The country is a month out from an extremely consequential midterm election for which pollsters and analysts have no historic comparison.
The election follows an unprecedented attack on the nation's Capitol fueled by a wannabe despot who also happened to be the sitting president.
It comes on the heels of a wildly unpopular Supreme Court ruling upending 50 years of reproductive rights law—immediately rendering anyone of childbearing age a second-class citizen and stripping them of both bodily autonomy and the right to make personal health care decisions.
The Biden administration has created more jobs in two years than at any time in history and yet inflation continues to be at its highest point (about 8.3%) since 1981 (8.9%).
Whether or not voters realize it, we are at a potential turning point in history in which American freedoms are beginning to contract after a relatively steady expansion over the past century.
Campaign Action
CNN analyst Ron Brownstein has cast the central question driving the election as: Your money or your rights?
Empirically speaking, Democratic presidents have proven to be better for the economy than Republican ones by most measures. But most voters think otherwise and, given the stubborn inflationary pressure on their pocketbooks, they are likely to blame the party in power.
On the other hand, women—particularly younger women—are grappling with the realization that the course of their lives could be irrevocably changed due to government intrusion on their personal freedoms and health care decisions. Many minority groups are also clear that gutting Roe v. Wade is just the first step in a long journey in which right-wing Republicans plan to impose their puritanical views on the masses.
“In large part that’s why this election is super weird,” Bryan Bennett, lead pollster for the progressive consortium Navigator Research, told Brownstein. “People are having to make this trade-off between the immediate economic concerns [where]…they might blame the incumbent party in power. But at the same time, they know that same incumbent party is the one that is going to protect that fundamental human right” on abortion.
At base, the midterms outcome will likely hinge on: Which issue will prove most compelling, and which potential voters motivated by which issue will actually cast a ballot?
To some extent, that dilemma sums up the unique challenges pollsters and prognosticators are facing as they try to analyze the race.
All this is a long way of saying that all the polls over the next several weeks, particularly the "likely voter" polls, should be taken with a big grain of salt.
As one pollster told me, it's not actually that weird for an outfit to have a 5-point difference in a race between its "registered voter" and "likely voter" polls. That differential tends to vary by pollster, with some having a larger variation around a handful of points and others having a smaller variation of just one or two points.
For instance, a widely circulated Marquette Law School poll this week showed GOP incumbent Sen. Ron Johnson of Wisconsin leading his Democratic rival, Lt. Gov. Mandela Barnes, by 6 points among likely voters, 52% - 46%. However, among registered voters, the same poll showed the race tied at 47%.
But as data analyst and TargetSmart CEO Tom Bonier noted, last month the same pollster put Johnson up just 1 point among both likely voters (LVs) and registered voters (RVs).
"So among RVs, the race has not changed," Bonier tweeted. "The biggest difference over the month is the sample of who is considered a likely voter."
So how exactly do pollsters decide who is "likely" to vote and who isn't? The short answer to this question is that it's a lot of guesswork.
Pollsters are trying to answer two basic questions without much supporting evidence: 1) How do you filter out respondents from your survey based on whether they’re likely to vote or not?; 2) What is the population of LVs going to be demographically on Election Day?
Once pollsters decide the first question and get their pool of likely voters, then they have to figure out how they should weight them? On the second question especially, everyone is just making an educated guess about who will turn out on Election Day.
But guess what? We are living in ahistorical times with unprecedented electoral dynamics and, man, what a total shitshow.
Which brings us to another weird thing Bonier pointed out (it's true, I think he's doing inspired work this cycle, so I'm a fan).
CNN released a poll Thursday that found Democrats leading the national generic ballot among likely voters by 3 points, 50% - 47%. But in a series of competitive House districts, the survey found likely voters preferred Republicans by 5 points, 48% - 43%.
So the national generic of LVs favors Democrats by 3 points, but polling of competitive districts favors Republicans by 5. Perhaps not impossible, but also a pretty jarring disconnect.
In fact, a decent amount of recent movement in some polls is attributable to pollsters adjusting who they think is a "likely voter."
Bonier had something to say about the state of the midterms when he joined Joy Ann Reid on MSNBC Wednesday.
"The polls are asking people what [issue] they think is most important, I think that's interesting," Bonier said, "We're looking at the actions people are taking, and that's really compelling. For a while it was registering to vote—and we're seeing this huge gender gap among voters registering to vote... really around the country. Now it's voting."
Over 1 million Americans have already cast votes in the midterm general election, and Bonier noted there's been a sizable gender gap in the people requesting ballots.
In Georgia, for instance, 60% of the people requesting ballots are women. “It's a bigger gender gap than they've had in previous elections, including in 2020, when Democrats obviously did well statewide,” Bonier noted.
Based on the actions people are taking, Bonier believes abortion is still very much motivating voters despite the fact that some analysts think the issue has peaked. In fact, the more front-and-center the issue is in certain states, the more it has continued to drive female registrations, he said, mentioning Idaho, Michigan, Wisconsin, Pennsylvania, Ohio, Nevada, and Arizona.
Bonier explained that female registrations spiked immediately after the Dobbs ruling in June, then it slowed a bit—though women registrants were still outpacing previous years. But after Sen. Lindsey Graham of South Carolina unveiled his national 15-week abortion ban, registrations among women ramped up again. (Thanks, Linz.)
Bonier's observations are based on what he has seen, which corresponds with the Democratic overperformance we have observed in a handful of specials elections this year.
So as you’re looking at the final weeks of polling this cycle, regardless of whether you like what you see or don't, just keep in mind that likely voter models are much more art than science, especially this year in a political environment that’s as alien to us as Mars.
And remember this friendly reminder from Daily Kos founder Markos Moulitsas.
"We're down to 4 weeks, this is a time for everybody to start thinking about ways they can help Democrats win this election," he said this week on The Brief. "We don't win this election with everyone sitting at home, looking at polls."
So visit http://dailykos.com/gotv to get your activism on.
Help Democrats keep control of the Senate with a $3 donation.
Help Democrats keep control of the House with a $3 donation.
And definitely vote early!