A village charges a boy with watching the flock and raising the alarm if wolves show up. The boy decides to have a little fun and shout out false alarms, much to the chagrin of the villagers. Then an actual wolf shows up, the boy shouts his warning, but the villagers are proper Bayesians who, having learned from their mistakes, ignore the boy. The wolves have a field day, eating the flock, the boy, and his entire village.

I may have augmented Aesop’s classic fable with that last bit.

The boy is certainly a crushing failure at his job, but here’s the thing: the village is equally foolish, if not more so. The boy revealed his type, he’s bad at his job, but the village failed to react accordingly. They updated their beliefs but not their institutions. “We were good Bayesians” will look great on their tombstones.

They had three options.

A) Update their belief about the boy and ignore him.

This is what they did and look where that got them. Nine out of ten wolves agree that Good Bayesians are nutritious and delicious.

B) Update their beliefs about the boy, but continue to check on the flock when the boy raises the alarm.

They should have weighted their responses. Much like Pascal taking religion seriously because eternal torment was such a big punishment, you have to weight you expected probability of truth in the alarm against the scale of the downside if it is true. You can’t risk being wrong when it comes to existential threats.

C) Update their beliefs about the boy and immediately replace him with someone more reliable.

It’s all fine and good to be right about the boy being a lying jerk but that doesn’t fix your problem. You need to replace him with someone who can reliably do the job.

So this is a post about fascism. Some think that fascism is already here, others dismiss this as alarmism, others splititng the difference claiming that we are in some state of semi- or quasi-fascism. Within the claims that it is all alarmism, what I hear are the echoes of villagers annoyed by 50 years of claims that conservative politics were riddled with fascism, that Republicans were fascists, that everything they didn’t like was neoliberalism, fascism, or neoliberal fascism. Get called a wolf enough times and you might stop believing that wolves even exist.

Even if I am sympathetic, that doesn’t get you off the hook. It hasn’t been fascism for 50 years will look pretty on your tombstone.

Let’s return to our options

• A) Don’t believe the people who have been shouting about fascism for years, but take seriously new voices raising the alarm.
• B) Find a set of people who, exogenous to current events, you would and do trust and take their warnings seriously.
• C) Don’t believe anyone who shouts fascism, because shouting fascism is itself evidence they are non-serious people.
• D) Start monitoring the world yourself

Both A) and B) are sensible choices! If you’ve Bayesian updated yourself into not trusting claims of fascism from wide swaths of the commentariat, political leaders, and broader public, that’s fine, but you’ve got to find someone you trust. And if that leads you to a null set, then D) you’re going to have to do it yourself. Good luck with that. It takes a lot of time, expertise, and discipline not to end up the fascism-equivalent of an anti-vaxxer who “did their own research.”

Because let me tell you, C) is the route to perdition in all things Bayesian. Once your beliefs are mired in a recursive loop of confirmation bias, it’s all downhill. Every day will be just a little dumber than the one before. And that’s the real Orwellian curse of fascism.