Washington Post: Clinton 46, Trump 45

Whoa.

With nine weeks until Election Day, Donald Trump is within striking distance in the Upper Midwest, but Hillary Clinton’s strength in many battlegrounds and some traditional Republican strongholds gives her a big electoral college advantage, according to a 50-state Washington Post-SurveyMonkey poll.

The survey of all 50 states is the largest sample ever undertaken by The Post, which joined with SurveyMonkey and its online polling resources to produce the results. The state-by-state numbers are based on responses from more than 74,000 registered voters during the period of Aug. 9 to Sept. 1. The individual state samples vary in size from about 550 to more than 5,000, allowing greater opportunities than typical surveys to look at different groups within the population and compare them from state to state.

The massive survey highlights a critical weakness in Trump’s candidacy — an unprecedented deficit for a Republican among college-educated white voters, especially women. White college graduates have been loyal Republican voters in recent elections, but Trump is behind Clinton with this group across much of the country, including in some solidly red states.

The 50-state findings come at a time when the average national margin between Clinton and Trump has narrowed. What once was a Clinton lead nationally of eight to 10 points shortly after the party conventions ended a month ago is now about four points, according to the RealClearPolitics polling average. A number of battleground states also have tightened, according to surveys released from other organizations in recent days.

The Post-SurveyMonkey results are consistent with many of those findings, but not in all cases. Trump’s support in the Midwest, where the electorates are generally older and whiter, appears stronger and offers the possibility of gains in places Democrats carried recently. He has small edges in two expected battlegrounds — Ohio and Iowa — and is close in Wisconsin, Pennsylvania and Michigan, each of which Democrats have won in six consecutive elections.

At the same time, however, Trump is struggling in places Republicans have won consistently and that he must hold to have any hope of winning. These states include Arizona and Georgia, as well as Texas — the biggest surprise in the 50-state results. The Texas results, which are based on a sample of more than 5,000 people, show a dead heat, with Clinton ahead by one percentage point.

[…]

Of all the states, Texas provided the most unexpected result. The Lone Star State has been a conservative Republican bastion for the past four decades. In 2012, President Obama lost the state by 16 points. For Democrats, it has been among the 10 to 15 worst-performing states in the past four elections.

The Post-SurveyMonkey poll of Texas shows a dead heat with Clinton at 46 percent and Trump at 45 percent. Democrats have long claimed that changing demographics would make the state competitive in national elections, but probably not for several more cycles.

A comparison of the current survey with the 2008 Texas exit poll (there was no exit poll there in 2012) points to reasons the race appears close right now. Trump is performing worse than 2008 GOP nominee John McCain among both whites and Hispanics, while Clinton is doing slightly better than Obama.

Among men, Trump is doing slightly worse than McCain did eight years ago. The bigger difference is among women. McCain won a narrow majority of women in Texas while Trump is currently below 40 percent. That’s not to say Texas is turning blue in 2016. Given its history, it probably will back Trump in November and possibly by a comfortable margin. But at this stage, the fact that it is close at all is one more surprise in a surprising year.

You can see the Texas numbers here, though there’s not much to see. In the four-way race, it’s tied at 40-40, with 11 for Libertarian Gary Johnson and 3 for Green Jill Stein. The 46% figure is on its own the highest number I’ve seen for any Democrat in a statewide poll in Texas since I started paying attention. If it holds up, then no matter what else happens, she will have accomplished something here.

I should note that there are some concerns about the methodology used for this poll:

The Post-SurveyMonkey poll used an online-based sampling methodology that differs from previous polls by The Washington Post. Those are telephone surveys based on random samples of cellular and landline phones.

The new poll was conducted online as part of SurveyMonkey’s 2016 Election Tracking project, which recruits respondents from the large number people who take polls on the company’s do-it-yourself survey platform, roughly three million each day. A subsample of respondents to this range of surveys — which includes formal and informal polls of community groups, companies, churches and other organizations — were invited to participate in a second survey with the prompt, “Where do you stand on current events? Share your opinion.” The survey was not advertised on any website, so individuals could not “click-in” in an effort to influence results. A survey invitation could be used only once.

From Aug. 9 to Sept. 1, the survey asked the sample of 74,886 registered voters about their presidential support, including between 546 and 5,147 respondents in each state. The final sample was weighted to the latest Census Bureau benchmarks for the population of registered voters in each state.

