Choice, Rationality and Social Theory (RLE Social Theory)

Barry Hindess

• Publisher: Routledge
• Year: 2014
• Print ISBN: 9781138782259
• eBook ISBN: 9781317652137
• Edition: 1
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

Choice, Rationality and Social Theory is a powerful rebuttal of the remarkably influential theories underlying 'rational choice analysis'. Rational choice analysis maintains that social life is principally to be explained as the outcome of rational choices on the part of individual actors. Adherents of this view include not only philosophers, political scientists and sociologists, but also prominent politicians in Western governments – notably of the United Kingdom and the United States. Rational choice analysis is said to be rigorous, capable of great technical sophistication, and able to generate powerful explanations on the basis of a few, relatively simple theoretical assumptions.

Barry Hindess argues that the theory is seriously deficient, first, because there are important actors in the modern world other than human individuals, and second, because it says nothing about those processes of deliberation that play an important part in actors' decisions. The use of highly questionable assumptions about actors and their rationality has the effect of closing off important areas of intellectual inquiry and ignoring the reality of certain forms of thought and the social conditions on which they depend. These points are established through detailed examination of the concepts of the actor and of rationality – providing an overall argument that constitutes a serious challenge to any adherent of rational choice analysis.

1 Introduction

DOI: 10.4324/9781315763729-1

The terms ‘choice’ and ‘rationality’ are used in many different ways in the social sciences, but one of the most influential contemporary usages is to identify rationality with behaviour that maximizes the satisfaction of preferences. The rational choice approach proposes to analyse human behaviour on the assumption that actors are rational in just that sense. Much of social life is then to be explained as the outcome of the rational choices of individual actors. The argument of this book is directed against that and related explanatory usages of the notion of individual rationality. There are, of course, other influential conceptions of rationality (for example, in critical theory) but they are not my concern here.

Models of maximizing behaviour are widely used in economics, and rational choice analysis can be understood as extending that economic approach to other areas of human behaviour. It is in this spirit that Gary Becker informs us that the economic approach

is applicable to all human behaviour, be it behaviour involving money prices or imputed shadow prices, repeated or infrequent decisions, large or minor decisions, emotional or mechanical ends, rich or poor persons, men or women, adults or children, brilliant or stupid persons, businessmen or politicians, teachers or students. (Becker, 1976, p. 8)

Becker goes on to suggest that too many social scientists are tempted to hide their own lack of understanding of their subjects’ behaviour behind allegations of ‘ignorance and irrationality, values and their frequent and unexplained shifts, custom and tradition, the compliance somehow induced by social norms, or the ego and the id’ (ibid., p. 13). Here those of us who fail to adopt the economic approach are accused of a kind of intellectual deceit, hiding our own ignorance behind talk of the irrationality of others. Becker is far from being alone in claiming the moral high ground for his own theoretical assumptions and condemning those who disagree, and not all who do so favour his economic approach¹ – but it is a style of argument that deserves little consideration.

Becker’s argument on this point represents the weakest case for rational choice analysis, but there are more serious arguments to be considered. Adherents of rational choice analysis have claimed that it is rigorous, capable of great technical sophistication and able to generate powerful explanations across a wide range of situations on the basis of a few, relatively simple theoretical assumptions.² It also appears to take more seriously than sociological functionalism or structuralism the point that actors do indeed make decisions and act on them. It insists that actors are not mere creatures of their position in some overarching societal totality. No one, of course, would be so foolish as to maintain that action is always rational. The claim of rational choice analysis is not so much that the assumption of rationality is descriptively accurate but rather that it performs an important heuristic function: it is a useful simplification and it provides the means of identifying the place of non-rational elements in human behaviour.

On this view, non-rational elements may be introduced in our analyses only when explanation in rational terms has clearly failed. Jon Elster, for example, clearly recognizes that action is not always rational and that there are problems with the identification of rationality with maximizing behaviour. He nevertheless insists that ‘explanation in terms of optimization remains the paradigm case of intentional explanation in the social sciences outside psychology’ (Elster, 1983a, p. 75). All other cases of intentional explanation are then to be understood as involving specific departures from the paradigmatic norm.

This book is a critical discussion of some of the more serious claims of rational choice analysis. The clarity and technical sophistication of much rational choice analysis seem to me undeniable, and I do not dispute its claim to a certain kind of rigour. Unfortunately, those positive features of rational choice analysis depend on the adoption of a number of highly questionable assumptions about actors and about their rationality. Those assumptions are the target of my critical discussion. I argue that the theoretical parsimony and explanatory power of rational choice analysis are bought at far too high a theoretical cost, and I dispute both the assumption of actors’ rationality and the paradigmatic status assigned to it.

It is important to be clear what is at stake in this last point. It would not be difficult to find instances of behaviour that seem far from rational, or areas of social life where the assumptions of the economic approach appear not to be appropriate. Hardin, for example, suggests that those assumptions yield ‘a notoriously poor explanation of voting behaviour’ (Hardin, 1982, p. 11) and Barry (1978) makes a similar point for political participation in general. Or again, sympathetic critics like Sen and Elster, and others who are not so sympathetic, have argued for a more complex view of actors’ rationality than is normally provided in rational choice models.

