Theories of Justice

Alejandra Mancilla, Tom Campbell, Tom Campbell

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781351879705
• Edition: 1
• Topic: Droit
• Subtopic: Jurisprudence

Forty years ago, in his landmark work A Theory of Justice, the American philosopher John Rawls depicted a just society as a fair system of cooperation between citizens, regarded as free and equal persons. Justice, Rawls famously claimed, is 'the first virtue of social institutions'. Ever since then, moral and political philosophers have expanded, expounded and criticized Rawls's main tenets, from perspectives as diverse as egalitarianism, left and right libertarianism and the ethics of care. This volume of essays provides a general overview of the main strands in contemporary justice theorising and features the most important and influential theories of justice from the 'post Rawlsian' era. These theories range from how to build a theory of justice and how to delineate its proper scope to the relationship between justice and equality, justice and liberty, and justice and desert. Also included is the critique of the Rawlsian paradigm, especially from feminist perspectives and from the growing strand of 'non-ideal' theory, as well as consideration of more recent developments and methodological issues.

Part I
Theorizing Justice

[1]
Two ways to think about justice

David Miller

Nuffield College, Oxford University, UK

abstract

This paper contrasts universalist approaches to justice with contextualist approaches. Universalists hold that basic principles of justice are invariant—they apply in every circumstance in which questions of justice arise. Contextualists hold that different principles apply in different contexts, and that there is no underlying master principle that applies in all. The paper argues that universalists cannot explain why so many different theories of justice have been put forward, nor why there is so much diversity in the judgements that ordinary people make. Several strategies open to universalists are considered and found to be wanting. Contextualism is defended against the charge that it cannot explain why contextually specific principles are all principles of justice, the charge that it can offer no practical guidance when principles conflict, and the charge that it inevitably collapses into a form of conventionalism.

keywords

justice, universalism, contextualism, conventionalism, Rawls, Walzer

Introduction

Justice is one of the oldest topics in political thought, but as anyone working in contemporary political theory will testify, it shows no sign of becoming exhausted. The last quarter of the 20th century produced a cornucopia of new theories, the landmark work of Rawls being followed by the alternative accounts of social or distributive justice proposed by Barry, Cohen, Dworkin, Nozick, Sen, Steiner, Walzer and many others. Pessimists might want to argue that the Owl of Minerva was going about its business in the usual way, and that philosophizing about justice reached its peak at precisely the moment when the capacity of the modem state to implement any recognizable scheme of social justice was about to disappear. But I shall put that pessimistic thought aside¹ in order to draw attention to a second feature of this literature on justice, namely that there is no sign of an emerging consensus on how justice is to be understood—no sign that, say, a suitable refined version of the Rawlsian theory can yield principles of justice that everyone recognizes as valid. So the situation we are in is one of intense interest in the topic of justice (and I am concerned here primarily with distributive justice in the broadest sense, that is, the just allocation of resources and benefits of all kinds, rights, freedoms, opportunities, property, income and so forth between persons) together with fairly radical disagreement as to which theory of justice is actually correct.

How should we respond to this state of affairs? One answer was provided a short while back by Alasdair MacIntyre, when he argued that modern liberal societies lacked the social forms (he called them ‘practices’) within which ideas of justice could be given a determinate content. As a result, he said, ‘we have all too many disparate and rival moral concepts, in this case rival and disparate concepts of justice, and… the moral resources of the culture allow us no way of settling the issue between them rationally’.² I believe that MacIntyre's diagnosis was wrong,³ but the phenomenon he was drawing attention to is certainly one that deserves our sustained attention. What, after all, is the point of theorizing about justice if the only result of our efforts is to produce yet one more theory to lay alongside the existing array, with no real prospect of converting anyone who already adheres to a rival theory?

