61 pages 2 hours read

Daniel Kahneman, Olivier Sibony, Cass R. Sunstein

Noise: A Flaw in Human Judgment

Nonfiction | Book | Adult | Published in 2021

A modern alternative to SparkNotes and CliffsNotes, SuperSummary offers high-quality Study Guides with detailed chapter summaries and analysis of major themes, characters, and more.

Summary and Study Guide

Overview

Noise: A Flaw in Human Judgment is a 2021 New York Times bestselling nonfiction book that explores the concept of noise or unwanted divergences in decision-making. It is researched and written by Israeli American Nobel Prize-winning psychologist and economist Daniel Kahneman, American legal scholar Cass R. Sunstein, and French business professor Olivier Sibony. While the authors point out that bias which causes people to make inconsistent discriminations is acknowledged by companies, the problem of noise remains largely invisible and unexamined.

New York Times reviewer Steve Brill has praised the work as a “welcome handbook for making life’s lottery a lot more coherent” (Brill, Steve. “For a Fairer World, It’s Necessary First to Cut Through the Noise.” The New York Times. 18 May 2021). He adds that the book is highly relevant to the lack of credibility in today’s institutions, writing that “we are living in a moment of rampant polarization and distrust in the fundamental institutions that underpin civil society. Eradicating the noise that leads to random, unfair decisions will help us regain trust in one another.” Thus, Brill believes that the book and the noise-reducing strategies it advocates are accessible and essential.

This study guide uses the Harper Collins May 2021 Ebook Edition.

Summary

The authors argue that noise, or unwanted divergence, is present wherever judgments are being made. Noise can take the form of two individuals in the same profession disagreeing about a proposed course of action. In the area of medical diagnoses, at least one of the disagreeing doctors is wrong, resulting in serious consequences for the patient. In the law, the years a defendant may spend in jail can depend on the severity of the judge rather than the crime.

The authors identify different types of noise, which include level noise, meaning a person’s consistent tendency to act in a certain way; occasion noise, where random external factors such as the weather and the performance of a local sports team could influence a judge’s decision; and pattern noise, where a particular judge has a tendency to act in a particular way when certain situations occur. For example, a typically harsh judge may deviate from their pattern and be especially lenient towards young women. The authors found that pattern noise was the most disruptive of the three types, and that it can only be measured statistically through the examination of multiple cases.

The authors examine different industries including law, medicine, and forensic science, in addition to common situations such as hiring practices and performance evaluations. They look at the typical prevalence of noise in each instance and discuss proven strategies for noise reduction. They find that even where an outcome cannot be easily predicted—for example, how well a candidate will perform in a new role—the process of gathering evidence matters and can reduce error and disagreement. Consistently, the authors found that a balance of being open to new information while not being distracted by seemingly related but irrelevant material was an important criterion in noise-reduction, They also recommend tackling different parts of a decision individually and aggregating the independent opinions of staff. In contrast to business leaders who claimed that intuition was their primary resource, the authors recommend delaying the use of intuition until all other parts of the judging process are complete. This is because intuition is prone to being influenced by noise.

The authors recognize the problems of implementing noise-reduction strategies: They can be time consuming and expensive; they remove initiative for employees and so make their jobs uninspiring and tedious; and worst of all, algorithms used to replace human judgment can augment the worst of human biases against already marginalized groups. However, the authors argue that when noise-reducing strategies fail, it is often because they are not complex enough to suit the requirements of the decision to be made. They add that judgment should be more about accuracy than personal expression and that failure to be precise can contribute to inequities in many areas of life.