A Reply to Sleat

Matthew Sleat and I agree that the question of how we, as fellow citizens, should proceed in the face of radical disagreement over matters of justice is a serious challenge that cannot adequately be addressed by any ideal theory which assumes that in a perfectly just society citizens would converge on a particular conception of justice. I do not treat this as grounds for rejecting the project of ideal theory, but I argue that this question—which I refer to as the political question—urgently needs addressing within a non-ideal theory that takes seriously what Jeremy Waldron calls “the circumstances of politics.” My proposal is that within non-ideal theory we need to develop a notion of legitimacy that does not presuppose the possibility of converging on a particular conception of justice, or even a family of such conceptions, and which explains how, despite their disagreements over matters of justice, fellow citizens have a reason to abide by the outcome of inclusive procedures. This is where Sleat and I part company: he thinks that in developing this alternative notion of legitimacy, I presuppose agreement of a kind that is still not available. Sleat’s criticism is a powerful one, but my claims about the possibility of agreement are more modest than he acknowledges. I recognise the possibility of disagreement that goes all the way down, but attribute normative significance to the wide degree of agreement on political procedures that I argue is to be found in mature liberal democracies: I claim that, in non-ideal theory at least, differences in degree can be normatively important.

Let me emphasize that my claim is not that citizens in mature liberal democracies can agree on a particular account of legitimacy. As Sleat points out, there is no more reason to think that fellow citizens can agree on a particular conception of legitimacy than there is to think that they can agree on a particular conception of justice. My claim is rather that, from the standpoint of non-ideal theory, a particular conception of legitimacy explains how the vast majority of citizens might have a reason to abide by the outcomes of a set of procedures even though they disagree deeply over matters of justice and legitimacy—and indeed disagree about what democracy is and why it is valuable. More specifically, I claim that the vast majority of citizens in mature liberal democracies can agree that the inclusion of all citizens in the decision-making process is normatively significant, at least when it is above some threshold level, and thereby provides a reason for that vast majority to abide by the outcomes of procedures that achieve this minimum standard of inclusion. Citizens themselves do not need to converge on a particular conception of legitimacy in order for that to be the case.

To this Sleat would reply that those who subscribe to different conceptions of democracy, or different conceptions of what full inclusion in the political process would mean, will disagree over what counts as a minimum standard of moral acceptability: deliberative democrats, for example, will require that procedures must be minimally deliberative, whilst direct democrats might insist that they meet a minimum standard of direct participation. ¹ Whilst not wishing to deny that these disputes are live and important, they do not subvert the point I am making: that the vast majority of citizens, whether their sympathies lie with demanding ideals of deliberative democracy or direct democracy, or more modest forms of representative government, can agree that any political system that includes all citizens in the decision-making process—in a way that goes beyond what is involved in societies that possess what Rawls calls a consultation hierarchy ² —have a pro tanto reason for abiding by the outcomes of those decision-making processes even if for other purposes they may deny that these processes are morally acceptable. (The precise strength of that reason will of course vary amongst them, for it will depend upon the particular ideal of democracy or full inclusion to which they adhere.)

This is not to ignore one of Sleat’s other important points, namely, that there are serious disagreements amongst theorists and indeed citizens more generally over who should be included in the political process. He mentions disagreements over the issue of whether prisoners should be included and in relation to the age at which children should be given political rights. ³ And of course there are also disagreements over the inclusion of residents who are not yet citizens, and over whether naturalisation should be regarded as a condition of political inclusion or whether immigrants should automatically be given political rights after a period of residence. ⁴ My response, though, is essentially the same: despite these important disagreements, given their normative commitments, the vast majority of citizens have a pro tanto reason to abide by the outcome of procedures that are broadly inclusive, even if these procedures exclude some that they think should be included, such as sixteen- to eighteen-year-olds, prisoners, or immigrants who have been resident in the polity for four years or more but have not naturalised. (To say that they have a pro tanto reason for abiding by the outcome of these procedures is not to deny that they may also have reasons for criticising these procedures or that sometimes they may have even stronger reasons for resisting the legislation that emerges from them and for engaging in civil disobedience. Nor is it to deny that were a group to be excluded on, say, grounds of sex, religion or ethnicity, then this would deprive many—perhaps most—citizens of any reason at all to abide by the outcomes of these procedures.)

