The planning fallacy: plans and forecasts that are unrealistically close to best-case scenarios, improvable by consulting statistics of similar cases. Coined by Kahneman and Tversky.
Inside view vs outside view
Kahneman's curriculum team estimated two years; Seymour's reference class: 40% never finish, survivors take 7–10 years. They ignored the outside view and finished in eight. The inside view uses unique project details; the outside view uses reference class forecasting — baseline from similar projects, then modest adjustment.
Scottish Parliament (£40M → £431M), rail projects (106% passenger overestimate), kitchen remodels ($19K estimate → $39K actual) are the pattern.
Causes
- WYSIATI and unknown unknowns excluded
- Optimistic bias and incentive to lowball for approval
- Sunk costs and embarrassment after investment begins
Cure
Reference class forecasting (Flyvbjerg): identify class, get distribution of outcomes, adjust for this case. Pair with premortem.