Media Decides 'Suburban Women' Suddenly Love GOP, Is Very Very Wrong
You gals weren't all that big on bodily autonomy, were you?
Oh boy! Time for more Garbage Polling, with just days to go before the election! Daily Kos data nerd Kerry Elevald, who first brought attention to all the junky GOP polling that's been distorting national poll averages, is back with a brief dissection of a truly crap poll that the Wall Street Journal trumpeted as a huge setback for Democrats.
Previously: 'Red Wave' Narrative May Be Built On Crap Polling? Color Us Shocked.
According to the Journal's polling, white suburban women have allegedly had a complete change of heart on the generic congressional ballot, going from preferring a generic Democrat by 12 points in August to preferring Republicans by 15 points in their most recent poll, allegedly because of worries about the economy and inflation. That would be a 27-point swing, a massive shift in sentiment in just a bit over two months. If true, that could be terrible news for Democrats next week.
Spoiler warning: As you may have gathered, not very bloody likely. But let's see how dire the Journal wants readers to think the situation is for Democrats now that the price of frozen peas has outstripped any concerns about reproductive freedom:
“We’re talking about a collapse, if you will, in that group on the perceptions of the economy,” said Republican pollster Tony Fabrizio who conducted the poll with Democratic pollster John Anzalone. The poll showed that 54% of white suburban women think the U.S. is already in a recession and 74% think the economy is headed in the wrong direction.
Views of the economy among the group were substantially more negative than in the Journal’s most recent prior survey. In August, 43% thought the economy had entered a recession, and 59% said the economy was headed in the wrong direction
As Elevald notes, the data are pretty darn suspicious. Out of the 1500 participants in the poll, only 150 white suburban women took part, and the margin of error for that subgroup was nearly plus or minus six points overall. But wait, there's more! The article also notes that the margin of error on some of the policy questions was as much as eight points.
I'll just add that anytime you get a graph that looks like this, maybe you should question your own data, because unless something explosive has happened, the numbers probably shouldn't move like this:
Elevald notes that the poll gives Republicans a two-point advantage overall, "which is totally within the realm of possibility." But the idea that white suburban women suddenly reversed their preferences so quickly didn't sit right. Data from the Civiqs tracking poll, which doesn't use "suburban" as a descriptor but does include "College educated," showed a 4-point lead for Democrats, "a far cry from R+15 among the Journal’s sample of 150 white suburban women."
Here, enjoy a screenshot of Elevald's graph, although it lacks the fun slider that lets you see the advantage for Dems widen a bit since the summer, as the "undecided" numbers decline.
In what could be even better news for Democrats, when the measure becomes Independent white college-educated women, the preference for a generic Democrat improves to 53 percent, over 36 percent for Republicans, a 17-point gap. We'd like to see the numbers for non-college-educated white women, though, since they probably lean more Republican. Maybe even in suburbs.
Elevald throws some logic at the matter, too, questioning whether economic concerns are all that likely to completely displace other reasons suburban women might be wary of Republicans:
Are suburban women worried about the economy? Yes, basically every voter has inflation on the brain right now. It's a question of whether the group actually believes Republicans are better for the economy and will prioritize that issue over a loss of critical freedoms such as making one's own reproductive decisions regarding family planning, pregnancy, and birth control. That tension may play differently in different states depending on whether abortion bans are in the offing if Republicans win governors’ races in places like Michigan, Pennsylvania, and Wisconsin.
But given the WSJ poll’s subsample and margin of error, a "collapse" among 150 white suburban women isn't much more meaningful than a fart in the wind
Also too, in an update to the article, Elevald notes that the latest NPR/PBS NewsHour/Marist Poll doesn't show anything like the "collapse" of support for Democrats among suburban women, as NewsHour correspondent Lisa Desjardins tweeted Wednesday. In August and early September, the poll showed suburban women's generic preference at 52 percent Democratic, 35 percent Republican. The November poll was nearly the same: 53 percent Democratic and 34 percent Republican, effectively identical within the margin of error. (The Marist results weren't differentiated by race, but if there had been a massive defection by white suburban women to Republicans, the topline numbers would have shifted, no?)
In conclusion, the Wall Street Journal poll's conclusions are weak sad poop, the end, have a nice weekend.
