Judges should judge carefully, reliably and fairly.
At least that is what we expect from the case law.
But in the United States, judges punish African-American defendants more severely than average, according to large-scale investigations.
The supposedly impartial are apparently subject to a "bias", a distortion of judgment.
Such prejudices, whether consciously or unconsciously, are not the only source of error in judgment.
Because regardless of them, the sentence varies considerably.
In one investigation, for example, the prison terms for extortion ranged from three to twenty years, even for comparable offenses.
Things don't look any better with asylum applications.
Some judges grant 88 percent of the applications, others grant residence status in only five percent of the cases.
Our mind is a measuring instrument
The judgments of the American judges are not only distorted, but they are also "noisy".
They show a high degree of variation.
This noise is the title and subject of Daniel Kahneman's new book, which the American psychologist and Nobel Prize winner in economics wrote together with the lawyer Cass Sunstein and the former McKinsey partner Olivier Sibony.
Human judgments are almost always noisy.
Regardless of which industry you look at, be it medicine, training or social work, people always have to judge others: determine care levels, make recommendations for secondary schools, clarify questions of custody.
The judgments can make successful careers possible or close life paths forever.
Most people realize that prejudice leads to distortion of judgment.
But very few people are aware that noise can be a much stronger factor, because the dispersion of the judgments remains invisible.
Dramatic consequences in large organizations
Our mind, as Kahneman and colleagues explain, is a measuring instrument with which we rate things and events on a scale. In our metric era, we judge everything: people, movies, books, and the phone provider's customer service; in everyday life with asterisks and likes, at work with reports and job references. HR experts predict how successful an applicant will be in their new job, and doctors assess how likely it is that a tumor will develop. Ideally, everyone would always judge precisely and consistently. But the practice is different.
This can be illustrated well with a stopwatch on the smartphone. If you try ten times in a row to stop at exactly ten seconds with your eyes closed, you will notice that the results fluctuate. If all values are around eleven seconds, the internal clock has a bias. If, on the other hand, the results are spread between eight and twelve seconds, it is noise. Those who can reduce their bias will achieve better results, but also with a noise reduction. According to Kahneman and colleagues, this simple example can be applied to many decisions in professional life.