Saturday, January 2, 2016
Use common sense with care
Many essays and journalistic opinion pieces are based on common sense. For example, common sense dicates that in most situations more of something good should always be positive. However, human societies are complex adaptive systems where non-linearities are pervasive, so that more of some good behaviour by individuals does not necessarily translate into a better aggregate. More use of majority rule (generally something good) may be desirable in some contexts but not in others, where more consensual or unanimous decision making processes may be desirable because of their robustness. Emerging macro behaviour is difficult to predict from individual behaviour, which is a reason why exporting institutions or policies that have worked in some context is not always a good idea. The difficulties of predicting aggregate behaviour are compounded by the fact that human individuals have several selves, that is, we behave differently depending on our mood, or depending on our time perspective, or on our role. For example, we have a set of preferences for the short run and another set of preferences for the long run (ask compulsive eaters). Or we behave with different degrees of rationality when we take decisions as consumers, as investors, as voters or... as researchers. I know of researchers who have done very serious work on social identities pointing out that narrow national identities are bad for redistributive preferences, and then I have seen the same researchers defending nationalist positions politically. In our complex reality, several selves combined with uncertainty make aggregate behaviour very difficult to predict. Another problem with common sense is the abundance of circular arguments. For example, by claiming that one country is not developing fast enough because it has "bad" institutions, is attributing causality to something that is just part of the description of the problem. Most probably, the country is not developing well AND does not have good institutions for similar underlying and difficult to observe reasons. To dig deeper into these reasons, probably we need more than common sense. We need science.