On the Efficiency of AC/DC

Bon Scott or Brian Johnson, Brian Johnson or Bon Scott. AC/DC fans debate endlessly about who was a better lead singer, dividing the world in two. In order to shed some light on this discussion, Robert J. Oxoby from the University of Calgary explored which of the two is more efficient in the terms of making agreements more likely. In order to do that, he ran an experiment in which subjects where randomly and anonymously matched and took part in a simple bargaining game while they were listening to a song performed by one of the two vocalists. The game was in fact the famous Ultimatum Game, in which one of the members of each pair must make a proposal on the division of 10$. The other party must decide whether to accept the offer, in which case subjects were paid in cash accordingly, or reject it, in which case both walked away empty handed. In this interaction, each time an agreement is not reached money is left on the table (the 10$) and inefficiency ensues. The question was then, which singer generates more agreements? The results (On the Efficiency of AC/DC: Bon Scott versus Brian Johnson, Economic Inquiry, 2008) were shocking: Brian Johnson is a better singer than Bon Scott in efficiency terms. The offers made by subjects who were listening to Shoot to Thrill were more generous and were rejected less often than those made by subjects who were listening to It’s a Long Way to the Top.

Surprised? The truth is that this paper was a supposed to be a joke and was never intended to be serious. The original data came from a PhD student of Oxoby who allegedly wanted to study the effect of music on the outcome of the ultimatum game (this may be seem to you a bit over the top but it has been shown that subjects are more likely to cooperate with strangers after watching a comedy than after watching a drama). However, the student played two different songs, making the data obtained useless. The student left the PhD unfinished (uhm, I wonder why...), Oxoby found the data and thought it could be fun to write a bogus paper with it. In fact, the moral of the story for you should be that correlation is not the same as causation. It is true that the results under the two songs were statistically different. But that in itself does not mean that there is a direct causation or that we have learned anything useful from the exercise.

However, Steven D. Levitt, world famous for his Freakonomics book but not precisely for his sense of humour, took the paper seriously and condemned it as a very bad example of what Economics can be. What do you think? Fun exercise? Bad science?