[WSJ / Daily Kos / Marist Poll / Lisa Desjardins on Twitter / Photo: Ted Eytan, Creative Commons License 2.0]
Yr Wonkette is funded entirely by reader donations. If you can, please give $5 or $10 a month to help fund the very best news, satire, and statistical grumbling.
New York Times Will Have Its Red Wave If It Has To Make It Up Itself!
Good polling for Democrats in swing districts somehow means the GOP will take the House.
The New York Times sure knows how to be counterintuitive — to the point of not making any damn sense at all. For instance, in a story last Thursday on polling the Times and Siena College conducted in "four archetypal swing districts," the second paragraph soberly proclaims that the polling "offers fresh evidence that Republicans are poised to retake Congress this fall as the party dominated among voters who care most about the economy."
That sure sounds like bad news for Democrats, especially in those four bellwether districts. At least it sounds like bad news until you get to the third paragraph, which notes that the actual polling results show the Democratic candidates "were still tied or ahead in all four districts — three of which were carried by Mr. Biden in 2020." What's more, we learn in the 11th paragraph that although Biden's approval rating "does not top 44 percent in any" of the four districts, the four Democrats who currently hold the seats are "running ahead of the president’s poor ratings."
Good thing for those four candidates that they aren't Joe Biden, isn't it?
We should probably start out by saying that the poll results we're talking about all come from Times/Siena College polling, so we can at least be sure the data isn't being skewed by the flood of late GOP polling that's screwing with national polling averages.
Read More: 'Red Wave' Narrative May Be Built On Crap Polling? Color Us Shocked.
So how does the polling that shows Democrats either ahead or tied in these four districts suggest there'll be a Republican takeover of the House? Well, you see, it's because maybe those leads somehow don't matter, because what if they don't hold up or voters in other districts not covered in the article go red?
But the party’s slim majority — control could flip if just five seats change hands — demands that it essentially run the table everywhere, at a moment when the economy has emerged as the driving issue in all but the country’s wealthier enclaves.
Weirdly, not one of the charts in the piece includes the top-line polling results for all four districts. We get breakdowns of some demographics, like the districts' education levels and ethnic makeup, and how they voted in the 2020 presidential election (before redistricting). But you need to dig into the text to find the actual poll numbers. We're told that Kansas's Third District is an outlier among the four, since it's the only one where "a majority of voters hold a college degree," and also where a majority of voters said they're more motivated by social issues than by economic matters in this election.
Oh, yes, and the actual polling margin? Rep. Sharice Davids, the Democrat, is pretty much stomping her Republican challenger, Amanda Adkins, 55 percent to 41 percent — a 14-point lead. The district went to Joe Biden by a far closer margin in 2020, just 51 percent to 47 percent.
By contrast, there really is a close race in Nevada's First District, where the poll has incumbent Democrat Dina Titus tied with Republican Mark Robertson with 47 percent of the vote each, in a newly redrawn district.
In New Mexico's Second District, the two candidates are also pretty much tied, well within the margin of error; the district was redrawn to give Democrats an advantage, and Republican incumbent Yvette Herrell, with 47 percent, is just a point behind Democratic challenger Gabriel Vasquez (48 percent). The polling shows that economic concerns are motivating lots of voters, and the Times dutifully includes quotes from voters who say they plan to go with Herrell. But if concerns about the economy are so overwhelming, shouldn't Vasquez be in more trouble, instead of in a statistical dead heat?
Finally, in Pennsylvania's Eighth District, near Scranton, the Times explains, the education stats and the fact that voters were the most concerned about the economy of any in the four districts "should benefit Republicans," especially considering that Donald Trump won it in 2020 and Joe Biden has the lowest approval rating of the four districts, too. So how are those factors killing the Democratic incumbent this year?
Yet Representative Matt Cartwright, a Democrat, leads his repeat Republican challenger, Jim Bognet, 50 percent to 44 percent.
Mr. Cartwright held 13 percent support among voters who said they backed Mr. Trump in the last presidential election, the most crossover support of any candidate in the polls. And he pulled off something else unusual: Across all four races, Mr. Cartwright was the Democrat who was winning the largest share of the vote both among voters focused on social issues and those focused on the economy.
Huh! Dems in Disarray, we guess?