Now, such points may dispose of Becker’s more extravagant assertions but in themselves they do little damage to the more serious claims of rational choice analysis. These treat the assumption of rationality as a simplifying assumption, on the grounds that while it is not always realistic it does provide a paradigm for the analysis of other cases. In effect, we start from the assumption of rationality and consider other possibilities only when explanation in rational terms has clearly failed. The result of this manoeuvre is that criticism based on cases of non-rational behaviour or on a more sophisticated notion of actors’ rationality can often be deflected without much difficulty. Of course, the argument goes, people are not always rational, but the assumption of rationality allows us to identify cases where other elements affect their behaviour. Of course, people are not always narrowly self-interested, but the assumption that they are allows us to identify the role of other motivations in their behaviour.

The point of claiming paradigmatic status for the model of optimizing behaviour is not then to deny the existence of other forms of behaviour. On the contrary, it recognizes cases where the assumption of actors’ rationality does not apply and identifies their theoretical location – precisely as specific complications of or departures from the paradigmatic norm.³ I argue that important questions of the forms of thought employed by actors and the social conditions on which they depend are obscured by the paradigm of optimizing behaviour.

The chapter following this short introduction provides a preliminary sketch of the basic structure of rational choice models and the variety of their applications, taking examples from a number of areas – discussions of government taxation and spending policies, explanations of economic growth and decline, and attempts to elaborate a rational choice Marxism. Many authors claim to draw political conclusions from their abstract analyses. In The Rise and Decline of Nations, to take just one example, Olson concludes that ‘the best macroeconomic policy is a good microeconomic policy. There is no substitute for a more open and competitive environment’ (Olson, 1982, p. 233). Not all supporters of the rational choice approach would agree with Olson on this point. Nevertheless, his attempt to draw political conclusions from abstract analyses of the behaviour of rational actors raises important questions of what we should look for in discussions of social phenomena. I return to that issue in the concluding chapter.

Chapter 3 continues this preliminary discussion by examining the distinctive model of the actor employed in the rational choice approach. First, actors are assumed to act rationally in terms of a relatively stable set of beliefs and desires. Secondly, it is often supposed that they are narrowly self-interested. We shall see that this second assumption can be relaxed without serious damage to the rational choice approach. Finally, there is an explicit methodological individualism which presents the structural features of social life as if they were reducible to the actions of rational individuals and their (often unintended) consequences. I agree that actors’ decisions should not be seen as effects of their positions in some overarching social structure, but it does not follow that social life is therefore reducible to the actions of individuals. Against that reductionism I argue in later chapters that actors’ decisions and actions depend on conditions that are external to the actor concerned.

These two chapters prepare the ground for the critical argument that follows. The model of the actor employed in rational choice analysis involves specific refinements of a model that is widely used in philosophy and the social sciences. In this more general model, actors are assumed to be human individuals and their actions are supposed to follow from their beliefs, desires and other states of mind. In effect, the actor carries a portfolio of beliefs and desires around from one situation to another. Given the situation of action, the actor selects from its portfolio those elements that seem relevant and uses them to decide on a course of action. I call this the portfolio model of the actor. Rational choice analysis modifies the portfolio model in two important respects. First, it treats the actor’s desires as exhibiting a utilitarian structure so that an optimal outcome can normally be defined in most situations confronting the actor. Secondly, it assigns a paradigmatic status to the assumption of rationality. I argue against the rational choice model both by questioning its refinements of the more general portfolio model and by disputing the portfolio model itself.

Chapter 4 discusses models of the actor. I begin by presenting an abstract concept of actor as locus of decision and action. Actors do things as a result of their decisions, and we call those things actions. Actors’ decisions play an important part in the explanation of their actions. Actors may also do things that are not the result of any decision, and they must be explained in some other way. Now, the portfolio model builds far more into its concept of actor than is provided in the minimal concept presented here. First, it assumes that actors are human individuals. I argue on the contrary that there are important actors in the modern world other than human individuals. Capitalist enterprises, state agencies and political parties are all actors in the minimal sense that they have means of reaching decisions and of acting on some of them. Any analysis of modern societies that treats human individuals as the only effective actors must be regarded as seriously incomplete.

Secondly, the portfolio model treats action as a function of belief and desire. Intentional analysis then requires that we work back to actors’ beliefs and desires by constructing an interpretation of their behaviour, including, of course, what they say. Disputes over rationality have played an important part in discussion of what is involved in understanding the behaviour of others and especially of other cultures.⁴ Davidson and other philosophers have presented powerful arguments to the effect that the process of constructing an account of actors’ beliefs and desires from observation of their behaviour in various contexts requires us to presume a fair degree of rationality and consistency in their behaviour. The assumption of rationality and consistency, which Davidson calls the principle of charity, would then be an essential part of any attempt to understand the behaviour of others. No discussion of the place of rationality in the analysis of human behaviour can afford to ignore these arguments.

