One of the storylines that emerged following Donald Trump’s victory on November 8 was that President-elect Trump outperformed Mitt Romney’ s 2012 showing with Hispanic voters. Trump also approached the 31 percent of Hispanic vote that Senator John McCain received in 2008, after McCain had spent the three years prior to the election pushing a massive amnesty bill in the Senate.
Donald Trump was predicted to “hit historic lows” with Hispanic voters since the media was constantly replaying some of his early, clumsy statements on illegal immigration, and the pundits predicting the outcome of the election had a constant positive feedback loop assuring them that Hispanic voters would guarantee a humiliating Trump loss.
The problem was that Trump won handily, and he performed better than expected with Hispanic voters, according to exit polling. The predicted surge did not happen, and most political observers are trying to figure out why. Except for pollsters from Latino Decisions, an organization which received a lot of attention pre-election for predicting that Hispanic voters would carry Clinton to victory. They are pretending their polling was spot on, and the reason Trump unexpectedly won was – well, they haven’t come up with a good explanation for that, yet.
Any poll can produce inaccurate results depending on the sample, the wording of the questions, and the honesty of the respondents. Since the question in a Presidential poll is straightforward (“who do you plan to vote for?"), clearly the samples (likely voters) most pollsters selected were unreliable, and it is plausible that "shy Trump voters" did exist to an extent.
There is a real difference, however, between a pre-election poll and a post-election exit poll. An exit poll is conducted during the actual process of voting, the outcome of which soon becomes known, and it is designed to be much more likely to capture a sample of actual voters, not just people who say they likely will vote.
Latino Decisions, which was employed by the Clinton campaign, released an “election eve” poll that was already declaring a Clinton victory with the largest share, by far, of the Hispanic vote ever in a Presidential race. The timing of the release and the p.r. campaign behind it may have given some the idea that it was an exit poll and had captured the behavior of actual voters. It had not. It was simply another inaccurate pre-election poll, albeit one explicitly cheerleading for one of the candidates.
Even after it became clear that almost all pre-election polls had been severally flawed, Latino Decisions maintained that we shouldn’t believe the exit polls. Their reasoning boils down to two main arguments:
- Latino Decisions polling of how Hispanics voted is more accurate because it is based on their methodology, which proved to be widely inaccurate in predicting the outcome of the 2016 election.
- Latino Decisions refuses to accept the reality that Trump won with nearly 30 percent of Hispanics voting for him because they don’t like the outcome of the 2016 election.
FiveThirtyEight, the website devoted to predicting elections, which gained significant attention for accurately calling 2008 and 2012, got the 2016 race badly wrong. But, in contrast to Latino Decisions, it is trying to understand how. It looked at the claims made by Latino Decisions that the exit polls are exaggerating Trump’s share of the Hispanic vote and found them unconvincing.
Harry Enten of FiveThirtyEight points out that if it is reasonable to assume that exit polls finding Trump did better overall with voters than pre-election polls suggested, which was the case, it is also likely that he did better with Hispanic voters, too.
Enten went down to the county level to examine areas where Clinton overperformed with Hispanic voters and those where she underperformed, relative to Obama in 2012. He concluded very simply that Clinton didn’t do sufficiently better than Obama in some areas to make up for her doing less well in other areas. Given turnout and exit polling, Enten’s take is that:
…the results do suggest that if nearly 80 percent of Latinos voted for Clinton, as Latino Decisions argues, then Latino turnout must have been down in many counties, or Clinton must have done much worse than Obama among non-Latinos in those counties. Otherwise, the overwhelming pro-Clinton Latino vote would have swung heavily Latino counties more dramatically toward Clinton. The evidence, then, suggests that Clinton fell short among Latinos in one of two ways: Either she didn’t win as large a share of them as Obama, or she didn’t convince as many of them to turn out to vote. Since both the exit polls andLatino Decisions agree that turnout among Latinos was up, the latter explanation doesn’t seem likely.
Trump winning 29 percent of the Hispanic vote, while not a significant increase over Romney, is significant in that Trump did do better while running on a promise to end illegal immigration. Latino Decisions may refuse to accept this, but it is not surprising to those who understand that Hispanics are not a monolithic group, and Trump’s positions on immigration are popular with a majority of Hispanic voters.
ERIC RUARK is the Director of Research for NumbersUSA
