Entrance poll results that show Bernie Sanders winning Latinos in Nevada are causing controversy. GOP voters increasingly say they'd support Donald Trump as the Republican nominee. And a polling experiment shows just how much anticipated turnout matters. This is HuffPollster for Monday, February 22, 2016.
ENTRANCE POLLS SAY SANDERS WON LATINOS IN NEVADA - Elise Foley: "Hillary Clinton easily bested Bernie Sanders in the Nevada Democratic caucuses on Saturday. But whether that was thanks to Latino voters -- or in spite of them -- isn't clear. That's because of contradictory results from the event. Entrance polls found that 53 percent of Latinos were supporting Sen. Sanders (I-Vt.), compared to 45 percent supporting former Secretary of State Clinton. But Clinton won caucuses in heavily Latino areas, casting doubt on whether the polling presented an accurate picture of how Latinos voted." [HuffPost]
But results in heavily Latino areas indicate Clinton had the advantage - Nate Cohn: “[T]here are a lot of reasons to question the findings from the polls. They have a small sample of precincts and voters, and they simply were not devised to provide precise estimates of the Hispanic vote….The actual election returns in Las Vegas’s Clark County hint at a different story. Analyzed neighborhood by neighborhood, they suggest that Mrs. Clinton might have won the Hispanic vote by a comfortable margin. She won about 60 percent of delegates in heavily Hispanic areas, a result that calls the finding of the polling into question.” [NYTimes]
Exit polls are ill-equipped to measure minority voters at the precinct level - More from Cohn: “In general, entrance/exit polls are not well suited to measure the Hispanic vote. This one is particularly problematic. Over all, the poll included just 1,024 respondents — and just 213 Hispanic respondents…. an entrance-exit poll isn’t a random sample of the population like a normal poll. It’s a random sample of precincts, usually between 15 and 60 in a state exit poll. This one had just 25 precincts. Race is not usually used as a criteria for selecting precincts….The exit polls have also been criticized for other issues measuring Hispanic voters, like whether pollsters conduct Spanish-language interviews. They generally do not….All evidence considered, and although we can’t know for sure, I’d err on the side of a Clinton win among Hispanic voters. But it would be hard to argue that she won Hispanic voters by a lot.” [NYTimes]
UPDATE: Pollster Joe Lenski noted on Twitter that the entrance/exit polls are in fact offered in Spanish to caucus-goers and voters in states with high proportions of Hispanics. Lenski confirmed to HuffPollster via email that they have always made Spanish questionnaires available, and that Spanish was an option in Nevada on Saturday, and will be available in Nevada on February 23, Texas on March 1 and Florida on March 15.
Pollster cites Latino age gap as reason for Sanders advantage - Anita Kumar: "The company that conducted the entrance polls of Nevada’s Democratic caucuses showing Hillary Clinton lost the Hispanic vote stood by its research Sunday in the face of an aggressive push back by the former secretary of state’s campaign. ‘Like any other poll, each campaign is going to try to pick out data that helps their cause,’ said Joe Lenski, executive vice president of Edison Research….He said Sanders’ numbers were driven by dramatic differences in Hispanics by age. According to the poll, Sanders won Hispanic caucus-voters aged 17-29 by 83-12 percent, and Clinton won those aged 30 and older by 65-34….Ultimately, there is no way to ever know for sure who won the Hispanic vote because official voting results don’t break down votes by race." [Miami Herald]
There’s evidence that Sanders is gaining among Latinos nationally - Mark Blumenthal and Jon Cohen, prior to the caucus: "While our national surveys have shown little discernible trend among all Democrats since the Iowa Caucuses, the movement among Latino voters suggests that a critical part of the so-called firewall of support that Clinton's campaign had hoped to rely on among non-white Democrats may be crumbling…. [A]mong Hispanic voters, what had been a two-to-one or better Clinton lead for the past six months has narrowed dramatically in the two weeks following the Iowa Caucuses. Where our surveys had shown Clinton leading by 20 percentages points or better during December and January, her advantage had reduced to just three percentage points (46 to 43 percent) on our most recent week of tracking (from February 8-February 14)." [NBC]
SANDERS NEEDS HIGHER VOTER TURNOUT TO GET NOMINATION - Michael McDonald: "If Sander's success depends on high voter turnout, his pathway to the Democratic nomination only gets bumpier in the coming contests. Voter turnout (statistics here) usually declines after Iowa and New Hampshire. In 2008, the last time the Democratic Party had a contested presidential nomination race, New Hampshire had the highest primary turnout rate of 54%. In the following primaries, the turnout rate averaged 30%, with only a few states rising above 40%. For the caucuses, which are low turnout affairs by their very nature, the Iowa caucus had the highest turnout rate of 16%, and the following caucuses averaged 5%.... So, if Sanders campaign relies on high voter turnout, he has a monumental challenge in front of him. He must truly transform politics in a way that has not been seen before in America." [HuffPost]
HUFFPOLLSTER VIA EMAIL! - You can receive this daily update every weekday morning via email! Just click here, enter your email address, and click "sign up." That's all there is to it (and you can unsubscribe anytime).
