President Donald Trump told Mexican President Enrique Peña Nieto in a phone conversation that he won 84 percent of the Cuban-American vote in the November elections.
“In the latest election, I won with a large percentage of Hispanic voters. I do not know if you heard, but with Cuba, I had 84 percent, with the Cuban-American vote,” Trump said during the Jan. 27 call, according to a transcript published Thursday by The Washington Post.
But the best estimates of Trump's share of the Cuban-American vote in November are far, far lower — 50 to 58 percent — and experts say they have no idea where Trump could have gotten his number.
The White House did not immediately respond to a request for comment.
Premium content for only $0.99
For the most comprehensive local coverage, subscribe today.
No one knows exactly how Cuban Americans vote because the process is secret, but many voters are regularly surveyed at exit polls to determine their preferences.
Another way to estimate preferences is to look at the residents of voting precincts and try to align them with vote results. Such estimates, however, depend in part on how the residents define themselves — just as Hispanic or specifically Cuban American, for example.
But none of the estimates of Cuban-American votes for Trump reached 84 percent.
Let’s take a look at exit polls first. An exit poll by the non-partisan Edison Research, which does polling for CNN and Fox, gave Trump 54 percent of the Cuban-American vote in Florida. About 67 percent of the 1.2 million voters of Cuban origin live in Florida, according to the Pew Research Center.
Latino Decision, a Democratic polling firm, gave Trump 52 percent of the Cuban-American vote in Florida, based on a November election eve poll. The firm estimated that Democratic candidate Hillary Clinton won about half the Cuban-American vote nationwide, indicating that Cuban-American voters outside Florida tend to be less conservative.
Immediately after the election, Republicans and Democrats clashed over the numbers.
Mauricio Claver-Carone, former executive director of the U.S.-Cuba Democracy PAC who was appointed by Trump to the Treasury Department, estimated that Trump won 58 percent of the Cuban-American vote based on his review of results from about 30 Miami-Dade precincts with large Cuban-American populations.
Democratic strategist Giancarlo Sopo and Florida International University professor Guillermo Grenier estimated that Trump won 50 percent of the Cuban-American vote after reviewing results from Hialeah, Westchester and West Miami.
Democratic pollsters and analysts said exit polls also showed Trump did not do as well as Mitt Romney, the Republican candidate in 2012, who won an estimated 65 percent of the Cuban-American vote in Florida.
Sopo and Grenier concluded that the Cuban-American vote in Florida was “important but not decisive” for Trump's victory.
“Even under Mauricio's (Claver-Carone) analysis there wasn't a single Cuban precinct where Trump got 84 percent of the vote,” said Sopo.
A more extensive analysis by University of Florida professor Daniel A. Smith also showed that Trump fell far short of his alleged 84 percent. Smith didn’t find “any empirical evidence” to support that claim, he told the Herald.
Smith said he plugged voter registration rolls into his calculations in order to reach a better estimate of how many Cuban Americans voted in each district. His results are conservative, he added, because the country of origin of some voters is unknown.
Smith then analyzed not 30 but 326 precincts that met several requirements: at least half the registered voters were Hispanic, at least 20 percent of those who voted were Cuban American and at least 100 Cuban Americans voted.
His analysis, available in his blog, showed that as the number of Cuban-American residents of the precincts grew, the tendency to vote for Trump also increased but not in an absolute or massive manner.
Even in precincts where more than half the Hispanic voters said they were born in Cuba Trump won only 55 percent.
Daniel A. Smith, UF
“Even in precincts where more than half the Hispanic voters said they were born in Cuba Trump won only 55 percent,” Smith wrote. Clinton also did well in other precincts, he added.
“For example, Clinton won nearly 55% of the more than 1,100 ballots of voters registered in Johnny L. Cotson Sr. Park, deep in the heart of Hialeah,” he wrote. “Hispanic voters cast over 82% of the total ballots in the precinct, and at least 60% of those Hispanic voters are naturalized US citizens who were born in Cuba.”
Official results show Clinton crushed Trump in Miami-Dade County, with 63 percent of the vote, and nearly tied Trump in Hialeah, which has an overwhelmingly Cuban-American population.
“The seemingly indelible narrative — that Cuban-Americans in Florida since the botched Bay of Pigs invasion have staunchly backed Republican presidential candidates — has faded,” Smith added.
Smith said the one precinct where Trump won the highest percentage of the Cuban-American vote in Miami-Dade — 68 percent — is Precinct 778 in Kendall.
Numbers aside, the belief in Trump's team that the Cuban-American vote helped its candidate win Florida's electoral votes nevertheless appears to have been decisive in Trump's decision to stiffen U.S. policy on Cuba.
The one indisputable fact?
Whether Trump won the Cuban-American vote by the 54 percent estimated by the polls or that unexplained 84 percent, the Cuban-American vote was decisive in changing U.S. foreign policy.
Follow Nora Gámez Torres on Twitter: @ngameztorres