POLITICS

HUFFPOLLSTER: Donald Trump Is Floundering In The Midwest

But he’s still running a tight race with Hillary Clinton in Florida.

New polls in Rust Belt states show Donald Trump falling behind Hillary Clinton. One third of likely voters say the election will be “rigged.” And Americans are more interested in the election than the Olympics. This is HuffPollster for Wednesday, August 10, 2016.

TRUMP IS DOWN BY DOUBLE DIGITS IN PENNSYLVANIA - Janie Velencia: “Donald Trump has placed a big bet in his quest for the presidency: that his appeal to white working-class voters can make him competitive in Rust Belt states that traditionally vote Democratic. Increasingly, it looks like he’s going to lose that bet. A new set of battleground polls released on Tuesday show Trump trailing Hillary Clinton in the most critical states he’s hoping to flip. The Republican nominee sits 11 percentage points behind Clinton in Pennsylvania, taking just 37 percent of the vote to Clinton’s 48 percent, according to a NBC News/Wall Street Journal/Marist poll of registered voters. Similarly, a Quinnipiac University poll of likely voters conducted over the same week shows Trump 10 points behind, at 42 percent to Clinton’s 52 percent. Two other Pennsylvania polls released this week show Trump trailing by 10 points and 11 points. The HuffPost Pollster Pennsylvania chart, which aggregates publicly available polls, also shows a widening gap for Trump in the Keystone State.” [HuffPost]

Trump is also behind in Ohio, Iowa, with a close race in Florida - More from Velencia: “In Ohio, Trump is trailing Clinton by 5 points, 38 to 43 percent, in the NBC/WSJ/Marist poll and by 4 points in the Quinnipiac poll, which has the race at 45 percent to 49 percent….Trump trails by 4 points in Iowa, where he’s polling at 37 percent to Clinton’s 41 percent, according to the NBC/WSJ/Marist poll. The HuffPost Pollster Iowa chart shows Trump about 3 points behind Clinton....Trump is behind by 1 point in Florida, pulling 45 percent of the vote to Clinton’s 46 percent, according to Quinnipiac.” [HuffPost]

National polls look good for Clinton, too - HuffPollster: “New polls out this week show Hillary Clinton maintaining a dominant lead over Donald Trump nearly two weeks after the Democratic convention ended and in the wake of Trump’s worst week ever….these polls are giving hope to Democrats that Clinton’s lead expansion isn’t just a convention bounce, but could be more permanent. Some have gone so far as to say the race is over. It’s a bit early for that ― convention bounces sometimes last a few weeks, and Election Day is still three months away. Clinton also still has big unfavorable numbers. We could be only one big economic crash in September or October away from President Trump. Anything could still happen ― and in a year when anything has happened, caution in drawing early conclusions is warranted. For now, though, these polls are undeniably good news for Democrats.” [HuffPost]

VOTERS ARE NEGATIVE ABOUT THE COUNTRY, 1/3 SAY ELECTION IS RIGGED - John McCormick: “Nearly seven in 10 adults, 68 percent, say the country is headed in the wrong direction [in a new Bloomberg Politics national poll], with almost half of that group blaming Democrats and 22 percent pointing the finger at Republicans. Thirty percent said they aren’t sure who is to blame. A solid majority of likely voters, 56 percent, say the U.S. is in a dark and dangerous place, a figure that includes 87 percent of Trump supporters. Four in 10 say the country is in a strong position for progress on the economy and national security, including two-thirds of Clinton supporters…. A majority of likely voters also don’t agree with Trump’s suggestion that the election is ‘rigged,’ with 60 percent rejecting that view and 34 percent agreeing. Among Trump’s supporters, however, 56 percent say it’s rigged.” [Bloomberg]

