A mounting array of evidence shows that the 2010 national exit poll failed to capture a representative sample of Latino voters and, resultantly, misstated levels of Latino support for candidates. However, research by Latino Decisions, conducted bilingually during the early voting period, provides a more accurate picture of Latino voter sentiment in the 2010 elections. When it comes to Latino voters in 2010, the national exit polls should be marked with a big asterisk, and the Latino Decisions figures used more reliably.
According to Frank Sharry, Executive Director of America’s Voice:
The national exit poll numbers for Latinos are wrong both intuitively and quantitatively, and deserve a big asterisk in the 2010 election analysis. Thankfully, we have an alternate set of information about Latino voters that stands up to the precinct-specific scrutiny and makes sense in light of what we know transpired on the campaign trail.
In-depth new analysis of the Latino vote in Nevada underscores that the national exit polls badly misstated Latino voter support in Nevada and elsewhere. Using precinct-specific data and based on “actual vote results from the state of Nevada, Harry Reid is estimated to have won 94% of the Latino vote” instead of the 68% estimated level of support figure generated by the national exit polls. Of note, the Latino Decisions data show Reid winning support from 90% of Nevada Latino voters – much closer to the official statement of votes from actual 2010 precincts. These figures also make intuitive sense, given the way Nevada Senate candidate Sharron Angle (R) engaged in anti-Latino race-baiting to a shocking degree.
The problems with the national exit polls’ estimates of Latino voters – which Latino Decisions accounts for in their methodology – are myriad. The pollsters rely on small sample sizes of Latino voters to draw conclusions; decide where to poll based on the percentage of the overall electorate that lives in rural, suburban and urban communities-without regard for the fact that the Latino electorate is concentrated in more urban locations; and do not accurately capture Spanish-dominant Latino voters. In fact, they under-survey Spanish-dominant Latino citizens “by almost 10-fold” according to Latino Decisions.
While the national exit poll will hopefully correct its methodological flaws with respect to Latino voters in future election cycles, pundits and analysts should know of the problems with the 2010 numbers and where to turn for accurate information on Latino voters in the interim.
New Analysis of Latino Voters in Nevada: http://latinodecisions.wordpress.com/2010/11/15/proving-the-exit-polls-wrong-harry-reid-did-win-over-90-of-the-latino-vote/
Latino Vote Stats: Latino Decisions’ Methodology More Accurate than National Exit Poll: http://americasvoiceonline.org/press_releases/entry/latino_vote_stats_latino_decisions_methodology_more_accurate_than_nati/
Latino Decisions’ 2010 Election Results: http://latinodecisions.wordpress.com/2010/11/02/latino-election-eve-poll-results-november-2-2010/