Philosophical Problems of Behavioural Economics

Stefan Heidl

• Publisher: Routledge
• Year: 2016
• Topic: Business
• eBook ISBN:
  9781317267836
• Subtopic: Business General
• Index: Business

The goal of behavioural economics is to improve the explanatory and predictive power of economics. This can be achieved by using theoretical and methodological resources of psychology. Its fundamental idea is that the relationship between psychology and economics cannot be subsumed under standard philosophical accounts of intertheoretical relations.

Philosophical Problems of Behavioural Economics argues that behavioural economics is best understood as an attempt to deidealize economic theory guided by psychological research. Behavioural economics deconstructs the model of decision-making by adding different elements. Based on this understanding behavioural economics has a number of tasks: first, it has to identify which economic theory needs to be challenged; second it aims to identify factors which need to be modelled within economic theories of choice and modify the theory accordingly; and finally, it has to create models that explain economic phenomena based on the new theory. This book analyses the different stages of this deconstruction process and shows how the scientific disciplines of economics and psychology are connected by it.

This volume develops a new account of intertheoretical relations based on the idea of deidealization and thus contributes to debates within the philosophy of social science. It is suitable for those who are interested in or study economic theory and philosophy, economic psychology and philosophy of social science.

1 Introduction

DOI: 10.4324/9781315637372-1

1.1 Introduction and motivation

This book takes as its starting point a methodological debate among economists about the relation of economics and psychology. This debate has grown in intensity in recent years because of the rise to prominence of the economic subdiscipline of behavioural economics. Adherents of behavioural economics claim that it is characterised by the application of psychological insights to economics (Camerer and Loewenstein, 2004: 8).

I want to analyse the methods and theories of behavioural economics to determine its nature and work out how behavioural economic theory is related to standard economic theory and psychology.

In this chapter, I will give an overview of behavioural economics, show that the relationship between psychology and economics as it is featured in behavioural economics cannot be subsumed under the standard account of intertheoretic relations, and then present a sketch of my analysis of behavioural economics.

I start in Section 2 with an overview of ‘new behavioral economics’ (Sent, 2004), which is the behavioural economic research project that I will analyse in the following chapters. The term ‘behavioural economics’ is applied to a number of different research projects that have all in common that they attempt to integrate psychological and economic research.

New behavioural economics started in the late 1970s with the work of Daniel Kahneman and Amos Tversky. Other types of behavioural economics try to develop new assumptions for economic theory based on descriptions of the processes of human decision-making. In contrast, new behavioural economics investigates how the behaviour of people deviates from the behavioural assumptions of standard economic theory and modifies these assumptions based on the resulting pattern of deviations (Sent, 2004: 742). In what follows, when I talk about behavioural economics I refer specifically to the research project of new behavioural economics.

In Section 3 I introduce the classical model of intertheoretic relations and argue that behavioural economics cannot be understood according to this account. I also discuss a study from the philosophy of biology that investigates non-reductive intertheoretic relations. I illustrate the character of my research by comparing it to this study. I argue that behavioural economics is best understood as an attempt to deidealise economic theory. As understood by philosophers of science an idealisation is a ‘deliberate simplification of something complicated with the objective of making it more tractable’ (Frigg and Hartmann, 2012, sect. 1.1). I call standard economics idealised because it uses a ‘highly simplified and stylized model of human cognition, preferences, and behaviour, that in every instance, omits a tremendous amount of psychological reality’ (Rabin, 2002: 672). Behavioural economics can be understood as deidealising the model of the economic agent in the sense of adding elements to the model of human cognition, preferences and behaviour from which the standard theory abstracts. These attempts at deidealisation are guided by experiments conducted by both psychologists and economists and the hypotheses investigated in these experiments are inspired by theories from cognitive and social psychology.

In Section 4 I summarise the content of the chapters that follow.

