Corona forecasts are far apart - and often wrong in the end.

Before the federal emergency brake, the pessimists had the upper hand.

They predicted a dramatic increase in deaths.

In fact, there were most corona deaths on January 22nd with 819 cases.

On April 21, the day before the brake came into force, there were 331. Of course: every death is one too many.

Rainer Hank

Freelance writer in the economy of the Frankfurter Allgemeine Sonntagszeitung.

  • Follow I follow

    But the declining death rate makes it difficult to make an argument for tightening the measures. In February 2021, Germany recorded an under-mortality rate, i.e. fewer deaths on balance than in the years 2016 to 2019. Academic heretics like the Mainz epidemiologist Sucharit Bhakdi have been annoying the mainstream with such comparisons for months. The worst weeks were the days around Christmas 2020. From then on we felt better. The emergency brake was pulled when the infection train was rolling in the right direction.

    Why do we make completely different decisions based on the same factual basis?

    Why do two experts who have identical information come to completely different conclusions?

    And why can even an expert change his mind at lightning speed, such as the federal chief virologist Christian Drosten, who, from pessimist to optimist, is now promising us a great summer?

    This is due to the many vaccinations, of which it was still said at the end of April that one should in no way hope for a rapid relaxation.

    Which distorts our decisions

    The new book by the psychologist and Nobel laureate in economics, Daniel Kahneman, which, written together with Olivier Sibony and Cass Sunstein, has just been published under the title “Noise”, offers exciting answers to our questions.

    For Kahneman, it's about "what distorts our decisions - and how we can improve them".

    Kahneman is a co-founder of behavioral economics.

    This stands for the realization that people behave far less rationally than they would like to do it themselves.

    Just for example: We postpone important decisions or tend to constantly overestimate ourselves.

    We can even give rational reasons for our failure.

    Kahneman calls this "bias," a consistent departure from a goal.

    “Bias” is common to our entire genre.

    With "Noise", Kahneman's current focus, no law can be seen behind the errors. The spread of hits is wide. They are “unpredictable errors”: our graphic shows the difference between “bias” and “noise” using the example of bullet holes on a target. "Bias" shows a regular deviation to the lower left. “Noise” is a chaotic distribution of failed hits; the result is "noisy" or "noisy". In the fourth diagram, the shooters systematically missed, but their hits are still widely spread ("bias" and "noise").

    "Bias" has long been known;

    “Noise” is overlooked, underestimated, denied.

    We pretend that all scientists have to agree in their judgments because they all interpret the same facts and derive instructions for action from them.

    Deviants are "canceled" and ultimately stigmatized.

    “The ignorance is denied,” writes Kahneman - because it would be a serious offense for the scientist.

    Independence is important

    People believe that their fellow human beings see the world the same way they do. Kahneman has many examples of the damage that “noise” can do. People who commit criminal offenses receive completely different sentences for exactly the same offense - one is imprisonment of five years, the other remains free on probation. One judge relies on rehabilitation and imposes a mild sentence, the other relies on deterrence and goes to the upper limit. Moods matter: On Monday after a local football team's defeat, judges tend to impose tougher sentences.

    In insurance companies, so-called underwriters calculate the premiums for the policies. You yourself think that a 10 percent deviation in the premium calculation is normal and tolerable. In reality, Kahneman comes up with differences of 55 percent. If one underwriter sets a premium of $ 9,500, the contract for the second is $ 16,700 (not just $ 10,500). Mind you: Both calculate on the basis of identical data and use the same risk models. Too low premiums lead to bankruptcy. If the premiums are too high, customers will migrate. Doctors' diagnoses differ from one another in a similarly wide spread. The therapies differ accordingly. It is a lottery that determines the fate of people.

    Back to Corona: The experts' deviations are characterized by both "bias" and "noise". Experts working directly or indirectly on behalf of the government tend to be optimistic or pessimistic ("bias") depending on the political goal. Medical professionals focus on health; Social scientists recommend broader cost-benefit considerations. This results in a "noise" spread.

    If you want to improve the world, you have to reduce "noise". Kahneman has no magic formula for this. But his recommendations are tough: not to deny “Noise” would be a lot. All judgments should be transparent in their full range independently of one another (“noise audit”). Independence is important. Because team building, which is very popular in Corona times, unfortunately increases the "noise" effect: pessimists and optimists radicalize each other. Intuition is just as dangerous as the dominance of domination knowledge. “Look at the case from an outside perspective,” advises Kahneman: “Try to think against yourself!” It's not easy, but it's worth it: Skepticism and modesty should be combined with one another.