Famous blunders and their causes

Tuchman set three tests before she would call a decision a folly rather than a mistake. The policy had to be seen as counterproductive in its own time, not only in hindsight; a feasible alternative had to be available; and it had to be the policy of a group, carried on past any one leader. Her cases were the Trojan horse, the Renaissance popes who provoked the Reformation, the British ministers who lost the American colonies and the United States in Vietnam. As a model, the pattern predicts something you can look for in any disaster: a warning on the record before the event, from people with the facts, which the decision-makers knew about and overrode. Reading a blunder, then, means finding the warning and asking why it lost.

The Challenger launch shows the recurring causes together. Booster O-ring erosion was first seen on the second Shuttle flight in November 1981; as later flights confirmed the damage, the Rogers Commission found, NASA and Thiokol responded by raising the amount of damage they considered acceptable, which is confirmation bias. A system the Commission called strained by nine flights in 1985 had 15 scheduled for 1986, and a launch constraint imposed in mid-1985 was waived for every flight after it: schedule as sunk cost. The contractor whose engineers objected depended on the customer who wanted to fly, and its managers were told to think like managers rather than engineers, which is incentive-caused bias. And the objection never travelled up to the people who gave the final go. Each cause is ordinary alone; together, under a deadline, they are how a known risk gets taken.

The trap in reading history this way is hindsight. Baruch Fischhoff showed in 1975 that people told how an event turned out rate that outcome as having been more likely all along, and do not notice the knowledge working on them. Gordon Wood made the same objection to Tuchman: whether an action ends up as an example of Churchillian determination or of wooden-headedness, he wrote, seems to depend on the unpredictable outcome of events. Her first test is the defence. If you cannot find a contemporaneous warning from someone who had the facts, you are probably looking at a bet that lost, not a folly, and the ignored warnings that turned out fine never make it into the books. Hold the standard, or every failure becomes someone's folly.