It is difficult to criticize fact-checking, because the term itself is so positively loaded. But here goes: The declining influence of designated fact checkers on social media is a good, and necessary correction. This is not a critique of the idea, but the lion’s share of current fact-checking practice. What began as an noble effort to combat disinformation gradually expanded into fact checkers arbitrating complex debates and matters of opinion – territory that belongs to traditional journalism. At Open Mind, we believe that the vast gray area between true and false requires the nuanced, detailed and continuous analysis and judgement of professional journalists and editors, not the verdicts of fact-checking entities. Only through this open and ongoing process and exchange of opinions, can we have a well informed and critically thinking society.
In the past decade-and-a-half, fact-checking organizations have proliferated, gaining influence as they became instrumental in social media content moderation. However, this era may be drawing to a close. Elon Musk’s purchase of X (formerly Twitter) brought a more permissive approach to moderation, and Mark Zuckerberg recently announced that Meta will discontinue their work with independent fact checkers. While some lament that “facts are out of fashion!” I remain optimistic, because today’s fact-checking practices have evolved in problematic ways that undermine their intended purpose.
The term “independent fact checkers” is itself a misnomer for two reasons:
- Independence: While many fact-checkers claim independence and adhere to ethical guidelines, their funding typically comes from two main sources: direct ownership by media outlets or compensation from tech companies such as Meta and Google. This raises legitimate concerns about both inherited editorial bias and skewed monetary incentives from the tech companies’ compensation models, which are often very opaque. Indeed, a Norwegian fact checker Faktisk.no was claimed to be paid per fact check but was revealed by the newspaper Minerva to receive compensation from Meta only when they flagged articles as false. This led to the fact checker rabidly flagging everything they could find, including opinion pieces. Reducing the reach of serious newspapers, like Nettavisen without disclosing this information to the newspapers.
- The term “fact” implies a binary – something is either true or false: Washington DC is the capital of the US, 1+1=2, etc. However, very few fact checkers operate within this binary understanding of facts and instead rate statements on a scale, including loaded terms like “mostly false”, “mostly true” or “lacking context”.
While nuance is important in verifying claims, an undesirable side effect of “fact checking” is that it allows room for interpretation. And interpretation is at odds with the binary term “fact” and opens up possibilities for both editorial and personal bias in the field of fact-checking.
Fact checkers lose credibility when they attempt to fact-check opinions rather than limiting themselves to verifying the facts underpinning those opinions. By nature, opinions are interpretations of data, not statements of fact, and evaluating them is inherently subjective. A notable example of this overreach occurred when The Washington Post fact-checked Bernie Sanders’ interpretation of a Mercatus Center working paper. The paper estimated that Medicare-for-All could save over $2 trillion over a decade under specific assumptions. Sanders accurately cited this figure, but the fact-check criticized his interpretation of the data rather than addressing the validity of the underlying facts. Assigning Sanders three “Pinocchios” for his opinion not only overstepped the role of fact-checkers but also blurred the line between objective fact-checking and subjective critique. Fact-checkers should focus exclusively on verifying the data itself, leaving interpretations to public debate.
This leaves the idea of fact checking in a bit of a conundrum. While a binary approach limits what can be fact checked, if you engage with the 80% in between true and false, you engage in opinion exchange. Consequently “fact-checkers” might be poorly suited to the task.
For fact checking to be effective, these organizations need broad public trust – something they currently lack. According to a 2019 Pew Research survey, Americans are deeply divided in their trust of fact checkers, with the split falling sharply along political lines: 70% of Republicans viewed fact checkers as biased and unfair, while nearly the same proportion of Democrats (69%) considered them fair and unbiased. This mirror-image pattern raises two possible interpretations. One could conclude that conservative politicians make more false statements, and their supporters resist being confronted with these facts. Alternatively, fact checkers themselves might harbor biases, particularly when evaluating complex issues that fall into the gray area between true and false. While both explanations have been proposed, verifying either creates a circular problem – who fact-checks the fact checkers?
So what does the way forward look like? X has implemented “community notes,” where registered users provide contextual information and vote on the best context to be displayed. Meta has suggested a similar system for their platforms. These community-driven approaches, reminiscent of successful collaborative platforms like Stack Overflow, show how efficient a community voting system can be.
The fact checkers that survive will likely be those that focus exclusively on verifiable binary facts, having the discipline to abstain from judgment in the vast majority of cases where absolute truth is elusive. This restraint may help them rebuild trust across the political spectrum, except among the most extreme.
As for the gray area of complex issues and interpretations, this is where journalism must reassert itself. The cure for bad information isn’t fact-checking alone – it’s better journalism, rigorous analysis, and open dialogue. While Large Language Models and AI tools can help process information at scale, they remain black boxes with embedded biases we don’t fully understand. This technological reality underscores why Open Mind champions the role of keen-eyed journalists who can discover biases and falsehoods wherever they appear, while remaining accountable to their readership.
Ultimately, we believe the path forward lies in combining technology with human judgment: community-driven systems for context and fact-checking for binary truths, all undergirded by professional journalism that can navigate the nuanced territory in between. Through this open and ongoing process of analysis and exchange, we can foster a well-informed and critically thinking public.
P.S. I welcome any and all fact checking of this opinion.
Sources:
https://www.bbc.com/news/articles/cly74mpy8klo
https://theweek.com/articles/791236/fact-checkers-have-medicareforall-problem
https://www.theatlantic.com/podcasts/archive/2024/09/fact-check-polarization-opinion-politics/679898