Nieman Foundation at Harvard
HOME
          
LATEST STORY
Just how broken is our political information ecosystem, anyway?
ABOUT                    SUBSCRIBE
Nov. 4, 2020, 2:56 p.m.

Bad polls, fake news, and that damned needle: Here’s how the media covered a twisty Election Night

“Polling seems to be irrevocably broken, or at least our understanding of how seriously to take it is.”

That was quite a ride, wasn’t it? Or, rather, this election is quite a ride, given that it’s still going on. The quadrennial media rituals — “too early to call,” a sudden awareness of the nation’s counties, Steve Kornacki — all returned for the 2020 presidential election. As of Wednesday afternoon, no news organization has been able to declare an overall winner, and there are still meaningful votes left to count.

There weren’t a ton of meaningful innovations in last night’s media coverage — unless you count “caution” as an innovation, which maybe you should. The added complexity of massive early voting and the still-fresh scars from 2016 generally led to outlets more willing to let the night develop. But it certainly wasn’t a happy night for many viewers, readers, and obsessive doomscrollers. Here are a few of the things we noticed.

The needle and the damage done

The New York Times’ angst-producing needle didn’t just return, it multiplied.

There were three needles for the 2020 election cycle, representing live estimates for presidential race results in the battleground states of Florida, Georgia, and North Carolina. (The Times skipped a national needle this cycle, citing mail voting and the many different timetables and methods states had for counting ballots and reporting results this year.)

The Times led with the controversial visualization, encouraging readers to “bookmark the needle” even before pointing to their main results page. The needle — which we were told was “way smarter” than 2016’s iteration — did get some early, positive reviews,  including when it was one of the first to report that Trump appeared to have sewn up Florida.

But the night was long. And those who stayed up late on Tuesday saw more of the windshield wiper-like wavering that made people crazy four years ago.

Nate Cohn, a correspondent for The New York Times’ The Upshot, was giving insightful behind-the-scenes information and commentary on the needle’s performance throughout the night. But if that much meta-explanation is necessary, how useful is the needle to readers?

This time, though, the Times didn’t build constant quivering into the needle, as it had in 2016. This year’s needle only moved with changing data, not to evoke the basic unknowability of the universe. Readers hated “the jitter” with a fiery passion when first introduced; in 2018, the Times gave readers the option to turn the jitter off — though it cautioned: “Switching it off only hides the uncertainty — it doesn’t make it go away.”

Patience, please

Despite ample warnings that it would take a long time for votes to be counted and that we’d be unlikely to know the outcome of the presidential race on Election Night, the advice proved difficult for people to internalize.

Between 8 and 10 p.m. ET, as it became clear that Trump would win Florida and that we were in for a long night, folks began to feel the echoes of 2016 — and news organizations saw that reflected in their traffic, as readers looked to what happened four years ago to try to get a sense of whether history was about to repeat itself.

Misinformation and the platforms

After all that preparation for fighting misinformation on Election Day, how did Twitter and Facebook do? Well — there have been no major disasters, but platforms did intervene to block early claims of victory — which mostly, but not entirely, came from Trump’s side.

But the closeness of the race and the fact that ballots are still being counted create a gray zone that’s ripe for the spread of misinformation. Pennsylvanians, in particular, are seeing a lot of it. And as Biden performed much worse than projected in Florida’s Miami-Dade county, some wonder if Spanish-language misinformation was a contributing factor.

(Remember how the Miami Herald’s Spanish-language paper ran a racist and anti-Semitic insert for months without noticing, because no one in management had read it? Or the Miami radio station that aired 16 minutes of paid programming that said a Biden win would lead to dictatorship by “Jews and Blacks” and control “by racial minorities, atheists and anti-Christians”?)

And Facebook’s algorithms did what Facebook’s algorithms do.

Margin of error = ∞

In the weeks leading up to the election, many polls showed Biden up by 10 or more percentage points nationwide and up by 6 or 7 points in “tipping point” states like Pennsylvania. There was talk of another blue wave.

While the continued counting of mail-in ballots will likely pad Biden’s popular vote edge — the so-called “blue shift” — right now Biden is only up about 2 points, 50.3% to 48.2%, nowhere near landslide territory. That’s led to more complaints about the quality of political polling — and of how news outlets cover them.

Washington Post media columnist Margaret Sullivan put both on blast, saying “we should never again put as much stock in public opinion polls, and those who interpret them, as we’ve grown accustomed to doing”:

Polling seems to be irrevocably broken, or at least our understanding of how seriously to take it is.

