Ever wondered why so few people saw the rise of Donald Trump, or Giorgia Meloni, or the vote for Brexit? People have an innate tendency to muddle up what they want to happen with what they expect to happen.

The German poet, Christian Morgenstern, who lived at the turn of the 19th and 20th centuries, wrote in his poem The Impossible Fact* about the final thoughts of a man after a fatal accident:

"And he comes to the conclusion:

His mishap was an illusion,

for, he reasons pointedly,

that which must not, can not be

Behavioural psychologists have a name for this: confirmation bias. We blend out information not consistent with our beliefs, and only take in what we agree with. Everybody does that. I do it too. Social networks are places where this is happening on steroids. They are bubbles of people who reinforce their own biases. 

An illuminating recent example was the response to Elon Musk's takeover of Twitter. A British newspaper reported a forecast from a network engineer who had just been fired. He predicted that the network would collapse within the next few hours. 

Some Twitter users, too, expressed concern that Musk would kill Twitter. So they migrated their account to Mastodon, a Twitter-like, open-source decentralised network. I personally think Mastodon is great, especially for scientific communities because it allows them to post type-set formulas. But it does not replace Twitter. Linux was also great. But it did not replace Windows.

It is interesting to study the examples of the engineer and the outraged Twitter user from the perspective of cognitive biases. 

Was the engineer really afraid that Twitter would collapse? Or was he afraid that Twitter could function without him? Remember the air traffic controllers who went on strike in the US in the early 1980s. Ronald Reagon dismissed all of them. They, too, thought they were indispensable.

The pathologies of the Twitter users are particularly insidious. Do they really fear that Twitter would collapse under Musk? Or rather, do they fear that it would not? I think most of those who panicked were outraged about Trump's reinstatement.

Twitter is, of course, not the most important example where cognitive biases cloud our judgement. Right now that would be the Russia-Ukraine war. 

I have heard a lot of armchair generals - literally generals who retired from active service not too long ago - who predicted at various stages of the war that Ukraine was only a few weeks away from victory. This is separate from the rational observation that Ukraine will eventually win the war if the west continues to supply weapons. The forecasting error about the timing has nothing to do with stupidity. Instead, it's about bias. They want Ukraine to win. We should also recognise that we have asymmetric information. We know what the Ukrainians are doing, what we are doing to help them, but we don't know nearly as much about the Russians. 

The western media highlight Russia's military failures, and report military successes as human rights violations. We keep telling ourselves the old adage that you cannot compensate for military failures on the battlefield by attacking civilian targets. Attacks on civilians constitute a war crime. They are morally repugnant. But they can be effective. The Dresden bombings in February 1945 were effective. 

I ask myself the following questions: if Putin expects he can cut off large parts of Ukraine from electricity during the entire winter, would that not impact Ukrainian battlefield troops? Don't they need electricity for communication, trains, and field hospitals? 

Our debate on economic sanctions followed a similar pattern. Remember the predictions that Russian GDP would contract by 30% after the west imposed sanctions? Russia's recession, measured in GDP, has turned out to be relatively mild. Russia is running a record current account surplus this year, of some $250bn. Now we are predicting that the $60 oil price cap would do him in. Really? Urals crude is trading at $65.

Biases are everywhere. Their list is long. Economists, sociologists and political scientists are particularly prone to model bias. No macroeconomist has ever dropped their allegiance to a school of thought, represented by a model, just because facts intruded. They tilt the facts to suit their model. This is why we have a divide in the debate about the future path of inflation between Team Transitory and Team Persistent.

Why would a rational person want to commit themselves to a team over a prediction? Would it not make more sense to keep an open mind and analyse data as they come in? Only a minority of economists and central bankers take such a dispassionate view. Most economists I know communicate like followers on social media. They form cliques, retweet each other's messages with approval, and express outrage at those they disagree with.

Here is my advice to anyone who engages in public debates and in particular policy analysis. Before you make your next prediction, check whether it accords with what you want to happen. If it does, stop right here, and engage in the following intellectual exercise: take the other side of the debate and try to demolish your own arguments. See how well you do. It is not a guarantee of success, but it may help you see your own cognitive biases in public discourse. It has helped me.

*The original title of Christian Morgenstern's poem is Die unmögliche Tatsache. The quote above is an English translation of the poem's last paragraph:

Und er kommt zu dem Ergebnis:

„Nur ein Traum war das Erlebnis.

Weil“, so schließt er messerscharf,

„nicht sein kann, was nicht sein darf.“