In the aftermath of election night 2018, New York Times columnist Bret Stephens wrote a column arguing that the Democrats had kind of blown it. Under the headline “The Midterm Results Are a Warning to the Democrats,” Stephens argued that the Democrats’ electoral haul was “meh” and evidence that “while ‘the Resistance’ is good at generating lots of votes, it hasn’t figured out how to turn the votes into seats.”
It was, on its face, a curious argument, given that the Democrats had wrested control of the House of Representatives, snagged a healthy number of governorships, flipped hundreds of state legislative seats, and triggered a down-ballot political “earthquake” in Texas.
The 28-seat swing that gave Democrats control of the House wasn’t even half the 63 seats Republicans won in 2010. Yet even that shellacking (to use Barack Obama’s word) did nothing to help Mitt Romney’s chances two years later. The Republican gain in the Senate (the result in Arizona isn’t clear at this writing) was more predictable in a year when so many red-state Democrats were up for re-election. But it underscores what a non-wave election this was.
There are many problems with that analysis, but the most significant among them is the number 28. At the time Stephens wrote this column, votes were still being tallied and many close races remained uncalled. That 28 was the lower boundary of the Democratic gains. Now we know that the Democrats have actually picked up 40 seats -- a tally that is about 43 percent higher than the number Stephens based his insta-analysis on.
Since then, the Times has quietly updated Stephens’ column to reflect the growing number of Democratic pickups, but the analysis on which it’s based remains glaringly static. Here’s how the updated version reads, as of this moment:
A 37-seat swing gave Democrats control of the House—a definite gain, but still less than the 63 seats Republicans won in 2011. Yet even that shellacking (to use Barack Obama’s word) did nothing to help Mitt Romney’s chances two years later. The Republican gain in the Senate (the result in Arizona isn’t clear at this writing) was more predictable in a year when so many red-state Democrats were up for re-election. But it underscores what a non-wave election this was.
The 28 jumped up to a (still low) 37, and the “wasn’t even half” has been disappeared. But the “non-wave election” diagnosis remains preserved in amber. With each update to the number of flipped seats, Stephens’ already bad analysis just gets worse.
What we see here are two distinct flavors of bad punditry.
The first is an issue of confirmation bias. Prior to the election, Stephens had already concluded that the Democrats had squandered their opportunity. In his October 12 column “Democrats Are Blowing It, Again,” Stephens wrote that House Republicans “now have at least a fighting chance of holding on to a majority despite the widely anticipated blue wave,” hanging his argument on a 10-day window of not-terrible polling for GOP candidates in battleground districts. A few days before the election, Stephens wrote that “Democrats should be walking away with the midterms. That they are not is because they have consistently underestimated the president’s political gifts, while missing the deeper threat his presidency represents.”
When the actual voting results badly undermined Stephens’ preformed conclusion, he tortured that inconvenient reality into a shape that roughly conformed to what he had already decided was true.
The second is the temptation and perceived need to provide instant, authoritative analysis of still-uncertain outcomes. Stephens was hardly alone in this failing -- early on election night, cable news was awash in dour, speculative takes about how the Democrats were going to come up short, causing attendant panic and depression among liberals on Twitter. By the time Stephens’ column was published, Democrats had taken the House but several races (including many toss-ups in California that would end up going to the Democrats) had yet to be called. The Democrats’ 40th pickup was called just two days ago.
The still more serious problem that flows from both of these issues is that pundits who fail in these ways are hardly ever held to account, and thus self-serving, half-baked analyses keep showing up in op-ed pages and on cable news. For Bret Stephens, “accountability” takes the form of quiet changes to his column that leave its conclusions intact even as it gets more wrong with each update.