Consumption and fitness

After posting Friday’s piece on Hansson and Stuart’s paper on natural selection and savings, I realised I had not commented on one of the most important assumptions made by the authors. To get their result that people would save such that they maximise consumption across generations, Hansson and Stuart assumed that consumption corresponded with fitness (a relative measure of reproductive success). To maximise fitness, one would need to maximise consumption.

This makes some sense over the very long-term, regardless of whether you consider that consumption contributes to fitness directly, by (say) improving health, or indirectly as a signal of quality. In the signalling case, it has been shown that conspicuous consumption as a handicap can be consistent with the standard utility approach to consumption used in economics.

While I can hand-wave my way through that explanation, some of the attempts in economics to more explicitly consider fertility and fitness have run into a larger problem. In a presentation by Alan Grafen on Sunday (at this conference), he went through four papers on the boundary of evolutionary biology and economics, and one of those he picked was by Robert Barro and Gary Becker. In Barro and Becker’s model, they created a utility function which included both consumption and fertility. An agent in the model would be interested in increasing consumption and their number of offspring. But as Grafen asked, what is the purpose of consumption in the model from an evolutionary perspective?

The interesting thing about Barro and Becker’s paper is that by adding an evolutionary biology flavour to the model, they have created a trade-off between fitness and consumption. Higher consumption reduces fitness, and no hand waving like I used for Hansson and Stuart’s model can justify consumption biologically. Any defence of Barro and Becker’s model will have to look to its predictive power, with the evolutionary explanation to come later. Given that most people in developed countries consume much more than can be justified biologically and do not seem to be maximising their fitness (sperm donation anyone?), there is room to defend it on this ground (in at least the short term).

The other issue that I should mention about Hansson and Stuart’s paper is that they implicitly assume that if one saves, the saver has effective protection of their private property and is able to pass it down to their children without fear of appropriation. As I have posted about before, violence may be a significant obstacle to the development of behaviour such as saving. While Hansson and Stuart saw differences in saving and labour preferences arising from environmental conditions, another plausible driver of differences would be the probability of any savings being stolen.