aka something to chew on for Thanksgiving – have a great holiday!
I’m back doing a little blogging while Chris is in Liberia (in part because our joint project there is finally moving along!), although he will also continue to post.
A new working paper uses a particular notion of rationality to compare individuals. Here rationality is basically internal consistency of choices across situations – is this person making mistakes with respect to their own utility measure, whatever that might be? [It’s possible to show that someone is inconsistent without knowing what their ‘true’ preferences are.]
The authors show that many people are consistent, but that there is variety and certain groups are more ‘rational’ than others: higher income; more education; males; non-elderly. Of course this is only correlational, as they point out, but they seem to think of rationality as an outcome (left-hand side of the regression) wrt income as well as gender and age. Finally, they use rationality as an input (right-hand side) to explain wealth differentials, conditioning on the socio-demographic variables above. Wasn’t clear to me why they decided it was a cause in some cases and an effect in others, but purely as correlations this is useful and suggestive. They are pursuing similar work with risk preferences, which I have also looked at in my research.
A coauthor of mine (David Wolpert) pointed me toward the paper and speculated (don’t take this too seriously) on what this might imply for politics. If — everything else being equal — we want people in power to be predictable / consistent / rational / stable, it might behoove us to choose those who are high-earners and educated. Of course this assumes the causality above, and it ignores all kinds of other factors that are [more] important, but at the margin it seems plausible.
5 Responses
Actually, the last thing we want from our leaders is for them to be rational — we want their utility functions to be highly altruistic, which would almost certainly show up as inconsistency in other circumstances.
So your conclusion is reversed, interestingly.
@George: very nice paper – I hadn’t seen that one. I completely agree that consistency should not always be the goal, although I think for elected officials I’d prefer it to profit-maximization.
And then there are papers like: http://mpra.ub.uni-muenchen.de/26589/1/MPRA_paper_26589.pdf
This post seems to leap at conclusions with dramatic moral and social content (we should elect a government of rich males from good schools) based on evidence that you recognize as weak. Your willingness to entertain that leap despite its lack of factual basis seems to be unhelpful, at best.
I haven’t read the paper, but it seems to me that what appears to be inconsistent behaviour could turn out to be consistent, given a better understanding of the underlying conceptual schemas involved. Maybe the behaviour of highly educated males with relatively high incomes is used as the standard of rationality in studies like this simply because the people who carry out the studies tend to be highly educated men with relatively high incomes. (I note that all the authors of this paper are male.)