Whether you are a judge, doctor or executive of a large insurance company, you surely know that the decisions of the people around you, and your own, are in many cases biased, that is, constantly deviate from their goal and always in the same way. And you will know this because over the past few years psychology, economics and, of course, the publishing industry have painstakingly pounded with their warnings about biases, about our statistical inability, for example. But it is likely that he is not so aware of another failure of human judgment, surely more sibylline but perhaps for that reason also more dangerous: noise. And what is noise? Noise is the unwanted variability, that people who would have to make similar decisions, such as what is the specific disease that you suffer or the penalty that the crime you have committed deserves, however make catastrophically different decisions.
To draw attention to noise, explain what it consists of, make the problem an object of urgent public debate, and propose mechanisms to mitigate it, three of the world’s leading specialists on these issues have just published ‘Noise: a flaw in human judgment’ (Debate). They are the Nobel Prize in Economics Daniel Kahneman, the consultant specialized in strategic decision making Olivier Sibony and constitutional law expert and advisor to the Biden Administration Cass R. Sunstein. We talk precisely here with the latter about noise, bad decisions and the danger of being judged by a starving judge.
QUESTION. In recent times, cognitive science has repeatedly warned about the importance of bias in our decisions, but has dismissed noise. Why?
ANSWER: The bias has a kind of charm, even charisma. It attracts the human mind. In a way, it’s fun. If we discover biases, it is as if we have solved a mystery and found the killer. Rather, the noise is like the character in a movie that you never pay attention to, and who happens to be the villain. Noise is unwanted variability, and that seems a bit technical, but it can be really bad, a source of unevenness and terrible mistakes.
P. Public and private institutions are tremendously noisy. Is the noise really a bug or is it somehow part of human nature shaped by evolution?
A. If it results in unwanted variability, if one doctor says “I probably have cancer” and another says “I probably have indigestion”, someone is really making a mistake! If one official gives you disability benefits and another doesn’t, something has gone wrong. We can blame evolution for many things (anger, for example), and perhaps it is also responsible for the noise. But even so, the noise is a mistake.
People should not depend on some kind of lottery, such as the assignment of a particular judge
Q. If a judge judges me just before I eat, hungry, I have a better chance of being convicted. And something like that in all kinds of areas. Believing in the neutrality and justice of institutions without trying to understand the human brain, is it the great liberal self-deception?
R. You can also see it from another point of view: the aspiration to justice is honorable, even essential. Let’s keep trying! Noise can dampen that aspiration, so let’s try to reduce it. People should not make their lives depend on some kind of lottery, such as assigning a particular judge to their case. We have a variety of proposals to reduce noise and they could help reduce injustice.
Q. I am very interested in predictive judgment. We human beings love to predict the future, we know that we are never right, but we do not stop trying. Is there a way to better predict or we better forget it?
A. Sometimes our predictions are pretty good. I predict that Spain and the United States will continue to exist tomorrow. Weather forecasts are usually reliable. I predict that if I play tennis tomorrow against Rafael Nadal, he will win. There are many better ways to make predictions; some people are ‘super forecasters’ and we know why. Everyone could use their approach and make better predictions. Still, it is true that some things are impossible to predict because they depend on many variables that cannot be identified in advance. Like who will be the best tennis player in the world in 2040.
P. They expose an important problem. If we want to reduce noise with quantitative measures, we run the risk of arousing the rejection of human beings who do not like to feel like cogs. Is it possible that an erroneous but cohesive response that builds community rather than undermines it is sometimes better?
A. Yes, that is true. However, we need to know the number and size of the errors. A happy company that keeps losing money won’t last long.
Q. Is the so-called political polarization a form of noise? Why has this polarization grown so much in the US and in the rest of the world and what can be done to combat it?
A. A great question! Polarization in political lines is not noise, although noise can cause polarization. The causes of polarization deserve a complete book; social networks contribute to this. One of the central ideas of our book, ‘decision hygiene’, would greatly reduce polarization and noise. But I won’t spoil the surprise here!
One of the central ideas of our book, ‘decision hygiene’, would greatly reduce polarization and noise
Q. You are part of the Biden Administration. How noisy is it and what actions seem most appropriate to reduce noise in public management?
A. I speak here as the author of a book (completed before I joined the Government), not as a civil servant, and I will simply say that it is a great honor for me to work in the Biden Administration.
Q. You are an expert in institutional law. In recent years, numerous voices in the US have called for a reform of the Constitutional Court and, in general, of the general judicial structure of the US. Would it be good to avoid the court being dominated by a specific party (in this case, the Republican) or is it part of the rules of the game?
A. Diversity in court is a good idea, as American history tends to show; the legal system benefits greatly from a variety of perspectives and approaches.