Red Waves Are No Match for Black Swans
The 2022 midterms are a lesson in improbability—and what happens when unlikely voters decide to vote
Once thought to be nonexistent, the discovery of the black swan in Australia gave rise to the bird’s use as a metaphor to describe an unexpected event of significant consequence.
The financial website Investopedia defines a Black Swan as “an unpredictable event that is beyond what is normally expected of a situation and has potentially severe consequences.” In his 2007 book The Black Swan: The Impact Of The Highly Improbable, written just before the Great Financial Crisis, professor and Wall Street trader Nassim Nicholas Taleb introduced "black swan theory,” which attempts to explain hard-to-predict events and our inability to compute their probability using scientific methods.
By definition, these events are rare and nearly impossible to predict. That said, while we consider the 2006 housing collapse and the subsequent financial crisis a black swan event, a few “saw the pattern in the Matrix” and invested accordingly (see: The Big Short by Michael Lewis). Earlier this year, I wrote about the hedge fund manager who made $15 billion in 2007 by betting on a crash in the housing market.
While the term “black swan” has come to be associated with improbable financial calamity, it is an apt descriptor for any rare, unforeseen event. Occurrences of this nature are often incorrectly rationalized after the fact with the benefit of hindsight.
Such a rationalization is currently taking place regarding the spectacular failure of the polling firms, and by extension, the media, to correctly forecast the most recent black swan — the 2022 midterm elections.
Inflation, gas prices, and crime—oh, my
In the run-up to the midterm elections, nearly all of the traditional media sounded the same alarm: Democrats should brace themselves for the proverbial shellacking. The D.C. intelligentsia warned that the midterms would at least be a Red Wave, if not a full-blown Red Tsunami.
To bolster their assessment, pundits pointed to historical data as empirical evidence that the party in power loses control of the House and Senate.
In the 22 midterm elections from 1934 -2018, the president's party has averaged a loss of 28 House seats and four Senate seats. The president’s party gained seats in the House only three times but gained seats in the Senate on six occasions. The president’s party has gained seats in both houses only twice.
The most recent exception to this eighty-year midterm election trend is George W. Bush, who, thanks to 9/11, added a few seats on both sides of the aisle in his first midterm. But since the country isn’t involved in a wartime situation (our Ukrainian support notwithstanding), the general prognosis was that Democrats were doomed to lose dozens of seats.
Why the impending bloodbath? All we needed to do was to look at the polls, we were told. The reasons for the coming wave were apparent: 1) inflation, 2) Biden’s dismal approval ratings, and 3) crime. The prevailing talking point was, “Sure, women are annoyed about losing their right to bodily autonomy, but have you seen the price of a gallon of gas?” (Not really, but the logic was almost that tortured.)
Unsurprisingly, Fox News took Red Wave hype to another level, as MSNBC’s Chris Hayes illustrated in this post-election montage:
But while Fox News engaged in the kind of propagandizing we’ve come to expect, they weren’t alone. Commentators on MSNBC presenters described the inevitable Democratic losses as a fait accompli, directing us to poll aggregators such as RealClearPolitics and FiveThirtyEight to support their assumptions.
The Sunday before the election, CNN’s Hilary Rosen lamented that while she was a loyal Democrat, she had a bad feeling about the party’s midterm prospects because of Democrat’s ill-advised focus on saving democracy from would-be autocrats and election deniers.
Rosen drove the point home by adding that “when voters tell you over and over again that they care mostly about the economy, listen to them. Stop talking about democracy being at stake.” The network’s Chris Cillizza characterized Democrat’s focus on abortion as “a losing bet.” Legacy pollster Frank Luntz predicted a Democratic debacle not seen since the Republican Revolution of the nineties.
According to Luntz, who was quoted in a BBC interview as saying, “The only person more unpopular in America than Joe Biden is Donald Trump,” the top issues for voters in the upcoming election were the economy, inflation, and crime. Luntz ranked abortion as the 6th most crucial election issue.
Former New Jersey governor turned ABC News pundit Chris Christie went a step further than Luntz, saying that abortion rights don’t matter to most Americans compared to the poor economy, drugs, and crime. In a cringeworthy display of mansplaining, Christie made his declaration surrounded by ABC’s Martha Radditz and two other women, self awareness be damned.
The Cook Political Report projected that Republicans were poised to win as many as 25 seats in the House of Representatives. Meanwhile, in an article published on the day of the election, RealClearPolitics predicted a slew of Democratic Senate losses in Arizona, Nevada, and Pennsylvania.
The piece, entitled “Why Midterm Voters Will Put Republicans in Power Across the U.S.,” pointed to Biden’s alleged missteps to support its conclusion:
Two years ago, Democrat Joe Biden beat Republican Donald Trump in large part because he promised a calmer political climate and claimed he would work across the aisle to achieve bipartisan change.
But when the former vice president kicked off his presidency on Day One by shutting down the Keystone Pipeline, many who put him in office were concerned. Their worries deepened when he botched our exit from Afghanistan in August 2021. And, this year, when he called skyrocketing inflation "transitory," lots of voters decided it was time to choose new leadership in the midterm elections on Nov. 8.
Curiously, the piece did not mention the Republican Party’s rampant election denialism, its lack of policies, or the Supreme Court’s Dobbs decision, which overturned Roe vs. Wade. In their estimation, voters were much more concerned about an unfinished oil pipeline halted two years ago.
The wave that never was*
But despite the flood of calamitous predictions, Democrats flipped governors’ seats in Maryland and Massachusetts. In Pennsylvania, Democrats won the governorship and flipped a Senate seat, winning by 5 points.
