Sen’s ‘The Idea of Justice’ (Chapter 3, Institutions and Persons)

Sen’s purpose in this comparatively short chapter seems to be to draw a distinction between principles of justice that focus on institutions and those that focus on behaviour and so on consequences, and then condemn “institutionally fundamentalist” principles for failing to take account of what actually happens. This distinction is supposed to be illuminated or perhaps even typified by a contrast Sen begins the chapter with, between the policies of two figures from Indian history, Ashoka and Kautilya. Ashoka, having seen at firsthand the horrors of coercion and violence during a campaign to extend his empire, apparently renounced the normal means of exercising power and instead exhorted his subjects to behave virtuously in a way that Sen reads as showing that he equated moral knowledge and moral motivation in a rather simplistic way. Indeed, as Sen notes, part of the reason that the political order did not totally collapse once Ashoka gave up on enforcing his will through force seems to have been that the administrative reforms implemented by his grandfather’s advisor, Kautilya, had a life of their own. The idea appears to be that focussing on institutions exhibits Ashoka’s utopian idealism about the possibility of spontaneous moral reform whereas focussing on behaviour and its consequences is more like Kautilya’s pragmatic acceptance of human fallibility.

This contrast though, is a different one from the one between institutions and consequences as the relevant units of moral assessment. Since, as Sen makes clear, Kautilya’s reforms were institutional reforms, it could hardly be the same one. The reason for this is that what one thinks are the relevant units of moral assessment and how one thinks about the possibilities for human motivation are two, perhaps related, but nonetheless clearly distinct questions. Blain I think has already mentioned that Sen’s reading of Rawls in this kind of area may not be entirely sympathetic, but there seem to me other cases where this claim borders on the bizarre. Hobbes, for example, is in Sen’s typology an institutionally-focused theorist, as his focus on the figure of the Sovereign presumably justifies. Yet Hobbes is clearly not of the view that human moral motivation can be improved by moral education in anything like the way that Sen presents Ashoka as being. A focus on institutions rather than behaviour or consequences in stating principles of justice may be for a number of reasons. One is that institutions provide and in a certain sense are stable patterns of coordinated behaviour enforced through sanctions, and so can constrain human behaviour in ways that ensure that it remains within at least an acceptable set of outcomes. This, for example, seems to be Hobbes’ reasoning. Not only does that depend on views of human behaviour that Sen at least implies are anathema to institutionally-focussed theorists and so demonstrate the failure of his attempt to align views of human behaviour with the focus of principles of justice, but it also casts doubt on the way that Sen wants to exclude institutional theorists from having a concern with consequences.

This is, on Sen’s account, not a symmetrical exclusion. Theories which focus on realization may take account of the ways in which the consequences they are supposed to be focussing on are produced; the means to these ends can be incorporated into their assessment. The effect of this definition is to make the struggle that institutionally-focussed theories face more difficult by allowing their opponents access to all the resources they can draw on without any parallel expansion of the tools they can make use of. The way Hobbes, though, thinks about the value of institutions is clearly dependent on their consequences. It is not clear that the consequences have any existence independent of the institutions - part of Hobbes’ case is clearly that the only way out of the State of Nature is through a near-absolute sovereign - but neither is it as if the institutions have virtues independent of the production of those consequences.

Sen acknowledges this at times, as when he observes that the difference principle is clearly an institutional requirement related to consequences: it is the production of a particular set of consequences that mean that institutions fulfil the difference principle, after all. Yet he seems determined to force institutionally-focussed theories into a deontological straitjacket, where his paradigms of institutionally-focussed theory seem to be Nozick and Gauthier, despite the fact that his real target, Rawls, clearly does not think about duties in the same sort of way. Indeed, Rawls’ own critique of libertarianism is precisely that it is not institutionally-focussed in the right way, that libertarians do not have a theory of the basic structure, thinking of it not as the exercise of public power but as analogous to private contract (see for example PL, Lecture VII, §3).

One way of thinking about this would be to think of the sorts of theories that Sen wants to condemn as failing to take proper account of consequences as procedural: a libertarian entitlement theory is roughly procedural, just as Rawls claims that institutions which meet the difference principle are examples of pure procedural justice. Yet there are a variety of different forms of procedural justice, all of which may take some interest in consequences. Whether a procedure’s outcome is appropriate because of the procedure itself or because the procedure tracks some independent criterion, considerations about the state of affairs they realize can be incorporated into their justification. As far as perfect and imperfect procedures go, the independent criterion could relate to outcomes just as it could to deontological considerations, while pure procedures may have their constitutive rules justified on grounds of their consequences; presumably competitive sports are if played to the rules pure procedures, yet changes to their rules can be and are justified on grounds of improving the spectacle.

The pair of contrasts Sen draws and claims are aligned in this chapter then seem to me unhelpful. They do not align, and the polemical use Sen wants to put them to I think relies on skewing the deck against his opponents. Part of this seems to be because of Sen’s insistence that asking “how things are going and whether they can be improved is a constant and inescapable part of the pursuit of justice”. That is why it would be wrong not to make the realization of states of affairs the focus of principles of justice. Notice though, that if the means by which states of affairs come about can be incorporated into their assessment, that this tells us nothing about whether consequences or the means of bringing them about are what can be improved. As long as deontological restraints are part of what matters from the perspective of justice, then until we know what deontological restraints are properly included in assessments of justice, this assertion is perfectly compatible with more or less any theory of justice. More, it is unclear why a perfectly general concern with the state of the world is always and everywhere a part of justice. That it would make you happy and cost me very little to say how much I liked your cooking does not mean that it is a question of justice whether I do. There are ways that things can be good or bad without being about justice, and I am not sure Sen is prepared to acknowledge that.