I don't see how it can simultaneously be the case that: Though Cooper concedes that there could be exceptions at specific companies (the study didn’t measu

I don't see how it can simultaneously be the case that: Though Cooper concedes that there could be exceptions at specific companies (the study didn’t measure individual firms) and How could this be? In a word, overconfidence. CEOs who get paid huge amounts tend to think less critically about their decisions. “They ignore dis-confirming information and just think that they’re right,” says Cooper. If he didn't look at individual people, how can he speak to the psychology of the effect? Are we suggesting that overconfidence shows up statistically as differentiated from other poor performance? What is the statistical proxy being used here?