So you've got a redrawn Nevada district where the Democratic incumbent is tied with the Republican, a redrawn New Mexico district where the Democratic challenger is ahead by a point against the Republican — although statistically, the race is too close to predict — and two districts where the Democrats are well ahead — one of whom, by the Times's metrics, ought to be in big trouble.
In a Twitter thread on the data that went into the article, coauthor Nate Cohn said that "on balance, the polls are better for Democrats than I would have guessed given our national polling," and repeated the point that Democrats can't afford to lose more than five seats and still hang on to control of the House.
The article's conclusion that the national polling still seems to show a GOP advantage drew a skeptical reply from the Other Data Nerd Named Nate, Nate Silver, who objected,
\u201c@Nate_Cohn Dude, the Ds in your polls are collectively outperforming Biden, who won the popular vote by 4.5 points! I don't know if this is a case where you don't want to throw your editors under the bus, but these are good polls for Ds and the narrative in the story doesn't match the data.\u201d— Nate Cohn (@Nate Cohn) 1666916716
Dude, the Ds in your polls are collectively outperforming Biden, who won the popular vote by 4.5 points! I don't know if this is a case where you don't want to throw your editors under the bus, but these are good polls for Ds and the narrative in the story doesn't match the data.
Cohn stammered a bit about the "overall House picture" and suggested Silver hadn't considered it, and suggested that the strong showing for Davids in Kansas was an "outlier," but couldn't quite explain why data showing the House will likely be a squeaker means a Red wave.
For that matter, in his own standalone blog post on the data, posted the same day as the story he co-wrote, Cohn said that the polling showed "some decent results for Democrats" and that the Kansas and Pennsylvania polling "count as great results for Democrats." In fact, Cohn said, he'd "expected a bit better for Republicans after our last national poll showed them up by three percentage points on the generic ballot." He also cautioned, as did the main story, that four districts aren't enough to make a judgment about the overall results in the House elections. Notably missing from the blog post, though, is any mention of that line about "fresh evidence that Republicans are poised to retake Congress" — instead, Cohn says the four polls "can’t really change my view of the race for the House overall," a far more cautious prediction, if it's a prediction at all.
Which leads us to our eternal question: New York Times politics editors, go fuck yourself.
[NYT / Nate Cohn on Twitter / NYT / Jimmy Baikovicius, Creative Commons License 2.0]
Yr Wonkette is funded entirely by reader donations. If you can, please give $5 or $10 a month and we'll read the New York Times for you.
'Red Wave' Narrative May Be Built On Crap Polling? Color Us Shocked.
Dems may be on to something if Fox News is calling this a 'conspiracy theory.'
in the final weeks of the 2022 midterm campaign, national polling averages appear to show a number of close races for the House and Senate tilting toward likely Republican wins. Very serious analysis pieces attempt to explain what's going on in the national mood — maybe it's Republicans deciding to stick with their party as the election gets nearer? People getting tired of hearing about abortion rights? Anger over declining gas prices, maybe?
Or perhaps the polling averages are being skewed by a lot of garbage data from GOP-friendly polling groups that have injected polling results that don't have much to do with actual voter opinion. Democratic strategist Simon Rosenberg notes on Twitter that there appears to be a "ferocious" GOP effort to "flood the zone with their polls, game the averages, declare the election is tipping to them." He says that while it's entirely possible for Republicans to win in many of the elections next weeks, the polling and early turnout numbers so far suggest there's not really any sudden shift to the GOP — especially not if there's strong turnout by young voters.
Rosenberg warned Sunday that media organizations are being "played" if they uncritically report polling averages like those from FiveThirtyEight and RealClearPolitics, given the number of GOP-aligned polls being added in recent weeks in key states. Just look, he says, at the percentages of polls the GOP-friendly groups have been doing in states Republicans consider competitive, compared to polling in states where Republicans don't think there'll be much movement:
\u201cHere are the percentages of polls released by GOP allied groups ranked A/B in @fivethirtyeight of all polls taken in the state in October: \n\nAZ - 75%\nGA - 55%\nNV - 60%\nOH - 50%\nPA - 60%\nWA - 50%\n\nand in states they think are not competitive: \n\nCO - 20%\nNC - 29%\nWI - 0%\n\n2/\u201d— Simon Rosenberg (@Simon Rosenberg) 1667167274
How much might the influx of GOP polling be skewing the polling averages? Rosenberg notes that there's a "3.3 pt difference between the generic on Real Clear [Politics] and one without any partisan polling."
