MINDSETOngoing practice92% confidence

When the Facts Change, Change Your Mind

Separate your public identity from your beliefs so you can update when evidence demands it.

Problem it solves

Doubling down on a failing position because your identity is fused with it

Best for

Investors, analysts, and decision-makers who have taken a public position and feel trapped by it.

Not ideal for

Decisions where consistency itself is the product being sold — e.g. a brand promise or legal contract.

Overview

Why this framework exists

Tim Harford opens the episode with a defining historical contrast: Irving Fisher and John Maynard Keynes both failed to predict the 1929 Wall Street Crash. Both made identical forecasting errors at the same moment. Yet Fisher died broke and disgraced while Keynes died a celebrated millionaire. The difference was not intelligence or access to information — it was the ability to change.

Fisher had built his entire public identity around his investment forecasts. His syndicated newspaper column, his public reputation, and his personal wealth were all anchored to the same thesis. When the crash invalidated that thesis, backing away would have required publicly dismantling his own identity. He couldn't do it. He doubled down, lost everything, and spent his final years explaining away the evidence.

Keynes operated privately. He managed money for a Cambridge college with no public performance record to defend. Before the crash he had privately acknowledged to a friend that he was 20% behind the market — he already had the humility to admit underperformance. When the facts changed, he was psychologically free to change. Harford's insight is structural: it was not superior character that saved Keynes, but the absence of a public identity fused with a specific belief. The framework suggests a proactive design principle — keep your core beliefs private enough that updating them doesn't destroy your persona.

Core principles

5 total
  1. Public commitment to a belief makes it psychologically expensive to update, regardless of the evidence.
  2. Acknowledging small underperformance privately is the early warning system that enables later flexibility.
  3. The smartest person in the room can still be destroyed by their inability to change their mind.
  4. Separate the belief from the believer — a wrong forecast does not make you wrong as a person.
  5. The longer you wait to update, the more sunk cost you are defending and the harder the update becomes.

Steps

5 steps
  1. Audit your public positions
    List the investment or financial beliefs you have stated publicly — on social media, to friends, to colleagues. Note which ones your identity or reputation is now attached to. These are your highest-risk beliefs for doubling-down failure.
    Pro tipThe ones that get the most social engagement are the most dangerous — the audience reinforces the position.
    WarningDo not confuse 'having conviction' with 'being unable to change'. Conviction is fine; identity-fusion is not.
  2. Create a private performance log
    Track your investment or decision performance privately, separate from what you share externally. Keynes wrote to a friend confessing he was 20% behind. Having that honest private record was what made updating possible when the moment came.
    Pro tipA simple spreadsheet or journal entry works. The point is that you see reality before having to defend it publicly.
  3. Establish pre-defined update triggers
    Before a market move or decision outcome, write down: 'If X happens, I will reconsider Y.' Defining the trigger in advance removes the in-the-moment identity defence. You're not changing your mind because you lost — you're executing a pre-agreed protocol.
    Pro tipPhrase the trigger in terms of evidence, not emotion: 'If active fund performance exceeds my index fund for 3 years running, I revisit the allocation.'
    WarningDon't set triggers so far out that you're never forced to review. Fisher's problem wasn't that he had no triggers — it's that his public reputation made him override them.
  4. Update incrementally and document the reason
    When evidence demands a change, make it in small steps and write down what changed and why. Incremental updates are less threatening to identity than sudden reversals. Documenting the reasoning turns the update from a defeat into intellectual progress.
    Pro tipHarford notes that Keynes changed his investment strategy after the crash, not his economic worldview — targeted updates on the failing piece, not wholesale collapse.
  5. Communicate updates in terms of new information, not error
    When sharing a changed view publicly, frame it around what the new evidence showed rather than how wrong you were. 'Given what's happened with X, I now think Y' lands better than 'I was wrong about everything'. This reduces the psychological cost of updating and makes the next update easier.
    Pro tipThis is not spin — it's accuracy. Your view was reasonable given what you knew; new information changed it. That is what happened.
    WarningDo not use this to avoid accountability for wilful ignorance. If the evidence was available and you ignored it, say so.

Checklist

Saved in your browser

Examples

4 cases
Irving Fisher's public ruin

Two weeks before the 1929 crash, Fisher published on the front page of the New York Times that 'stocks have reached a new and permanently high plateau'. He had borrowed heavily to invest, was publicly associated with the bull thesis, and could not retreat. The Dow fell 89%. Fisher was financially wiped out and his reputation destroyed.

OutcomeFisher never recovered. His identity was so fused with his forecast that when facts changed, he could not.
Keynes's private pivot

Keynes made the same forecasting error as Fisher — he also missed the crash. But he managed money for a Cambridge college with no public track record to defend. Before the crash he confided to a friend he was 20% behind the market. When the crash hit, he changed his investment strategy.

OutcomeKeynes died a millionaire and was celebrated at Bretton Woods as an architect of the post-war economic order.
The Chicago cult doubles down

A 1950s doomsday cult predicted the world would end on a specific date. Academic psychologists infiltrated the group. When midnight passed without destruction, the deeply committed members — those who had quit jobs and families — invented a new explanation: their faith had saved the earth. They then called a press conference.

OutcomeThe most invested believers doubled down rather than face the social and psychological cost of being wrong. Only the peripherally involved walked away.
Great Western Railway investors

Investors in the 1840s railway bubble who correctly picked the best-run company — Great Western Railway — and bought at the bubble's peak still did not outperform Treasury bills until the company was nationalised after World War II — nearly a century later.

OutcomeBeing right about the technology does not mean the investment price reflects fair value. The principle applies equally to dot-com and AI bubbles.

Common mistakes

5 traps
Fusing identity with a specific prediction
Fisher's entire public brand was built on his ability to forecast markets. When the market proved him wrong, backing down meant destroying not just the forecast but the brand. Anyone who has built a public persona around being right about something faces the same trap.
Doubling down after early losses to protect sunk costs
The cult members who had quit their jobs and abandoned their families could not accept the aliens hadn't come — the emotional cost of admitting failure was greater than the cost of continuing to believe. Harford draws the explicit parallel to investors who hold losers to avoid crystallising a loss.
Treating humility as weakness
Keynes privately admitted underperformance to a friend. That acknowledgement looked like weakness but was the act that kept him flexible. Fisher's public bravado looked like strength and made him brittle.
Waiting for total vindication before updating
The question is not 'was I completely wrong?' but 'has the evidence shifted enough to warrant a recalibration?' Fisher needed total defeat before he would consider changing, by which point it was too late.
Confusing being right about the technology with being right about the investment
Fisher was correct that American industry was transforming in the 1920s — productivity grew at 7% a year. He was wrong that this justified any price for stocks. Railway investors in the 1840s were correct about the importance of railways and still failed to get their money back for a century.

Origin story

How this framework came to be

Harford has told this story across his writing and the Cautionary Tales podcast, but revisits it here as the anchor for a broader discussion about index fund orthodoxy. His interest is not just biographical — he uses Fisher vs Keynes to pose a direct question to the audience: which investor are you most like? Have you publicised your investment thesis so widely that admitting you were wrong now feels impossible? The story also foreshadows the episode's later discussion of cult psychology — when people quit their jobs and leave their families for a belief, they cannot let it fail.

Source

Traced to primary
Source · PODCAST
Do We Need To Change Our Minds About Index Funds?
Tim Harford · 2025
Open source →

Related frameworks

Browse all Mindset →