Daniel Kahneman – Noise: Book Review & Audio Summary

Book Info

• Book name: Noise: A Flaw in Human Judgment
• Author: Daniel Kahneman, Olivier Sibony, Cass R. Sunstein
• Genre: Business & Economics, Social Sciences & Humanities
• Pages: 416
• Published Year: 2021
• Publisher: Little, Brown and Company
• Language: English
• Awards: Winner of the 2022 National Academy of Sciences Best Book Award

Audio Summary

Synopsis

In “Noise,” Nobel Prize winner Daniel Kahneman and co-authors Olivier Sibony and Cass R. Sunstein expose a hidden flaw in human judgment that’s been hiding in plain sight. While we’re familiar with bias—systematic errors that skew our thinking in predictable directions—noise represents the random variability that makes identical situations produce wildly different outcomes. From judges handing down inconsistent sentences to doctors making contradictory diagnoses, this groundbreaking book reveals how noise corrupts decisions in medicine, law, economic forecasting, hiring, and beyond. Drawing on decades of research, the authors demonstrate that reducing noise could dramatically improve the quality of our judgments and offer practical strategies for creating more consistent, fair decision-making systems.

Key Takeaways

• Noise is the unwanted variability in judgments that should ideally be identical, distinct from bias which represents systematic errors in a consistent direction
• Occasion noise occurs when the same person makes different decisions in similar situations due to irrelevant factors like weather, mood, or hunger
• System noise reveals itself when different professionals within the same organization reach vastly different conclusions on identical cases
• Reducing noise through structured decision-making processes, guidelines, and aggregating multiple judgments can significantly improve outcomes in critical fields
• Most organizations dramatically underestimate the amount of noise in their systems, often assuming much greater consistency than actually exists

My Summary

Why We’ve Been Obsessing Over the Wrong Problem

I’ll be honest—when I first picked up “Noise,” I thought I was in for another rehash of Daniel Kahneman’s brilliant work on cognitive biases from “Thinking, Fast and Slow.” Boy, was I wrong. This book opened my eyes to something I’d never really considered: that random variability in judgment might be just as problematic as systematic bias, if not more so.

Here’s what struck me immediately. We’ve spent decades, maybe centuries, worrying about bias. Is this judge too harsh? Is that admissions officer prejudiced against certain groups? These are crucial questions, don’t get me wrong. But Kahneman and his co-authors—Olivier Sibony and Cass R. Sunstein—make a compelling case that we’ve been so focused on bias that we’ve completely overlooked its equally destructive sibling: noise.

Think about it this way. If you’re aiming at a target and all your shots land consistently to the left, that’s bias. You can correct for it once you recognize the pattern. But if your shots scatter randomly all over the place with no discernible pattern? That’s noise, and it’s much harder to fix because there’s no consistent error to correct.

The Invisible Problem Hiding in Plain Sight

What makes noise so insidious is that it’s largely invisible until you specifically look for it. The authors describe a fascinating phenomenon they call “noise blindness”—our tendency to assume that professionals in the same field, using the same information, will reach similar conclusions.

Spoiler alert: they don’t.

The book opens with examples that honestly made my jaw drop. In one study, when insurance companies asked their underwriters to estimate the appropriate premium for the same cases, the judgments varied by a factor of three. Not 3%, mind you—three times. In another study, experienced forensic psychiatrists were asked to evaluate the same offender’s risk of violence. Their estimates ranged from 1% to 50%.

As someone who’s worked in publishing for years, this hit close to home. I remember attending editorial meetings where we’d discuss manuscript submissions. The same manuscript that one editor called “brilliant and groundbreaking” another would dismiss as “derivative and poorly executed.” We always chalked this up to “different tastes” or “editorial vision,” but reading “Noise” made me wonder: how much of that was just… noise?

Understanding the Anatomy of Noise

Kahneman and his team don’t just identify the problem—they dissect it with surgical precision. They break noise down into distinct categories, each with its own characteristics and implications.

