I come with a warning.
I read dKos multiple times every day. I have since 2003! But haven’t written much in the past few months. I had been spending most of my time over on election Twitter (until space Karen rendered the site largely unusable). Although my field isn’t political science, I am proud to have earned a Ph.D. and am intimately familiar with the data science used by everyone from elite media election analysts on down to just random election twitter users such as myself.
At the beginning of the summer, some of us, under the guidance of Simon Rosenburg, Christopher Bouzy, Tom Bonier, et al., started to notice that the vaunted red wave was … not materializing in the data.
That’s … not how things turned out. President Biden is on track for the 5th best midterm performance since 1930. That’s right behind FDR after the Great Depression, George W. Bush after 9/11, and John F. Kennedy after the Cuban Missile Crisis.
My final prediction, 22 days before the election, was that, based on data science, Democrats would win 51 Senate Seats (through PA, GA, NV, and AZ) and lose 10 House Seats.
I should buy a lottery ticket.
No, it wasn’t 100% accurate(1)
But how could I, a nobody from Daily Kos, be orders of magnitude closer to the result than the experts who are paid millions of dollars to get this right? Only through a colossal failure of our elite media institutions. It will take more than one diary to fully explain how they could be so wrong, and the nobodies could be so right. The first installment is on what motivated them.
Motive
In short, our media elites were motivated to get this result wrong. People, this is really scary.
“So what?” you say. They look like fools.
Yes, they look like fools. The Dave Wassermans, the Charlie Cooks, the Larry Sabatos, the angry white men who run RCP, the Nate Silvers, all of CNN, they all look like fools.
But a breakdown of this magnitude is more evidence we are living through one of the greatest media failures of the past two decades (and there is some stiff competition on that front). A media this degraded cannot protect us from autocracy.
So what do I mean by media elites being motivated to get this result wrong? Shouldn’t they be motivated to get the result right? If you nailed this election, you would’ve looked like a genius! Everyone would click on you and you’d get wealthy. Behold, the concept of motivated reasoning!
Motivated reasoning is the phenomenon in cognitive science and social psychology in which emotional biases lead to justifications or decisions based on their desirability rather than an accurate reflection of the evidence...
Media elites identified a red wave (in contrast to contradictory evidence) because they wanted a red wave. And there was not a single media elite who dissented.
If you think this has a connection to the media’s odd negative obsession with President Biden, you’d be right. The media has been relentlessly negative towards President Biden. So much that it caught the attention of the … Beltway media, a year ago.
So why did the Beltway media want a red wave? Aren’t they all liberals themselves? Remember, stories of a red wave were often accompanied by stories of voters prioritizing economic concerns. (See also: here, here, here, here, yada yada yada). Despite inflation, the economy is really good. Here is the New York Times on October 17:
Republicans Gain Edge as Voters Worry About Economy, Times/Siena Poll Finds
With elections next month, independents, especially women, are swinging to the G.O.P. despite Democrats’ focus on abortion rights. Disapproval of President Biden seems to be hurting his party.
The Times Sienna poll for which the above article headlines was a major component of the red wave snowball, and I will come back to it (and Nate Cohen) in later diaries. But just as the media was mad at Biden for taking away their forever war, supposed liberal elites wanted to take Biden down a peg for relentlessly pursuing pro-worker and fiscally expansionist policies. Remember, these same elites are often unfortunately seated next to Larry Summers at dinner parties, and according to people like Larry Summers, Biden’s pro-worker and expansionist policies are the cause of inflation.
I’m not saying Larry Summers is right or wrong (he’s wrong), but the desire to see Biden up there offering a mea culpa Wednesday morning for writing everybody a $1400 check overwhelmed any and all sense of reason.
But believe it or not, this was not the only motive.
Another part was just general fear of Republicans. Polls had missed in the direction of Democrats famously in 2016 and also 2020. We will talk more about how the miss in 2020 in particular is especially dangerous, as there were serious signs of herding (pollsters adjusting their results to match others). The greatest sin in Beltway media is being perceived of as “partisan” (which translates as “a liberal” to us). Here is Dave Wasserman from just today.
Good Lord! How out of touch can you be!
Coverage of the 2022 midterms paralleled coverage of Hillary Clinton in 2016. The beltway media didn’t want to be perceived of as partisan, and considered this a good opportunity to take Hillary Clinton down a peg.
And lastly, there was a purely financial motive. What we would consider nonpartisan media (i.e., elite media) is consumed almost exclusively by liberals. Conservatives don’t read the Washington Post or New York Times. They won’t watch CNN or PBS News Hour. They have their own media ecosystems. And elite media long ago discovered that negative stories about Democrats / positive stories about Republicans generate clicks. Scaring the libs is a major business element of elite media.
So you add all these motives up, and you get motivated reasoning.
* * *
I’ve been working on additional diaries, going over what the data actually said, herding phenomena, how we knew rightwing pollsters were full of it, and more importantly, what we can do going forward.
Notes:
(1) I was wrong in that OH would go for Vance by more than projected (6.5%, not 3% as I predicted), although Cheri Beasley lost by the 3% predicted. I didn’t give a margin for Mandela Barnes’ loss in that tweet, but in other tweets I had predicted larger than the ultimate margin of 1%. I was also wrong on the magnitude of the House gains by Republicans (it appears they may gain 6 or 7, not 10).
