A critical discussion of Jon Elster's arguments about rational choice, infinite regress and the collection of information

SUMMARY Jon Elster has argued that (a) the optimal collection of information creates an infinite regress (we have to collect information about how much information to collect about how much information to collect …) and (b) it is often very difficult to estimate the net value of more information because we have to make decisions in situations that are fast changing, unique or novel. Taken together these two arguments imply, according to Elster, that it is often impossible to make a rational choice and this, in turn, means that economists should focus more on theories of satisficing (people acting when they have an alternative that is "good enough") and less on theories based on optimization. The purpose of the dissertation is to examine these arguments. The answer to whether rational choice implies an infinite regress and whether it is impossible to rationally collect the optimal amount of information clearly depends on the definition of rationality (Chapter 2). To examine this question, I first present the standard economic theory (expected utility theory). This theory, however, places few restrictions on what probabilities the agents are allowed to use (and still be labeled rational). In response to this some authors (like Jon Elster) argue that rationality implies not only that we chose the optimal action for given beliefs, but also that these beliefs are optimal for given information and that the collection of information must be optimal. Others, like Russell Hardin, argue that it is not intuitive to demand that a decision must be based on an optimal amount of information before it is labeled rational. The issue is important because if Hardin is right (that rationality does not demand optimal collection of information), the problem of infinite regress in the collection of information does not even get off the ground. I argue, however, that Hardin is wrong on the intuitiveness of excluding information-gathering (Is it rational to buy a house without collecting information?). There are also good methodological reasons (like parsimony) for making the optimal collection of information a requirement of rationality (Why make a different assumption about the choice of information than the choice of action?). Chapter 3 is mainly descriptive. I first present the standard economic theory for determining the value of information (based on Hirshleifer and Riley). The presentation is not only descriptive since it turns out that there are several factual mistakes in their account. Next follows an …