The rise and fall of "fact-checking"

"Fact-checkers" as the high priests of journalism had a political beginning at Facebook — and have met a political end.

On Tuesday, Meta CEO Mark Zuckerberg announced a series of changes to Facebook and Instagram that will considerably dial down the level of content moderation on those platforms. As someone who tries to be non-hypocritically pro-free speech, my inclination is to welcome the changes. But Zuck’s motivations are questionable: there’s no doubt that Meta and other media companies are under explicit and intense political pressure from the incoming Trump administration. So perhaps it’s the right move for the wrong reasons.

Just so that this doesn’t turn into another 5,000-word monster of a post, I’m going to focus solely on the first of five new policies that Zuckerberg announced. Here it is in his own words:

First, we're going to get rid of fact-checkers and replace them with community notes similar to X starting in the US. After Trump first got elected in 2016, the legacy media wrote nonstop about how misinformation was a threat to democracy. We tried in good faith to address those concerns without becoming the arbiters of truth, but the fact-checkers have just been too politically biased and have destroyed more trust than they've created, especially in the US. So, over the next couple of months, we're going to phase in a more comprehensive community notes system.

So, the news is that Facebook is eliminating a partnership that began in December 2016 with independent fact-checking organizations and replacing them with a Twitter/X style Community Notes program.

To which I say: that sounds fine, actually.

The fact-checking partnership was implemented during what was something of a moral panic on the left about “fake news” and the role it played in Trump’s first election win. I’ve always thought this was at most a minor factor in Trump’s victory over Hillary Clinton.

Still, the term “fake news” took off after 2016 — and so did the notion of perpetual, round-the-clock fact-checking. “Fake news” waned in popularity once Trump embraced the term, but right on cue, the more scientific-sounding terms “misinformation” and “disinformation” displaced it. Before Trump’s election, meanwhile, broader public interest in fact-checking was pretty much limited to election years. But by 2017, Trump’s first year in office, Google searches for “fact check” were almost three times higher than four years earlier in 2013.

Fact-checking has always been an intrinsic part of journalism

The notion of “fact-checking” as a separate subfield within journalism has always been strange. Fact-checking has long been an essential part of every journalist's job to the point where it doesn’t really need a name. A writer or reporter might be permitted some clunky prose or an unclear thesis in the first draft of their story. But they’re expected to have made every effort to get the facts right from the outset.

Meanwhile, there always have been such things as professional fact-checkers, but the term had a different connotation than how it’s used today. For instance, even though I had both a great editor and a great research assistant, I also hired a “fact checker,” paid out of my own pocket, for On the Edge, who focused especially on the parts of the book that could be legally or politically sensitive. He was incredibly meticulous, having a sixth sense for cases where, say, a fact was correctly cited from the source material, but the source was wrong. Again, though, this was part of the editorial process before the copy went to press — rather than having the fact-checker as some sort of journalistic high priest or arbitrator.

Because organizations like FactCheck.org initially especially focused on presidential election campaigns, scrutinizing claims made in speeches, ads and debates, it made some sense for them to play a designated role. Elections are all-hands-on-deck mode for newsrooms where the editing process is otherwise stretched thin. Essentially, it was a seasonal job, like gift-wrappers during Christmas season at a busy department store.

Following the 2016 election, though, a group of fact-checking organizations wrote an open letter to Zuckerberg to volunteer their services, claiming they had unique expertise that Facebook lacked internally to solve the “scourge” of “fake news” :

Last week you wrote that the problem of fake news and false information online is particularly complex. In your words: “Identifying the ‘truth’ is complicated.” We agree. It also cannot be the exclusive responsibility of any one organization.

As a network of independent fact-checking organizations set up to promote accuracy in public debate and the media everywhere from South Africa to Nepal, Argentina to the United Kingdom, and adhering to an open code of fact-checking principles, this is a challenge we deal with daily.

Long having an ambivalent relationship to the news and under political pressure — Facebook was blamed for the influence of Russian disinformation on the 2016 campaign — the company green-lit the partnership. Now, after getting called into Congressional hearing after Congressional hearing, making huge missteps on the Hunter Biden laptop story and suppressing legitimate discussions of COVID origins — and then being threatened by Trump — you can see why Zuckerberg feels burned. If this week’s decision was motivated by political and business incentives rather than high-minded principles — and it almost certainly was — so was adopting the fact-checking program eight years ago.

Zuckerberg’s claim is probably right, even if his intentions are poor

But is Zuckerberg’s claim that “fact-checkers have just been too politically biased” correct?

