Why a meaningful shift in the Trump-Clinton race might be at hand

By Nate Cohn, New York Times News Service

Saturday, Aug. 6, 2016 | 2 a.m.

Hillary Clinton has a large and perhaps growing lead in the nation and in many of the predominantly white battleground states where Donald Trump was thought to have his best shot, according to a wave of new surveys released in the past two days.

Three national surveys — from Fox, NBC/WSJ and Marist/McClatchy — showed Clinton ahead by big margins: 10, 9 and 15 percentage points. It’s the worst polling stretch for a presidential candidate at this stage since John McCain in mid-October 2008.

It’s a little hard to make sense of it all, in part because the timing is a little unusual. This ought to be the heart of Clinton’s post-convention bounce, when polling analysts generally preach caution. On the other hand, Trump had a tough week of his own making — drawing condemnation from Republican leaders and even causing a few high-profile defections from donors and the conservative media.

The prudent approach is to wait, and see whether Clinton’s lead endures for another week or two — after convention bounces usually fade. In the interim, we can cautiously say that there is more reason than usual to think Clinton’s newfound lead represents a meaningful shift in the race, one that would make a comeback for Trump seem daunting if it holds.

In general, it’s not worth overthinking post-convention bounces. They don’t necessarily even reflect real shifts in voter attitudes, just changes in how likely people are to respond to surveys. But strong conventions can speed up unification of divided parties and usher in a real shift.

Part of the reason Clinton’s bounce seems more likely to last is that it seems to be coming from greater party unity: growing support from Bernie Sanders’ backers. The Marist poll showed Clinton with the support of 90 percent of his voters, while a CNN poll gave her 91 percent. These numbers had usually been in the 60s or 70s.

Clinton’s support appears to be underpinned by her improving image among Democratic-leaning voters, according to Gallup data. She could still lose some of those voters to third-party candidates, but Trump faces similar risks.

These gains have allowed Clinton to move up to around 47 or 48 percent of the vote in recent polls. This is not an especially impressive tally, but for that same reason it seems reasonable to expect she could keep it: She’s not winning the support of a huge number of voters whom you would expect to abandon her.

At the same time, the unusually large number of voters considering a third-party candidate make Clinton’s tallies seem likelier to be enough to win.

The polling numbers for Gary Johnson or Jill Stein will probably decline — especially once likely-voter screens are added. Clinton’s tally would probably improve in the process; some of those voters would drift to her, and her voters would become a larger share of likely voters.

Trump, on the other hand, faces the opposite issue. His support has plummeted: In fact, he didn’t even breach 40 percent in any of the national surveys that were released Wednesday or Thursday.

His huge edge among white working-class voters has all but vanished, at least temporarily. In the last four national surveys, he has held just 51.5 percent of white voters without a college degree — worse than the 57 percent he held in June and July polls conducted before the Democratic convention, and even worse than the 55 percent of the vote held by Mitt Romney in pre-election polls in 2012.

The effect has been obvious in many states with large numbers of white voters without a degree, like New Hampshire and Pennsylvania. A MassINC/WBUR poll showed Clinton up by 15 points in New Hampshire — the third-whitest state in the country. She led by 11 points among likely voters in a Franklin and Marshall poll in Pennsylvania.

She has not secured these white working-class voters for herself. She’s averaging 34 percent of white voters without a degree over the last four surveys — less than the 37 percent President Barack Obama held in 2012. Or put differently: Trump, even at his nadir, appears to be winning white working-class voters by about the same margin as Romney did in pre-election polls.

Realistically, Trump will win many of these voters back under more favorable circumstances. But that doesn’t mean that his decline is a nonissue. There hasn’t been a presidential candidate with so many poll numbers in the 30s at this stage of the race in a long time.

And while it may just be a coincidence, his numbers have gotten worse with every survey this week — even as the convention fades in the rearview mirror. It at least raises the possibility that he is being damaged as much by himself as the Democratic convention.

The concern for Trump isn’t that he can’t recover to 42 or 43 percent of the vote. It’s that the same stories making fairly reliable supporters reconsider their allegiances, at least momentarily, are solidifying Clinton’s hold on the 47 or 48 percent or so who support her in the polls.

Perhaps even worse, he could be hardening the resistance of the remaining undecided voters who don’t have a favorable impression of either candidate but who do not believe that he is qualified to be president or that he has the temperament to do the job. Altogether, he appears to be in danger of becoming unacceptable to a clear majority of voters. Indeed, just 39 percent of voters said he was “acceptable” in the Marist poll. No, this is not a normal result of a convention bounce.

Polls conducted a few weeks after the conventions have proved to be generally accurate. They’re not perfect, but no modern presidential candidate who has trailed a few weeks after the conventions has gone on to win the popular vote. Trump is in serious danger of needing to be the first candidate to pull it off — with more baggage weighing him down, for good measure.