2022 U.S. Congressional (and Other) Elections

[N.E. Brigand]

Looking ahead to 2024: Florida's governor, Ron DeSantis, is closing in on or even edging out Donald Trump in some recent Republican polling about who that party's nominee should be two years from now.

DeSantis won the governorship in 2018, defeating Tallahassee mayor Andrew Gillum in a very close race (49.6% - 49.2%). I'm sorry to note that earlier this week, Gillum was indicted on 21 federal felony counts stemming from his allegedly "diverting money raised during the campaign to a company controlled by one of his top advisors."

[N.E. Brigand]

A new NPR-Marist poll finds that Democrats hold a 48%-41% lead nationally on House races this fall.

Even if the poll is correct, it doesn't mean Democrats are likely to hold the House, because the national figure is offset by Republican structural advantages in districting (rather like Al Gore and Hillary Clinton winning the most votes nationally but still losing in the Electoral College).

Still, this is potentially good news following on the CBS poll that says the Roe decision has energized Democrats much more than Republicans. Like that poll, this poll finds Americans don't approve of the Court's decision. In this case, the margin was 56% opposed and 40% in favor of overturning Roe. The breakdown is 53%-41% among Independent voters.

[Voronwë the Faithful]

Yeah, I think it will take a lead of more like 8-9 points in the generic congressional poll to hold the House, and I don't see that happening. Still, an encouraging sign. There are a few wild cards (Dobbs the 1/6 hearings and the generally unpredictability of Trump and the candidates he has spawned, issues related to gun control and mass shootings, etc., gas prices and inflation in general) that make this election pretty variable.

[N.E. Brigand]

On the other hand, this misleadingly-titled AP article indicates that:

The AP examined nearly 1.7 million voters who had likely switched affiliations across 42 states for which there is data over the last 12 months, according to L2, a political data firm. L2 uses a combination of state voter records and statistical modeling to determine party affiliation, meaning that the switchers include both those who have formally changed their registration and those who L2 estimates have shifted toward the GOP.

While party switching is not uncommon, the data shows a definite reversal from the period while Trump was in office, when Democrats enjoyed a slight edge in the number of party switchers nationwide.

But over the last year, roughly two-thirds of the 1.7 million voters who changed their party affiliation shifted to the Republican Party. In all, more than 1 million people became Republicans compared to about 630,000 who became Democrats.

That's a net gain of about 400,000 voters for Republicans. The article focuses more on people who switched from Democratic to Republican than the reverse, and of the few individuals the reporters contacted are some who who actually voted Republican in 2016 but hadn't changed affiliations before this year.

One Colorado voter described in the piece says he switched from Democrat to Libertarian six years ago. He switched recently to Republican because he became "increasingly concerned about the Democrats’ support in some localities for mandatory COVID-19 vaccines, the party’s inability to quell violent crime, and its frequent focus on racial justice."

Edited to add: This psephologist from the University of Florida pokes some holes in L2's methodology, but can't say whether the study overstates or understates the net Republican gain.

[N.E. Brigand]

A new NPR-Marist poll finds that Democrats hold a 48%-41% lead nationally on House races this fall.

Even if the poll is correct, it doesn't mean Democrats are likely to hold the House, because the national figure is offset by Republican structural advantages in districting (rather like Al Gore and Hillary Clinton winning the most votes nationally but still losing in the Electoral College).

Still, this is potentially good news following on the CBS poll that says the Roe decision has energized Democrats much more than Republicans. Like that poll, this poll finds Americans don't approve of the Court's decision. In this case, the margin was 56% opposed and 40% in favor of overturning Roe. The breakdown is 53%-41% among Independent voters.

Two more polls: a Morning Consult poll finds Democrats with a three-point advantage (45%-42%) over Republicans in the generic House ballot, and a Yahoo/Yougov poll find Democrats with a seven-point advantage (45%-38%). Three points is definitely not enough, and even if it's seven to eight points (which might be enough), there's lots of time for Republicans to turn things around, but I think it's fair to say that November's outcome is much less certain than it seemed a week ago.

