The Needle Is Moving

A week ago, I noted that the Real Clear Politics popular-vote polling average had Kamala with a 1.8% lead. Just this past Friday, I noted that her lead was down to 1.3%. As new polls were released over the weekend, it continued to fall, until by this morning, it's down to 1.0%.

On one hand, this is a lagging indicator, since the average includes old polls that haven't been updated. On the other, it includes polls that have proved highly inaccurate and biased toward Democrats in the past. But on top of that, a "national popular vote" average is Democrat-leaning, since it totals votes in high-population, Democrat-leaning states like New York and California, when the Electoral College is meant to limit big-state advantage in elections.

So far, I've seen basically no cogent commentary on what this decline in Kamala's popular vote averages means. However, it comes in the wake of the vice presidential debate Octobrr 1, her Call Her Daddy interview October 6, her 60 Miniutes interview October 7, ansd her Fox News interview with Bret Baier October 16, but before Trump's visit to a Bucks County, PA McDonald's yesterday. The size of the crowds out to see Trump makes me wonder if the polls are missing something important:

Just LOOK at this crowd outside the Trump McDonald’s in PA

Holypic.twitter.com/IuQgpqXaOb

— DC_Draino (@DC_Draino) October 20, 2024

Because the RCP average is a lagging indicator, it doesn't fully reflect the trend in the most recent polls:

A new Fox News poll, conducted between October 11 and 14 among 1,110 registered voters and 870 likely voters, showed that Trump is leading Harris by 2 points among registered and likely voters, on 50 percent to her 48 percent, which falls within the poll's margin of error, a 4-point swing from when Harris was leading Trump by 2 points a month ago.

Additionally. the latest ActiVote poll, conducted from October 3-8, showed Donald Trump holding a 1.2-point lead nationwide with a 3 percent margin of error. That came after a September poll from ActiVote had Kamala Harris ahead by 5.4 points.

Trump has also seen positive signs in the swing states. RealClearPolitics' poll tracker last week showed that Michigan had flipped in favor of Trump for the first time since July 29. Nevada and Pennsylvania have also gone Republican.

But neither of these polls was done in the wake of the Fox News Bret Baier interview. Nevertheless, The Hill reported today,

Former President Trump overtook Vice President Harris in the Decision Desk HQ/The Hill election forecast Sunday, the first time he has been deemed the favorite over her this cycle.

The model predicts Trump has a 52 percent chance of winning the presidency, while Harris has a 48 percent chance, as of Sunday.

Since late August, the election forecast put Harris’s chances of winning between approximately 54 percent and 56 percent. In early October, however, those dynamics began to shift, and the election forecast predicted both candidates’ chances to be closer to 50 percent.

On Oct. 17, the model predicted the two candidates were equally likely to win next month, and Trump took the advantage Oct. 20.

This represents a shift of as much as 8%, but the piece nevertheless concludes,

[T]he race remains a toss-up, according to the election forecast, since the polling in all seven states remains within the margin of error, meaning typical polling inaccuracies could shift the results in either direction.

As of this morning, most respectable opinion is still withholding judgment on whether Kamala's October shift in strategy to do interviews has been effective. For instance, at The Guardian:

Once interview-shy VP plunged into podcast populism and interview with shock-jock Howard Stern after polls showed her slipping against Trump

First came “joy”, with some cosy but unmemorable TV sit-downs with sympathetic hosts, a Vogue cover, and a billion dollars to spend on TV ads. Then reality hit: Kamala Harris’ strategy to win over a mysterious sliver of undecided US voters was not working and she was slipping back in the polls.

So the vice president went on Fox News, part of a pitch to white working-class women, who voted for Trump more strongly in 2020 than 2016. It was a win for Fox’s news division – 7.8m viewers, or four times host Bret Baier’s nightly average – but was it a win for the Harris campaign?

Well, was it? All we get is a shrug:

But will any of it matter? There are no guarantees that undecided or unengaged voters today will be any less undecided in two-and-a-half weeks’ time.

It says a lot that the pundits are willing to say so little, when it's indisputable that the numbers are actually shifting. And especially in the past week or ten days, something's clearly changed.