Better polling for Clinton, Trump still in striking distance

New York Times News Service

Donald Trump had a difficult week after the first presidential debate, and, as you might expect, his polling was not great.

The national polls, and most polls in battleground states, released since the debate show Hillary Clinton with a comfortable advantage, perhaps by four or five percentage points nationwide among likely voters in a four-way race. She has led by even more in the head-to-head matchup.

The polls suggest that Clinton’s performance has energized Democratic-leaning voters, helping to reduce the big gap between registered voters and likely voters that plagued her in many September polls.

But Clinton’s big weakness among white working-class voters is still enough to keep Trump fairly close — perhaps closer than he’s been for much of the year.

Voters: On average, Clinton has fared about two or three points better in post-debate polls than in similar surveys conducted in September. Yet she isn’t doing much better than she did in August or July — an indicator that Trump has made at least some meaningful gains over the last few months, if one assumes that the post-debate polls represent a relatively good moment for Clinton.

The three high-quality, live-interview national surveys that have been released since the debate — from Fox News, CNN/ORC and CBS News — tell a fairly consistent story about Clinton’s renewed strength.

Before the debate, all three polls showed Clinton ahead among registered voters. But, on average, they showed Trump ahead among likely voters (pollsters use models to determine which registered voters are likely to vote on Election Day).

The post-debate surveys did not show Clinton doing much better among registered voters in the four-way race. But Clinton took a comfortable lead of three to five points among likely voters, as the gap between registered and likely voters vanished.

Swing: It’s a big swing, but not an uncommon one. The period after high-profile media events can often lead to big shifts in enthusiasm, causing one side’s supporters to be disproportionately considered likely voters.

But there’s no guarantee that the shift holds. Sometimes it’s temporary, like a convention bounce. Other times, it’s a longer-lasting shift in voter interest, which really does pick up ahead of an election.

It’s worth noting that the CNN, Fox News and CBS News polls were among the best surveys for Trump before the debate, in no small part because they showed Clinton facing the largest penalty among likely voters. Other surveys — like those from ABC/Washington Post and NBC/WSJ before the debate — actually showed Clinton faring better among likely voters than registered voters.

If most of Clinton’s post-debate gains are driven by higher enthusiasm, she might not make as many gains in those surveys as she did in the CBS, Fox and CNN polls. Even so, Clinton would probably lead by four or five points in an average of high-quality, live-interview national surveys.

Battleground states: That surge in Democratic enthusiasm has translated into gains in the battleground states. The bounce is most obvious in Sun Belt states like Florida, North Carolina, Colorado and even Nevada, where Democrats are most dependent on nonwhite and irregular voters to win elections. In general, the post-debate polls in these states are among Clinton’s best polls of the year.

They show that Florida, in particular, might be a big problem for the Trump campaign. Clinton has led in all four post-debate surveys of Florida. The two most highly regarded surveys of the bunch — from Quinnipiac and Mason-Dixon — showed Clinton with four- and five-point leads. There is nearly no path to the presidency for Trump without Florida.

Clinton also led by one to three points in three polls of North Carolina, a state that Mitt Romney won in 2012. Two polls showed Clinton ahead by 11 points in Colorado.

But Clinton did not post such impressive margins in the North, where Democrats have traditionally depended on winning considerable support among white working-class voters. Quinnipiac showed Clinton still trailing by five points in Ohio, where she led in most polls until Labor Day. The polls in Pennsylvania, Michigan and New Hampshire were better for her — with leads of four to nine points — but they are not among her best surveys of the year.

Support: Clinton’s debate win may have energized her supporters, but she does not appear to have made much progress shoring up her weakness among white voters without a college degree.

Donald Trump meets with military veterans at the Retired American Warriors Town Hall  in Herndon, Va., Oct. 3, 2016. National polls, and most polls in battleground states, released since the presidential debate on Sept. 26 show Hillary Clinton with a comfortable advantage, perhaps by four or five percentage points nationwide among likely voters in a four-way race. (Stephen Crowley/The New York Times)

Trump led by an average of 56-28 among white voters without a degree in the three national surveys, nearly the same as the 58-29 margin he held in pre-debate surveys in September. Both were better for Trump than Romney’s 58-35 lead with that group in pre-election polls in 2012. They’re also better for Trump than many surveys after the Democratic convention.

Trump’s strength among white working-class voters isn’t enough to put him in the lead — not nationally, not in the Sun Belt and not even in the relatively white states of the North. But it does keep him close.

The danger for Trump is that there’s nothing he can do to fix his problem in Florida or North Carolina. If the Democratic turnout is strong, as post-debate polls imply, and Clinton makes gains among well-educated white voters, as polls have shown all year, there just isn’t much room for him to fight back with additional gains among white working-class voters.

But if he can dodge Clinton’s shot at a knockout blow in North Carolina or Florida, his strength among white working-class voters gives him a chance to defeat her.

To do it, he would need Ohio, Iowa, Maine’s 2nd Congressional District, where Trump leads — and then just nine more electoral votes to reach 269, the number needed to throw the race to the Republican-held House. He could do it by winning both Nevada and New Hampshire, or by winning Pennsylvania, or a tougher state like Wisconsin or Michigan.

It wouldn’t be easy. Clinton has a clear polling edge in the states he needs. But it’s still possible. The Upshot’s model still gives Trump a 21 percent chance of becoming president.