[Cerin]

I wonder whether the Democrats switching to Republican are doing so in order to vote against Trump-supported candidates in the Republican primaries. Cheney is openly appealing by mail to Democrats in Wyoming to switch parties and vote for her in that primary. I believe it was 37,000 Democrats who voted for Raffensperger in Georgia; I don't know if that was enough to push him above 50%. The number of Democrats voting against Cawthorne in North Carolina was many more than the margin of his defeat.

I don't know how it works in all states, if you have to officially switch parties, or just request the particular party ballot, etc. I guess it would all depend on how crucial the Dem. race is considered to be in those instances, and how hard the establishment Repubs are pushing in those races. I know they went all out in Georgia to defeat Trump's picks and I would imagine Cheney is getting unlimited support in her race.

[River]

I wonder whether the Democrats switching to Republican are doing so in order to vote against Trump-supported candidates in the Republican primaries. Cheney is openly appealing by mail to Democrats in Wyoming to switch parties and vote for her in that primary. I believe it was 37,000 Democrats who voted for Raffensperger in Georgia; I don't know if that was enough to push him above 50%. The number of Democrats voting against Cawthorne in North Carolina was many more than the margin of his defeat.

In Colorado there was some party-switching talked about, especially to remove Boebert. But, in the CO primaries, you don't need to completely switch parties. Unaffiliateds are allowed to select one ballot and vote. We voted that in when we voted in having Presidential primaries because unaffiliateds represent at least a third of the state electorate. It's probably better for the parties themselves to allow these people a voice early on because you absolutely cannot win a statewide race in Colorado without the support of unaffiliated voters. So if a party wants to win those races, said party needs to allow the unaffiliateds a chance to vet their first round of candidates. Anyway, a large fraction of the unaffiliateds returned GOP ballots this time around. This kept statewide races are free of Big Liars and we'll be heading into November talking about 2022 instead of 2020. Boebert won her primary so the gambit didn't necessarily work at the Congressional district level.

[Frelga]

I'd like the sauce for the 37K figure, please?

Until recently, Democrats had an open primary in California - unaffiliated voters could request a ballot for Democratic party. GOP ballot was only available to registered GOP voters.

Now we have a free for all primary, where everyone picks a candidate regardless of the party, and the top two go on the general ballot, also regardless of the party.

[Voronwë the Faithful]

I'd like the sauce for the 37K figure, please?

Here's an AP article that cites that number (you can also find it at NewsMax and a number of other right-wing sites).

Some Democrats voting in GOP primaries to block Trump picks

An Associated Press analysis of early voting records from data firm L2 found that more than 37,000 people who voted in Georgia’s Democratic primary two years ago cast ballots in last week’s Republican primary, an unusually high number of so-called crossover voters. Even taking into account the limited sample of early votes, the data reveal that crossover voters were consequential in defeating Trump’s hand-picked candidates for secretary of state and, to a lesser extent, governor.

[Alatar]

Is Cerin going to be expected top provide citations for everything she says? I don't see any other posters here being held to that standard, and I know I haven't been.

If we want to hear opinions other than our own, shouting "prove it!" every time someone says something you disagree with is a really bad way of going about it. Not aimed specifically at you Frelga but a noticeable increase in these requests for citation seems to be incredibly coincidental with the arrival of a Republican in the discussion.

[Voronwë the Faithful]

I could be wrong, and would be happy to be corrected if I am, but I'm not sure that Cerin would describe herself as a Republican. That having been said, while I value having facts backed up by specific citations (and try to provide them wherever possible), I think your point has merit.

[Frelga]

Oh no, I am being singled out, it's an outrage.

If something is stated as a fact, I want to know where it came from, yeah. No one is under any obligation to provide any information just because I am asking for, just as I am free to dismiss unsubstantiated statements as potentially inaccurate.