In my view, it’s at least pointing in the right direction, in line with my Indigo Blob theory about how the lines between nonpartisan institutions and partisan actors have become blurred. In the B.T. days — Before Trump — journalists who were appointed (or who appointed themselves) as fact-checkers tended to be experienced generalists with a scrupulous reputation for nonpartisanship — a sharp contrast to edgier and less experienced journalists in the Trump era who would later claim to own the disinformation beat. Perhaps because demand for fact-checking was coming overwhelmingly from the left — part of what I’ve called “The Big Cope” —- the journalists who selected into the subfield tended to be especially left of center:

There’s also the question of what claims are deemed as requiring a “fact check” or scrutinized for containing “misinformation” instead of being handled in the ordinary course of journalistic business. I suspect these are often precisely those claims that are either unresolved or unresolvable. Matters of opinion more than facts qua facts. But claims that are politically inconvenient to one side. (As an empirical matter, usually inconvenient to Democrats because the errors from fact-checkers nearly always seem to run in that direction.) The scrutiny of Biden’s age was one such example. Though obviously a suitable matter of journalistic inquiry, claims that the White House was covering up Biden’s deficiencies were often treated as “conspiracy” theories even though subsequent reporting has borne them out.

The common thread here is doth protesting too much. If a claim were easily refuted through regular journalistic methods, it would be. What filters through to the fact-checkers, who are rarely the journalists on the front lines of a story, are often the edge cases: half-truths and political hyperbole, or claims for which there’s “no evidence” either way, but a particular null hypothesis is privileged. Labeling these claims as “dangerous” misinformation or otherwise cordoning them off as out of bounds is essentially a bluff. But news consumers often fall for bluffs, and college-educated progressives — the main consumers of mainstream political journalism — are particularly unlikely to question the wisdom of self-proclaimed experts with the “correct” political opinions.

Of course, most claims that the fact-checkers make — the overwhelming majority, I’d guess — are narrowly true, or at least not provably untrue. The problem is these claims are nevertheless often strung together to advance a particular narrative — they’re spin, in other words.

Ironically, this was true even for how fact-checkers responded to Zuckerberg’s announcement. Understandably, fact-checking groups were not happy about it — it will cost them a considerable amount of funding and influence. But in responding to the news, they made a series of technically accurate statements that weren’t germane to his critique:

“We did not, and could not, remove content,” wrote Lori Robertson, the managing editor of FactCheck.org, which has partnered with Meta since 2016, in a blog post. “Any decisions to do that were Meta’s.”

The fact-checking organizations couldn’t remove content at Facebook, that’s true. Instead, content flagged by the fact-checkers was “demoted” and required you to click through some cumbersome warning screens to view it. Still, even in the aftermath of the week’s decision, the fact-checkers were clear enough about what their intent was: to put a heavy finger on the scale.

Under the Meta program, we provided links to our articles to Meta, which used them to direct users to our work and reduce the distribution of falsehoods.

And while the fact-checkers protested that their work “isn’t about censorship”, this also isn’t responsive to Zuckerberg’s critique. In fact, he didn’t mention “censorship” in explaining why he ended the fact-checking partnership. (That’s much more pertinent to the other changes that Meta announced, like increasing the range of permissible political speech on LGTBQ issues.) Rather, his critique of the fact-checking partners is they promised to solve a problem for him by providing non-partisan expertise, but instead they were too politically biased and just added another set of headaches.

Claims of political bias are hard to refute — bias is inherently somewhat in the eye of the beholder. But to the extent the fact-checkers have responded directly all, their response has been unpersuasive:

“I don’t believe we were doing anything, in any form, with bias,” said Neil Brown, the president of the Poynter Institute, a global nonprofit that runs PolitiFact, one of Meta’s fact-checking partners. “There’s a mountain of what could be checked, and we were grabbing what we could.”

Undoubtedly, there is a vast array of false claims made on Facebook every day, many of which originate with Trump and other politicians. So fact-checkers have to pick and choose their battles. That doesn’t absolve them of bias, however. The most persistent biases in the mainstream media are rarely sins of commission — the media rarely just makes stuff up — but sins of omission: which stories are deemed “newsworthy” or not, and then which receive greater editorial resources.¹ If you’re mostly grabbing stories that are low-hanging fruit from your vantage point because you’re overwhelmed, that’s probably going to reflect your biases more than if you make a concerted effort to reach for the higher branches.

And look, I obviously have my own biases, too. My impression is that journalists who label themselves as misinformation experts or fact-checkers have a relatively poor capacity for self-reflection, perhaps because these are such self-aggrandizing labels to begin with. But that’s a hard claim to prove. It might be wrong, and even if it’s mostly right, I’m sure you could point to counterexamples. Conversely, I have a relatively favorable impression of Community Notes on X. But it’s a relatively young program, and community-driven moderation can be hard to scale and can eventually develop its own toxic hierarchies — and Facebook is a much bigger platform than X.

I just don’t think it has done journalism much good to have a group of people specifically designated as misinformation experts or fact-checkers — that should be everyone’s job. And although I don’t really trust Zuckerberg’s motivations, it was fact-checkers who pressured Facebook for the partnership in the first place, not the other way around. It’s another chapter in the long history of journalists trying to sew ground with Meta and not liking what they reaped.

1. There is plenty of progressively-tinged misinformation out there — I’m not saying precisely as much as among conservatives, but plenty. To be fair, these sometimes do get attention from fact-checkers — they’re not batting .000.