We develop theories of justice, I shall assume, because we want to order and explain our intuitive beliefs about what fairness requires in different situations—beliefs that are at least to some degree uncertain and conflicting, whether within each person's thinking, or between different people. If we all knew exactly what justice required of us, and our conclusions were all the same, the intellectually curious might still want to produce explanatory theories, but the need to do so would be far less pressing than it is for us now. A good theory, I shall also assume, is one that fulfils Rawls's condition of reflective equilibrium.⁴ According to Rawls, theory construction must begin with our ‘considered judgements’ of justice—those judgements about the justice of practices or institutions in which we feel confidence and that, as far as we can tell, are not distorted by emotion, self-interest and so forth. The theory should systematize and explain those judgements, although it may also cause us after due reflection to abandon some of them if they turn out to be inconsistent with the theory itself. This requirement that there should be a reasonably close fit between the theory and our pre-theoretical considered judgements can, I think, be defended in two ways. First, our aim should be to develop a theory of justice specifically, not a theory of political morality in general, and to succeed in this aim we must make use of the conceptual markers that are present in everyday judgement. People think and say that this is a matter of justice but that is not, and unless we pay attention to such distinctions we can have no warrant for saying that what we have produced is indeed a theory of justice.⁵ Second, our aim is to develop a theory that has practical force, in the sense that people will be motivated to act on its requirements. Sooner or later, we hope, the principles laid out in the theory will command widespread consent, and we can aspire to create a political community effectively governed by principles of justice that its members themselves endorse. But unless we envisage a Damascene conversion on a mass scale, such a hope is forlorn unless the theory can incorporate in the right kind of way the considered judgements that people already make. It may be possible to shift quite radically people's concrete judgements about particular institutions or practices (say about the justice of income distribution in capitalist societies), but not, I am assuming, the underlying principles that inform these judgements.

Universalists and contextualists distinguished

I have briefly described the purpose of setting out a theory of justice. My aim in this article is not to put forward a new substantive theory, but to lay out and examine two broad ways of thinking about justice—two contrasting assumptions about the kind of theory we are looking for. On the one hand, we have the view that I shall label ‘universalism’. On this view, our aim is to discover principles of justice that can and should guide our judgement and our behaviour in all circumstances—principles that are universally applicable in the sense that whenever we have to assess the justice of a decision, or an institution, or a policy, we appeal, at base, to the same considerations. Universalists do not deny that the way these principles are applied may depend on circumstances, and this may make it seem as though, when we think about what justice requires, we appeal to different principles on different occasions. But on the universalist view, this appearance is misleading, because the basic principles of justice are invariant.⁶ Our aim in producing a theory, therefore, should be to lay out and explain those basic principles, from which it follows of course that the theory itself is circumstance independent: it tells us what justice is, not merely what justice requires given a particular set of circumstances.

In contrast, there is the view that I shall call ‘contextualism’. Contextualists assume that principles of justice are context specific rather than invariant across contexts. In other words, we must first ask questions of the general form ‘Who is distributing what to whom and in what circumstances?’, and only once that is resolved are we in a position to say which principles of justice are relevant to the decision or policy in question. Moreover, this context dependence of principles of justice goes all the way down, so that the contextually specific principles are free-standing, and not simply derivatives of invariant basic principles. On this view, the kind of theory we should be looking for is one that connects principles to contexts in a systematic way. Rather than laying out principles PI… Pn as constituting justice in all circumstances, it should take the form ‘In Cl, PI; in C2, P2;… in Cn, Pn,’ where the Cs are the distributive contexts in which principles of justice may be applied.

This contrast between universalism and contextualism needs to be distinguished from two others. It is not the contrast between monism and pluralism—between the view that justice can be captured in a single principle and the view that justice comprises two or more independent principles. Of course, any interesting form of contextualism must also be pluralist, since there is no point in distinguishing contexts unless one thinks that different substantive principles of justice apply in each. On the other hand, universalists may be monists, but they need not be. A monistic form of universalism would hold that justice in all its applications can be captured by a single principle, such as a principle of equal rights or a principle of desert. But universalists may also believe that justice comprises a number of independent principles that are either to be applied in lexical order, or perhaps traded off against each other. What makes this a universalist position is that the same principles are applied in all situations. Admittedly, there is something a little puzzling in the idea that justice might be pluralistic all the way down, and yet not contextual in nature. One wants to know what the source of the pluralism is, and pressing this question may have the effect of revealing a would-be universalist as a closet contextualist (I shall come back later to ways of dressing up contextualist theories as universalist). Yet, formally at least, the distinction between universalist and contextualist theories is not the same as the distinction between monistic and pluralistic theories of justice.