Especially since I frequently fact check information my parents get from right wing sources and 80% of it is either plain not true or creatively interpreted.

[Sunsilver]

What Frelga said...

I hang out with a group of very conservative right wingers on another board, and I make sure I fact check nearly everything I post there, because I know my viewpoint is in the minority and is likely to be challenged.

I'm not as careful when I post here, but I still frequently fact check stuff, as I've goofed up a few times. And that can be pretty embarrassing!

And I'm going to add it's human nature to want proof when someone tells you things that don't jibe with what YOU have been told is true.

[N.E. Brigand]

Is Cerin going to be expected top provide citations for everything she says? I don't see any other posters here being held to that standard, and I know I haven't been.

If we want to hear opinions other than our own, shouting "prove it!" every time someone says something you disagree with is a really bad way of going about it. Not aimed specifically at you Frelga but a noticeable increase in these requests for citation seems to be incredibly coincidental with the arrival of a Republican in the discussion.

I provide citations for about 95% of what I post here, and when I don't, I'm sometimes asked to do so anyway -- and then I try to comply with that request. (But I almost never ask others for their source, or their sauce.)

[N.E. Brigand]

According to a new report in the Atlanta Journal-Constitution, the former football player Herschel Walker, who is the Republican candidate challenging Sen. Raphael Warnock in Georgia, claimed in speeches in 2017-2019 that he had been an "FBI agent" and "worked in law enforcement, so I had a gun" and specifically that "I work with the Cobb County Police Department." He's also claimed to have earned a degree from the University of Georgia. None of this is true. His campaign says he attended a week of training with the FBI and that his reference to Cobb County meant only that he had been given an honorary position by that department (but the AJC thus far has been unable to get confirmation of those claims).

When I posted to encourage skepticism earlier this evening, I was not referring to this story in The Daily Beast, a legitimate news source, which I only encountered after that: Scoop: Herschel Walker Has a Secret 10-Year-Old Son, Who He Doesn't See. According to the story, Walker is paying child support and does send gifts to his son. But he's still a huge hypocrite, given that he has repeatedly criticized what turns out to be his own behavior in public statements in recent years, for example saying that if a man has "a child with a woman, even if you have to leave that woman . . . you don't leave the child" and "if you got a child, hug your child every day," and much more along those lines.

Oh my. Herschel Walker, Critic of Absentee Dads, Has Yet ANOTHER Secret Son—and a Daughter (The Daily Beast) Two days ago, when The Daily Beast reported on Walker's first-known secret child, his campaign responded:

"Herschel had a child years ago when he wasn't married. He's supported the child and continues to do so. He's proud of his children. To suggest Herschel is 'hiding' the child because he hasn't used him in his political campaign is offensive and absurd."

Prior to that, Walker was only known to have one child, 22-year-old Christian Walker, who just last week mocked a celebrity for father multiple children by multiple mothers: "Nick Cannon is everything wrong with our culture. Knocking women up and then being too busy banging other women to take care of his children. Fathers, control your ding-a-ling, get home, and raise your damn kids."

I'm not sure if it's because of the news of Herschel Walker's three previously unacknowledged children, and it's just one poll, but Quinnipiac finds him trailing Sen. Raphael Warnock by a 54%-44% margin, and a narrow plurality of those voters says that Walker is not honest. It's probably not a poll whose respondents are lopsided in favor of Democrats, although it may skew slightly in that direction, given that in Georgia's governor race, Brian Kemp and Stacey Abrams are running about even -- and Kemp beat her by 1.4% four years ago.

[Cerin]

That 37,000 was apparently from early voting only, according to this Vice article. The actual estimated number of Democratic cross-overs was 77,000.