Occasion noise is perhaps the most unsettling because it reveals that we’re not even consistent with ourselves. The same judge might sentence one defendant to three years in the morning and another defendant who committed an identical crime to five years in the afternoon. Why? Maybe they’re hungry (the famous “hungry judge effect” is real and documented). Maybe it’s hotter in the afternoon. Maybe their favorite team lost over the weekend.

The weather example from the book particularly fascinated me. Research shows that college admissions officers pay more attention to academic credentials on cloudy days and give more weight to non-academic factors on sunny days. I mean, can you imagine? Your entire future potentially hinging on whether the sun was shining during your interview?

System noise operates at a broader level, revealing variability across an entire organization or profession. This is where things get really troubling. The book cites the case of two immigration judges in Miami where the chance of winning asylum varied from 5% to 88% depending on which judge heard your case. These aren’t small differences—they’re life-altering disparities based essentially on the luck of the draw.

The Psychology Behind Random Errors

What I found particularly valuable was the authors’ exploration of why noise exists in the first place. It’s not just random chaos—there are psychological mechanisms at work.

One major contributor is what they call “matching”—our tendency to match the intensity of our response to the intensity of the stimulus as we perceive it. But here’s the catch: we all perceive intensity differently. What seems like a moderate infraction to one person might seem severe to another, not because of bias necessarily, but because we’re all calibrated differently.

The book also delves into the role of substitution, where we unconsciously answer an easier question than the one we’re actually facing. A loan officer might be asked “What’s the probability this person will default?” but unconsciously answer “Do I like this person?” instead. This substitution happens differently for different people and even differently for the same person at different times, creating noise.

Reading this section, I couldn’t help but think about all the hiring decisions I’ve been part of over the years. How many times did I think I was evaluating someone’s qualifications when I was really just reacting to their likability or confidence? And how much did my evaluation change based on whether I’d had my morning coffee or was dealing with a headache?

Why Organizations Vastly Underestimate Their Noise Problem

One of the most eye-opening sections of “Noise” describes what the authors call “noise audits”—systematic examinations of variability within organizations. Time and again, when they conducted these audits, executives were shocked by the results.

In one memorable example, a large insurance company was confident that their underwriters would agree within about 10% on policy pricing. The actual variability? Around 55%. The executives literally didn’t believe the results at first and insisted the study must be flawed.

This pattern of underestimation happens for a clear reason: we rarely get direct feedback about noise. If you’re a judge sentencing criminals, you don’t get to see what sentence your colleague down the hall gave for a similar case. If you’re a doctor making a diagnosis, you usually don’t know what diagnosis three other doctors would have given the same patient.

We operate in what the authors call “parallel universes,” making our judgments in isolation and assuming that other qualified professionals would reach similar conclusions. But they wouldn’t, and that’s the problem.

Making Better Decisions in Your Daily Life

Now, you might be thinking, “This is all very interesting, but I’m not a judge or a doctor or an insurance underwriter. How does this apply to me?” Trust me, I had the same question, and the book delivers some genuinely practical insights.

First, recognize that you’re noisier than you think. We all like to believe we’re consistent in our judgments, but we’re not. I’ve started noticing this in my own life. My enthusiasm for a book pitch varies depending on what time of day I read it, whether I’ve eaten recently, and even what I read immediately before it. Acknowledging this variability is the first step toward managing it.

Second, use decision hygiene. This is the authors’ term for practices that reduce noise without requiring you to identify its specific sources. One simple technique: delay your intuitive judgment. When I’m evaluating something important now, I try to first break down the decision into specific criteria and evaluate each one independently before forming an overall impression. This prevents my immediate gut reaction from contaminating my assessment of individual factors.

Third, aggregate multiple judgments when possible. The book shows convincingly that averaging the independent judgments of several people almost always produces better results than relying on a single expert, even if that expert has more experience than the others. In my own work, I’ve started seeking out multiple perspectives on important decisions rather than trusting my own judgment alone.

