In defense of factchecking

Getting past "he said, she said" means accepting that reporters' conclusions won't always agree with our own

When it comes to factchecking, sooner or later, everyone’s a critic. And those criticisms come mainly in two flavors. One line of attack holds that the factcheckers at PolitiFact,, and their sister sites are slaves to semantics, issuing narrow technical rulings that ignore some more fundamental truth. After all, politicians know how to deceive us with carefully worded cant. Stamping “True” on that rhetoric is like giving them a license to lie.

The more common critique points in the opposite direction. It argues that these so-called factcheckers do violence to the English language by ignoring plain fact and finding ways to complicate what should be black-and-white questions. “PolitiFact, you are fired,” Rachel Maddow declared last year when the site applied too much nuance to President Obama’s State of the Union address. “You are a mess. You are fired. You are undermining the definition of the word ‘fact’ in the English language by pretending to it in your name.”

I’ve had both reactions to rulings that cut against my own (very left) politics, and that’s the point: It’s remarkable, and revealing, how easily we glide from one camp to the other, demanding context and nuance one day and slavish literalism the next. Take the apoplecty among liberal critics this week after PolitiFact awarded a “Half True” to House Majority Leader Eric Cantor’s claim that the federal deficit is growing. In fact, as the site explained in detail, the deficit has been falling rapidly as the economy recovers. But according to the Congressional Budget Office, health care costs and interest on the debt will start to drive it up again in 2016 and beyond.

That didn’t satisfy Paul Krugman, who warned the ruling would give ammunition to misguided budget-cutters on the right. “This should be simple,” he concluded. “PolitiFact should just rule on the facts … But apparently it can’t do it.” The Washington Post’s Greg Sargent argued that the factcheckers should have ruled the claim “plainly false” with a caveat noting deficits would rise again. MSNBC’s Steve Benen found a colorful analogy:

Imagine we’re driving down the highway in a car and I step on the accelerator. I then assure you, “Don’t worry, the car is slowing down,” despite the fact that the car is speeding up. PolitiFact would apparently say my claim is “half true” because sometime soon, the car will probably decelerate.

Now, you can quibble with the comparison: Do we really have gas-pedal control over the budget? Isn’t the larger point that under current commitments deficits are expected to go up, and keep going up? (Benen also offered a heat wave analogy, which seems more apt.) The language and framing matter a lot, because at root this is an argument about what Cantor meant and how people will interpret his words. Was he trying to mislead voters about the state of the deficit this year, or expressing concern—however misguided his remedies—about the longer-term trend? Mother Jones’s Kevin Drum acknowledged that the kinder interpretation seems reasonable in light of Cantor’s other remarks, even as he called factchecking a “doomed enterprise.”

But compare this episode to the epic backlash after the top factcheckers singled out as one of the most egregious lies of 2011 the Democratic claim that Rep. Paul Ryan’s budget would “end Medicare.” More context, the left demanded. Krugman explained that the essential nature of the health entitlement would be lost in converting it to some kind of subsidy for private insurance plans. Benen, master of automotive metaphors, likened the change to taking the metallic badge from a Ferrari and attaching it to a golf cart:

“Where’s my Ferrari?” the owner would ask.

“It’s right here,” I’d respond. “This has four wheels, a steering wheel, and pedals, and it says ‘Ferrari’ right there on the back.”

By PolitiFact’s reasoning, I haven’t actually replaced the car—and if you disagree, you’re a pants-on-fire liar.

It’s tempting to throw up one’s hands and say there’s no role for the factcheckers to play on questions like this. As Sargent argued at the time, “this disagreement ultimately comes down to differing interpretations of known facts—and not to a difference over the facts themselves.” At New York, Jonathan Chait ran through the proposed changes and then asked, “Is that ‘ending Medicare?’ Well, it’s a matter of opinion.”

That’s the wrong word, though, just as “opinion” is the wrong word for a lot of the deeply reported work that appears on the Op-Ed page. To say the Who are better than the Stones is an opinion. To say the Ryan budget wouldn’t “end” Medicare—and that it’s dishonest to claim it would—is a factual argument. (PolitiFact founder Bill Adair says the site practices “reported conclusion” journalism.)

To be sure, any factual arguments reflect particular values or perspectives. And critics from the right, like those on the left, say that undermines the factcheckers’ claim to neutrality. In response to PolitiFact’s previous “Lie of the Year” selection—conservative claims that Obamacare was a “government takeover of healthcare”—the Wall Street Journal editorial page complained that, “in reality PolitiFact’s curators also have political views and values that influence their judgments about facts and who is right in any debate.”

