0

« PolitiFact's Truth-O-Meter tackles a claim about fat turkeys | Main | Obamacare tax-credits uncertain in Supreme Court case. What will FL do? »

Rick Scott campaign: post-election survey shows why exit polls on Cubans, Hispanics were wrong

@MarcACaputo

Though exit polls indicated Gov. Rick Scott lost Hispanics by a 20 percentage-point margin, the Republican’s campaign conducted its own post-election survey that showed he might have almost tied Democrat Charlie Crist with these voters.

Scott’s survey, conducted by OnMessage Inc., shows Scott earned 47 percent of the Hispanic vote compared to Crist’s 49 percent, unlike the exit polls that had the Democrat leading the Republican 58-38 percent. The 2010 exit polls had Scott winning 50 percent of the Hispanic vote to Democrat Alex Sink’s 48 percent.

“While an array of news articles point to a Rick Scott victory ‘despite losing ground with Hispanics,’ that’s simply not true,” wrote OnMessage’s Wes Anderson and Kayla Dunlap in a polling memo.

One potential problem with the surveys from OnMessage and Edison Research (which conducts the exit polls for media groups): Their Hispanic samples were relatively low. OnMessage’s sample was 304 and Edison’s was 367. So the error-margins of the results will vary widely. (UPDATE/aside: A few readers have pointed out it's important to note that some voters in post-election surveys have a tendency to say they backed the winner).

A third survey, conducted on the eve of the election by the premier Hispanic polling firm of Latino Decisions, had 400 Florida respondents and found Crist leading Scott 52-45 percent -- results that fall somewhere in between the OnMessage and Edison surveys.

OnMessage’s polling also took issue with Edison’s results for Cuban-American voters. It’s always a contested topic because Cubans (especially those in Miami-Dade) tend to vote Republican and are the most-reliable of Hispanic voters. Also, because Cuban-Americans are a subset of Hispanic voters, the margin of error in surveying this demographic group is even greater.

OnMessage said Scott won Cubans over Crist 65-30 percent. Edison showed Crist ahead of Scott, 50-46 percent. Unfortunately, Latino Decisions didn’t report Cuban-voter results.

So who’s right? Who knows? When the Florida voter file is finally updated next month, we can examine voting patterns of heavily Cuban-American precincts to get a better idea of how the vote broke.

My guess is OnMessage is more right on Cuban voters. It’s tough to believe that Crist, who made little outreach with Spanish-speaking voters and who called for an end to the Cuban embargo, would have attracted majority support from Cubans. Yes, it’s true that younger Cuban Americans tend to vote more Democrat or independently and aren’t as hardline about Cuba policy, but most election data indicated this was an older electorate.

Scott, meanwhile, had a Cuban-American running mate in Carlos Lopez-Cantera and the support of Miami-Dade’s Cuban-American legislative delegation. Scott got just 39 percent of Miami-Dade’s vote in 2014 and, considering 72 percent of the county’s registered Republicans are Cuban-Americans, it’s reasonable to guess that an outsized portion of the Scott vote was among Cuban Americans.

“When the Hispanic vote is broken down by county of origin, we find that Governor Scott won a sizable majority of Cuban voters as well as more Puerto Rican voters than many expected,” Anderson and Dunlap wrote. “In the end, most Hispanic voters were focused on the economy, and they decided that under Governor Scott’s leadership, the state’s real estate and job markets are headed in the right direction.”

Download Scott poll

Comments