A few weeks ago we released a new problem profile focussed on improving decision-making in major societal institutions:
In 2003, the United States chose to invade Iraq. Most now agree this decision was deeply flawed, costing trillions of dollars and thousands of lives.
Exactly what went wrong here is a contested and controversial issue. At best, the decision-making process severely lacked rigour, and at worst, it was heavily biased.
The government justified the invasion thanks to the intelligence community’s claim that it was “highly probable” that Iraq possessed Weapons of Mass Destruction (WMD) – but this statement was ambiguous. Policymakers took that to indicate near-100% certainty, and made decisions accordingly.1 But “highly probable” could easily also be interpreted as 80% certainty, or 70% – carrying very different practical implications. Those involved didn’t really think through the relevant probabilities, or consider how likely the estimates were to be wrong, or the implications if they were.
Others have suggested that the US had already decided to invade Iraq, and that this decision influenced intelligence collection – not the other way around. This a particularly extreme example of what’s known as motivated reasoning – a tendency to reason in ways that support whatever conclusion one wants to be true.
The call to invade hinged on the subjective impressions of a few key people – subjective impressions that later turned out to be wrong,