How To Poll A New Sort Of Election

Gender is an enormous part of the election campaign, in a way that we have never seen before. It is tied up with racism, as well. From what I can see, what is happening overlaps with gender in Hillary Clinton’s campaign and race in Barack Obama’s, but is different from both.

One way to look at this is that the candidates quite literally and obviously embody their platforms. An old white man, loaded with all the negatives that can imply about race and gender, opposes a younger woman of color. We have heard from the old man for a long time now; the younger woman is new to the campaign and to the electorate. Christian nationalism against the promise of a country that includes everyone.

How this will play in terms of votes, though, is an open question.

There are significant currents of both sexism and racism in the US, and they continue loud and clear in Donald Trump’s campaign. The Republican Party is going heavy on advocating traditional roles for women, with J.D. Vance leading the charge.

A couple of polls recently have tried to get at the role of gender in the campaign. It’s not easy. Harris is also a woman of color, and that is also a factor. Acknowledging that, and that the effects of racism and sexism are interrelated, I’m going to emphasize sexism and gender in this post.

A poll by Fairleigh Dickinson University shows Harris ahead, 50-43. The poll embedded an experiment to try to determine effects of race and gender. Respondents were primed by other questions in the poll. There’s a fuller description at the link. I recommend reading it before drawing conclusions.

When voters are made to think about the race or gender of the candidates, Harris’ lead grows substantially; when they’re not, support is essentially tied. Harris is also helped by strong support among the slightly less than half of men who reject traditionally masculine identities. Trump’s strongest support is among men who hold traditionally masculine identities, while women and other men strongly favor Harris.

….

“When voters are thinking about race or sex, Trump’s support just plummets,” said Cassino. “All the time, we hear strategists and pundits saying that Democratic candidates shouldn’t talk about identity, but these results show that making race and sex salient to voters is bad for Trump and boosts Harris.”

This is the opposite of what I would have expected, but the times, they are a-changin’. And it’s only one poll.

Another study shows that sexism is baked into partisanism and therefore won’t help or hurt Harris. The article, however, does not tell us how levels of sexism were determined beyond “agreeing with sexist statements.” The Fairleigh Dickinson poll measures something more subtle, something that I think is likely to be important: sexism and racism that are unacknowledged and unnoticed, even by their bearer. These attitudes are part of living in a society that contains sexist and racist structures. Again, this is only one study. The differences between these two need to be examined.

More explicitly masculinist is Nate Silver’s coming book, which describes his approach to analyzing polls through the game of poker, and this response to it, which says, no, let’s look at it through horse racing. Such approaches, without questioning the assumptions behind them, allow, even encourage, full play of unacknowledged masculinist biases.

Most polls are proceding as usual. Tom Bonier and Simon Rosenberg are noting, and trying to include in their analysis, the great influx of new voters. Bonier tends to publish via Twitter/X threads, which I no longer pick up, but Heather Cox Richardson summarizes:

Pollster Tom Bonier noted what he called “the Harris Effect”: the 13 states that have updated their voters files since July 21—when Biden said he would not accept the Democratic nomination for president—have seen “incredible surges in voter registration relative to the same time period in 2020, driven by women, voters of color, and young voters.”

The registrations of young Black women have almost tripled compared with the same period in 2020. The registrations of young Hispanic women are up by 150%. “Black women overall have almost doubled their registration numbers from 2020,” Bonier wrote. 

These changes benefit Democrats, Bonier noted. “Democratic registration has increased by over 50%, as compared to only 7% for Republicans. These new registrants are modeled as +20 pts Dem, as compared to +6 during the same week in 2020.”

To deal with this unusual campaign, polling organization 538 has changed their presidential prediction model from one based on “fundamentals” to one based on polls. Which acknowledges that the old “fundamentals” aren’t working.

The applecart has been upset. The old ways are dying. The polls probably aren’t worthless, but I think that looking at their movement makes more sense than looking at them in some absolute way. And the related prediction models should be taken with more than the usual grain of salt.