DOES WEIGHTING TO PAST VOTE BIAS POLLS? -   Nate Cohn, on Monday: “There’s an interesting new entry in political polling: the U.S.C. Dornsife/Los Angeles Times ‘Daybreak’ poll. It’s different from other surveys because it’s a panel, which means it recontacts the same voters over and over. In 2012, a similar panel study done by RAND was considered a big success. But so far, the U.S.C./LAT panel has consistently been far out of step with other surveys. Donald Trump has led in nearly every survey it has conducted in the last few months, by as much as seven percentage points….One factor that could be contributing to the panel’s tilt toward Mr. Trump is its decision to weight its sample according to how people say they voted in 2012….The pollsters ask respondents whether they voted for President Obama or Mitt Romney. They then weight the sample so that Obama voters represent 27 percent of the panel and Romney voters represent 25 percent, reflecting the split of 51 percent to 47 percent between the two among actual voters in 2012….People just don’t seem to report their past vote very accurately. Answers tend to wind up biased toward the winner; often, people who vote for the loser say they ‘can’t remember’ or say they voted for someone else….If the panelists, like those in other surveys, are likelier to recall voting for the winner (Mr. Obama), then the poll is unintentionally giving extra weight to Republican voters.” [NYT]

USC’S POLLSTERS EXPLAIN THEIR DECISION - In a call with reporters earlier this week, Arie Kapteyn, the executive director of the USC Dornsife Center for Economic and Social Research, acknowledged Cohn’s point but said the use of similar weighting in the 2012 election, which moved the RAND panel’s results about a point toward the Democrats, produced an accurate estimate. Kapteyn also referred to research showing that shifts in the polls are sometimes attributable to which candidate’s supporters are more eager to answer a survey at any given moment, rather than a genuine change in opinion. “If you don’t weight with previous voting, you’re also losing information,” he said. “Those are two sides of the coin. On one hand, you may indeed introduce some bias. On the other hand, by not doing that you may overestimate swings.”

The “Daybreak” panel differs in other ways from traditional polling. It asks voters to estimate both the percent chance that they’ll support a certain candidate, and the percent chance that they’ll vote at all. This allows the pollsters to account for the possibility that some Americans considered likely voters by traditional polls won’t show up, while some who wouldn’t be included in most likely voters will end up making it to the polls, but also makes comparisons to other surveys apples-to-oranges.

So far, only about half of all voters on the panel are 100 percent sure about their candidate of choice, lower than in 2012. “When you look at the people who are 100 percent made up at this time, you actually do see numbers that are a lot more reflective of what you would see in a traditional polling setting,” Jill Darling, the  survey director for the Understanding America Survey at USC, said. As both Cohn and the panel’s directors note, one of the prime benefits of a panel like this one is the ability to track change over time ― in this case, a shift of about 3.5 or 4 points toward Clinton over the past month.

AMERICANS ARE MORE INTO THE ELECTION THAN THE OLYMPICS - HuffPollster: “Forget the literal sparring and horse racing happening in Rio ― most Americans are far more interested in the political competition shaping up at home, according to a new HuffPost/YouGov survey…. Sixty-seven percent of those polled say they’re very interested in the election, while just 21 percent are very interested in the Olympics….And by a 9-point margin of 40 percent to 31 percent, Americans say they’d rather watch news about the election than news about the Olympics. (Another 23 percent are futilely trying to tune out both.) In a rare moment of political unity, Republicans, Democrats and independents all say they prefer campaign coverage, although the margins differed by party. Democrats chose election news by only a narrow 5-point margin, while Republicans preferred it by a more substantial 23 points.” [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).

WEDNESDAY’S 'OUTLIERS' - Links to the best of news at the intersection of polling, politics and political data:

-Samantha Neal explains why third party candidate Evan McMullin could cause Donald Trump trouble in Utah. [HuffPost]

-Charles Stewart III and Stephen Pettigrew note a study that finds the U.S. improving at running elections. [WashPost]

-An international survey finds many people still concerned about their national economies. [Pew Global]

-Kathleen Weldon looks at the history of public opinion on the Family and Medical Leave Act of 1993. [HuffPost]

CONVERSATIONS