Gayle –
Thanks for forwarding the excellent article from The Economist on the evidence about human irrationality and how that is being accommodated (by some) and resisted (by others) in the social sciences.
One of the most striking ideas in the article got only a few sentences of attention. This was Bryan Caplan’s remark about treating irrationality as a consumption good.[1] He added that, so treated, it’s not surprising that people are least rational about religion and politics, since the costs of irrationality in these arenas is so nearly zero. I’m going to limit my comments to the way Caplan’s hypothesis pertains to politics.
First, it needs to be understood what “rationality” in the article and as Caplan uses it means. What is being talked about is the notion of rationality defined and formalized by decision theory. This notion (or slight variations upon it) is pervasive in the social sciences. Quoting Robin Dawes (from Rational Choice in an Uncertain World) and then me (in Sunk Costs):
“A rational choice can be defined as one that meets three criteria:
1. It is based on the decision maker’s current assets....
2. It is based on the possible consequences of the choice.
3. When these conditions are uncertain, their likelihood is evaluated without violating the basic rules of probability theory.” (Dawes 1988, p. 8)
An additional condition is needed to the effect that these considerations (and only these) are to be used in ranking possible outcomes of one’s actions with respect to the extent to which those outcomes satisfy one’s preferences. Then, the action correlated with the most highly ranked outcome (or, in the case of ties, one of those tied for highest ranking) is to be selected. These conditions attempt to spell out what is involved in the intuitive idea of selecting the best means to one’s ends.”
The important point for present purposes is that it only appeals to agents’ actual goals and preferences and prescribes how best to satisfy those. Accordingly, when Caplan talks about rationality, he’s (usually) talking about actions that are best suited to the preferences of the agents. (I’ll say more about this below, especially about the fact that the irrationality he also talks about is not just an absence or failure of rationality in this sense.)
Caplan suggests that people’s political actions (presuming that the effort involved in coming to political conclusions, as well as more mundane things like voting and support for candidates can be counted among their political actions) are likely – likelier than most other kinds of action – to be infected by irrationality, since the cost of irrationality is so nearly zero.
What could he mean by that? The idea is related but not identical to what Public Choicers call rational ignorance. The basic idea is the pretty obvious point that, even when a better option is in some sense available, it may not be enough better for you to invest in the search costs (time, money, cost comparison, and so on) that would be necessary for you to find out about it. For example, there may be some grocery store in your area where, on average and over the long term – say, a year or so, you could save on your purchases. But if you don’t already know about it, the difference is not likely to be worth the effort of finding it. You’re actually better off not knowing about the better option and also not having expended the resources you would need to expend in order to find out about it. A bit more precisely, you couldn’t reasonably expect the pay-off from finding out about the better option to outweigh what you could reasonably expect the costs of finding out about it to be. Your ignorance ... is rational.
This really comes into its own, though, when applied to politics. Most people, probably, do not have much in the way of direct or for-its-own-sake interest in political participation: they care about politics as it affects their jobs, families, communities, etc., but not just because they think being politically involved is a good thing. For people like that – which is most of us – the expected pay-off of coming to correct conclusions is tiny. There’s almost no chance that, in any large-scale election, one person’s vote will make the difference. Rational voters (not to mention that it’s hard to see why rational people would vote at all!) will budget accordingly the time and attention and care that they devote to considering political issues, candidates, their promises and likely (or unlikely!) performance.
If candidate A promises a program that will benefit his constituency by a million dollars and candidate B says it will cost a million instead, then, supposing they’re seeking office in a district of a million voters, the per capita cost or gain is only a dollar. Whatever happens, the voter is probably better advised to spend a few extra minutes at work than to spend time on figuring out which candidate, if either, is telling the truth about the program. To even begin to seriously consider the issue, investigate the merits of the program, consider the likelihood that the candidates are both honest and will be able, once elected, to do what they say – would require far more effort than any pay-off the average voter can reasonably expect. What you get is, more or less, what we’ve got. Voters make their decisions based on scant investigation, plus whatever information (or misinformation) gets handed to them for free. It’s not surprising, then, that their political preferences have more or less random relations to the merits of the candidates and policies.
So far, this is a more or less standard Public Choice approach to explaining political failure. It’s not the only one, but most of the others seem to rely upon it at some point. But on closer examination, it looks more troublesome. Even if people aren’t very well-informed about political matters, it’s not obvious why their views would be systematically mistaken – that is, mistaken in the same way on the part of enough people to drive the political process in the direction of perverse outcomes. Why shouldn’t we expect instead a random scattering around the correct conclusion – and therefore, since presumably there is more evidence for the correct conclusion than for any alternative, why should we not expect that the median view will tend to be correct? Or, why aren’t people who are assumed to be merely rationally ignorant also highly tentative in the way they hold their conclusions, since they will recognize that their conclusions were formed on the basis of little information and could therefore well be mistaken? But if they did hold their political beliefs in this tentative way, it shouldn’t be hard to persuade them to abandon mistakes with more evidence and better arguments.
So Caplan is going beyond this in interesting ways. He’s not just saying that people may be rational to be ignorant of factors that bear on their political decisions. The same factors that can make it rational, for most people, to be ignorant – high costs of acquiring information and low pay-off for getting it right – can also make it rational for them to be irrational about how they come to their political conclusions in the first place. That’s not paradoxical. “Irrational” in that sentence is an epistemic notion. It has to do with failure to adhere to correct standards of reasoning and evidence-gathering. But since being irrational, in that sense, is an action, it makes sense to ask questions about whether it is rational – that is, about whether a person’s ends are well-served – by being irrational. The argument is just an extension of the argument for rational ignorance in politics. Being rational in the epistemic sense has costs. It requires thought, discipline, careful consideration. These are not free goods. They compete with other things we also care about – like ease, popularity, the indulgence of whim and so on. Besides that, many of us – maybe most of us – have preferences about what we would like to be true, what would make us feel good (or better) about ourselves and our place in the world. Since the chance of a single normal citizen’s affecting political outcomes is negligible, being rational (epistemically) about politics may well be too costly. Even tiny gains to be had from not thinking clearly about politics, being swept away by a candidate’s charisma or caught up in mass excitement, are probably more than enough to outweigh any gains from being epistemically rational about the subject. For most people, it can be rational – can best serve their ends – to be not just ignorant but irrational about politics.
Or ... when irrationality is cheap, people buy more of it.
Rob
_____
Rob Bass
http://personal.bgsu.edu/~roberth
