When we combine anchoring, with intellectual investment, with the status benefits, and with the high cost of changing frames, it’s not rational to expect intellectuals to act otherwise.

What we find is this: intellectuals who specialize in very narrow domains and are current with the data tend to be very good. Whenever intellectuals try to work on generalizations it at least appears that we do better asking people on the street.

this is why economists supposedly disagree: most economic knowledge of any discreet topic is almost always counter-intuitive. So if you ask a thousand economists a question you will get terrible results, but if you as the top 12 in any niche you will get awesome results.