Ratio Juris. Vol. 18 No. 3 September 2005 (315-45)
Authority, Equality and Democracy
ANDREI MARMOR*
Abstract. The purpose of this essay is to argue that considerations of fairness play an essential role in the justification of democratic decision procedures. The first part argues that considerations of fairness form part of a practical authority's legitimacy, and that in the political context, those considerations of fairness entail a principle of equal distribution of political power. Subsequently, the article elaborates on the kind of equality which is required in democratic procedures, arguing that different principles of equality should apply to the deliberation and the decision stages of democracy. Finally, the article concludes with a few sketchy remarks on the possible relations between considerations of fairness and soundness of democratic procedures.
Democratic regimes and democratic cultures instantiate many goods that we value, both intrinsically and instrumentally. Political participation may be intrinsically valuable for many people. A culture of free and open public deliberation is perhaps a good in itself, and it may also greatly facilitate a robust civil society. It seems to be the case that democratic regimes are conducive, perhaps even essential, to economic prosperity. Thus in different ways and to various degrees, democracies instantiate a large variety of goods that we value. The core idea of a democratic decision procedure, however, consists in the practice of reaching authoritative resolutions on matters of public concern by a majority vote. But this core aspect of democracy is famously puzzling: The relationship between the legitimacy of an authoritative resolution and its democratic credentials is far from clear. Few things become good or desirable only because most people want them. How can such accidental, almost arbitrary, facts as the majority-minority divide, determine the legitimacy of an authoritative resolution? Is democracy justified only because, and to the extent that, it leads to good government? Or
* I am grateful to Scott Altman, David Enoch, David Estlund, Elizabeth Garrett, Chaim Gans, Alon Harel, Gregory Keating, Charles Larmore and Joseph Raz for very helpful comments on drafts of this essay.
© Blackwell Publishing Ltd 2005, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Maiden 02148, USA.
316
Andrei Marmor
is it also required by considerations of fairness?1 My purpose in this essay is to argue that considerations of fairness play an essential (but not exclusive) role in the justification of democratic decision procedures. In the first part of this essay I will argue that considerations of fairness form part of a practical authority's legitimacy, and that in the political context, those considerations of fairness entail a principle of equal distribution of political power. Subsequently, I will elaborate on the kind of equality which is required in democratic procedures, arguing that different principles of equality should apply to the deliberation and the decision stages of democracy. Finally, I will conclude with a few sketchy remarks on the possible relations between considerations of fairness and soundness of democratic procedures.
A clarification is in order before we proceed. When I say that a majority decision procedure is the core aspect of democracy, I have already taken sides in an important debate within democratic theory. Roughly put, contemporary theorizing about democracy is divided between those who emphasize the deliberative aspect of democracy, and those who regard democratic procedure as mainly an aggregative form of collective decision making. Deliberative democratic theory strives to articulate a fundamental political ideal centered around ideal deliberative procedures which aim at the common good. Such theories regard deliberation as an essential, not instrumental, aspect of public reason (see Cohen 1999, 67). Deliberation can be essential in two possible ways: Either an ideal model of deliberation can be seen as yielding decisions which are constitutive of the common good, or else they can be seen as models of deliberation aiming to discover truths about the common good.2 In both cases, however, the focus of the theory is on models of ideal deliberation. On the other hand, aggregative models of democracy take voting, not deliberation, as essential to democracy. On this view of democracy, collective decision making is not an ideal in itself, but only a practical need forced on us by the necessities of our communal lives. Such theories take the need for authoritative decisions as basic, and seek to work out a theory of democracy as the best political process for legitimizing authoritative decisions.3 Whether these two models of democratic theory are necessarily mutually exclusive is a difficult question I will not try to answer here. Nor will I have anything to say against the deliberative models of democracy. My main concern in this essay is how to justify the crucial role of majority rule in the political domain.
1 This is an over-simplification. Voting in a democratic decision procedure may have expressive and cognitive values over and beyond its contribution to good governance or political fairness. Although I will largely ignore these aspects of democracy, I do not intend to underestimate their importance.
2 The epistemic role of democratic deliberation is emphasized by Estlund; see, for example, Estlund 1999, 173; cf. Richardson 1999, 349.
3 This type of democratic theory is defended, for example, by Christiano 1996, and Waldron 1999a.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
317
1.
