Join us
The Media Today

In the Eye of the Needle

November 5, 2024 By Jon Allsop
County employees process mail-in ballots at a Clark County election department facility in Las Vegas, Nevada, in 2020. AP Photo/John Locher

Sign up for The Media Today, CJR’s daily newsletter.

Early yesterday, the union that represents tech staffers at the New York Times walked off the job, citing allegations of unfair labor practices amid stalled contract talks with management. (The union claims that bosses imposed return-to-office mandates without negotiation and also interrogated members about their strike plans.) The strike would have been disruptive at the best of times—the Times’ tech workforce ensures the smooth running of the paper’s digital output, including push alerts and live blogs—but could be particularly problematic this week, for reasons that are likely obvious unless you’ve been living on Mars. (And even then, you should be paying attention.) In particular, various observers have emphasized that the work stoppage could imperil the Times’ election night “Needle,” which seeks to project the outcome in real time based on evolving data. The union leaned into such concerns yesterday, mocking up a fake Needle showing “Strike Certain” in bright-red font.

If the Times’ election week labor dispute is unusual, the Needle itself is no stranger to election week chatter: in 2016, it was memed into oblivion after its debut coincided with Donald Trump’s unexpected victory; in 2020, the Times benched its national-level Needle altogether, citing the difficulty of projecting the overall outcome given pandemic-related changes to advance voting behavior, and instead put up three mini-needles showing the evolving direction of the vote in three states—Florida, Georgia, and North Carolina—that offered detailed data and were expect to count quickly. This year, the national Needle will apparently be back (perhaps accompanied by individual counterparts in the seven expected swing states), though in a preview article published yesterday, the Times said that, in light of the strike, it would only publish a live version of the Needle if it’s confident that the computer systems underpinning it are stable. (If they aren’t, Times journalists are instead planning to run their statistical model periodically and live-blog any insights about the output.)

Memes and strikes aside, the point of the Needle, the Times wrote yesterday, is to “put election results in proper context as they come in” and counter any “mirages” in favor of one party or another, since early returns can be unrepresentative of the likely outcome. There are various mundane reasons for this, but we live at a time, of course, when these mirages can really matter—in 2020, Trump seized on seemingly positive (yet actually meaningless) scores among votes that were counted first to call for the count to be stopped and his victory confirmed. Something similar could happen again this year; the pandemic, which contributed heavily to the precise patterns of vote-counting last time, is no longer a factor, but the result could still take days—or even, perhaps, weeks—to come into focus, and we are now four more years into Trump and his allies’ all-out assault on the integrity of elections. With that in mind, and voting almost over, it’s not just the Times that has had to make contingency plans.

Back in 2020, there was some concern that news executives weren’t adequately prepared for how different that election was going to look amid the ravages of COVID; a few months out from Election Day, Ben Smith, then the media columnist at the Times, canvassed leading journalists and was struck by their “blithe confidence.” As November neared, however, various major outlets outlined plans to inject more “caution, patience, and transparency” into their coverage, as I wrote at the time—from the Times’ Needle changes to pledges at major networks to be clear about possible mirages on air. The Associated Press, which plays a central role in calling election results, said that it would be more open about what it was calculating and why, even though, as the AP media reporter David Bauder noted at the time, going public in such a manner conflicted with the company’s unshowy culture.

This year, we’ve seen more such promises of, well, caution, patience, and transparency—especially on the latter front, in light of Trump’s lies last time around and the related erosion of trust in both electoral processes and the mainstream media. The AP has already run stories making public “the kind of granular vote-count information that used to be circulated just within the AP and other newsrooms,” as the Washington Post’s Elahe Izadi reported recently, and is also planning to stream live video from voting locations across the country. (According to Axios, the AP will have north of five thousand staffers on duty on election night, a company record; internally, the organization refers to its election operation as “the single largest act of journalism that exists.”) “We need to be better and faster in explaining what is happening” in key moments during the count, Julie Pace, the AP’s senior vice president and executive editor, told Bauder this year, “as opposed to saying, effectively, ‘We’re the AP, we have a 99% accuracy rate, of course we’re right.’”

