Abstract
The author of this article describes a game-theory-based economics class on how people should, and do, form beliefs, communicate, and make decisions under uncertainty. Topics include Bayesian and non-Bayesian belief updating, the value of information, communication games, advertising, political media, and social learning. The only prerequisite is introductory microeconomics. The course also seeks to convey both the value of thinking in probabilities and awareness of mechanisms of strategic influence in students’ everyday lives.
Notes
1 For example, suppose you face a 50–50 gamble that either pays you $1600 and or costs you $1000. With a square root value function and coefficient of loss aversion of 2, prospect theory predicts rejection of this gamble. So, if you were offered the same (independent) gamble two days in a row, you would reject it each day. But if you pool the two gambles, you accept the gamble, even if you are a value function-maximizer and ignore initial wealth (see, e.g., Barberis Citation2013). Moreover, you “usually” accept each individual gamble (and the pooled gamble) if you are an EU maximizer, even if initial wealth is relatively low and you are fairly risk averse (e.g., with initial wealth of $2,000 and utility function of square-root of wealth), implying that accepting the gamble is normatively optimal.
2 Helpful and engaging additional readings that can be used here or elsewhere in the course on the merits of pushing one’s self to be more Bayesian are selected chapter(s) on Bayesian updating from Tetlock and Gardner (Citation2016) and Galef (Citation2021).
3 Online versions of this game are possible too, although they do not implement the original model quite as cleanly.