One of the main puzzles about democracy begins with the general conditions for the legitimacy of a practical authority. Why should one ever have a reason to do something only because she has been told by an authority to do it? I will assume here that Joseph Raz suggested the most plausible answer: The role of authorities in our practical reasoning is to mediate between the putative subjects of the authority and the right reasons which apply to them in the relevant circumstances. An authority is legitimate if its putative subjects are likely better to comply with the relevant reasons which apply to them by following the authoritative resolution than by trying to figure out or act on those reasons by themselves. This is, basically, what Raz called the Normal Justification Thesis for the legitimacy of an authority (henceforth, NJT; see Raz 1986, chaps. 2-4). Roughly, then, Raz's conception of practical authority maintains that the justification for compliance with a practical authority must rest on the likelihood that the person subject to the authority will do better by following its provisions than by attempting to rely on his or her own judgment. There are two main types of situation in which this may happen: Either the authority knows better what ought to be done in the relevant circumstances, or else it is just better situated to solve a collective action problem.4 In both cases, however, the NJT would seem to render the legitimacy of democratic decision procedures purely instrumental. As Raz himself recognized, "it is the truth or soundness of the decisions which counts ultimately. Truth and soundness provide the argument for the legitimacy of the authority." Therefore, Raz concludes, democracy is justified only instrumentally, "only if it leads, by and large, to good government" (Raz 1989, 778f).5
I will suggest, however, that this view about the legitimacy of practical authorities is too narrow: The legitimacy of a practical authority, particularly in the public-political domain, rests on a combination of the soundness of its decision and the fairness of the process which has led to it. The NJT captures the essence of the former: An authoritative decision is sound when the authority's subjects are likely better to comply with the reasons that apply to them by following the authoritative resolution, than by attempting to figure out or act on those reasons by themselves. But, as I will argue below, the fairness of the decision procedure is also an essential element of a public authority's legitimacy.6
Let me answer an immediate objection before we proceed. Suppose that a putative authority renders a decision which is, under the circumstances, perfectly sound as required by the NJT, though it did not result from a fair
4 On the distinction between these two components of the NJT I have elaborated in Marmor 2005, chap. 9.
5 Richard Arneson 2004 defends a very similar conclusion.
6 An argument in this spirit has been advanced by Hershovitz 2003.
© Blackwell Publishing Ltd 2005.
318
Andrei Marmor
procedure. Let us suppose that the decision actually violated the requirement of procedural fairness, whatever we take it to be, that should have applied in the circumstances. Still, would it not be irrational on the part of the authority's subjects to decline to follow the decision, sound as it is, only because the requirement of fairness has been violated? Under normal circumstances it would be irrational indeed, but this only means that not every aspect of the legitimacy of an authoritative decision bears on the subjects' reasons for following it. Raz is right to maintain that the subjects' main reasons for following an authoritative decision consist in the soundness of the decision. The concept of legitimacy, however, need not be exhausted by the reasons for following an authority's decree. The idea of the legitimacy of an authority may comprise elements which do not have a direct bearing on the reasons to follow the authority's decisions, though they may have an indirect and secondary role to play in such reasons for action.
Consider, for example, the analogy of a parental authority. Suppose that George has a teenage daughter, Sarah, and they usually take an annual vacation together. George decided that they should go for a ski vacation, as he reasoned that Sarah would enjoy learning to ski, and that she would have the opportunity to acquire a new skill that will enable her to take many more enjoyable vacations in the future. Let us assume that under the circumstances, this decision was sound. But if George did not consult with his daughter on any of this, would she not have a legitimate reason to complain? Note that Sarah could not complain that the decision was made without regard to her interests. On the contrary: We have assumed that the decision is sound precisely because it adequately reflects the interests of Sarah. Nevertheless, it would seem quite understandable if Sarah complains that she should have been consulted on this issue. From her perspective, the soundness of the result is not all that matters, and understandably so.
We have not yet explained, however, why the component of fairness is an essential part of the legitimacy of public authorities. Why would a sound but procedurally unfair authoritative decision be tainted with illegitimacy? As we have seen, the answer cannot be derived from the thesis that authoritative resolutions must be responsive, in the appropriate ways, to the interests of the subjects. Authorities can achieve this goal without following fair procedures.
