In the aftermath of Obama’s decision to halt deportations and provide temporary work permits to certain young undocumented immigrants, Latino Decisions released a new poll on how Latinos in five battleground states are reacting to Obama’s move. Unsurprisingly, the poll found that Latinos support Obama’s move and are more enthusiastic about his presidency. A few hours later, PPP released a poll conducted on behalf of DailyKos/SEIU showing Obama up to 61 percent with Hispanics, compared to just 53 percent last week. Some have characterized these polls as evidence that Obama has received a “boost” among Latino voters, but that may be wishful thinking. Neither poll demonstrates that Obama has made meaningful gains with Hispanics or Latinos.
Although the Latino Decisions poll showed that 49 percent of Latinos were “more enthusiastic” about Obama, there is no way to know which Latino voters are more enthusiastic. Were those the Obama supporters already likely to support the President, or did Obama build his support among Latinos who were relatively unlikely to vote? We also don't know the magnitude of the increase in enthusiasm. Is the move the difference between voting and staying home? None of this indicts the poll, but these basic questions do indict those claiming that the poll is proof of big Obama gains.
Yesterday afternoon, DailyKos released PPP's weekly numbers on Hispanics, which showed Obama up to 61 percent from 53 percent last week. As a general rule, I'm skeptical of relying on small sub-samples, since the margin of error is unusually high. In the last PPP poll, they appear to have surveyed or weighted the survey to just 90 Hispanic respondents. For a national poll of 1,000 registered voters, 90 Hispanics is about right. The issue is that relying on a small sample produces large margins of error, and the history of PPP’s tracking poll proves that. Since PPP started its tracking poll, Obama’s share of the Hispanic vote has varied between 49 and 72 percent. Not only do wild swings demonstrate the error involved in small sub-samples, but Obama’s 61 percent looks less impressive in broader context.
So what would demonstrate whether Obama’s move improved his standing with Latinos? At the very least, there are three indicators to watch over the next couple of weeks. First, Latino Decisions is publishing a poll with a full account of the horserace in five key battleground states later this month. Unlike most polls, Latino Decisions focuses on Latino respondents and consequently generates big sample sizes. Conveniently, Latino Decisions recently released a poll, providing a solid baseline for comparison. Second, you should follow Obama’s standing in Gallup’s tracking poll, which is notable for its large sample sizes (445 Hispanics in the last four week sample). Because of this, Gallup has shown considerably less variance than PPP, so it will be much easier to distinguish real movement from static. Third, Gallup’s tracking poll asks voters whether they will “definitely vote.” That will provide the most relevant indicator of whether Obama’s decision is likely to turn voters out to the polls. Currently, 66 percent of Hispanics say they will “definitely vote,” compared to 81 percent of whites and 76 percent of blacks. Unfortunately, it will take a while for Obama’s move to trickle through Gallup’s month-long numbers, but if these indicators eventually yield similar answers, we will be able to confidently say that Obama’s gambit paid off. Realistically, one should not expect clear and substantial movement toward Obama, since polls show that Obama has already consolidated most of the Obama 08' Latino vote. Further gains won't be easy.