A Washington Post analysis conducted in the immediate wake of the election estimated that 77,000 Democrats voted in the GOP primary last Tuesday. They may not have all voted for Raffensperger, but subtracting them from his total would have left him with 49 percent of the vote.

https://www.vice.com/en/article/wxdwax/ ... rats-voted

When I was a liberal, I didn't identify as a Democrat, and now as a conservative, I don't really think of myself as a Republican, because of a dislike for the political establishment. In a two-party system, though, it's not hard to peg which party my views are more aligned with.

[Frelga]

I want to acknowledge that Cerin has been a good sport about being asked for sources, so I'm not sure why this even came up.

Also, I gave up on the math in that article, and why there are simultaneously 37K and 77K crossover votes (and I treasure my ignorance). Regardless, I now have a greater appreciation of California's top-two primary system. I don't like how it condemns 3rd party candidates to languishing in obscurity, and we could some day see two candidates from the same party in the general election, but at least it does not lead to the appearance of shenanigans.

[N.E. Brigand]

On the other hand, this misleadingly-titled AP article indicates that:

The AP examined nearly 1.7 million voters who had likely switched affiliations across 42 states for which there is data over the last 12 months, according to L2, a political data firm. L2 uses a combination of state voter records and statistical modeling to determine party affiliation, meaning that the switchers include both those who have formally changed their registration and those who L2 estimates have shifted toward the GOP.

While party switching is not uncommon, the data shows a definite reversal from the period while Trump was in office, when Democrats enjoyed a slight edge in the number of party switchers nationwide.

But over the last year, roughly two-thirds of the 1.7 million voters who changed their party affiliation shifted to the Republican Party. In all, more than 1 million people became Republicans compared to about 630,000 who became Democrats.

That's a net gain of about 400,000 voters for Republicans. The article focuses more on people who switched from Democratic to Republican than the reverse, and of the few individuals the reporters contacted are some who who actually voted Republican in 2016 but hadn't changed affiliations before this year.

One Colorado voter described in the piece says he switched from Democrat to Libertarian six years ago. He switched recently to Republican because he became "increasingly concerned about the Democrats’ support in some localities for mandatory COVID-19 vaccines, the party’s inability to quell violent crime, and its frequent focus on racial justice."

Edited to add: This psephologist from the University of Florida pokes some holes in L2's methodology, but can't say whether the study overstates or understates the net Republican gain.

More elections analysts have looked at this study and found it wanting. Here's G. Elliot Morris discussing the problem:

This means it’s possible that instead of the AP counting just people who have actually changed their party registration, they could also be counting:

• Voters who were modeled by L2 to be Democrats or Other but registered as Republicans; and

• Voter who registered as Democrat or Other but who get modeled by L2 as leaning Republican, perhaps because of either:

• • Polling data that indicated a shift away from Democrats among certain demographics or in particular geographies; or

• • Election results that showed shifts from Democrats in certain jurisdictions

[...]

A closer look at the data reveals the pattern is more noise than signal. While voters may be more likely to vote for Republicans now compared to 2020, they are not registering as Republicans in greater numbers. They certainly haven’t surged in the millions.

[N.E. Brigand]

Looking back at the final polling averages from 2020, I find that they showed Joe Biden beating Donald Trump by 51.4% to 43.5% with 5.1% undecided.

The actual numbers in the election were 51.3% to 46.9%.

In other words, it looks like Biden voters were accurately represented by the national polls, but Trump voters were understated by 3.4%.

That might be a statistic worth bearing in mind as you see polls for the upcoming midterm elections.

(And as always, it's better to look at the averages for a given race than at any one poll for that contest.)

[Voronwë the Faithful]

The 538 Midterm forecast is here. The bottom line is that they forecast an 87% chance of the GOP taking the House, and a 54% chance of the GOP taking the Senate. So, with regard to the possibility of the Democrats retaining control of Congress, it's not quite in "so your saying there's a chance" territory, but it's not that far off.

https://projects.fivethirtyeight.com/20 ... -forecast/

ETA: It is interesting to me to see that they give Walker a 55% chance of winning in Georgia, despite the recent poll that N.E.B. linked to showing him behind by 10 points.