It's not just Rosenberg, either; Tom Bonier, CEO of Progressive data firm TargetSmart, pointed out Friday that an "avalanche" of GOP polling — some relying on an "older, whiter, more male" sample of voters than in the actual electorate — was making it look like Republicans were moving ahead.
\u201cLast week: polls show GOP leads nationally\nEarly this week: polls show Dem leads nationally\nLate this week: GOP pollsters release an avalanche of polls showing GOP leads. \n\nThis one, like the PA poll yesterday, shows an older, whiter, more male electorate than we've seen in GA.\u201d— Tom Bonier (@Tom Bonier) 1666990303
Digging further into that particular one-day poll is eye opening; it shows Republican Gov. Brian Kemp with almost 20 percent of the Black vote, which ... nah, not bloody likely. A recent survey of Black voters in Georgia showed only 15 percent had a "favorable" view of Kemp, and that's not even the same as saying they'd vote for him.
And wouldn't you know it, the influx of GOP polling showing Herschel Walker suddenly up three points — which has brought him within a point of Sen. Raphael Warnock in the FiveThirtyEight aggregate — all appeared well after news stories about Walker's having paid for an abortion. As Kerrey Elevald explains at Daily Kos, until around October 21, Warnock had a fairly steady three or four-point lead, but then, voila!
Of the seven aggregate polls taken since Oct. 21, five of them were conducted by either GOP-aligned groups or pollsters that use friendly GOP modeling: Trafalgar Group, Rasmussen Reports, Moore Information (Walker poll), co/efficient, and InsiderAdvantage. All of them put Walker in the lead by anywhere from 2 to 5 points.
What's more, the New York Times/Siena College poll from the same period still showed Warnock leading by three points, while an Atlanta-Journal Constitution poll showed the two in a tie.
Elevald continues:
Perhaps the most deceptive polling outfit is InsiderAdvantage, which polls for Fox affiliates across the country, giving the surveys the veneer of being even-handed media-sponsored polls. But a quick google search of the pollster finds their handiwork generating GOP-friendly headlines in several of the key contests they have polled:
• FOX 5 Atlanta: Kemp, Walker hold leads in major Georgia races in new InsiderAdvantage/Fox 5 poll
• FOX 29 Philadelphia: InsiderAdvantage/FOX 29 poll: Fetterman, Oz neck and neck as Shapiro’s lead over Mastriano narrows
• FOX 10 Phoenix: 2022 Arizona Election Poll: Lake leads governor's race, Senate race tightens
Wow, Republican prospects are really improving across the board. Amazing.
The Fox News website even ran a story this morning about the influx of GOP-friendly polls and their effects on polling, although the piece frames the data as a nutty "conspiracy theory" being pushed by MSNBC host Joy Reid, don't you know.
If you ignore all the accusations that Reid is a dangerous crazy, the story actually reports Rosenberg's concerns, noting his observations — or "claims" — that roughly half of the polls included in recent swing-state averages have come from Republican-aligned firms. Rosenberg even gets a fairly substantial blockquote (we've corrected typos in the Fox News transcription):
"What’s really unfortunate is that the places we rely on to help us tell us what’s going on in the election have been corrupted by a flood of Republican polling in the last few weeks. Now, in six major battleground states, more than half the polls conducted in October have been conducted by Republican firms. That means basically we can’t trust the data on Real Clear Politics or FiveThirtyEight any longer. It’s essentially Republican propaganda," Rosenberg claimed. [...]
"Listen, these are junk polls. The Republicans, this is part of the information war. They’re trying to suppress Democratic turnout, create more negative sentiment for Democrats and more positive sentiment for them. What I think is disappointing, many of the people who do the analysis on elections, should’ve caught this. This is an unprecedented massive campaign by the Republicans to game the polling average. And it’s disappointing to me this wasn’t caught earlier by many of the people that do this that are on TV and do this for a living. But it has to be understood now that the polling averages have been corrupted. We now need to look in my view towards the early voting," Rosenberg said.