The Post-SurveyMonkey poll employed a “non-probability” sample of respondents. While standard Washington Post surveys draw random samples of cellular and landline users to ensure every voter has a chance of being selected, the probability of any given voter being invited to a SurveyMonkey is unknown, and those who do not use the platform do not have a chance of being selected. A margin of sampling error is not calculated for SurveyMonkey results, since this is a statistical property only applicable to randomly sampled surveys.

The ability of random samples to represent the overall population is grounded in probability theory. With non-probability samples, testing is necessary to ensure a particular sampling strategy, along with adjustments to match population demographics, can consistently produce accurate estimates.

See here for more, including full question wording. I’ve seen a number of critiques of the poll, from folks like RG Ratcliffe on Facebook, which hold me back from embracing this as anything more than just another data point. Five Thirty Eight includes this result in its tracking page for Texas, but grades the poll at a C-, and the overall 538 assessment of Texas is more influenced by three weekly Ipsos Reuters polls that have shown Trump with much more typical double-digit lead. They can’t both be right. There’s a Texas Lyceum poll coming out next week, so perhaps we’ll get some corroboration for one or the other.

The Chron rounds up some local reactions.

Matt Angle, executive director of the Lone Star Project, an organization dedicated to making Democrats competitive again in Texas politics, noted that the SurveyMonkey poll’s web-based methodology has yet to be proven. He remains hopeful, but unconvinced, of a Clinton tie with Trump.

“I haven’t seen anything else to make me think that,” Angle said. “But I do think the atmosphere (for Democrats) is good. It’s better than I’ve seen it in recent elections.”

Chris Perkins, a GOP pollster and political consultant in Austin, said he, too, was skeptical. “Internet polling has been proven over and over to be wildly inaccurate. Can some of them be correct? Yes. But then there’s the old saying, ‘even a broken clock can be right twice a day.'”

Recent internal polls by both Democratic and Republican-leaning pollsters have tended to give Trump a strong lead in Texas, though the margins have fluctuated significantly in what many analysts rate as an unconventional election year.

[…]

The length of time covered by the poll also has raised questions, since it does not reflect potential changes in respondents’ attitudes over a three-week period while the campaigns ebbed and flowed.

“I’m surprised they would put that much stock in an online survey, based on the fact that internet surveys have a tendency to be inaccurate and that it took nearly a month to complete,” Perkins said.

Some academics say that as the science of polling evolves, however, each new poll has to be seen as a piece of a puzzle.

Rice University political scientist Robert Stein, a polling expert, said that well-designed internet polls can be as reliable if not more than traditional telephone polls. “You have to treat this is one more data point, and it’s a big data point,” he said. “I wouldn’t disregard it.”

James Henson, director of the Texas Politics Project at the University of Texas, said he has used both types of polling, and found the Post’s results to be generally in line with the national trends – except in Texas.

“The most controversial result is the Texas result,” he said. “It’s an outlier.”

Henson noted, however, that a number of recent private polls he has seen have given Trump only single-digit leads in Texas. Four years ago, President Obama lost the state by 16 points to Mitt Romney; four years before that, he lost to John McCain by some 10 points.

“When you haven’t seen a double-digit poll in this matchup in a long time, it’s hard not to get some sense that there is some kind of movement, and consider what the narrative of that movement is,” he said.

Craig Goodman, who teaches political science at the University of Houston in Victoria, said that although the Washington Post-SurveyMonkey poll has to be “taken with a grain of salt,” it fits into the broader context of a tightening race. “Certainly, Donald Trump hasn’t been running away with Texas,” he said. “It certainly fits in with the broader narrative of where Trump has been struggling.”

Anyway. Juanita passes along a rumor that Clinton has a double-digit lead in Harris County, which if true (I’m not ready to believe it) likely suggests a high concentration of not-Trump Republicans here. Democratic Sheriff candidate Ed Gonzalez sent out a fundraising email yesterday afternoon that teased a poll saying “Sheriff Ron Hickman is vulnerable and we are well-positioned to win” and “the race is a dead heat and nearly a third of the vote is undecided”, which neither confirms nor denies Juanita’s rumor. Who knows? I’ve added the result to the right hand sidebar. We’ll see what the outfits that normally poll Texas have to say soon enough.

Related Posts:

• UH Hobby School (Harris County only): Clinton 43, Trump 36
• Two unsatisfying articles about the 2016 Democratic sweep in Harris County
• Dems sweep Harris County