Fourth, structure your decision-making process. The more structure you impose, the less room there is for noise. When I’m reviewing book proposals now, I use a consistent checklist of criteria and force myself to evaluate each one before moving to the next. It’s more tedious than going with my gut, but the results are more consistent and, I believe, more accurate.

Fifth, be aware of irrelevant factors. Since reading “Noise,” I’ve become almost paranoid about the potential influence of my mood, the weather, my hunger level, and other factors that shouldn’t matter but do. I’m not saying I can eliminate their influence entirely, but awareness helps. If I’m feeling irritable, I try to delay important judgments. If I’m unusually euphoric, I’m more skeptical of my positive assessments.

The Tension Between Noise Reduction and Other Values

To their credit, Kahneman and his co-authors don’t pretend that reducing noise is always the highest priority or that it comes without costs. They acknowledge several important tensions.

First, there’s the trade-off between noise reduction and flexibility. Rules and algorithms reduce noise, but they also reduce the ability to account for unique circumstances. A judge who must follow strict sentencing guidelines will be less noisy, but also less able to show mercy in exceptional cases.

Second, there’s the question of expertise and judgment. Professionals often resist noise-reduction measures because they feel these measures don’t respect their expertise. A doctor might resent being asked to follow a checklist, feeling it reduces them to a technician. The authors argue that this resistance is often misplaced—that expertise and structure can coexist—but they acknowledge the psychological reality of the resistance.

Third, there’s the simple reality that reducing noise takes effort. It requires organizations to implement new procedures, train people differently, and often invest in technology. In a world of limited resources, noise reduction competes with other priorities.

I appreciated this nuanced discussion because it prevents the book from becoming a simplistic manifesto. Yes, noise is a problem, but like most problems, addressing it involves trade-offs and judgment calls.

How This Book Compares to Kahneman’s Earlier Work

Having read “Thinking, Fast and Slow” multiple times, I was curious how “Noise” would compare. They’re complementary rather than redundant, which is exactly what you’d hope for.

“Thinking, Fast and Slow” is primarily about the architecture of the mind—how our two systems of thinking (fast, intuitive System 1 and slow, deliberate System 2) produce systematic errors. It’s about the machinery of thought and the predictable ways it malfunctions.

“Noise,” by contrast, is about variability in outcomes. It’s less concerned with why individual minds make errors and more focused on why different people (or the same person at different times) reach different conclusions when they shouldn’t.

If I had to choose which book to read first, I’d probably still recommend “Thinking, Fast and Slow” because it provides essential background on cognitive psychology. But “Noise” addresses a genuinely different problem and offers distinct solutions. Together, they provide a comprehensive picture of how human judgment goes wrong.

Another comparison worth making is to Cass Sunstein’s other work, particularly “Nudge” (co-authored with Richard Thaler). While “Nudge” is about designing choice architecture to guide people toward better decisions, “Noise” is about making professional judgments more consistent. The former is about helping laypeople make better choices; the latter is about helping experts make more reliable judgments.

Where the Book Could Have Gone Further

As much as I appreciated “Noise,” I did find myself wishing the authors had explored certain areas more deeply.

First, the book focuses heavily on professional judgment in organizational contexts—judges, doctors, corporate executives. I would have loved more discussion of noise in personal decision-making. How much noise is there in our choices of romantic partners, careers, or places to live? The book touches on these questions but doesn’t fully explore them.

Second, while the authors discuss various noise-reduction strategies, they don’t provide much guidance on how to choose among them. Different strategies have different costs and benefits, and I would have appreciated more practical advice on matching strategies to situations.

Third, the book is quite dense and technical in places. Kahneman and his co-authors clearly did their homework—the research base is impressive—but at times I felt like I was reading an extended academic paper rather than a book for general readers. Some sections could have benefited from more examples and less statistical detail.

