WHAT IS THE BEHAVIORAL ECONOMICS “REVOLUTION”?

For me, economics is a collection of ideas and conventions that economists accept and use to reason with. Namely, it is a culture. Behavioral economics represents a transformation of that culture. Nonetheless, as pointed out by Camerer and Loewenstein (2003), its methods are pretty much the same as those introduced by the game theory revolution. At the core of most models in behavioral economics there are still agents who maximize a preference relation over some space of consequences and the solution in most cases still involves standard equilibrium concepts. However, the behavioral economists are not committed to what is usually referred to as rational motivations. An economic fable (or a model, as we would call it) that has at its core fairness, envy, present-bias and the like, is by now not only permitted but even preferred.

Why now? Perhaps, economists have finally realized that orthodox economic models are too unrealistic and dogmatic. And perhaps it is the result of our constant search for new directions in research. One might also ask why other ideas (such as those of bounded rationality) are less welcome than those of behavioral economics. I think that this is because the profession prefers progress in small steps. The models of behavioral economics are not that different from those of applied economics and thus are not perceived as a threat.

Introduction

This chapter contains some comments on the language used by game theorists. As such, it is quite different from the rest of the book. Why am I interested in the study of the rhetoric of Game Theory? Admittedly, “I am not thrilled” with the fact that Game Theory is viewed by many as “useful” in the sense of providing a guide for behavior in strategic situations. This view is reinforced by the language we game theorists use.

Consider, for example, John McMillan's Games, Strategies and Managers (1992). On the cover of the paperback edition is a quote from Fortune praising the book as “the most user friendly guide for business people,” and Akira Omori of Arthur Andersen Co. is quoted as saying that “[the book] will be helpful both for beginning managers and for advanced strategic planners. In fact, I would recommend that anyone engaged in the US–Japan negotiations read the book.”

Or consider Avinash Dixit and Barry Nalebuff's bestseller, Thinking Strategically. The Financial Times (December 7, 1991) says of the book: “Thinking Strategically … offers essential training in making choices and weighing possibilities not only in business but in daily life.” Schlomo Maital says in a review of the book published in Across the Board (June 1991): “Ever since 1926, when John von Neumann, a brilliant Hungarian mathematician and physicist, published his path-breaking paper on Game Theory, it has been known that analytical models of games could be built.

At the opening of this short series of lectures, I briefly presented five issues which fall under the heading of “economics and language.” By now you have realized that the issues were indeed quite distinct. I found it difficult to point out the common factor among the topics other than the title and the speaker.

I am deeply grateful to three people who have (surprisingly) agreed to comment on this manuscript. Tilman Börgers, who was also the original discussant in Cambridge in 1996, Bart Lipman, and Johan van Benthem. Tilman and Bart bring some perspectives from economists, and Johan brings his perspective as a logician. It must be hard to write such a review, especially when one wants to criticize the author. Thus, I wish that the reader will pull out from the comments especially the critical comments and the suggestions for further research.

The three discussants helped me out in finding the common features of the first four essays. As Bart Lipman puts it, the first four chapters touch three related questions: “What gives a statement its meaning?,” “Why is language the way it is?,” and “How does language affect action?”

The discussants also persuaded me that the last chapter is very different in nature. If it belongs to this book beyond my rhetorical use of the phrase “economics and language”, it is only because it presents my general approach to economic theory, which is a thread running through these lectures.

Introduction

This chapter will present a research agenda whose prime objective is to explain how features of natural language are consistent with the optimization of certain “reasonable” target functions. Rather than discuss the research agenda in abstract, I will begin with the specific argument and return to the general discussion at the end of the chapter.

This chapter discusses binary relations. A binary relation on a set Ω specifies a connection between elements within the set. Such binary relations are common in natural language. For example, “person x knows person y,” “tree x is to the right of tree y,” “picture x is similar to picture y,” “chair x and chair y are the same color,” and so on. I will avoid binary relations such as “Professor x works for university y” or “the Social Security number of x is y,” which specify “relationships” between elements which naturally belong to two distinct sets. I will further restrict the term “binary relation” to be irreflexive: No element relates to itself. The reason for this is that the term “x relates to y” when x = y is fundamentally different from “x relates to y” when x ≠ y. For example, the statement “a loves b” is different from the statement “a loves himself.”

Certain binary relations, by their nature, must satisfy certain properties.

The psychologist Joel Davitz once wrote: “I suspect that most research in the social sciences has roots somewhere in the personal life of the researcher, though these roots are rarely reported in published papers” (Davitz, 1976). The first part of this statement definitely applies to this book. Though I am involved in several fields of economics and game theory, all my academic research has been motivated by my childhood desire to understand the way that people argue. In high school, I wanted to study logic, which I thought would be useful in political debates or in legal battles against evil once I fulfilled my dream of becoming a solicitor. Unfortunately, I became neither a lawyer nor a politician, and I have since come to understand that logic is not a very useful tool in these areas in any case. Nonetheless, I continued to explore formal models of game theory and economic theory, though not in the hope of predicting human behavior, not in anticipation of predicting the stock market prices, and without any illusion about the ability of capturing all of reality in one simple model. I am simply interested in the reasoning behind decision making and in the arguments people bring in debates. I am still puzzled, and even fascinated, by the magic of the links between the formal language of mathematical models and natural language. This brings me to the subject of this lecture – “Economics and Language.”

Grice's principles and game theory

The study of languages (see Levinson, 1983) is traditionally divided into three domains: syntax, semantics, and pragmatics. Syntax is the study of language as a collection of symbols detached from their interpretation. Semantics studies the rules by which an interpretation is assigned to a sentence independently of the context in which the sentence is uttered. This chapter deals with the domain of pragmatics.

Pragmatics examines the influence of context on the interpretation of an utterance. An “utterance” is viewed as a signal which conveys information within a context. The context comprises the speaker, the hearer, the place, the time, and so forth. How the hearer views the intentions of the speaker and how the speaker views the presuppositions of the hearer are relevant to the understanding of an utterance. Thus, in game-theoretic terms, the way in which an utterance is commonly understood may be thought of as an equilibrium outcome of a game between speakers of a language.

To illustrate the type of phenomena explained by pragmatics, consider the following natural language conversations:

Example 1: A telephone conversation between A at home and his friend B, calling from a telephone booth:

1. A: “B, I am about to go for a walk. What is the weather like outside?”

2. B: “It is not raining heavily now.”

Normally, A concludes from B's statement that it is raining but not heavily.

Introduction

The starting point of this chapter is the view that decision makers deliberating before making a choice often verbalize their considerations. It follows that the “language” a decision maker uses to verbalize his preferences restricts the set of preferences he may hold. Thus, interesting restrictions on the richness of the decision maker's language can yield interesting restrictions on the set of an economic agent's admissible preferences.

Before we proceed to the main investigation, some background is required. An economic agent in a standard economic model possesses a preference relation defined on a set of relevant consequences. The preferences provide the basis for the systematic description of his behavior as well as for welfare analysis. We usually assume that an economic agent is “rational” in the sense that his choice derives from an optimization given his preferences. Given that we adopt the rational man paradigm, the other constraints imposed upon an economic agent's preferences are often weak. For example, in general equilibrium theory, we usually only impose conditions such as monotonicity, continuity, and quasi-convexity. On the other hand, we restrict the discussion in many economic studies to some family of preference relations which share a simple utility representation. Several questions thus arise: What is the reason that we do not further restrict the set of preferences in the many cases where generality prevents us from obtaining stronger results?