Encouraging polls are hard to come by lately but he’s had a couple. Trafalgar gave him a good one a few days ago when they found him a point up in Wisconsin, matching his margin from 2016. That was an outlier, as the previous three surveys of the state had Biden ahead by eight, eight, and 11 points. But if anyone’s entitled to a benefit of the doubt on a Rust Belt poll it’s Trafalgar, which detected Trump’s impending upsets there four years ago when no other pollster did. So maybe Wisconsin’s tighter at the moment than the conventional wisdom believes.
The president got another pair of good results today. A University of Texas survey has him ahead of Biden by four points there, which is half the margin he enjoyed against Hillary. But no one expects Trump to build on his 2016 margins in the electoral college this fall; if he wins, it’ll be by the skin of his teeth. Holding Texas is the first baby step towards that goal. UT says it’s shaping up that way at the moment despite a 46/48 job approval for POTUS in the state. Among independents it’s at 36/50, although Trump enjoys a 41/27 lead over Biden among that group. They’re the difference for him right now — but 32 percent of indies remain undecided. If most of them tilt Democratic Texas would be a toss-up.
Which, actually, is how other recent polls of the state have seen the race. Fox News found Biden up a point a few weeks ago. Early last month Quinnipiac found Trump up a point. It’s in play, but POTUS probably retains a small advantage. UT’s result isn’t an outlier.
This Gravis poll of Arizona for Trump-friendly OAN *is* an outlier. A big one.
Biden led in the last four polls of Arizona before this one, twice by margins of seven points. How’d Gravis get a Trump +4 margin? It could be that they’re right and all the other pollsters are wrong, of course, but I notice that their sample contains a *lot* of less educated voters. Normally it’s the opposite: One of the reasons why pollsters missed Trump’s big wins in battleground states four years ago is that they tend to include too few people in their samples who lack college degrees, which is Trump’s base. In this case it looks like Gravis did the opposite. Four years ago, the exit polls in Arizona found more than a third of the electorate had a college degree (36 percent). In Gravis’s poll just 21 percent did, and college grads are a key Biden demographic. It’s possible that they won’t show up this fall in the same numbers as they did in 2016 but there’s no obvious reason to think so.
The “tell” that something’s off about Gravis’s data is that they also have Martha McSally leading Mark Kelly by four points, 46/42. I can believe that Trump is leading Biden narrowly in AZ, although it seems unlikely. I can’t believe that McSally is leading Kelly. Until this survey, she hadn’t led him in any survey since May of 2019. She’s trailed him in the last three polls of the state by at least nine points. There’s no reason to think there’s been a sudden double-digit swing there towards a Republican in the Senate race at a moment when Trump and the GOP are struggling nationally.
Arizona was one of six swing states polled recently by the lefty firm Change Research. Note the trend in AZ in particular:
The biggest shift towards Biden over the past two weeks was in AZ, coincidentally a state with one of the worst COVID outbreaks in the country over the same period. Among the six polled, Arizona’s Doug Ducey easily had the worst job approval of any governor. It may be that discontent about the threat from coronavirus is (temporarily?) driving voters there away from the GOP, which would make Gravis’s result even harder to believe. In fact, according to Politico, Democratic ad teams are zeroing in on ads that link Trump’s personal foibles to his handling of the pandemic as their best bet to persuade undecideds:
“You can’t chase the Trump clown car,” said Bradley Beychok, president of the progressive group American Bridge. “Him drinking water and throwing a glass is goofy and may make for a good meme, but it doesn’t matter in the scheme of things … What people care about is this outbreak.”…
“One thing we saw in polling a lot before the coronavirus outbreak is that people didn’t think he was a strong leader or a good leader, they complained about his Twitter,” said Nick Ahamed, analytics director at Priorities USA. “But they had a hard time connecting those character flaws they saw in him with their day-to-day experience.”
Trump’s response to the coronavirus pandemic and recent protests, he said, “really made concrete for people the ways in which his leadership has direct consequences on them and their loved ones … It’s easier to make ads that talk about his leadership than before the outbreak.”
One last data point. Monmouth has a new national poll out today showing the race essentially unchanged from last month, with Biden up 53/41 — a gain of a single point. One interesting question, though, had to do with whether each candidate had the “mental and physical stamina necessary to carry out the job of president,” an allusion to concerns about Biden’s mental wherewithal. More people were “very confident” in Trump’s stamina (33 percent) than in Biden’s (23 percent). But more people were at least “somewhat confident” in Biden’s stamina (29 percent) than Trump’s (12 percent). Add it all up and there’s more overall confidence in Biden’s stamina at the moment (52 percent) than the president’s (45 percent). One bad debate by the Democrat could change that for the worse, but it’s a big clue as to why the “Sleepy Joe” attacks aren’t working so far.