Should social media dampen uncertain stories?

27 October 2020

In the last weeks of the 2020 presidential campaign, the New York Post broke a story alleging corrupt behavior by the family of Joe Biden, a candidate for President. The story wasn't confirmed by other media organizations. Supporters of Donald Trump acted to spread the story on social media, but both Twitter and Facebook took unprecedented efforts to block the story. That action to block the story became a story in itself, and there's been much discussion about whether the social media giants should block a story like this. Reading this discussion I think there's an important nuance that's been missed, one that applies in general to cases like this. [1]

In this kind of situation media organizations act as an amplifier, deciding to what degree a story should reach more people. Newspapers make this decision a human one, the editors decide what readers should pay attention to. In this case, most of them decided that this story was too likely to be false, and thus shouldn't get attention - at least without further investigation. For a social media organization, however, the notion is that this human editing role should be much smaller. They do edit material - violent images, child abuse, terrorist organizations and the like do get blocked. But how should they react to an unverified story with strong partisan political overtones, close to an election?

Much of what I've read on this is about whether they should block it or let it flow. I think there is another option to consider, which is dampen its flow by blocking such a story temporarily. The idea here is to consider the old saying that "a lie can travel halfway around the world before the truth can get its boots on". Sensational stories travel fast, even before the internet, and sensational stories often require time for proper investigation before we can have a decent idea of where the truth may lie. In cases like this it makes sense for social media to look at the provenance of the story. Instead of asking "is this true", they ask "can we tell yet if this is a damaging lie?". They can look at how media organizations other than the original breakers of the story are reacting - are they confirming it yet, or waiting for more information. If the picture is sufficiently uncertain, then there is justification for blocking it provisionally. Part of the decision here is considering the short and long term impacts of the story. A story breaking just before an election, that is likely to change votes, is one to be careful of because any refutation could easily come too late. Similarly social media should be wary of inflaming raw emotions that could lead to violence.

Recognizing uncertainty is a key element of this approach. People are often troubled by probabilistic reasoning. We see this in election forecasts, modeling the spread of covid-19, and the planning of software projects. There's a desire for certainty which is really a failure to face up to the reality that there is so much we don't know. If there's a sensational story, one whose veracity is uncertain - it's perfectly reasonable to wait for more information before giving it the powerful amplification that a popular social network delivers.


1: With this post, I'm talking more about the general principle of dealing with news stories that have some serious doubt to their veracity, and don't want to get into the particular details of this individual case. But for future reference, here's a few links that discuss the background. The Washington Post did a general explainer on the background to all this. The New York Times reported that the newsroom of the Wall Street Journal passed on the story and that the New York Post reporter didn't want to have their name on their story. The Guardian summarized the overall social media reaction