1.2 Behavioural economics

Behavioural economists want to improve the explanatory and predictive power of economics by using the theoretical and methodological resources of psychology. Behavioural economics is said to provide economics with ‘more realistic psychological foundations’ (Camerer and Loewenstein, 2004: 3), to improve ‘the realism of psychological assumptions underlying economic theory’ (Camerer, 1999: 10575) or to emerge out of ‘efforts to incorporate more realistic notions of human nature into economics’ (Rabin, 2002: 657).

Improving the explanatory power of economics by integrating new psychological assumptions is a feasible strategy because explanations of economics employ models of economic agents to explain how economic phenomena arise out of the interaction of these agents. These models are constructed with the help of theories of rational decision-making. Psychological studies of individual decision behaviour seem relevant to economics because the knowledge gained in these investigations can be used to improve the theories of choice on which models of economic agents are based.

Such a view is taken by Colin Camerer and George Loewenstein, who think that assumptions about the behaviour of economic agents in economic theory are implicitly ‘behavioral’ or psychological, and that economic theory should be connected to psychological theories and data (Camerer and Loewenstein, 2004: 39). Psychological theories of individual behaviour should underlie economics because economics analyses how individuals allocate resources. It can thus be improved by including more ‘realistic’ descriptions of individual decision behaviour (Camerer, 1999).

An example of such an implicitly behavioural assumption is that some features of financial markets are explained by the attitudes of investors to risk. These attitudes influence how the investors choose among financial products. Behavioural economists argue that assumptions about the risk behaviour of agents should be grounded in empirical results concerning how people evaluate risky financial products.

In the next two sub-sections, I will describe which parts of economics and psychology are connected within behavioural economics. I am going to start with the economic theories that are used to create models of economic agents.

1.2.1 Models of economic agents

The standard model of the economic agent assumes that agents have preferences according to which they rank bundles of goods. Standard assumptions about the preferences of an agent are that they are complete and transitive. Completeness means that the agent can compare any two bundles of goods. Transitivity means that where we have three bundles a, b and c, and a is preferred to b and b is preferred to c, that a is also preferred to c.

An agent is called rational if her preferences are complete and transitive and she chooses her most preferred bundle of goods. If the preferences of the agent fulfil these conditions one can write down the available bundle of goods in a list in which the most preferred is at the top and the least preferred at the bottom. If one has such an ordering of the bundles of goods, it is possible to assign numbers to each line of the list so that bundles higher on the list get a higher number. Such a number is called the utility of the specific bundle and represents its positions in the total preference ordering of the agent.

This simple theory of the agent can be employed to describe consumer choices. Given her preferences for goods, prices and her income it is possible to predict which goods the agent demands and how her demand changes as a reaction to changes of prices or her income.

Expected utility theory

This basic model of the rational economic agent can be extended in various ways. An important extension is the one to decision-making under uncertainty. Decision-making under uncertainty describes situations in which the agent is unsure about the outcomes of her actions. Examples of such actions are the purchase of an insurance or an investment in risky stocks. The extension to decision-making under uncertainty is important because many economic decisions contain an element of uncertainty.

Such decisions are modelled as choices between different lotteries. A lottery is described by its possible outcomes and the associated probabilities. An example would be the lottery (x, p; y, 1 – p). By playing this lottery one gets either the outcome × with probability p or the outcome y with probability 1 – p.

The standard theory of choice under uncertainty is expected utility theory. According to expected utility theory, an agent chooses between lotteries by comparing the expected utility of the lotteries. The expected utility of a lottery is the sum of the utilities of the outcomes each weighted with its associated probability. The following is a simple example of such a choice in which the agent chooses between either taking part in a lottery or keeping her current wealth:

The agent has a current wealth w of 10. Her preferences for money are described by the utility function:

She can participate in the following game:

0.5 chance of winning 6, 0.5 chance of losing 6.

