The benefits of competition


Jason Collins


October 8, 2012

I recently came across a review of Robert Frank’s The Darwin Economy by Ted Bergstrom. Frank’s argument is largely based on the concept that a person is made worse off when they respond to someone else’s consumption choices, as it often turns into a winner takes all arms race. Bergstrom makes an important point that people may wish for someone else to increase their level of conspicuous consumption or competitive output. Bergstrom writes:

Suppose, for example, that neighbors A and B each consume two goods, x whose consumption is publicly observed and y whose consumption is not. Suppose that A gets an income windfall and buys more x. Shortly thereafter, we observe that B buys more x, although his income hasn’t changed. Can we conclude that B has been made worse off by A’s good fortune? We know that B now chooses a bundle that he rejected before A’s windfall, which would be bad for him if he were indifferent about A’s consumption. But, in fact, he is not indifferent about A’s consumption. It is not hard to construct “realistic” vignettes in which B would be pleased to see the increase in A’s consumption of x and pleased to increase his own x in response. For example, A if paints his house, or improves the appearance of his garden, B might enjoy the neighborhood improvement and although he could still afford his old combination of x and y he would prefer to complement his neighbor’s action by his own home improvements. Other examples come from athletic endeavors. Suppose that A and B frequently play tennis together and traditionally win about equally often. For some reason, A’s game improves, and he begins to win more than half the time. This induces B to play harder or perhaps purchase costly tennis lessons so that once again they win about equally often. Are we to conclude that B is worse off and A is no better off than before the improvement of their games? Not necessarily. Both may be evolutionarily programmed by our hunter-gather past to enjoy the challenge of succeeding at a difficult task. After all, they do not play each other for prize money, they play for the pleasure of competing.

I have some sympathy for Bergstrom’s argument, although I wonder how many examples of this type might be explained by competition in other domains. For example, if someone is pleased that their neighbour renovates their house and then increases the upkeep on theirs, is this because there are people besides their neighbour with whom they are engaging in positional competition? Does it increase the prestige of their neighbourhood?

I am also not convinced that the pleasure of competing is materially improved through your playing partner in social tennis becoming better than you, particularly given experimental evidence on changes in testosterone and cortisol from losing “fun” competitions. Being at the bottom of the pile is bad for your health. Is the additional training simply intended to increase the probability of winning against the improved player? Or is the training against the better competitor yielding benefits in more success against other potential competitors?

Despite my doubt about whether these examples are useful or representative, I am hesitant to ignore them in drawing policy conclusions that may affect competition. As I suggested recently, there may be spillovers from competitive activity that yield broader benefits. Plus, do I have the required level of insight about someone’s enjoyment of tennis to be making assessments on their behalf? It’s a reasonable assumption that the mismatch between private and social benefits means that private investment in winner takes all competition exceeds the socially optimal level, but quantifying or controlling that mismatch with imperfect information about a third party’s intentions is a difficult task.