Keynes and the Limits of Quantification
John Maynard Keynes' sharp distinction between calculable risk and genuine, irreducible uncertainty — a direct challenge to the whole quantifying project covered so far.
Having traced several centuries of progress in quantifying risk, Bernstein devotes a chapter to economist John Maynard Keynes' influential pushback against the idea that everything genuinely important about the future can be captured by probability calculations at all. Keynes drew a sharp distinction between "risk" — situations where the range of possible outcomes and their probabilities are genuinely knowable, like a roulette wheel or an insurable event with abundant historical data — and true "uncertainty," where the very structure of possible future outcomes is not knowable in advance at all, making any specific probability assigned to it more a matter of psychological confidence and convention than genuine calculation.
Keynes applied this distinction directly and skeptically to long-run economic and investment forecasting specifically — arguing that many of the most economically consequential questions (will there be a major war, what will interest rates be in twenty years, will a specific new technology succeed) belong to the genuine-uncertainty category rather than the calculable-risk category, meaning the confident-looking probability estimates sometimes attached to such forecasts can be a kind of false precision, dressing up genuine not-knowing in the reassuring clothing of a calculated number.
| Risk | Uncertainty | |
|---|---|---|
| Example | A roulette wheel; a large, well-documented insurable event | Will there be a major war in 20 years? |
| Possible outcomes | Genuinely knowable in advance | Not knowable in advance — the range itself is unclear |
| Probability estimate | A genuine calculation | More a matter of psychological confidence than real calculation |
Bernstein is candid that Keynes' distinction sits in real tension with the triumphant, centuries-long story of quantification told in this book's earlier chapters — it is a direct, credentialed challenge to the implicit assumption that more mathematical sophistication always means better handling of an uncertain future. The book does not fully resolve this tension in Keynes' favor or against him, instead treating it as a genuine, permanent complication that anyone applying probability and statistics to real-world decisions — especially in economics and investing — needs to hold onto rather than resolve away, echoing the same humility this Book Club's The Black Swan course argues for regarding genuinely unprecedented events.
- Keynes distinguished "risk" (genuinely knowable outcomes and probabilities) from "uncertainty" (where the very structure of future outcomes isn't knowable at all).
- He argued many consequential economic and investment questions belong to the uncertainty category, where a confident-looking probability estimate can be false precision.
- This directly complicates the book's own earlier chapters — a genuine, unresolved tension between quantification's real power and its real limits.