The supposedly commanding lead that Joe Biden carried for weeks didn’t last very long into Tuesday evening. This was a lead, remember, that many predicted could result in a landslide Biden victory, help turn the Senate blue, and bring the Democrats amazing victories in red states like Ohio and Florida.

It didn’t take long for that dream to dissipate into a much more typical process of divvying up the states into red and blue, with a lot of unknowns added in. But none of it amounted to the clear repudiation of Trump that a lot of the polling caused us to think was coming.

Much of the polling was, in hindsight, awful. Adjustments that pollsters made after 2016 — like reducing the weights given to college-educated respondents — apparently didn’t seem to prevent some source of systemic error.

To be fair, poll aggregators, who came in for a lot of criticism after 2016, did seem to be more rhetorically responsible in this cycle than in 2016, doing more to emphasize uncertainty and how any poll or prediction was in reality a range of potential outcomes. In FiveThirtyEight’s final forecast, Nate Silver said that:

…because of Trump’s Electoral College advantage, which he largely carries over from 2016 — it wouldn’t take that big of a polling error in Trump’s favor to make the election interesting. Importantly, interesting isn’t the same thing as a likely Trump win; instead, the probable result of a 2016-style polling error would be a Biden victory but one that took some time to resolve and which could imperil Democrats’ chances of taking over the Senate. On the flip side, it wouldn’t take much of a polling error in Biden’s favor to turn 2020 into a historic landslide against Trump.

Reality ended up somewhere pretty close to the first of those two possibilities, if you were paying attention.

The Atlantic’s David Graham examined the larger problem that bad polling poses for society:

Pollsters and analysts are unlikely to get much sympathy, especially today. But the train wreck of their industry has consequences that run deeper than its impact on their own professional lives, or even having set incorrect expectations for the presidential race. Much of American democracy depends on being able to understand what our fellow citizens think. That has become a more challenging task as Americans sort themselves into ideological bubbles—geographically, romantically, professionally, and in the media they consume. Parties are now mostly ideologically homogeneous. We no longer spend much time around people who disagree with us. Public-opinion polling was one of the last ways we had to understand what other Americans actually believe.

If polling doesn’t work, then we are flying blind. That is an especially acute problem at the moment, because the coronavirus pandemic has made the old way the media got at this—shoe-leather reporting, despite its many shortcomings—much harder to pull off. The Trumpist alternative of simply trusting gut feelings isn’t any better. Gut feelings have failed plenty of candidates before; they may still prove to have failed the president this time.

Fair, balanced

Fox News isn’t known for its evenhanded coverage of Donald Trump or Joe Biden, and while the network tries to differentiate between its “news” and “opinion” sides, the line is often difficult to see.

But the Fox News Decision Desk, like its polling unit, has maintained a strong reputation for independence. And it maintained that reputation with some election calls that were frustrating for Republicans. Fox called the House for Democrats early, at about 9:15 p.m., before other networks. (Though it also said Democrats would add at least five seats, which now seems unlikely.) Fox News was also the first network to call the House for Democrats in 2018.

More significant was its early call of Arizona for Biden — soon after 11 p.m., more than three hours before the Associated Press did the same — which led to Republicans and Trump himself calling for a retraction.

For months, Trump and his campaign have been laying the groundwork for false claims of “voter fraud” and a “stolen” election. The early Fox call of Arizona blunted that momentum.

(This being 2020, a rumor spread on social media that Fox had retracted its Arizona call. It had not.)

Photo of “I Voted” stickers at Langley High School in McLean, Virginia, by Megan Lee of VCU Capital News Service used under a Creative Commons license.

POSTED     Nov. 4, 2020, 2:56 p.m.
SHARE THIS STORY
   
 
Join the 50,000 who get the freshest future-of-journalism news in our daily email.
Just how broken is our political information ecosystem, anyway?
Nearly half of Trump supporters surveyed still believe he’ll be sworn in for a second term in January. Not that he should be — that he will be.
“Whoa!” “I’m crying!” “Worrisome!” “Buckle up!” The swift, complicated rise of Eric Feigl-Ding and his Covid tweet threads
The scientist has gained popularity as Covid’s excitable play-by-play announcer. But some experts want to pull his plug.
Parler is bringing together mainstream conservatives, anti-Semites, and white supremacists as the social media platform attracts millions of Trump supporters
While it’s too early to tell if Parler is here to stay, it has already achieved a reputation and level of engagement that has overtaken other alternative platforms.