Friday, April 13, 2012

Talk is Trash


I recently studied a few cheap talk models, beginning with the seminal paper by Crawford and Sobel in 1982. By “cheap talk”, most economists really mean free talk. A cheap talk game is one where one agent can send a message without incurring any exogenous cost: the only cost incurred is endogenous, i.e. it affects the sender’s payoff by influencing the receiver’s action choice. In short, cheap talk is a costless signal.

From a linguistic perspective, the “talk is cheap” maxim is traditionally taken to mean that the words in a message could be meaningless and need not be supported by truth, actions or evidence. The rising popularity of sarcasm should lend some added meaning to the phrase: what is said is very different from what is meant because the inherent meaning of our words is often diluted by how we say it.

So is the game theoretic use of the phrase cheap talk itself cheap talk? No! Economists are nothing if not precise. But it is possible that calling an entire subset of signalling games “Cheap Talk” is an attempt to make it sound a lot more interesting than it really is. If cheap talk is enough to make a paper sound interesting, imagine how much more intriguing a trash talk model would seem. Measuring the cost of indulging in trash talk and weighing it against the psychological advantage it confers could help us calculate whether being courteous is worthwhile at all. Children waste much of their childhood learning social conventions and much of their adult life brimming with resentment as they teach their children to fight their instinct to be frank.

Such a model might do to social relations what the prisoner’s dilemma did to gang loyalty – although it is socially optimal for all of us to be nice and the outcome is Pareto superior to any other, if each of us acts in a self-interested manner, being nasty would be a dominant strategy unless social ostracism outweighs the personal gain from feeling smug. However, if the psychological cost of courtesy is greater than its social benefit, we could adjust social structures to eliminate the rationale for teenage rebellion and angst altogether. It would be fascinating to see teenagers rebel by being excessively nice to everyone because their parents were so blunt all the time. Gandhi would be so proud. 


No comments:

Post a Comment