In Michigan, where RealClearPolitics’s pre-election poll showed incumbent governor Gretchen Whitmer with a one-point advantage, Whitmer won by double digits. Democrats also won the Attorney General and Secretary of State races and both legislative chambers for the first time in decades.
In Minnesota, Governor Tim Walz won his election. Democrats also took back the Senate in the legislature, giving them full power for the first time in nearly a decade. A RealClearPolitics pre-election poll had Arizona gubernatorial candidate and election denier Kari Lake ahead of Katie Hobbs by over 3 points. Hobbs went on to win the governor’s race. (As of this writing, Lake has yet to concede.)
The Democratic Party also had a historic midterm performance on the state level. This is the first time since 1934 that a president’s party achieved a net gain of governorships in their first midterm.
When the “unlikely voter” decides to vote
Although Republicans managed to flip the House, their thin majority hardly constitutes a Red Wave. Democrats had such a rare level of success one might categorize the achievement as a political black swan. So why didn’t the experts see it coming?
One reason the media missed the boat is their dependence on aggregated polls like RealClearPolitics and FiveThirtyEight. Typically, these polls rely on a methodology based on individuals with a history of voting, known as “likely voters.” This model is considered the holy grail of polling practices.
Based on conventional wisdom, GenZ voters are an unreliable segment of the electorate. In retrospect, the polls failed to recognize that cohort’s enthusiasm level. Motivated by the abortion issue, not inflation or crime, young women voted in numbers that staunched the possibility of a Red Wave:
Exit polls show 72 percent of women ages 18-29 voted for Democrats in House races nationwide. In a pivotal Pennsylvania Senate race, 77 percent of young women voted for embattled Democrat John Fetterman, helping to secure his victory…
Remove young women from the equation, and neither women nor young people delivered much to the Democrats on Tuesday, according to exit polls conducted by Edison Research. Fifty-three percent of women overall voted blue in House contests. Women over 45 delivered the party no advantage at all. And Democrats won a comparatively low 54 percent of votes from young men…
Among young voters of both genders surveyed in Edison exit polls, 80 percent favored legal abortions, and 49 percent named abortion as the issue that sealed their vote. Sixty-two percent of women under 45 listed abortion as their top issue.
According to data compiled by the Center for Information and Research on Civic Learning & Engagement (CIRCLE), whose non-partisan research organization focuses on youth civic engagement, midterm election voters between 18 and 29 said abortion was their top issue, followed by inflation and crime.
Also, the polls used in aggregation models matter. For example, InsiderAdvantage, a polling firm ranked by RealClearPolitics as the 2nd most accurate national pollster for the period of 2016 through 2020, had Kari Lake (who ultimately lost the governor’s race) with a double-digit lead over Katie Hobbs in late October.
Atlanta-based Trafalgar Group, a Republican-funded pollster used in RealClearPolitics and FiveThirtyEight models, was on the wrong side of almost every race this election cycle. Slate’s post-election piece on Trafalgar’s Robert Cahaly, a frequent guest on Fox News, breaks down the problem with so-called junk polls, complete with receipts:
A 2020 New York Times article about the company—published before Biden’s victory, even—noted that “Trafalgar does not disclose its methods, and is considered far too shadowy by other pollsters to be taken seriously.” But it does release crosstabs, i.e., breakdowns of polling responses by subgroup, which allow observers to see under the hood at least a little bit to find out how it gets its overall numbers.
This process raises more questions than it answers. In this year’s Georgia race, for example, Trafalgar asserted that Herschel Walker led Raphael Warnock by a literally unbelievable 62–37 margin among voters 18 to 24 years old. Other polls had that margin completely reversed, and according to AP VoteCast, Warnock won the 18–29 group by a margin of 64–36 in his 2020 runoff against Kelly Loeffler. The company also said that Republican Don Bolduc was winning Hispanic voters in New Hampshire 70–17; while 2020 exit poll data doesn’t include Hispanic voters in New Hampshire because that’s such a small subgroup, Biden beat Trump among “Latino” voters in nearby New York state by a margin of 72–27.
One can only speculate as to where Trafalgar found a set of Gen Z voters—89 of them, according to the other data it provided—that broke 2-to-1 for the MAGA candidate in a swing state like Georgia.
Perhaps the best (or worst) example of Trafalgar’s polling ineptitude was the pollster’s pre-election prediction for the Vermont Senate race, where their poll missed by over 30 points.
The Washington Post’s Dana Milbank offered this post-election analysis to explain the polling debacle:
Political journalists were suckered by a wave of Republican junk polls in the closing weeks of the campaign. They were also swayed by some reputable polling organizations that, burned by past failures to capture MAGA voters, overweighted their polls to account for that in ways that simply didn’t make sense. And reporters fell for Republican feints and misdirection, as Republican operatives successfully created an artificial sense of momentum by talking about how they were spending money in reliably blue areas.
An extraordinary profusion of bad partisan polling flooded the media late in the campaign, coming from GOP outfits such as Trafalgar (which had Blake Masters over Mark Kelly in the Arizona Senate race, Don Bolduc over Maggie Hassan in the New Hampshire Senate race, among others) and Rasmussen (which gave Republicans a five-point edge in the generic ballot).
Useless polling notwithstanding, it’s comforting to know that, in the final analysis, the electorate cares more about their rights and democracy than the price of a gallon of gasoline.
*Fans of HBO’s House of the Dragon will recognize the Rhaenys Targaryen reference.
If you enjoy and want to support my writing, consider signing up for my weekly-ish as a paid subscriber. You’ll receive early access to my posts and subscriber-only podcast. You’ll also get invited to occasional Zoom events and paid-subscriber-only chats.