Chapters 4 and 5 pay particular attention to the work of Donald Davidson, who has provided one of the clearest philosophical investigations of the character of intentional explanation. Chapter 4 examines the portfolio model and its assumption of an holistic rationality in detail. The final section contrasts the treatments of rationality in rational choice analysis and the more general portfolio model. The argument that interpretation of the behaviour of others requires a presumption of rationality does not entail the paradigmatic status which rational choice analysis claims for what it describes as rationality.

Chapter 5 questions the portfolio model itself by disputing Davidson’s argument that intentional analysis must presuppose an holistic rationality. The problem with the treatment of action as resulting from belief and desire is that it says nothing about those processes of deliberation that sometimes play an important part in actors’ decisions. More precisely, it takes the rationality of those processes for granted by treating them as transparent intermediaries between belief and desire on the one hand and the action that results from them on the other. Once some definite process of deliberation is admitted as an element in the actor’s decision then that process and the techniques and forms of thought employed within it must be regarded as objects of investigation.

I consider two influential ways in which the rationality of actors’ deliberations has been brought into question. One is Simon’s concept of satisficing or bounded rationality, and the other is the claim that there are distinct ‘rationalities’ or styles of reasoning. The latter has to be treated with some care. A common way of looking at different ways of thinking is in terms of a distinction between conceptual schemes and reality: different schemes produce different perceptions of reality.⁵ Davidson has shown that the conceptual relativism involved here cannot be sustained. But there is another relativism concerning distinct styles of reasoning that does not fall under Davidson’s critique. This has devastating consequences for the assumption of an holistic rationality. First, if specialized techniques may be employed in actors’ deliberations then the problem for intentional analysis is to identify those techniques and to investigate the conditions on which their usage depends, not to take their rationality for granted. Secondly, there is no reason to suppose that the specialized techniques employed by actors in different areas of their activity will be consistent with each other. There is no reason, in other words, to suppose that actors will exhibit an holistic rationality. The final section of the chapter shows that these arguments undermine the characterization of action in terms of a more or less stable portfolio of beliefs and desires.

The final chapter considers the implications of these arguments, both for rational choice analysis and for the portfolio model generally. Human individuals and social actors use conceptual and other tools in their deliberations. Societies will differ in the range and variety of conceptual tools that are available to their members. In any given society the use of some tools will be commonplace, and the use of others (the poison oracle, cost-benefit analysis, geomancy) will be relatively specialized. The concerns and objectives that motivate actors and the results of their deliberations will depend on the techniques and forms of thought they are able to employ. The problem with the portfolio model is not so much that it denies the place of deliberation in actors’ decisions, but rather that it treats deliberation as if it were transparently rational and therefore of little explanatory significance. Rational choice analysis does the same, with a more restrictive notion of rationality. I argue on the contrary that the tools used by actors in their deliberations, their connections with other tools, and the ways in which they depend on social conditions are legitimate and important areas of investigation.

The concluding section returns to the claims of rational choice analysis to theoretical parsimony and explanatory power. I argue that many of the apparently powerful results of rational choice analysis depend on an implicit structural determination of the forms of thought employed by actors. The problem here is simply stated: the assumption of actors’ rationality tells us that there will be a certain consistency in their behaviour, but it tells us nothing of the substance of their concerns. The explanation of social conditions and structural features of social life as resulting from the rational actions of large numbers of individuals therefore requires some further assumption regarding the social distribution of actors’ concerns and objectives. The most common assumption is that what is to count as rational action for some particular actor is a function of that actor’s membership of one of the social categories recognized by the rational choice model in question; political leaders, say, behave as they do because they are political leaders and because they are rational. Here an important part of actors’ concerns is determined by the social conditions in which they find themselves. I argue on the contrary that actors’ concerns and objectives depend in part on the techniques and forms of thought they are in a position to employ. While there are always connections between actors’ social location and the forms of thought they employ in deciding on courses of action, there is no simple correspondence between them.

As for theoretical parsimony, rational choice analysis abstracts from the forms of thought employed by actors in their deliberations to produce an account of what hypothetical rational actors would do in their place. There are two important issues to be considered here. First, it is doubtful if any worthwhile political conclusions can be drawn from an analysis that deliberately abstracts from the forms of thought employed by the actors concerned in their deliberations. Secondly, parsimony in this case is bought at the price of closing off important areas of intellectual inquiry concerning the forms of thought employed by actors and the social conditions on which they depend.

2 The rational choice approach to social behaviour

DOI: 10.4324/9781315763729-2

This chapter provides a preliminary sketch of the basic structure of rational choice m...

Table of contents

• Cover Page
• Half Title Page
• Title Page
• Copyright Page
• Original Title Page
• Original Copyright Page
• Table of Contents
• Acknowledgements
• 1 Introduction
• 2 The rational choice approach to social behaviour
• 3 Rationality, egoism and social atomism
• 4 Models of the actor
• 5 Rationality, action and deliberation
• 6 Individualism and social structure
• Notes
• References
• Index