“Sixty-five percent of those attending the ABA Annual Meeting session said they were better than average at predicting the settlement value of a case, and 76 percent said they were better than average at predicting when a trial court judgment would be reversed on appeal.” But when asked a multiple-choice question on basic Bayesian statistics, only 34% of the attendees got it right–just nine points better than chance. (The most popular answer was also the most incorrect.) The ABA Journal blog also reports that attendees also suffered from severe cognitive bias on such issues as non-economic damages:
A hypothetical described a case involving a school teacher who lost his arm in an accident. Half were told that the plaintiff offered to settle for $100,000, and the other half learned of a $10 million settlement offer.
A majority of the $100,000 group said a judge would assess the value of pain and suffering between $500,000 and $2 million. But a majority of the $10 million group went higher, saying the value would be between $1 million and $5 million.
When even the lawyers and judges can’t accurately peg noneconomic damages, what other evidence do we need to show that uncapped and unscheduled noneconomic damages are unconstitutionally arbitrary and irrational?