Friday, February 10, 2023

Short Takes: Behavioural Economics, Monty Hall, and more

I find behavioural economics and other aspects of psychology interesting, but I often get lost between a study’s results and the conclusions people draw from these results.  A good example is the oft-repeated fact that most people believe they are above-average drivers.  I have no doubt that a large majority of people will consistently report that they are above-average drivers.  However, the tidy conclusion that these people are overconfident isn’t obvious to me.

There is no single measure of the quality of a driver.  Imagine two brothers where one believes that it is crucial to observe the speed limit at all times, and the other believes it is prudent to always stay up with the flow of traffic to minimize relative speeds.  These standards of driving skill are in conflict, and each brother judges the other to be a poor driver.  Each brother believes he is the better driver based in part on his view of what makes a driver good.

It may be that both brothers are overconfident as well, but we can’t necessarily draw this conclusion from the fact that each brother believes he is better than the other.  In the general case, it’s hard to say whether a high fraction of people think they are above-average drivers primarily because of differences in the standards they apply or primarily because of overconfidence.

This was just a single example, but I find such gaps come up often between a study’s results and the conclusion people draw from those results.

Here are some short takes and some weekend reading:

Annie Duke
has a plausible explanation of why people can’t learn to get the Monty Hall problem right but pigeons can.

Andrew Hallam looks at the mutual fund with “the best performance record of all American mutual funds” and makes a surprising comparison.

helps you control your spending with ideas from behavioural science.

Justin Bender
explains the two levels of foreign withholding taxes on dividends from international equity ETFs.

No comments:

Post a Comment