Confirmation Bias
The tendency to seek, weigh, and remember evidence that confirms an existing conclusion while ignoring or discounting what contradicts it. In Munger's framework it is not a standalone tendency but the compound output of doubt-avoidance and inconsistency-avoidance — and the error Darwin's discipline was designed to defeat.
“Almost any intelligent person with my checklist of psychological tendencies in his hand would, by simply going down the checklist, have seen that Milgram's experiment involved about six powerful psychological tendencies acting in confluence to bring about his extreme experimental result.”
Concept Analysis
Definition & Origins
Confirmation bias is the tendency to seek, weigh, and remember evidence that confirms an existing conclusion while ignoring or discounting what contradicts it. Munger used the textbook term sparingly — in his system the phenomenon is usually decomposed into its component tendencies — but his one explicit deployment of the phrase is surgical: in the autopsy of McDonnell Douglas's disastrous aircraft evacuation test, the company "ignored the strong disconfirming evidence from the failure of the first test because confirmation bias, aided by the triggering of strong Deprival-Superreaction Tendency, favored maintaining the original plan."
The phrase itself entered Munger's vocabulary with a specific valence: it is, he noted, a term of opprobrium — the name we give to the opposite of what Darwin did. That framing matters. Confirmation bias is not one more neutral heuristic in a catalog; it is the standard failure against which Munger's entire epistemic discipline is defined. The checklist, the inversion habit, the pre-committed falsifiers, the protection of bad-news bearers — each is a countermeasure aimed at this single, default behavior of the committed mind.
Core Ideas
In Munger's mechanics, confirmation bias is a compound, not an element. Doubt-Avoidance rushes the mind to a conclusion to remove puzzlement; Inconsistency-Avoidance then resists changing it. Once both have operated, the mind filters reality to fit the settled conclusion — decide first, then curate the evidence. The "bias" is the visible output of that two-stage machinery.
Its signature is the survival of a plan past its own failure data. The McDonnell Douglas case is Munger's paradigm: a near-idiotic test plan, a failed first test with serious injuries, and a continuation decided anyway — because the evidence against the plan was weighed by minds committed to it. Confirmation bias is why "being aware of psychological ploys is not a perfect defense": the filter runs on the awareness too.
Darwin is the antidote's patron saint. Munger's repeated prescription is Darwin's counter-habit: give disproportionate attention to disconfirming evidence, precisely because the mind naturally does the opposite. Extreme objectivity is not a talent but a procedure — write down what would prove you wrong, then go look for it.
The filter is invisible from inside. No one experiences confirmation bias as it operates; what the subject experiences is a reassuring accumulation of evidence that he was right all along. This is why Munger's detection method is procedural rather than introspective — the checklist of tendencies, run on schedule, catches what self-examination cannot. His standing proof: almost any intelligent person with the checklist in hand could have diagnosed the confluence in Milgram's experiment immediately, while the psychology professoriate needed more than a thousand papers to get it ninety percent understood. The professors were not dim; they were looking through the filter.
Practical Application
Invert the evidence search. Before committing, spend the majority of diligence on the case against. Munger's investment process — and his insistence that you must be able to state the other side's argument as well as its proponent — is confirmation-bias engineering.
Pre-commit to falsifiers. Decide in advance what evidence would kill the thesis, so that its later appearance cannot be reclassified as noise.
Institutionalize the messenger. The organization that punishes bad news is confirmation bias with a payroll: leadership's picture of reality degrades until reality arrives uninvited. Protecting the bearer of disconfirming evidence is a structural antidote.
Score yourself on belief changes. Munger's operational test for whether the countermeasures are working is brutally simple: when did you last change an important conclusion? A decision system that never produces reversals is not a system that is always right — it is a system whose filter is working perfectly. The track record of destroyed convictions is the only reliable evidence that disconfirming evidence is actually reaching you.
Separate the hypothesis from its advocate. Committees fall into confirmation bias collectively when the proposal and its sponsor become one object: attacking the analysis is then experienced as attacking the analyst, and the room politely converges. Munger's structural fix was to make the strongest statement of the opposing case a formal role rather than a social risk — assigning the bear case to a capable advocate whose job is to be wrong-seeking, so that no one's identity rides on the audit. The practice looks inefficient in calm meetings and is the only thing standing between the group and a McDonnell Douglas afternoon in the bad ones. Organizations that adopt it discover its second benefit: the quality of the original proposals rises, because authors write for a room that will actually test them.
Common Misconceptions
Misconception 1: Intelligence is a defense. Munger's cases involve engineers, executives, and professors — "almost any intelligent person with my checklist" could have diagnosed Milgram, yet a thousand papers missed what the checklist sees. Smart minds build better filters, not weaker ones.