Sleat attributes considerable significance to the point that we cannot reasonably expect universal agreement amongst fellow citizens. But I do not claim that we can. Indeed I am careful to say that it is widespread agreement—the agreement of the vast majority of citizens—that matters. ⁵ Here Sleat in effect poses a difficult dilemma for my position: either fully universal agreement does not matter, in which case why not suppose that widespread agreement on a more demanding standard is what is important for addressing the political question—perhaps even widespread agreement on substantive principles of justice, such as those that make up Rawls’s conception of justice as fairness; or fully universal agreement does matter, in which case the fact that there is merely widespread agreement that some minimum level of inclusion in the political process has normative significance does not enable us to answer the political question. But it seems to me that this dilemma can be overcome. If it were possible for the vast majority of citizens to converge on a more demanding standard that is met in their polity, then indeed that would furnish them with an even stronger reason to abide by the outcome of its decision-making procedures. I entertain this possibility, arguing that whether such agreement is possible depends ultimately on the particular society in question. I concede that there is some force to the idea that in mature liberal democracies there might be widespread agreement on the need for rights to protect freedom of conscience and prevent the state from imposing a particular religion on its citizens, ⁶ in which case if the agreement is wide-ranging enough there are grounds for including the protection of such a right in the specification of the minimum standard. Non-ideal theory of the sort that is required to answer the political question needs to be sensitive to the kind and extent of agreement available within particular societies, and in that way needs to be historically and empirically informed. Furthermore, the degree of agreement available within a particular polity makes a normative difference to what kind of answer can be given to the political question.

One of the reasons why Rawls’s brand of ideal theory is incapable of providing an adequate answer to the political question is because it places strict limits on what would count as a well-ordered society and what kinds of disagreement could arise there. A well-ordered society is one in which no one subscribes to a conception of justice that is grounded in an unreasonable comprehensive doctrine, such as those that would suppress freedom of conscience, and in which everyone accepts one of a family of reasonable liberal political conceptions. It is not only conceptions of justice that deny the fundamental equality of all citizens that are excluded on this basis but also conceptions that do not give the appropriate priority to basic liberties, and conceptions that do not involve a commitment to ensuring that citizens, whatever their social position, have adequate means to make intelligent and effective use of their liberties and opportunities. ⁷ In effect, this means that in a well-ordered society citizens cannot subscribe to communitarian conceptions of justice that allow community values to outweigh basic liberties, or to libertarian conceptions that forbid redistributive taxation to meet the needs of the worst off. For Rawls, ideal theory is concerned with the characterisation of a well-ordered society, and that in effect means a society that is regulated by one of the family of reasonable liberal political conceptions of justice, with each citizen subscribing to one or other members of that family. There is no room within ideal theory for adherents to other conceptions of justice or for the more radical forms of disagreement that Sleat and I emphasise. ⁸ The political question in its most challenging form—that is, how we, as fellow citizens, should proceed in the face of deep disagreement, a significant part of which is generated by these other conceptions of justice, simply does not arise in this context.

This is not in itself a problem, for Rawlsians can quite properly insist that ideal theory does not attempt to answer all the important questions that can be asked about a just society and how it is to be achieved. But whether the political question can be adequately addressed even within non-ideal theory will depend on how non-ideal theory is understood. For those who see non-ideal theory as always parasitic on ideal theory, it is tempting to treat the political question as if it could be satisfactorily answered by addressing the rather different question “What is the best way of achieving a well-ordered society in which basic institutions are governed by one or other reasonable conception of justice and in which everyone subscribes to one or other of these conceptions?” But in so far as Rawlsians treat this new question—call it the instrumental question—as equivalent to the political question, they misunderstand the latter. Answers to the instrumental question could not be regarded as a satisfactory response to the political question by any of the considerable number of citizens in mature liberal democracies who do not hold what Rawls regards as one of the family of reasonable liberal political conceptions of justice; or, at least, a citizen with views that lie outside of this family could not regard such answers as satisfactory responses to the political question without implicitly accepting his or her exclusion from the category of “fellow citizens.” When answers to the instrumental question are seen as answers to the political question, they are in effect declarations made by (philosopher?) rulers to their subjects. It is true that even in mature liberal democracies any answer to the political question will have that character for some because there is unlikely to be sufficient common ground for there to be any actual or possible set of procedures that each and every citizen has a reason to endorse given their diverse normative commitments. But in that case it seems to me that an adequate answer to the political question will have to seek as much common ground as possible given the facts of disagreement. When there is a set of actual or possible procedures that the vast majority of citizens have a reason to endorse, then it is this that must provide the basis for an answer to the political question. Contrary to what Sleat claims, widespread convergence is normatively significant in answering that question even though it falls short of universal agreement.