Romney versus 42 percent who supported President Obama. This is not consistent with eight recent national polls of Latinos, where an average of 70 percent of Latinos supported President Obama while just 22 percent supported Gov. Romney.
Nor does it make much sense when some state polls have Romney leading among Hispanics in Florida or Obama leading by just eight points among Nevada Latinos. These findings are reminiscent of the 2010 election when Nevada polls drastically underestimated Harry Reid’s lead among Hispanics voters and consequently had the election outcome completely wrong.
In addition to errors in the pre-election polls, exit poll samples of Latinos have been notoriously wrong. In 2004, national exit poll partner Warren Mitofsky noted after the fact that their Latino data was not representative, and included too few Spanish speaking Latinos. In 2010 the exit polls reported that anti-immigrant candidate Sharon Angle won 30 percent of the Latino vote, but post-election analysis revealed she likely won 10 percent or less.
Given how crucial the Latino vote will be in this election, pollsters need to do a better job of analyzing the Latino vote and ensure that they represent the voters’ true intentions. CAP Action polling expert Ruy Teixeira and Latino Decision’s Matt Barreto discussed how the problems and pitfalls of poor polling and poor exit poll data blur the true measure of the Latino voter in a press call you can listen to here.
Listen to yesterday’s press call, here.
Resources:
To contact experts on this topic, please contact Laura Pereyra at [email protected] or 202.203.8689.
###