Misconception 2: More data fixes it. Additional evidence processed through the filter strengthens the error — each confirming item adds confidence while disconfirming items are discounted. The search protocol must change before the volume helps.
Misconception 3: It only distorts weak conclusions. The McDonnell Douglas plan was maintained not despite being near-idiotic but because commitment had already been made — the bias defends exactly the conclusions that need defending least and protects the errors most expensive to admit. The bigger the commitment, the stronger the filter, which is why catastrophic institutional errors survive evidence that would have reversed a casual opinion in a week.
Munger's Own Words
"Mc Donnell Douglas ignored the strong disconfirming evidence from the failure of the first test because confirmation bias, aided by the triggering of strong Deprival-Superreaction Tendency, favored maintaining the original plan." — Charlie Munger, The Psychology of Human Misjudgment (Harvard, 1995)
"Almost any intelligent person with my checklist of psychological tendencies in his hand would, by simply going down the checklist, have seen that Milgram's experiment involved about six powerful psychological tendencies acting in confluence to bring about his extreme experimental result." — Charlie Munger, The Psychology of Human Misjudgment (Harvard, 1995)
"One of the most successful users of an antidote to first conclusion bias was Charles Darwin. He trained himself, early, to intensively consider any evidence tending to disconfirm any hypothesis of his, more so if he thought his hypothesis was a particularly good one. The opposite of what Darwin did is now called confirmation bias, a term of opprobrium." — Charlie Munger, The Psychology of Human Misjudgment (Harvard, 1995)
Thought Evolution
Legacy & Influence
Confirmation bias occupies a unique position in Munger's system: it is the one failure mode against which nearly all his epistemic machinery is aimed. The checklist, the inversion habit, the written thesis with pre-committed falsifiers, the assigned dissenter, the protection of bad-news bearers — read as a set, these are not miscellaneous good habits but a coherent engineering response to a single defect: the committed mind's tendency to curate reality in its own favor. Munger's phrase for the target state, "extreme objectivity," is defined procedurally rather than temperamentally — it is what you get when the countermeasures run on schedule.
The concept arrived in Munger's vocabulary from academic psychology, where Peter Wason's selection-task experiments in the 1960s had shown subjects overwhelmingly choosing tests that could confirm rather than falsify a rule. Munger's contribution was not the discovery but the relocation: he moved the bias out of the laboratory and into the boardroom, the evacuation test, and the investment committee, insisting that its most expensive expressions occur exactly where the most credentialed minds operate. His Darwin framing — the bias as "the opposite of what Darwin did," a term of opprobrium — remains the cleanest pedagogical inversion in the whole canon: objectivity is not the absence of bias but the presence of a counter-habit.
Its modern legacy is visible wherever decision quality is taken seriously. Red-team reviews, pre-mortems, kill criteria, and the short-seller's adversarial brief are all confirmation-bias countermeasures wearing different names, and the investment world's growing ritual of writing the thesis down before the outcome is known is Munger's paper-trail method institutionalized. The bias itself has not yielded to a century of awareness training — which is precisely his point: the filter cannot be educated away, only routed around by procedure. The organizations that learned this lesson no longer ask whether their people are objective; they ask whether their process gives disconfirming evidence a way in.
Related Concepts
Case Companies
McDonnell Douglas — The Test That Failed Twice. Directed to run a realistic evacuation test, the company chose an obviously too-dangerous method; the morning test failed with serious injuries; and the company ignored the strong disconfirming evidence, bent — in Munger's phrase — on maintaining the original plan like a gambler making his final big bet to get even. Confirmation bias turned one failure into a cascade.
Darwin — The Counter-Case. Munger's standing exhibit of the opposite discipline: Darwin trained himself to give special weight to evidence against his own theories, and "the life of Darwin demonstrates how a turtle may outrun the hares, aided by extreme objectivity." The point is procedural: objectivity of that grade is a habit installed against the filter, not a gift.
The Investor's Written Thesis — A Filter With a Paper Trail. Confirmation bias thrives in memory, where the original reasons for a decision quietly mutate to fit whatever has happened since. Munger's counter-procedure was to fix the thesis in writing at the moment of decision — the expected economics, the named falsifiers, the price that would prove the analysis wrong — so that the later mind, busily curating its narrative, collides with its own earlier testimony. This is the Darwin method ported to portfolio management: the notebook does not forget, does not rationalize, and does not share the owner's incentive to have been right. Investors who skip the writing step are not saving time; they are clearing the way for the filter, and their subsequent confidence in the position measures the filter's success rather than the position's quality.
Mentioned In
Source: Poor Charlie's Almanack, The Wit and Wisdom of Charles T. Munger