Hispanic voters were largely credited with President Obama’s victory in 2012, but they were not as crucial as many believed. Obama did not even need to win the Hispanic vote to put him over the top, thanks to high black turnout and support among white voters in the North. The turnout among Hispanic voters did not surge, even though exit polls implied that it had.
This year, Hispanic voters, perhaps motivated by Donald Trump’s harsh language against Hispanic immigrants, really might decide this election.
Early voting data unequivocally indicates that Hillary Clinton will benefit from a long awaited surge in Hispanic turnout, vastly exceeding the Hispanic turnout from four years ago.
It is too soon to say whether it will put her over the top. The geographic distribution of Hispanic voters means that many of her gains will help her in noncompetitive states like Texas and California, not Michigan and Pennsylvania.
But the surge is real, and it is big. It could be enough to overcome Trump’s strength among white-working class voters in the swing states of Florida and Nevada. If it does, it will almost certainly win Clinton the election.
In Florida, voters who indicated they were Hispanic on their voter registration form represent more than 15 percent of the early vote. In 2012, Hispanic voters were just 12 percent of the final electorate.
The numbers are striking in part because of the sheer magnitude of the early vote so far. Already, more than 6.4 million total voters have cast their ballots in the state — equal to 75 percent of the final turnout in 2012. In total, as many Hispanic voters have already cast ballots in Florida’s early voting period as cast ballots in all of 2012.
The Hispanic surge in Florida is not simply because Clinton has drawn typically reliable Election Day voters to vote early instead: According to Daniel Smith of the University of Florida, fully 36 percent of the Hispanics who have voted so far did not vote in 2012.
It is also striking because Hispanic voters are typically among the least likely to participate in early and absentee voting. If that pattern continues this year — suggesting a robust Hispanic turnout on Election Day — then Trump is probably in serious trouble.
The pre-election polls in Florida appear to have assumed a lower level of Hispanic turnout. The final Upshot/Siena poll in Florida suggested that the state’s electorate would be 67 percent white, by registration, and 14 percent Hispanic — just 2 percentage points higher than the 12 percent of 2012.
A new Quinnipiac poll today had Clinton ahead by 1 point in the state and put Hispanic voters at 16 percent of the electorate. But this was based on the race that registered voters self-reported to pollsters, not the race that they indicated on their voter registration form.
In our two Florida polls, registered Hispanic voters represented 13.6 percent of the electorate, but 16 percent of likely voters were self-identified Hispanic voters. About 22 percent of voters who were neither white, black nor Hispanic identified as “Hispanic,” even though it is not how they registered.
If registered Hispanic voters represent 15 percent of the electorate, self-identified Hispanic voters could be 18 or 19 percent of voters.
The data on Hispanic turnout is not as illustrative elsewhere, because most other states do not ask about race and Hispanic origin on voter registration forms. But this is a national trend.
The Hispanic vote in Nevada has propelled Democrats to a considerable lead in the early vote. Many analysts believe that it has already been enough to secure the state for Clinton. The turnout has surpassed 2012 levels in several of Las Vegas’ heavily Hispanic precincts.
The huge surge in Hispanic turnout is possible — and sustainable — in part because there was no surge four years ago. Even now, the turnout among white registered voters is at a higher percentage in the Florida early vote than among Hispanic voters because Hispanic turnout, historically so low, has a long way to go to catch up.
There is another possible error in the polls: Clinton’s share of the Hispanic vote. In general, the highest-quality polls of Hispanic voters give her a larger lead than the one Obama held with that group in 2012. But there are plenty of surveys where this does not seem to show up. There are very few surveys that show Clinton faring much, much better than Obama, suggesting an underlying bias in many public polls.
Why would the polls tend to underestimate Democratic strength among Hispanic voters? There is considerable evidence that pollsters tend to contact too many well-assimilated, English-speaking, high-turnout Hispanic voters who live in less Hispanic areas. These voters tend to be more Republican. If true, Clinton’s strength among less assimilated, Spanish-speaking and low-turnout Hispanic voters in heavily Hispanic and urban areas might be missed in the polls.
Clinton had a lead of 60-26 percent among Hispanic voters in Upshot/Siena Florida polls, which used English and Spanish interviews to get the right number of low-turnout voters and voters in heavily Democratic areas.
Whether Hispanic turnout will be enough for Clinton to win the presidency is hard to say. In the most contested states, Hispanic voters represent a larger than average share of the electorate only in Florida and Nevada.
They are just a fraction of the electorate in many of the states that could prove decisive — North Carolina, Ohio, Pennsylvania, New Hampshire or Michigan.
Hispanic voters can give Clinton a knockout blow with a win in Florida. It would be especially symbolic if it came early in the night.
But if Hispanic voters do not put Clinton over the top in Florida, she will need to cobble together enough strength among black voters and white Northerners in states like Pennsylvania and Michigan.
In that event, Hispanic voters would still be an important part of Clinton’s path — say, in helping win Nevada and Colorado — but not clearly decisive.
Either way, it is likely that the Hispanic vote will pad Clinton’s margin in the national popular vote and sustain her chances in the Electoral College.