Rather than address any of Rosenberg's substantive critiques, the story simply closes by insisting that "Reid has had a history of pushing bizarre and outright false conspiracy theories and misinformation," pointing to an erroneous 2020 Tweet in which Reid misstated the origins of the name "FiveThirtyEight." (She'd mistakenly said it was named for George W, Bush's margin of victory in Florida in 2000, when in fact it's named for the total number of seats in the House and Senate, so really, nothing she reports can ever be trusted again.)
In short, don't let stories about a coming "red wave" next week fool you into not voting. Most of the biggest races are so close that the polling is a tossup, so turnout is going to mean everything, the end.
UPDATE: Also too, speaking of turnout, especially by the youngs, there's this very encouraging information from the aforementioned Tom Bonier, in a short thread noting that if John Fetterman's debated performance had an effect on Democrats' enthusiasm, it's not showing up in turnout for the early vote.
\u201cBut the most astounding element of the PA early vote? Voters under the age of 30 returning ballots thus far are +69.2% D. At this point in '20 that same age group was +51.9D.\u201d— Tom Bonier (@Tom Bonier) 1667269783
Just 4 days ago, the Dem margin among those ballots returned so far was 2 pts wider than the same point in 2020. Now it's 3.6 points wider. Meaning the returns since the debate have gone even more solidly Dem than they were before.
But the most astounding element of the PA early vote? Voters under the age of 30 returning ballots thus far are +69.2% D. At this point in '20 that same age group was +51.9D.
He adds that one reason that the youth vote may appear lower this year is that unlike 2020, young Republicans don't appear to be turning out like they did in 2020. Hope that turns out to be the case!
[Daily Kos / Crooks & Liars / Simon Rosenberg on Twitter / Fox News]
Yr Wonkette is funded entirely by reader donations. If you can, please give $5 or $10 a month to help us help you sort through all the nonsense. Any nonsense we add is deliberate, of course.
Yes, Oklahoma And Other Red States Have Higher Violent Crime. It's A Fact, KEVIN.
It's the gun laws, dummies.
Joy Hofmeister, the Democratic running for governor in Oklahoma, said something OUTRAGEOUS during her debate Wednesday with incumbent Gov. Kevin Stitt:
"So let’s talk about facts: The fact is, the rates of violent crime are higher in Oklahoma under your watch than in New York or California," Hoffmeister said. "That's a fact." She's the state superintendent of public schools, and she has done her homework on this.
The claim was so obviously impossible that Stitt had to snort-laugh, interrupt Hoffmeister (twice) and point in disbelief at his Democratic challenger. It was also true, as the Washington Post pointed out (free gift link):
According to the Centers for Disease Control and Prevention, Oklahoma has a homicide rate of 9 deaths per 100,000, compared with California’s rate of 6.1 and New York’s rate of 4.7.
Yep, Oklahoma's murder rate is nearly double that of New York. That's the sort of thing you probably ought to have good information on if you're the governor of the state with more violent crime.
Stitt nonetheless tried to enlist the audience, mocking Hofmeister for such a ridiculous true statement: "Hang on, Oklahomans, do you believe we have higher crime than New York or California?"
Stitt's bit was met with hoots and applause from the audience, and the debate moderator said the matter would be fact-checked by nonprofit journalism outfit The Frontier (which hasn't yet published its fact check).
Nonetheless, people all over Twitter were happy to point out that Hofmeister was right. Here's video:
\u201cFinally. A Democratic candidate just correctly pointed out that red states like Oklahoma have HIGHER crime rates than New York and California.\n\nThe Republican candidate and moderator refused to believe it.\u201d— No Lie with Brian Tyler Cohen (@No Lie with Brian Tyler Cohen) 1666226386
Just to be scrupulously accurate, we'll note that, in a separate fact check, Oklahoma TV station KFOR, using FBI crime statistics, noted that Oklahoma had higher violent crime than New York and California in 2020, but that in 2019, the first year of Stitt's term, Oklahoma had "higher crime than New York, but not California." Going a bit further still, WaPo's Phili Bump notes that between 2019 and 2020, violent crime rates in Oklahoma actually increased more than either New York or California. Oh look, it's a chart!
