Waylaid Dialectic

August 5, 2012

Is it really rational to use rational choice models to explain political behaviour in Solomon Islands?

Filed under: Development Theory — terence @ 8:17 pm

I am currently doing two things that would seem to be mutually incompatible: reading, and being convinced by, Daniel Kahneman’s new book, and modelling (in my head at least) Solomon Islands electoral politics using rational choice models of human behaviour.

These two things ought to be mutually incompatible for the simple reason that Kahneman provides powerful evidence that human beings really are not rational utility maximisers much of the time. Which would seem to mean that rational choice models are wrong in their very essence.

So why am I still using them. I think a qualified defence of their use can be made as follows:

1. First, rational choice models are tractable and quite easy to use. What’s more they fit snugly with tools such as game theory that are very valuable aids to thinking about collective action. And at this point in time there aren’t any systematic, readily usable replacements that I am aware of.

2. Obviously, if the models are at total odds with human behaviour, then regardless of point 1, we should still discontinue their use. But this doesn’t seem to be the case. Quite a few social features do seem to approximate those that we would expect from a world of rational agents. Which is likely because we do, at least some of the time, reason our way though problems. And, sometimes, in the case of recurrent decisions, even if we never reason perfectly, trial and error likely gets us to a place where our decisions are much the same as those made by a reasoning machine. More than this though, in much of Kahneman’s book he discusses our non-reasoning side in terms of short-cuts, tricks that our brain uses to make decisions more efficiently than if we had to reason through them. Instincts in other words. Sometimes these tricks get tricked. And we end up with unreasonable outcomes, but much of the time they probably do a reasonable task of making reasonable choices. The short of choices we may well have made anyway, had we bothered reasoning through. And this means that rational choice models may well be accurate a lot of the time.

3. Moreover, by providing clear pictures of expected outcomes, rational choice models also become a very useful tool for highlighting when systematic irrationality is taking place. Spot the deviation and then find a way of explaining it. One good example from Solomons elections is the use of so-called ‘floats’ by the big candidates in Honiara electorates. These are large processions of cars and supporters that the candidates parade around with in the days leading up to the election. Why do they do that? Because they want to make themselves look like likely winners. Voters generally believe (accurately enough) that the candidates with the most money are most likely to win elections in Honiara. Also, everything else being equal, in the strongly clientelistic politics of Solomon Islands, and in an environment where, one way or another, candidates do seem able to find out who voted for them, if you know who’s going to win, it is a good idea to vote for them. If you don’t, they likely won’t help you once in power. So, if you are a candidate it makes sense to look like you are likely to win. Convince people of this and quite a few of them will vote for you for that reason alone. Hence the floats. So far so rational. However, the trouble is, everyone knows that floats are just a way of seeming like you are a likely winner, and not necessarily reliable, so you would expect rational voters not to be swayed by them. And you would expect rational candidates to stop using them once they stopped winning votes. And yet they persist. Why? This would seem to be an interesting puzzle, with an answer probably involving misplaced instincts from voters, who know what the game is all about, but who still can’t help feeling that it’s probably an idea to vote for the candidate with the biggest float. So, even when they are wrong, rational choice models can be useful pointers. One day, when elections are actually being systematically studied by well resourced researchers in Solomons, as opposed to study by one PhD student, these deviations from rationality ought to be a great avenue of inquiry. For now though, I’m sticking to the stuff that does fit rational choice models. Because there seems to be a lot of it.

Of course there are dangers in using rational choice models. One of them being that they are potentially un-falsifiable. Most outcomes can be explained in stories involving reason, even if they are very convoluted and wrong, and this needs to be avoided.

But, if I can do this, and be careful, and not overuse, them my instincts tell me rational choice models still ought to be very useful tools in analysing the data I have.

[Update: Irrationally, I forced myself to finish this post before blog surfing. First thing I did when I finally got there was read this at Roving Bandit:

There are good reasons for keeping prospect theory out of introductory texts. The basic concepts of economics are essential intellectual tools, which are not easy to grasp even with simplified and unrealistic assumptions about the nature of economic agents who interact in markets. Raising questions about these assumptions even as they are introduced would be confusing, and perhaps demoralizing. It is reasonable to put priority on helping students acquire the basic tools of the discipline. Furthermore the failure of rationality that is built into prospect theory is often irrelevant to the predictions of economic theory, which work out with great precision in some situations and provide good approximations in many others.

— Daniel Kahneman “Thinking, Fast and Slow”

1 Comment

  1. […] And while these models may hinge on problematic assumptions about human behaviour, as I explain here, it would be a mistake to dismiss them out of […]

    Pingback by Poor political governance in Solomon Islands – what use rational choice explanations? | Development Policy Blog — August 21, 2012 @ 7:05 am

RSS feed for comments on this post.

Blog at WordPress.com.

%d bloggers like this: