Opponents of President Obama’s Cuba policies argue that by changing his position on the issue, Donald Trump’s performance among Cuban-Americans improved by 21 points, and that this is what clinched him the Sunshine State.
To make their case, they point to a September 2016 poll by The New York Times / Sienna College that indeed showed Clinton and Trump at 41-33 percent among Cuban-American voters. They then contrast the 33% figure with the Edison Research exit poll showing Trump at 54% with Cuban-Americans.
There are three big problems with such a sweeping conclusion:
- First, The New York Times / Sienna College poll had a tiny Cuban-American sample of 40-50 people (for our purposes, we’ll just say 50). Such a small sample yields a margin of error of +/- 14 points. No serious analysis would rely so heavily on such a small sub-sample where the group in question was not the primary target of the research.
- Even the Times’ own polling guru, Nate Cohn, cautioned readers to take the Cuban sample with a grain of salt.
- This arguments cherry-picks (unreliable) survey results that support their thesis and ignores the overwhelming majority of (good) data that contradicts it. Sixty-two percent of Americans support ending the U.S. embargo on Cuba, according to The New York Times. Similarly, a Bloomberg Politics poll—that nailed the election results—found that two-thirds of Florida voters, including 57% of Hispanics, support ending the embargo. Moreover, a Florida International University survey found that 55% of Cuban-American registered voters (a 743-person sample size with a +/-3.6% margin of error) support President Obama’s Cuba measures. The only reason why the President’s detractors would argue that a 50-person sub-sample is more credible than an academic study where its 743 respondents were the primary research target is because the results of the latter discredits their theory.
Certainly, if Cuban-Americans had “repudiated” Obama’s Cuba policy as some argue, Democrats would not have made 9 and 14-point gains in Hialeah and Westchester, respectively, in these two heavily Cuban and traditionally GOP suburbs in Miami-Dade County.
The argument that Cuban-Americans tilted the state in the president-elect’s favor is further weekend by historical trends showing that Clinton’s performance was among the highest best showing ever for a Democratic candidate among this electorate (regardless of one’s preferred exit poll). Republicans are supposed to win Florida’s Cuban-American vote. The fact that Clinton ended up in the 40’s and made inroads in Hialeah and Westchester is a good sign for Democrats and should be troubling for Republicans.
Of course, the most obvious problem with the argument by the opponents of the President’s Cuba policy is that it overstates the size and weight of the Cuban-American vote in Florida while ignoring the shifts that occurred among other demographic groups, namely the white non-Hispanic electorate that is 10 x’s larger than the Cuban-American (62% vs. 6%).
As Dr. Grenier of FIU and I pointed out earlier this week, the Cuban-American vote is still important but it was not decisive in this election. Even if Clinton’s Cuban-American numbers had improved by 10 points, she still would have lost the state. Meanwhile, had she and Trump simply replicated President Obama’s share of the white non-Hispanic electorate, Clinton would have gained an additional 467,000 votes and won the state.
As Steve Schale detailed, it was Trump’s strong performance in heavily white non-Hispanic suburban and exurban counties in Central Florida’s I-4 corridor, such as Pasco, Volusia and Hernando that won him the state— not a questionable “shift” among Cuban-American voters.