Second, the contrast between universalism and contextualism is not the same as the contrast between objectivism and relativism. Although it is sometimes alleged that contextualist theories of justice must collapse into conventionalism or relativism (this allegation is one that I shall address in due course), a contextualist understanding of justice makes claims that are themselves objective and universal in character. As indicated above, a contextualist theory will take the form ‘In C1, P1; in C2, P2;… in Cn, Pn’ This entails that whenever a decision has to be made or a policy followed in context C1, the correct principle to apply is P1 Contextualists tend to assume that, when we look at different societies, we find many contexts of distribution reiterated across them, although no two societies are likely to be exactly the same in this respect. So a full account of justice in society S1 will reveal that its requirements there are not identical with those in S2, but this difference can in turn be explained by differences in the set of distributive contexts found in SI and S2. A further implication is that what justice requires in S1 is not necessarily what the inhabitants of S1 believe justice to require in their society. Contextualism leaves room for the claim that people's beliefs about justice may in certain respects be mistaken, for instance, if they hold that P2, rather than PI, is the appropriate principle to apply in C1.

I have presented universalism and contextualism as though they were sharply opposed ways of thinking about justice, but there are also different ways of holding the middle ground (which can be presented as universalist concessions to contextualism, or vice versa). For example, one may develop a theory of justice whose form appears to be universal, but then add that this theory applies only to societies of a certain type—modern liberal democracies, for example. Unless this theory can be represented as the application of a still more basic theory that applies universally, adding the rider involves injecting a contextualist element into the theory. Or again, one may draw a distinction between justice within political communities and justice between political communities (or international justice) and offer separate accounts of each, which again amounts to moving somewhat in the direction of contextualism, though not as far as someone who begins by distinguishing different sub-contexts inside political communities.⁷ Perhaps the most defensible theory will turn out to be located somewhere on this middle ground. Nevertheless, I want to continue with the basic contrast, and in particular to see what can be said in favour of the contextualist way of thinking, because I believe that this approach has been short-changed in recent political philosophy. Universalism comes naturally to philosophers as an account of what they are doing when they think about justice, or indeed about any moral or political concept, whereas anthropologists, sociologists and social psychologists are likely to assume as a matter of course that justice will take on different forms in different social contexts. Faced with this fact, philosophers are apt to respond that what these social scientists are up to is merely description, whereas their own concern is normative. I have tried to show elsewhere why I regard this antinomy as unhelpful.⁸ Here my aim is to examine problems facing universalist theories of justice, and to show how different ways of escaping from those problems concede a good deal to contextualism. Then I ask whether contextualist theories can deal successfully with the problems that they face.

Refining the distinction

But first, I need to say a little more about how the line between universalism and contextualism is to be drawn. There is a danger that the real division between these two ways of thinking about justice becomes blurred by presenting as contextualist theories whose real structure is universalist, and vice versa. So let me begin with a few clarificatory remarks about universalism. Universalists about justice believe that we can identify fundamental principles of justice that apply in every circumstance in which questions of justice arise, regardless, for instance, of the specific type of resource that is being distributed, the institutional setting in which the distribution is taking place, and so on. But this holds only at the fundamental level. At a more applied level, universalists can make room for variation. For instance, they can defend secondary principles that represent the best way of cashing out the fundamental principles in particular social circumstances. Suppose that a universalist takes equality of resources as her fundamental principle of justice. Following Ronald Dworkin, she might argue that in contemporary societies the best way to operationalize this principle is to treat resources as privately owned and to allocate them by means of a Walrasian auction.⁹ But she might also argue that in quite different circumstances (say, in a technologically undeveloped society) equality of resources might be best achieved by giving each person equal access to resources that remain collectively owned.

Furthermore, some principles defended by universalists may be of such a kind that their implementation necessarily depends on contingent features of the social context in which they are being applied. The Rawlsian difference principle (‘Inequalities of income and wealth are to be arranged to the greatest benefit of the least advantaged’) provides a good example. The shape of the monetary distribution that would result from applying this principle depends on whether and how economic inequalities have the effect of improving the position of the worst off, and this in turn depends on contingent facts such as the extent to which different jobs require special talents to be performed effectively, the extent to which material incentives are needed to attract talented people into those jobs, and so forth. Thus someone applying the difference principle might judge both egalitarian society A and highly inegalitarian society B to be just if it turned out that in each case the worst-off group were doing as well as they could given the relevant facts in each society.

Lastly, universalists may allow that the answer to the question ‘How much justice can be achieved in this society?’ will also depend on contingent circumstances. They may concede that it is simply impossible to implement the principles that they take to define justice fully in some societies, or that the costs of doing so would be too great in the light of other values. In other words, universalists need not hold a strong view about the priority of justice in order to be universalists. They can allow for trade-offs between principles of justice and other political principles. And so, they can judge that society S is as just as it reasonably can be, meaning that further attempts to make it conform to their conception of justice would be too costly in the light of these other principles.

With these clarifications in place, what...