In the case she wins her wealth will be 16; she attaches a utility value of 4 to this wealth. If she loses, her wealth will be 4 with an associated utility of 2. The expected utility of participating in this game is 0.5*4 + 0.5*2 = 3. The –– utility of participating in the game is 3 which is lower than , which is the utility associated with her current wealth. So expected utility theory predicts that the agent prefers keeping her current wealth to taking part in the gamble.

Like standard utility theory, expected utility theory assumes complete and transitive preferences. In addition, it demands that preferences among lotteries fulfil the independence axiom, which states that the preference between two lotteries that differ in only one outcome should be identical to the preference between the two different outcomes.

Game theory

A third important theory used in modelling the behaviour of economic agents is game theory. Game theory is used to model situations of strategic choice. These are situations in which the agent interacts with other agents and the overall outcome depends on the choices of all the agents. To model such situations one needs to assume that the agents have models of the other agents that they use to predict their actions. The action of a single agent depends on her preferences and her beliefs about the actions of the other agents.

Types of strategic interactions are modelled as games. A game is specified by the number of players, the actions available to the players and the preferences of the players over the outcomes of the game. The outcomes of the game result from the combination of the different actions. Games are solved by applying a solution concept. A solution specifies a strategy, which is a plan for action, for each player so that no player regrets choosing her strategy. A commonly used solution concept is the Nash-Equilibrium. In a Nash-Equilibrium no player wishes to choose another action than the one she has actually chosen given the action of the other players.

An example of a game is the prisoner’s dilemma. It is used to model social situations in which people would be better off if they cooperated, but such cooperation is hard to ensure because there are individual benefits of defection. The story connected to the prisoner’s dilemma is the following: two people who committed a crime together are arrested by the police. The police do not have sufficient evidence to prove that the two committed the crime, so they try to get at least one of them to confess. For that purpose, the police separate them and offer the one who confesses the crime a reduced punishment.

If both players choose to ‘cooperate’ with the other player, neither of them confesses the crime and the police can only prove a minor offence, which means that both get a small punishment. If one cooperates and the other defects, this means that one prisoner confesses the crime and the other does not. The one who did not confess gets the full punishment, while the one who confessed gets no punishment at all. If both defect, both will be punished although less severely than if only one person was found guilty. The prisoner’s dilemma is characterised by a specific structure of pay-offs that is set out in Table 1.1. The pay-offs describe the preferences of the agents. The outcome with the pay-off 3 is the most preferred, while the outcome with the number 0 is the least preferred.

Table 1.1The prisoner's dilemma

	Cooperate	Defect
Cooperate	2,2	0,3
Defect	3,0	1,1

With the help of standard game-theoretic reasoning one comes to the result that (Defect, Defect) is the only Nash-Equilibrium in the prisoner’s dilemma. It is the only equilibrium in which no player can improve her outcome by changing her action. Given every other outcome each player has an incentive to deviate. For example given the outcome (Cooperate, Cooperate) it would be advantageous for each player to play ‘defect’ given that the other player plays ‘cooperate’ because this gives her the outcome associated with a utility value of 3.

The prisoner’s dilemma is considered an important game because many real-life situations are thought to be instances of it. Cooperation in the prisoner’s dilemma seems desirable but is, according to the standard theory, impossible to attain. Experimental research on the prisoner’s dilemma often seeks to determine whether people manage to cooperate.

For the analysis of the methods of behavioural econo...

Table of contents

• Cover Page
• Half Title Page
• Title Page
• Copyright Page
• Table of Contents
• List of tables
• Acknowledgements
• 1 Introduction
• 2 The ontological status of economic preferences
• 3 Selectionist arguments against a deidealisation of economics
• 4 Aggregation and the deidealisation of economic theory
• 5 Explanatory autonomy and intertheoretic relations
• 6 Localising anomalies of economic theory
• 7 The structure of behavioural economic theories
• 8 Models in economics: standard and behavioural
• 9 Conclusion
• Index