Twitter | Search | |
Nick Rowe
I ran 2 identical Coronavirus polls, 4 weeks apart. Both show "Median Voter Self-Deception" (The median voter has biased beliefs about the beliefs of the median voter). **But the bias changed sign.** Thread 1/n
Reply Retweet Like More
Nick Rowe Mar 30
Replying to @MacRoweNick
Here is the 1st poll, March 3
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
Here is the 2nd poll, March 29
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
Usual caveats about small sample, unrepresentative sample (probably most important), and sample at least partly different between the 2 polls. But setting those caveats aside....4/n
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
This is how I interpret the poll results: 1. Wisdom of Crowds is unreliable Because if what happens depends on the crowd's beliefs about what happens (negative feedback), and the crowd is demonstrably biased about its own beliefs, the crowd will also be biased about what happens.
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
2. (I'm less sure about this bit) The March 3 poll results made me more pessimistic (which was right, in hindsight). The March 29 results make me more optimistic. Because concern over Coronavirus is a *negative* feedback process (people take offsetting actions). 6/n
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
3. (I'm very unsure, and unclear, about this bit) I *think* this sort of self-referential poll is telling us *something* about how irrational panics/manias (and whatever you call their opposite) work. The crowd's bias about itself flips from one direction to the other. 7/n
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
And *something* happened mid-March, where it seemed lots of people ("normies") suddenly flipped from being not very concerned to being very concerned, without a lot of new information about the fundamentals. 8/n
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
Anyway, those are *my* interpretations, and how I changed *my* beliefs on seeing these 2 poll results. What are yours? How did you change your beliefs? 9/end
Reply Retweet Like
Nick Rowe Mar 30
Replying to @MacRoweNick
Plus (slightly off-topic): the poll results make me think that Lucas 1972 "Expectations and the Neutrality of Money" *might* have been right after all. Because it has exactly the same mechanism: average beliefs about average beliefs being biased.
Reply Retweet Like