Because FBI crime stats take a while to compile, Bump adds, we simply don't have the data to fact check whether Oklahoma has higher crime rates this year, which is awfully convenient for Republican politicians, who rely on the perception that crime is increasing whether it really is or not:
Polling from YouGov conducted in August showed a pattern that’s recurred over and over since violent crime began to fall in the 1990s: People tend to think crime is rising nationally even if they don’t think it’s increasing where they live. So if you are a candidate or organization with a vested interest in amplifying concern about crime, it’s easy to cherry-pick incidents or isolated measures to reinforce that concern.
Notice what Stitt said to the audience. Not “her figures are wrong,” but “do you believe we have higher crime than New York or California?” Not fact but feeling.
In reality, Hofmeister has identified a reality that Republicans would rather not talk about: As a Third Way study published earlier this year points out, murder rates in the last couple years have increased in red states as much or more than in blue states. In fact, in 2020, "per capita murder rates were 40% higher in states won by Donald Trump than those won by Joe Biden." And eight of the ten states with the highest murder rates in 2020 have gone to the GOP candidate in every presidential election since 2020.
Third Way's Jim Kessler told Oklahoma TV station KOKH that it's mostly the guns, stupid:
"There's a couple of factors you can't ignore. Nearly all murders are committed by guns. In Republican states, gun ownership is far higher than in blue states," [...]
But that claim is not cut and dry. States like Utah, Nebraska, and the Dakotas, traditionally Republican states, have very low murder rates.
"This is not iron-clad laws of physics. What I'm saying about places like Utah, Nebraska, and the Dakotas, they're very sparse areas and I do believe you need some population density to make this work," said Kessler.
I'm not sure anyone wants to "make that work" better than it already does, but you get the point. Compulsive idiot and GW Bush administration torture defender Marc Thiessen attempted to debunk the Third Way study by claiming that the real cause of crime in those red states is that they have cities with Democratic mayors, so really it's still Democrats being "soft on crime," but that's ridiculous bullshit, since it's red state legislators and governors making the gun laws, not the mayors.
For that matter, the study notes that cities with Republican mayors have some of the highest murder rates in the country, which further supports the idea that it's gun policy, not mayoral party, that's contributing the most to high homicide rates:
For example, Jacksonville, a city with a Republican mayor, had 128 more murders in 2020 than San Francisco, a city with a Democrat mayor, despite their comparable populations. In fact, the homicide rate in Speaker Nancy Pelosi’s San Francisco was half that of House Republican Leader Kevin McCarthy’s Bakersfield, a city with a Republican mayor that overwhelmingly voted for Trump. Yet there is barely a whisper, let alone an outcry, over the stunning levels of murders in these and other places.
In that latter example, the high crime rate in Kern County, California, appears to have more to do with the high level of gang activity spilling into rural towns than with any leader's party affiliation.
And for all the Republican campaign rhetoric blaming progressive prosecutors for increases in violent crime in the last couple years, a recent study released by the Center for American Progress actually found something very different, as the Atlantic's Ronald Brownstein explains:
Countering conventional wisdom, the study found that homicides over recent years increased less rapidly in cities with progressive prosecutors than in those with more traditional district attorneys. It also found no meaningful differences between cities with progressive or traditional DAs in the trends for larceny and robbery. “I think it’s really important to emphasize the extent to which we looked for a relationship and found none” between a prosecutors’ commitment to reform and crime rates, Todd Foglesong, a fellow in residence at the University of Toronto and one of the co-authors, told me.
So hey, Democrats, y'all need to be making a lot more noise about the facts, because there's no reason to let Republicans run on pointing at you and saying, “Can you believe this?" The belief that Republicans are tougher on crime is just wrong.
In any case, we're delighted to see Joy Hofmeister getting national attention, especially since she's polling remarkably well against Stitt in the final weeks of the campaign. It certainly can't help Stitt that, thanks to his ongoing fights with Indigenous tribes over casino profits, hunting and fishing rights, and tribal sovereignty, the state's five largest tribes have for the first time ever endorsed a candidate for governor, and it's Hofmeister.
He's just such an asshole, and you should send Hofmeister a few bucks if you can, the end.
[WaPo (gift link) / KFOR-TV / WaPo / KOKH-TV / Third Way / Atlantic / AP / Frontier]
Yr Wonkette is funded entirely by reader donations. If you can, please give $5 or $10 a month to help us keep you up to date on what's true, what's false, and what's updog.
Subscribe to the Wonkette YouTube Channel for nifty video content!