Finally, I wished for more discussion of the political and social implications of noise reduction. When you start talking about reducing judicial discretion or implementing algorithmic decision-making, you’re wading into controversial territory. The authors acknowledge these issues but don’t fully grapple with them.

The Broader Implications for Society

Despite these limitations, “Noise” raises profound questions about fairness, expertise, and the nature of professional judgment.

Consider the justice system. We like to think that “justice is blind”—that similar cases receive similar treatment. But if noise is as pervasive as this book suggests, then justice is actually quite arbitrary. Your sentence might depend on whether the judge is hungry, what case came before yours, or whether it’s a sunny day. Is that acceptable? Most people would say no, but then we have to grapple with the solutions, which might involve reducing judicial discretion in ways that make us uncomfortable.

Or consider medicine. If different doctors reach wildly different conclusions about the same patient, that’s a serious problem. But the solution—more protocols, more guidelines, more algorithmic decision-making—threatens the traditional model of physician autonomy and expertise. How do we balance these concerns?

These questions don’t have easy answers, and to their credit, the authors don’t pretend they do. But they’ve done us a service by forcing us to confront them.

Putting Noise Reduction Into Practice

The most actionable part of “Noise” is the authors’ discussion of practical noise-reduction strategies. Let me highlight a few that I’ve found particularly useful.

The mediating assessments protocol (MAP) is a structured approach to complex judgments. Instead of making one holistic judgment, you break the decision into multiple independent assessments, evaluate each one separately, and only then combine them into a final judgment. I’ve started using this for manuscript evaluations, and it’s made a real difference.

Relative judgments reduce noise by asking people to compare options rather than evaluate them in isolation. Instead of asking “How good is this candidate?” you ask “Is this candidate better than the last three we hired?” This simple shift can significantly reduce variability.

Aggregating judgments is perhaps the most powerful noise-reduction strategy. Get multiple independent assessments and average them. The book presents compelling evidence that this “wisdom of crowds” approach consistently outperforms individual expert judgment, even when the experts are more experienced than the crowd members.

Using guidelines and checklists doesn’t eliminate judgment, but it structures it in ways that reduce noise. The key is to make guidelines specific enough to reduce variability but flexible enough to accommodate genuine differences in situations.

Questions Worth Pondering

Reading “Noise” left me with several questions I’m still mulling over. How much consistency should we actually want in human judgment? Is some noise actually valuable, representing appropriate responses to subtle differences we can’t easily articulate? When we reduce noise, do we risk losing something important about human judgment and expertise?

I’m also curious about the relationship between noise and innovation. If everyone in an organization makes judgments the same way, does that stifle creativity and novel approaches? Could some noise actually be productive, generating the variability that leads to breakthroughs?

These aren’t questions the book fully answers, but that’s okay. The best books don’t tie everything up in a neat bow—they open up new ways of thinking and leave you with productive puzzles to solve.

Final Thoughts From My Reading Chair

I finished “Noise” feeling both disturbed and enlightened. Disturbed because it revealed just how much randomness affects decisions I thought were based on expertise and careful analysis. Enlightened because it provided a framework for understanding and addressing this randomness.

This isn’t a book you read for entertainment—it’s a book you read to change how you think about judgment and decision-making. It’s dense, sometimes technical, and occasionally repetitive. But it’s also important, offering insights that could genuinely improve how we make consequential decisions in medicine, law, business, and beyond.

For anyone involved in professional judgment—and really, who isn’t?—this book offers valuable tools for recognizing and reducing unwanted variability. For general readers interested in psychology and decision-making, it provides a fascinating complement to the literature on bias and cognitive errors.

I’d love to hear your thoughts. Have you noticed noise in your own decision-making or in your organization? What strategies have you found helpful for reducing unwanted variability? Drop a comment below and let’s continue this conversation. After all, one of the best ways to reduce noise is to share perspectives and learn from each other’s experiences.