On The Relevance of Learning and Evolution to Economic Theory

On The Relevance of Learning and Evolution to Economic Theory

Tilman Börgers

ELSE working papers from ESRC Centre on Economics Learning and Social Evolution

Keywords: ract: The title of the discussion to which this paper is intended to contribute; "Rationality; Learning; and Social Norms" spans a wider field of questions than I shall cover. I shall focus on one particular development in the recent economics literature; na-mely the development of a literature on learning and evolution. This literature is concerned with dynamic processes which describe how economic agents adjust their behaviour over time; and how; after agents have gained experience; their behaviour may become (rational in the economists' sense of this word. Several important academic journals have recently devoted special issues to this literature. The purpose of this paper is to discuss the contribution of this literature to economic theory. I shall argue that the literature on learning and evolution has the potential to contribute substantially to our understanding of the economy. But I shall also criticize several aspects of the literature as it is currently developing. I shall suggest that the literature might benefit from a change of perspective. This change of perspective could in my opinion increase the impact of the literature on economic theory. Much of this paper will be taken up by a discussion of the potential contribution of the literature on learning and evolution. For this; I shall need to place the lite-rature in context. My starting point is that the literature provides an explanation of how rational (in the economists' sense of the word) behaviour comes about. I shall begin in Section II with a discussion of the interpretation of the rationality hy-pothesis in economics. The interpretation which I shall favour is that the rationality hypothesis is a positive hypothesis about observable individual behaviour. In this interpretation the rationality hypothesis simply says that economic agents behavi-our can be interpreted as the solution to some optimisation problem. It does not say that this is because economic agents consciously solve optimisation problems.

If the rationality hypothesis is interpreted in this way; it seems natural to ask next how well it does in practice. This will be discussed in Section III. I shall argue that the evidence is mixed. Economic agents typically act in some; but not in all situations rationally. This is shown both by experimental and by "real world" evidence. A question which then arises is what distinguishes situations in which agents behave rationally from those in which they don't. A central claim of this paper is that to find an answer to this question we need to do research on learning and evolution. This is because; to understand the distinction between situations in which agents do or do not behave rationally; one needs to understand the processes which bring about rationality; and learning processes and processes of evolution are among these processes. In Section IV I shall illustrate by means of examples how research on learning and evolution can illuminate empirical observations concerning the rationality of behaviour. The issue of rationality or irrationality of individuals' behaviour which I put at the centre of my argument is; however; not the issue on which the recent literature on learning and evolution has focused. Rather; it has given most attention to the problem of equilibrium selection in games. In Section V I shall suggest that this has been a mistake. I have three arguments for this.

First; the question of when rationality comes about logically precedes the question of equilibrium selection. Se-cond; from an empirical point of view; the question on which I suggest focusing is the more urgent question. Finally; the chances of making substantial progress with the equilibrium selection question appear to be relatively small as compared to the chances of making progress with the question on which I suggest focusing. Section V will contain a second point of criticism of the recent literature. It is that the literature has made too much use of evolutionary models taken from biology. If used in economics; such models must be interpreted as reduced forms of learning and imitation models. I shall argue that the most popular evolutionary models can indeed be interpreted in such a way; but that this interpretation relies on very special learning models. There are many other learning models which appear equally plausible; and which are not equivalent to biological models. The two criticisms of the recent literature on learning and evolution in Section V form the second step of the main argument of this paper. The paper is concluded with brief remarks in Section VI. (search for similar items in EconPapers)
View list of references

Downloads: (external link)
ftp://all.repec.org/RePEc/els/esrcls/relev.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: http://EconPapers.repec.org/RePEc:els:esrcls:050

Access Statistics for this paper

More papers in ELSE working papers from ESRC Centre on Economics Learning and Social Evolution
Contact information at EDIRC.
Series data maintained by s. malkani ().