Stop Obsessing about Kamala Harris’ (Polling) Bumps
I’m going to defend Jonathan Karl, who described ABC’s poll showing a six point lead over Trump as within the margin of error. Here is Dan Drezner’s complaint about Karl’s comment, which is similar to that of many other people.
Take, for example, Sunday’s ABC News/Ipsos poll of the national race. It showed Harris leading 52%-46% among likely voters, a six-point lead that was outside the margin of error. Given the closeness of the race, a national poll showing Harris ahead seems newsworthy.
That, however, is not how ABC’s Jonathan Karl chose to frame it:
Karl says that Harris’ lead is “just barely outside the margin of error,” which is just a weird way to describe one of the few polls where someone has a statistically significant lead. Karl could have simply pivoted from the poll result to talk about how it’s still very close in the Electoral College — but he didn’t. Instead, he described a poll in which Harris had a significant lead as a toss-up.
It’s absolutely right that this poll is outside the margin of error (it is unchanged since before the convention). But Karl is right that the race is much closer in swing states, where voters have been flooded with Trump attacks on Harris for weeks now.
I think Democrats are telling themselves a wildly overoptimistic story about this race. I’m grateful Kamala and her campaign manager keep warning that she remains the underdog in this race.
That’s because this race is unlike any normal race. That’s true not just because Harris is a mixed race woman, though both her gender and race should raise concerns that the polls are overestimating her support (we literally hear stories about Republican women wondering if their spouses will learn for whom they voted). But it’s true because Kamala is not yet halfway through her race, and she’s running against a former President over 90% of the way done.
Pundits are measuring this odd campaign rhythm according to normal rules, such as that conventions bring a bounce (neither did this year) or that Labor Day marks some line in a sand about the final stage of the race.
As one example, both Frank Luntz…
And Nate Silver…
Pointed to this single Michigan poll of 600 possible voters to defend their argument that the Vice President has not gotten a bump from the convention. Neither of these men — a Republican partisan and a guy whose gambling habit may be influencing his analysis — are reliable sources.
And this is a particularly bad poll on which to base such judgments. Polling in MI has been pretty shitty going back two decades (though it is true that Trump has underperformed in many of them). It took WDIV/Detroit News five days to release this poll as compared to one day for their July poll. It was all done post-RFK endorsement of Trump (and as such could reflect RFK’s Trump-leaning vote moving to the former president), but before his bid to be removed from the ballot failed. Because of Michigan’s significant Arab American population, it is the swing state most likely to be influenced by Biden’s failures on Israeli policy. The August poll has a Likely Voter category (the one they report) and a Definite Voter category, the latter of which Kamala leads by 1.6%.
And as my former blogmate Dana Houle (who has run statewide campaigns in MI) noted, this poll delays release of crosstabs, and when they released theirs in January, they showed wildly unlikely results (results equally inconsistent with July’s poll).
More importantly, both the WDIV poll and the ABC one show two things that many polls are reflecting: First, while overall support for the candidates may look the same, the nature of their support is changing, with a gender split growing for each.
More curiously, that’s happening even as Kamala Harris’ favorability is going up. Even Joe Biden is on course to tip over into favorable ratings by election day!
That’s also happening as the electorate, at least in the short term, is becoming more female, more diverse.
What’s going on with the race is that Trump has a ceiling of support. While more men may be saying they’ll vote for Trump, Trump is not getting more popular.
And so he needs to do something to increase Kamala’s negatives (the success of negative ads may explain the narrower polling in swing states, but Trump’s future ad payments may indicate he’s blowing the money it would take to keep that up).
And therein, I think, was the intent of the Arlington Cemetery stunt — where Trump’s people, invited in by a few people who lost family members in the Afghan withdraw — took video from the gravesides of people whose family did not give consent, and did so after physically shoving a cemetery staffer.
This is the Benghazi playbook. Trump’s attempt to politicize an Afghan withdrawal that he played an instrumental role, according to his own former National Security Adviser, in making chaotic. This is, as everything with Trump is, a planned stunt coordinated with the House GOP.
House Speaker Mike Johnson (R-La.) announced that Congress will honor the 13 American service members killed in the attack by presenting their families with the Congressional Gold Medal on Sept. 10.
“Congress has a duty to ensure these sacrifices are never forgotten, and it is my distinct honor to announce that Congress will bestow the families of these 13 heroes with the Congressional Gold Medal — the highest award Congress can present to any individual or group,” Johnson said in a statement released last week.
The ceremony, and remarks by a bevy of Republican lawmakers, will take place at the U.S. Capitol Rotunda the same day as the presidential debate between Harris and Trump.
Like Trump’s planned attack on Biden incorporating documents altered by DOJ, and like his hosting of Tony Bobulinski in October 2020, Trump is hoping he can use the debate to stage a Reality TV event that gives right wingers a hook for the remainder of the campaign.
The reason why right wingers still complain that 51 former spooks said, truthfully, that the Hunter Biden laptop looked like an Russian information operation is that it undercut Trump’s Reality TV show; Trump even tried to use that as his stunt for the debate with Biden.
Here, though, Trump doesn’t have the two to four years on which both the Hunter Biden laptop and the Benghazi attacks built. Plus, the Arlington stunt has begun to backfire, most notably with John McCain’s son publicly endorsing Kamala in response. If Jamie Raskin succeeds in getting answers from DOD about what happened before the debate, it risks upending Trump’s hoped-for attack by demonstrating the contempt in which he holds service members. This risks turning into yet another story on how Trump believes service members are suckers and losers.
There’s one more thing that remains unsteady in this race: The great disparity in most polls between statewide and presidential polling (one exception out today, CNN’s, shows at least two state races — the Senate races in AZ and PA — that are not remotely credible). That may reflect misses in the modeling of the race more generally.
Kamala Harris has not gotten the polling bumps where pundits are trained to look for them.
But even as they watch for those signs closely, they’re not contemplating how other nearly unprecedented movement might shape the race.
Update: One more point about the weird timing of this race. USAT has a poll (which finished fieldwork on August 28) showing that Kamala has significantly narrowed the margins on the two topics Trump wanted to run on: the economy and immigration.
Harris also has made inroads on which candidate would do a better job handling important issues.
- On the economy, voters’ top concern, Trump was favored over Harris by 6 percentage points, 51%-45%. That’s an asset, to be sure, but it is less than half the 14-point advantage he held over Biden in June.
- On immigration, an issue that energizes Republican voters, Trump was favored by 3 points, 50%-47%, down from the 13-point preference he had over Biden.
She has narrowed that gap, even while she’s still rolling out policy proposals, such as new tax credits to support small business formation.
Harris’ proposal, released on Tuesday, calls for significantly expanding the tax deduction for start-up expenses from $5,000 to $50,000, while also setting the goal of 25 million new small-business applications during her first term, according to a Harris campaign official granted anonymity to describe details of the plan. The plan also proposes reducing barriers to getting occupational licenses and developing a standard tax deduction for small businesses.
There’s a famous line Andy Card used when discussing the Iraq War in 2002: “From a marketing point of view, you don’t introduce new products in August.”
Whether by necessity or design, Kamala can offer news events like this for the next several weeks. And this one sets up a solid contrast before the debate, in which Trump will be forced to defend tax cuts for billionaires over support for small businesses.