Is it time to completely rethink fact-checking?
The basic format of political fact-checking is identical the world over: Start from a widely circulated claim by a public figure, then rate the truthfulness of its objectively verifiable components.
Tom Rosenstiel, executive director at the American Press Institute, thinks it may be time to rethink this model. Local newsrooms especially, he suggests, should make “the key unit of fact-checking not a claim or a fact, but an issue.”
The most obvious reason fact-checkers start from claims is that they are newsy statements made by powerful people that go on to shape news cycles, campaigns and public policy. In this sense, claim-centric fact checks are an instrument for accountability as well as of verification. Besides, fact checks in this format are straightforward for readers to digest and can be easily collected into a broader reference for each politician.
Related Training: Fact-Checking: How to Improve Your Skills in Accountability Journalism
Underlying Rosenstiel’s proposal for a Copernican revolution in fact-checking is an overarching concern about the trust and reach of fact-checkers in 2017. Poynter got in touch with Rosenstiel to discuss this idea; what follows is a transcript of our conversation, edited for clarity and length.
Without getting into the "why" just yet, could you explain what you mean by issue-centric fact-checking?
Fact-checking would start from the core issues that are being discussed in a campaign, or the big problems that face a community. The next step would be to find out what impressions people have that are correct and incorrect on those issues. Specific claims would be almost a third-order element in this system. The claims would be elements that have contributed to confusion or disagreement over an issue, but not the main story.
This approach would have a number of advantages...
Before we get to that, I’d like to talk a little bit more about the "how." How do you imagine fact-checkers collect these overarching issues? Through polling?
So there's some history here to consider. If you go back to the 1992 presidential election in the United States, there was an effort to identify what were the most important issues facing the country. The Washington Post, in particular, set out through various means to identify what the most pressing issues facing the country were.
That included in-depth interviews with people around their kitchen tables with reporters like David Broder, who was famous for that kind of work. There were surveys done and experts consulted. The publications made a pretty transparent and good faith effort to not necessarily rank these issues but to identify maybe five or six.
That was done in that particular election in reaction to the 1988 presidential election where the major issues in the campaign had been prison furloughs and whether Michael Dukakis supported the Pledge of Allegiance. Some people deemed these topics to be more symbolic that substantive and thought that maybe the discussion about the future of the country was dominated by political consulting strategy and not helpful to voters.
That was an early step towards fact-checking, in the sense that news organizations were saying ‘we're not just here to take a picture of what's happening.’ In the same way that fact-checking takes you off the sideline and onto the field, this identification of major issues was a way of doing that.
Locally, it might not have to be so formal. You just need some kind of predicate that is transparent.
So is this more of a fact sheet approach than a correction approach?
Let me mention the "why" a moment before I speculate on the "what." This was an idea we began to generate more than a year ago, but the election has made it seem like an even better idea.
First, we know from research we did with Brendan Nyhan and other scholars that there is a partisan divide to fact-checking and that people are more resistant to the fact-checking when it’s their guy who is being fact-checked.
The second thing we discovered, from the last six months, is that fact-checking can be too literal, too narrowly focused. You can fact-check a claim and have audiences say "OK he got those numbers wrong, but I am still not going to be shaken from my larger belief that we have too many immigrants or that the system is rigged or that people of a certain religion were actually happy about 9/11."
And then the third thing is: How do we currently decide what to fact-check? Are we just grabbing at random claims here or there that sound fishy? What fact-checkers choose to fact-check reflects a lot about their motive, or their perceived motive. It's probably an under-appreciated issue that how you go about selecting claims to fact-check has a big impact on whether people trust you as a fact-checker, regardless of how thorough you are afterwards.
Do you think there is more research that could be done to test your proposition that readers will trust fact-checking around issues more than fact-checking around claims? Or should someone just try doing it?
Probably a bit of both. I think there is a lot still to be learned about how people receive fact checks, but probably the most useful thing to do is to try this idea out and see where the rub points are.
Ultimately, the ethos of fact-checking should be to help the news consumer to decide for themselves how to think about an issue — rather than to wave a finger and say this is right and this is wrong. And I think this holistic issue approach really helps in that regard. It's this very rich version of fact-checking that gets at what fact-checking is for in the first place, which is to create understanding, not to say "Hey that guy is a 67 percent liar-type."
This sounds to me a bit like the model pursued, among others, by Full Fact in the United Kingdom. Their mantra is to “play the ball, not the man.” That seems closer to an issue-centric approach. Is that a cultural thing, too? Is discourse more adversarial in the United States?
No, actually, I think we are less adversarial. The American tradition is more of the scribe writing things down and passing them on rather than the interpreter.
There may well be a model to follow here, but the goal is also to empower local newsrooms to become leaders by conducting this approach in their communities. One of the things that contributed to political polarization in the United States is the nationalization of all our issues. We're a big country, used to thinking that all politics is local. But as the internet has come along, people have accessed and focused more on national media. They have spent more time looking at national issues and spent less time looking at local concerns. So this is a way of making local concerns matter again.
Readers are very much attracted to this claim-centric model. Especially at a time when we want fact-checking to perform well on social media, does the issue-centric approach lend itself a little less to that?
I think that's potentially true.
My final question is less relevant now that I see you see this first and foremost as a local prerogative rather than a national one. Still, I think one of the strengths of Factcheck.org, PolitiFact and The Fact Checker is that they have a large backlog of claims that remain relevant or become relevant again. Can you envisage their work being repackaged under this new approach?
Yeah and I think this is really important actually. Maybe it is one of the first places where it happens. In journalism in general, we do a pretty bad job of taking the small pieces and creating larger pieces of knowledge. We don't do much with our archives. There have been discussions about the living news page where you put together the many stories you've written and you kind of convert them into a form of Wikipedia.
I think that in fact, even though I've been pushing this as a local option, the model may really come from a Factcheck.org or one of the national fact-checkers saying: ‘OK we’re going to organize all of our knowledge around the healthcare system.’ It might take someone at Factcheck.org a week to do a couple of these and then we have a pretty nice model without much work.
The goal for me is to do two things for fact-checking.
First: how do we do it so that fact-checking continues after the election is over and when the real business of governing begins? Making it issue-oriented rather than campaign-claim oriented seemed like an obvious step towards that.
The second motivation is to make people more open to fact checks that challenge either what they thought or the point of view that their party was articulating.