RUBIO OUT-PERFORMED POLLING TO TAKE SECOND IN SOUTH CAROLINA - HuffPollster: "Businessman Donald Trump won the South Carolina primary Saturday night, but on Sunday morning he exaggerated the significance of that win. On MSNBC’s "Morning Joe," Trump said it was amazing to win by such a big margin -- 10 points -- and claimed that he won more support than the polls indicated he would….The problem with that claim? HuffPost Pollster’s polling average put Trump’s support at 33.9 percent, and he actually won 32.5 percent of the votes in South Carolina….If anything, Trump underperformed the polling average…. Sen. Marco Rubio (R-Fla.) is the only candidate with a legitimate claim of outperforming the polls in South Carolina. He averaged 15.9 percent support according to HuffPost Pollster, although the polls conducted closest to the primary showed him surging into the range of his 22.5 percent finish in the voting." [HuffPost]
GOP VOTERS INCREASINGLY BELIEVE TRUMP CAN WIN THE GENERAL ELECTION - Bill Barrow and Jill Colvin: "Donald Trump has emerged as the front-runner for the GOP nomination by winning over roughly a third of Republicans in the early voting states and in preference polls….But a new AP-GfK poll finds registered Republicans and GOP-leaning voters put Trump at the top of the still-unwieldy GOP field when it comes to which candidate fits best with their stand on the issues. They give Trump the best marks for competence and decisiveness. Far more Republicans than not say they'd vote for Trump in the general election, and 86 percent of Republican voters think he can win in November — giving him a 15 percentage point advantage over his nearest rival. If the number of Republican candidates shrinks as expected after Saturday's primary in South Carolina, Tuesday's Nevada caucuses and on Super Tuesday on March 1, the Trump coalition, it would appear, has plenty of room to grow." [AP]
Numbers from NBC/WSJ polls show the same pattern
POLLING EXPERIMENT SHOWS HOW TURNOUT CAN MAKE A DIFFERENCE - One of the biggest challenges pollsters face, especially during primary elections, is in predicting how many voters will turn out. An increasingly-common method for doing so is to incorporate data from voter files, publicly-available information on Americans' voting histories. If survey respondents can be matched up with their vote histories, pollsters may have a clearer sense of their likelihood to show up on Election Day than they'd get from simply asking about their intentions to vote. The GOP firm Echelon Insights conducted an experiment to test this, using automated phone calls to reach 935 voters in the two days prior to the South Carolina GOP primary. They also aimed to see whether using a voter list and weighting on turnout would be enough to address the bias caused by excluding cell-phone only households, which can't be reached through automated calls.
Trump fares slightly better in a high turnout scenario - Pollsters Patrick Ruffini and Kristen Soltis Anderson in a memo provided to HuffPollster before the election, which saw a record turnout of more than 730,000 voters: "Our goal was to show the various ways that turnout scores from the voter file could be used to construct different scenarios based on expected turnout. If we weight our data to traditional demographics, using age and gender, we find Trump leading with 31%, Rubio at 20%, Cruz at 18%, Bush at 10%, Kasich at 9%, and Carson at 7%. Introducing individual-level modeled turnout scores into the weighting yielded some modest differences…. Lower turnout hurts Trump, but we find the effect is perhaps not as great as people believe...Based on this, we are also able to say that our raw sample resembled a turnout universe of roughly 800,000, based on the distribution of turnout scores from people who responded to the survey. Campaign pollsters have been critical of media pollsters for building samples that allegedly represent a universe much larger than those that will actually vote. We think that using voter lists with turnout scores could be a good way of getting a handle on whether the universe we're talking to in our surveys reflects the universe that is likely to turn out to vote."
MONDAY'S 'OUTLIERS' - Links to the best of news at the intersection of polling, politics and political data:
-Donald Trump leads in 10 of the next 14 primary states. [NY Post]
-Polling suggests Jeb Bush voters won't be a solid voting bloc for any of the remaining candidates. [HuffPost]
-Philip Bump finds that Hillary Clinton's polling looks a lot like it did in 2008. [WashPost]
-Harry Enten foresees a series of victories for Clinton post-Nevada. [538]
-Nate Cohn says Hillary Clinton's win in Nevada suggests national strength. [NYT]
-Chris Cillizza explains why Jeb Bush never had a chance in the 2016 race. [WashPost]
-Democratic voters care deeply about the income gap and Wall Street regulation. [AP]
-A majority of Americans support raising the minimum wage. [AP]
-Americans view China as the world's leading economic power. [Gallup]
-Views on abortion have become increasingly polarized. [WashPost]