Per Bauder, the AP’s efforts “to be more systematic and thorough in its explanatory efforts this year” can be traced directly to its controversial, if ultimately correct, early call of Arizona for Joe Biden in 2020. Not that the AP was the first outlet to make the call, of course—that, famously, was Fox News, whose number-crunchers put the state in Biden’s column before midnight on election night, leading to on-air confusion and a furious response from figures in Trump’s orbit who lobbied (unsuccessfully) to have the call reversed. Two journalists involved in the call subsequently left Fox; the network insisted that a restructuring was at issue, but one of those affected, Chris Stirewalt, later said he had been fired. (Stirewalt is now an on-air analyst at the upstart network NewsNation, where he, too, is promising enhanced transparency this week.) And yet Arnon Mishkin—the widely respected and, judging by recent interviews, apparently unflappable head of Fox’s decision desk—has said that he feels no pressure to make calls a certain way. (Asked by Politico whether he’d seen an episode of the TV drama Succession in which a character resembling him is pressured to fudge a call at a fictional right-wing network, Mishkin said that he had not, but that one of his kids had told him that the character in question was better-looking.) This time around, Fox is promising sharper graphics and clearer on-air communication around calls, among other things.

Major outlets are also taking steps that go beyond race calls. CBS, for instance, says that it has ramped up election-related fact-checking; ABC and NBC are sending teams of reporters to observe ballot counting as it happens, as is CNN. And the latter network has reinforced its team of legal experts and created a version of its “Magic Wall”—an election night prop (and sworn rival of NBC’s “Big Board”) that enables journalists to show viewers what’s going on around the country—that can be accessed online. Ultimately, such preparations can only take news organizations so far. As CNN’s Jake Tapper put it to Vanity Fair’s Natalie Korach recently, “nobody has any idea how this is going to end.” 

Sign up for CJR’s daily email

Still, we can make some educated guesses as to things that might happen along the way. As Tapper told Korach, it’s reasonable to expect that Trump will once again declare victory before all the votes have been counted if it suits him to do so. (“They have to be decided by nine o’clock, ten o’clock, eleven o’clock on Tuesday night,” Trump said at a weekend rally, proving Tapper’s point. “These are crooked people.”) We also know that his allies and supporters will circulate nonsense about supposed irregularities online; indeed, they’re already doing it. As in 2020, major outlets’ promises to combat this type of behavior with transparency and factual clarity is welcome. And yet—as in 2020—it’s also a safe bet that media muscle memory from more quote-unquote “normal” election weeks will kick in, especially on TV, in ways that unintentionally muddy the waters, subtly or otherwise. Amid top outlets’ more specific technical plans, it’s to be hoped they’ve learned these lessons, too.

First of all, if Trump does prematurely declare victory, major networks are not obligated to air any remarks to that effect live, as they apparently felt they had to last time around; refusing to do so isn’t some form of censorship, and would allow journalists to report on what Trump has said in due factual context. Second, if counting drags on, journalists—and on-air talking heads in particular—should resist the temptation to use dynamic language (Harris has “gained” votes; Trump has “momentum”) to describe that count, which is merely the delayed reporting of fixed data; in the past, such language may have seemed innocuous, but in the context of insinuations about fake votes being “added” as counting progresses, precision matters. Third, and perhaps hardest of all, journalists should pace themselves. Once polls started to close in 2020, certain anchors and pundits on CNN, for example, entered special-coverage mode at breakneck speed and with the cadence of drill sergeants, barking out both meaningful and meaningless data with a uniform, exhausting degree of urgency. (Not all “KEY RACE ALERTS” are key race alerts.) Whatever you’re saying, just saying it more slowly and quietly can help viewers separate what matters from what doesn’t.

