Elster vs. Elster

It was a delight to have Jon Elster, a scholar of the highest stature, here at CIDE.  One thing that left me puzzled during his lectures, however, was that his methodological views seemed to be very different when he considered the status of social science in general, than the framework he actually applied in his own research. For instance, in his paper on the excessive ambitions of social science, Jon Elster criticizes both, rational choice models and statistical analysis. These are some notable quotations:

Cohorts after cohort of students are learning –and many of them subsequently hired to apply or teach– useless theories. Their efforts and talents would have been vastly more useful to society had they been harnessed to more productive purposes.

Rational-choice models may (…) help us explain, predict or shape behavior. Although simple and robust models may do this in a rough-and-ready sense, the sophisticated models that are the pride of the profession do not.

My claim is that much work in economics and political science is devoid of empirical, aesthetic or mathematical interest, which means that it has no value at all. I cannot make any quantitative assessment of the proportion of work in leading journals that fall in this category. I am firmly convinced, however, that the proportion is non-negligible and important enough to constitute something of a scandal.

I suggest that a non-negligible part of empirical social science consists of half-understood statistical theory applied to half-assimilated empirical material (the emphasis is Elster’s).

If many applications of rational-choice theory and statistical theory are wasteful or harmful, why do they persist? (…) I shall discuss two possible mechanisms: mind-binding and pluralistic ignorance.

Controversial (excessive?) as they may be, let’s assume that the jury is out on the above assertions. On the other hand, Elster’s paper on democracy and justice discusses the issues of compulsory voting and turnout under a rational choice framework, backed up with cross-country statistical evidence (the emphasis in what follows is mine):

As I just argued, it would be a danger sign if a large majority of the citizens chose not to exercise their right to vote. And in fact, why should they exercise it? The cost of voting, although small, is not zero. The benefit of voting, if measured by the chance of casting the decisive vote, is essentially zero. (…) modern game theory can be used to show that his conclusion is correct. If voters are rational and self-interested, only a small number will vote.

(…) compulsory voting might trigger a norm of fairness, in the following sense. (…) When voting is made compulsory, their estimate of the likelihood that others will vote goes up, and hence they are more likely to vote as well.  The alternative explanation is that the fear of sanctions provides a full account of the higher turnout under compulsory voting. In general, these sanctions are not severe and enforcement levels are low. Yet from an expected-utility point of view, even a small perceived risk of a small fine for non-voting might exceed the small cost of voting. Also, voters might not know how small the risk is, and they might not be rationally motivated to find out.

To my knowledge, there has been no systematic attempt to find out whether the effects of compulsory voting are due mainly to a stronger feeling of civic duty induced by the law or mainly to the deterrence effect of sanctions. I suspect that cost-benefit analysis fails to capture the psychology of voting, but I have no clear and clean alternative explanation to propose.

I want to pursue some further implications of the fact that voting is a collective action problem. To overcome a free-rider problem, institutional designers may use either the stick or the carrot – punish non-compliers or reward compliers.

The important point is that the requirement of compulsory voting tends to have two distinct effects. On the one hand, it consolidates democracy, by increasing turnout. On the other hand, it promotes justice, by improving the match between the outcome of elections and the popular will. When voting is voluntary, poor citizens are somewhat less likely to vote than the better-off. The tendency varies across countries, and is not always very strong, but it can be substantial. The impact of education on voting is even more substantial. In addition to findings by political scientists, evidence in favor of the correlation is provided by the strong resistance of right-wing parties to compulsory voting, as well as to many measures intended to facilitate voluntary voting.

So, the question is, which version of Jon Elster are we supposed to believe?