Ch3. Confirmation Bias and Herd Behavior — Why We Only Believe What We Already Believe

Do We See the World as It Is?

We assume that humans are rational — that we gather evidence, analyze it, and reach logical conclusions.

Reality is nearly the opposite. We tend to reach our conclusion first, then collect only the evidence that supports it.

This is Confirmation Bias.

---

What Is Confirmation Bias?

Confirmation bias is the tendency to seek out, interpret, and remember information in ways that confirm what we already believe.

Peter Wason’s Card Experiment (1960)

Four cards are on a table: E, K, 4, 7

Rule: “If a card has a vowel on one side, it has an even number on the other side.”

Which cards must you flip to test whether the rule holds?

Most people choose E and 4: E to check for an even number behind it, and 4 to check for a vowel in front.

The correct answer is E and 7. If the back of E is odd, the rule is violated. If the front of 7 is a vowel, the rule is violated. Card 4 tells us nothing — if a consonant appears behind it, the rule is fine.

We instinctively look for confirmation, not falsification.

---

Why Does Confirmation Bias Exist?

Cognitive Ease

Information that matches what we already believe is processed faster and feels pleasant. Contradictory information creates cognitive load and discomfort.

The brain optimizes for efficiency. Confirmation bias is the brain conserving energy.

Ego Threat

Evidence that we are wrong is not merely an informational update. It registers as a threat to the self. “My judgment was wrong” → “I am incompetent” → “My value is diminished.” This chain activates automatically.

Changing a belief is not a cognitive act — it is a psychological risk.

Group Identity

Many of our beliefs are tied to the identity of our group. Political views, religious convictions, brand loyalties — updating them implies leaving the tribe. Confirmation bias functions partly as a mechanism for preserving group membership.

---

Echo Chambers and Filter Bubbles

In the digital age, confirmation bias is amplified by technology.

Echo Chamber

In a group of like-minded people, shared beliefs are continuously echoed and reinforced. When everyone you follow on social media holds the same perspective, that belief is confirmed repeatedly and becomes more resistant to challenge.

Filter Bubble

A concept coined by Eli Pariser: algorithms analyze your past behavior and show you only what you want to see. Search results, YouTube recommendations, news feeds — all are optimized to reinforce existing beliefs.

The combination of echo chambers and filter bubbles produces fractured realities. People who have consumed entirely different information streams about the same event can no longer even agree on basic facts, let alone interpret them together.

---

Herd Behavior

If confirmation bias is an individual phenomenon, herd behavior is its collective expression.

Solomon Asch’s Conformity Experiment (1951)

Participants were given a simple task: judge the length of line segments. Seven confederates in the room deliberately gave the wrong answer.

Result: 75% of participants conformed at least once to the obviously wrong answer.

Two reasons drive conformity:

1. Informational influence: “If so many people agree, maybe I’m the one who’s wrong.”
2. Normative influence: “Disagreeing risks social rejection.”

---

Financial Bubbles: The Economic Consequences of Herd Behavior

Financial markets are where herd behavior produces its most dramatic results.

The Dot-Com Bubble (1995–2000)

Internet company stock prices soared regardless of earnings or revenues. “Everyone says it’s going up, so it will” → everyone buys → prices actually rise → even more people buy. This feedback loop inflated a massive bubble. The NASDAQ fell 78% from peak to trough.

The 2008 Financial Crisis

The belief that housing prices never fall (confirmation bias) + “everyone is buying, so I should too” (herd behavior). The result was the largest housing market collapse in US history.

Economist Robert Shiller identifies narratives — stories collectively shared by a population — as the core engine of financial bubbles. It is not data that moves prices; it is the story of “everyone thinks this way.”

---

Strategies to Reduce Confirmation Bias

1. The Falsifiability Test

“What evidence would prove my belief wrong?” If you cannot answer that question, your belief is closer to faith than to knowledge.

2. Devil’s Advocate

Before important decisions, deliberately argue the opposing side. Assign someone in a group to formally voice objections. This is especially effective in committee decision-making.

3. Seek Out Diverse Sources

Intentionally consume perspectives from sources that disagree with you. Understanding the logic of the opposing view — even if uncomfortable — is a necessary cognitive exercise.

4. Bayesian Thinking

Practice updating the strength of your beliefs in proportion to new evidence. Not: “Does this confirm or deny?” But rather: “How much does this evidence shift the probability?” Treat beliefs as probabilities, not certainties.

---

Chapter Summary

Concept	Description	Real-Life Example
Confirmation Bias	Seeking only confirming information	Finding only “buy” news for stocks you already own
Echo Chamber	Beliefs amplified in like-minded groups	Social media communities reinforcing one worldview
Filter Bubble	Algorithm reinforcing existing preferences	YouTube recommendation loops
Herd Behavior	Conforming to group judgments	Buying assets because “everyone else is”
Financial Bubble	Collective result of herd behavior	Dot-com crash, 2008 housing crisis

Countermeasures:

1. Falsifiability test (“What would prove me wrong?”)
2. Devil’s advocate (deliberately argue the other side)
3. Intentionally consume diverse information sources
4. Bayesian updating (adjust beliefs proportionally to evidence)

Next chapter: The Sunk Cost Fallacy and Regret Aversion — why we keep making bad decisions because of what we’ve already spent.