In a 2016 study entitled What Aren’t They Telling Us?, Chapman University surveyed American citizens on their belief in various conspiracy theories. They found that 30% of Americans believe in a government cover-up surrounding Obama’s birth certificate, 24% in a cover-up of the moon landing, and 33% in a cover-up of the North Dakota crash.
Except, there is no such thing as the “North Dakota crash”. It was invented by the Chapman university surveyors as a control, and yet ranked sixth out of the ten conspiracies shown to respondents. In a post on Slate Star Codex, Scott Alexander posited an explanation for such a high agreement rate:
I think this is probably a story about low-information voters. […] If your prior is “most conspiracy theory-ish things are probably true”, this sounds like the kind of thing that could be true.
In other words, people who believe in other conspiracy theories will be inclined to agree that there is a cover-up, having no information and a prior inclination towards governmental distrust. This is a perfectly reasonable explanation – after all, if most of us were confronted with the question “Agree or disagree: Fox News has lied about the North Dakota crash”, we would tend to agree, because we’ve developed rational priors to implicitly distrust Fox News. But Scott Alexander then goes on to analyze other statistics in the context of this “North Dakota Constant”:
Who believes Obama was secretly born in Kenya? Lots of people – including 28% of blacks. […] I have no idea what these people are thinking, except that 28% is pretty close to the North Dakota Constant and maybe we should just write this one off.
Scott sees the North Dakota Constant as proof that fake news is not as effective as we might assume from seeing statistics like “46% of Trump voters believe in Pizzagate!”. Instead, these statistics mostly show that a lot of people inherently tend to distrust the government and don’t have much more information on each specific topic.
But what Alexander fails to consider is that this low-information system is the model of modern American political discourse. With social media and modern technology, information availability is at an all-time high: but as the number of sources available to voters has grown, so has the total number of political issues. And it is the contemporary standard that once you have chosen to be politically active, you must have a stance on every single one of them. You can’t be a die-hard liberal without an opinion on abortion, nor a conservative who doesn’t care for economics. Of course, in reality it is nearly impossible to know enough about these issues to actually come to well-reasoned opinions on each of them. It therefore becomes necessary to adopt political stances not on a case-by-case basis, but in sweeping generalities – “I agree with liberal ideas”, “I believe in conspiracy theories.” It’s politics through priors, plain and simple. And as a result, the average voter on the average issue is passionate, but ill-informed.
This is how fake news works, and why it is indeed so effective. Pizzagate didn’t need any evidence, it just needed headlines. And even if they clicked on the article, the majority of people who already had the conspiracy-prior were convinced by the time they read the headline. The North Dakota crash is the perfect example of this: 33% of Americans didn’t need evidence, because indeed there is none – they just needed the headline to align with their priors, and they were convinced. Furthermore, they probably had no idea that their own prejudice was the primary reason for their agreement. In the comments of the Chapman University post, we can still see people actively seeking out and forming conspiracy theories around an obscure 2013 oil tanker crash in North Dakota. These people are falling for the exact trap that the North Dakota Constant exposes: they saw “North Dakota crash” on a list of conspiracy theories, chose to believe it, and yet still will double down after being told not only that there is no evidence for it, but it doesn’t even exist.
For this reason, I see the “North Dakota Constant” not as a harmless statistical baseline, but as perhaps the most powerful quantitative metric of the downfall of rational political discourse in this country. The higher the constant, the larger the proportion of the population that is willing to cast aside evidence in favor of their own prejudice. To a certain extent, this is human nature – but only recently have we seen politics overtaken by it, and entire creeds formed on unfounded hate, conspiracy, and belief in the inexplicable presence of evidence and studies supporting their views despite an actual lack thereof. As long as the North Dakota Constant still stands at a non-negligible percentage of our population, political discourse will continue to be steered by irrational and ill-informed opinions, and people will continue to be polarized across the lines of pre-existing bias. I say, the sooner we get rid of the North Dakota Constant, the better.