Well, yes. The shape of the world matters; facts take on particular meaning depending on the political context. But the “government takeover” analyses at PolitiFact and couldn’t have been clearer. They pointed out that the new health-care rules rely entirely on—and will actually expand—the private insurance market. Obamacare doesn’t feature public doctors like Britain’s NHS, or public insurance like Canada’s single-payer system. It doesn’t even have a “public option.” As welfare states go, that’s pretty weak tea. The factcheckers made a convincing argument about how reasonable people would interpret the phrase “government takeover” in light of conventional linguistic usage and the real political-historical context in which that language is embedded. We have lots of actually existing examples of what government-run healthcare looks like, and Obamacare ain’t it.

Factcheckers did the same thing in the “end Medicare” episode, applying a kind of Occam’s Razor to the phrase—though it can be hard for people (like me) who favor a larger government role in healthcare to see that. Here’s the logic: If Ryan’s privatization plan ends Medicare, what would we say of a proposal that actually ended it? Of course a scheme to give seniors two free aspirin a year and call it Medicare would count as killing the program, as Chait mused; but does everything short of that fall into some hazy, undifferentiated sphere of personal opinion?

It’s ironic that these journalists now take so much heat for making factual arguments. Not long ago one of the most common, and persuasive, critiques of the media—especially from the liberal blog set—was that journalists refuse to scrutinize political rhetoric beyond the “horse race.” The factchecking movement is a response not just to outlandish political claims but to the false balance and “he-said, she-said” journalism that publicized those claims uncritically. A kind of reporting that rethinks objectivity has to interpret evidence and draw conclusions. We can’t ask reporters to be more fearlessly assertive and analytical and then expect their conclusions always to agree with ours.

A subtler critique is that factcheckers go astray when they venture into “uncheckable territory,” as Greg Marx put it in a lucid take on the “end Medicare” brouhaha. Building on an analysis by Jay Rosen, Marx tries to mark off that territory by erecting a fence between arguments about truth and about legitimacy, which a project rooted in the language of fact and falsehood should steer clear of:

A project that involves patrolling public discourse, though, will inevitably involve judgments not only about truth, but about what attacks are fair, what arguments are reasonable, what language is appropriate.

But things get tricky here, too. Marx concludes (and Rosen agrees) that “you can’t report your way to the conclusion” that the GOP plan does or doesn’t end Medicare. The dispute turns on politics, not factual evidence. But by the same logic you can’t strictly report your way to the conclusion that the Affordable Care Act doesn’t take over the healthcare system. Your can’t report your way to finding that President Obama didn’t go on an “apology tour,” or that he didn’t propose to “gut welfare reform,” or (to take a contrary example) that Mitt Romney never “backed a law” to ban all abortion.

All of these conclusions rest on subtle judgments about what arguments are reasonable and what language is appropriate. Even in technical matters, “true” and “fair” quickly get tangled up. Take a simple question: How many millionaires don’t pay federal income taxes? Sen. Harry Reid said the number was 7,000 for 2011, and earned two Pinocchios from The Washington Post. It turns out there’s more than one way to count millionaires, so there’s no truly value-free way to rate the claim true or false. These are the kinds of unstable facts the factcheckers work with all of the time. They navigate a messy political economy of information in which there’s always more than one standard to choose from, and experts and data sources inevitably come with some bias.

The factcheckers do issue some rulings that don’t provoke much disagreement. In these cases, questions of legitimacy fade into the background. This may happen much less often than we would like to think, though, and it’s a litmus test that political actors can easily game. To apply it too rigidly—to say these are the only questions factcheckers should rule on—would provide a powerful shield for politicians’ most misleading claims.

More to the point, this is the wrong litmus test for a journalism that very deliberately rejects the “he said, she said” formulations that sustain the “View from Nowhere.” We can’t ask journalists to make the judgment that torture is torture—in the face of the rhetorical, political and legal apparatus that has been erected to redefine that word—and then also insist that they stick to pure “reporting.”

Or to put it the other way around: The fiction of a clear division of labor between reporting, on one hand, and interpretation or argument on the other is an artifact of conventional objective practice, which obscure the reporter’s decision-making and argument-building to make it seem “as if the facts speak for themselves.” We can’t have it both ways.

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

Lucas Graves is an assistant professor in the school of journalism and mass communication at the University of Wisconsin. Follow him on Twitter at @gravesmatter. Tags: , , , , , , ,