None of this is to say, of course, that election week will unfold exactly as it did in 2020: we could get a national race call much sooner, or it could take even longer; as far as mirages go, there’s evidence that early voting patterns have cleft less neatly down partisan lines this time. Mishkin, however, predicts that the “over/under” for a national call is Saturday, which would be the same delay as last time, and some of the counting procedures for early ballots are the same this year, too, especially in the key states of Pennsylvania and Wisconsin, which still don’t allow for mail ballots to start being processed until Election Day. Richard L. Hasen, a leading election-law expert, wrote recently for Slate that he suspects that “some Republicans in Pennsylvania and Wisconsin have resisted making the change because uncertainty and the blue shift is a feature, not a bug: It allows for calling into question the legitimacy and fairness of the vote count if Democrats win.” And yet, Hasen argues, the possibility of this blue shift—or vote counts appearing to shift toward Democrats as mail ballots are slowly tallied—has been curiously undercovered in the media this time.

Whenever the results become clear, the press will need to be prepared for weeks of procedural gamesmanship—and, perhaps, worse. The Succession episode featuring the Mishkin-like character revolved around an apparent arson attack on a facility holding ballots, and already rang true (even if it wasn’t quite) to anxious media types when it aired last year. (The episode “brought back tremors,” my colleague Bill Grueskin told me at the time. “The only thing it lacked was the New York Times’ Needle.”) Recently, several very real ballot drop boxes were set on fire in Oregon and Washington State—an isolated set of incidents, it would seem, but a warning of the sort of thing that could unfold in the coming days and weeks. (A suspect reportedly used incendiary devices marked with the slogan “Free Gaza”; it’s not clear why.) “At the very moment when a watchful press will be desperate for new developments, conspiracy theorists and Donald Trump’s allies will be intent on sowing chaos and doubt,” The Atlantic’s Elaine Godfrey noted yesterday—and even after the results are confirmed, once-obscure certification deadlines could turn into flash points well into December.

If it’s worth learning lessons from how we covered election week in 2020, it’s also worth learning from how we covered the subsequent two-month period, when too many media voices were complacent about the ongoing potential for election-related violence (perhaps due to the risible early nature of Trumpworld’s lies; Four Seasons Total Landscaping, anyone?) and the scenes of January 6 would have seemed a dystopian nightmare. It’s hard to imagine such complacency recurring this time—and yet, as I’ve written, the clarifying shock of January 6 has somewhat dissipated and focused coverage of election threats often gets divorced from, or overwhelmed by, business-as-usual chatter about the horserace. As voting ends and counting begins, it’s incumbent on us to stay focused, and to stay calm; no matter what happens, the next few weeks seem likely to be a marathon, not a sprint. It is, again, great to see major outlets acting on promises of transparency—but keeping promises to be more cautious and patient is always the harder part of the bargain. As Pace and others have acknowledged, the information environment surrounding the election is not entirely up to the media. But we can at least ensure that the information we put out isn’t muddied on the front end.

Also unclear for now: how long the Times’ tech staff will be on strike. (The stoppage is currently open-ended; the union wrote online this morning that while “Election Day is about democracy,” for striking staffers “today is also about democracy in our workplace.”) Bosses have said that they have “robust plans” in place to ensure uninterrupted coverage, but some staffers, on both the journalism and tech sides, have expressed concerns that the Needle and other key tools might buckle under enhanced election-night traffic. Even if the Needle survives the strike, it could be paused for a different reason, the Times says: the journalists behind it may at some point need to sleep. I, for one, plan on doing the same. 


Other notable stories:

  • Elisabeth Egan, of the New York Times, spoke with Craig Garnett, the owner and publisher of the Uvalde Leader-News, in Texas, who is out with a new book about the horrific shooting that claimed the lives of nineteen students and two teachers at a local elementary school in 2022. “While friends and neighbors were reeling, while lawmakers offered thoughts and prayers, the Leader-News staff put one word in front of the other, covering the shooting and mourning its seismic ramifications at the same time,” Egan writes. “They kept going when they learned that their colleague’s daughter was among the victims. They kept going when members of the national media went home.” Garnett’s book, Egan adds, is a “devastating account.”

Has America ever needed a media defender more than now? Help us by joining CJR today.

Jon Allsop is a freelance journalist whose work has appeared in the New York Review of Books, The New Yorker, and The Atlantic, among other outlets. He writes CJR’s newsletter The Media Today. Find him on Twitter @Jon_Allsop.