Admittedly, the fairness requirement applied to a decision procedure is often grounded on epistemic considerations: Procedural justice is often designed to enhance the reliability of the relevant procedure in yielding correct outcomes.7 Fairness may also have instrumental value which is not
7 Criminal justice is a good example: Many of the requirements of fairness that we apply to the procedures of criminal justice, such as a right to fair police interrogation and to a fair trial, the right to counsel, etc., are there to ensure correct results of the criminal justice procedures, namely, that only the guilty be punished and not the innocent. In these cases the value of procedural fairness is purely instrumental.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
319
related to the reliability of the procedure. The fairness of a decision procedure may enhance such goals as communal solidarity, inclusion, and cooperative participation. Regardless of such instrumental considerations, however, it is sometimes the case that procedural fairness is intrinsically valuable.
One of the main values of procedural fairness derives from the value of respect for people. Consider, for example, the case of a group of friends who meet once a month in a book club. Normally, they would have different preferences about the book they would want to discuss in each session. Thus, they would need some decision procedure to settle these issues. A fair decision procedure amongst them would manifest their mutual commitment to respect for each other. The fairness of the decision procedure in such cases ensures that everyone is duly respected (see Brighouse 1996,123). Note that in such cases, the value of fairness is intrinsic but not ultimate. Something is of ultimate intrinsic value if we value it for its own sake, regardless of anything else. Other intrinsic values, however, can be derived from ultimate values by practical reasoning. We justify their intrinsic value by non-instrumental derivation from other intrinsic values. The value of procedural fairness belongs to this latter category; it is intrinsic but not ultimate (see Raz 1986, 200; Marmor 2004, 138).
Thus, let me suggest that one line of thought which could justify the requirement of fairness with respect to authoritative political decisions is premised on the idea of fairness as respect for people's right to personal autonomy. The main value that we attach to personal autonomy is that people should create, as far as possible, their own lives through successive decisions and choices of their own. It is based on a vision of people controlling, to a considerable extent, their own destiny. As Raz put it, "[a]n autonomous person is in part author of his life. [. . .] An autonomous person's well being consists in the successful pursuits of self-chosen goals and relationships" (Raz 1986, 370). Therefore, Raz argues, "to be autonomous a person must not only be given a choice but must be given an adequate range of choices" (ibid., 373). Accordingly, the right to personal autonomy is the right to a social and political environment which enables people to lead an autonomous life. Now, assuming that political decisions potentially affect the lives of each and every individual in the community, one could claim that due respect for people's autonomy requires a fair decision procedure which would manifest respect for individuals' ability to make up their own minds about the right decision. We need not assume that everybody wants to participate in the political decision procedure. Surely, that would be a false assumption. But the value of respect for people's autonomy requires a political structure in which everybody has a fair chance to participate, and this is what democratic decision procedure aims to achieve (see Brighouse 1996).
Admittedly, however, this move from the value of respect for people's autonomy and the fairness of a democratic decision procedure is far from
© Blackwell Publishing Ltd 2005.
320
Andrei Marmor
straightforward. In particular, the egalitarian aspect of this respect needs to be explained and justified. In fact, there are two separate questions here: Why does respect for people's autonomy require equal respect? And even if it does, why would equal respect entail anything like an equal right to political participation?
Respect for people is not necessarily an egalitarian concept. On the contrary, respect for people is often a differentiating value: We typically assume that some people deserve more respect than others. Respect is normally a reaction to others' behavior, virtue, or ethical character. As these vary, respect owed to them may vary accordingly. For example, the conscientious, hardworking colleague would deserve more of our respect than the lazy and irresponsible one. Egalitarians have argued, however, that at some basic level, people, just in virtue of being humans, moral agents, or rational beings, deserve equal respect.8 Needless to say, it has proved extremely difficult to articulate the grounds for this basic egalitarian concern, and I have no hopes of resolving this issue here. Even if we assume, however, that at some basic level people deserve equal respect as moral agents, it remains to be explained why respect for their autonomy should be equal. After all, some people are better in making decisions than others. Peoples' deliberative rationality certainly varies a great deal; why assume, then, that we should respect their autonomy to an equal degree?
I believe that the answer is comprised of two considerations. First, as the example of the respect for a colleague suggests, respect for people is always relative to a certain capacity in which we regard them. The hardworking and conscientious colleague deserves our respect as a colleague but not necessarily as a sportsman, or as a spouse. In fact, he may deserve our respect for his work even if, as it happens, he is otherwise a pretty contemptible person. Therefore, the question arises in what capacity people deserve respect for their autonomy in the political context. Plausibly, we may hold that in the political domain people deserve respect as the subjects of political authorities and as members of a body politic, namely, as citizens? Citizenship I take to be a morally significant concept: From a moral perspective citizens are bearers of rights and duties vis a vis the state and each other. Any liberal conception of citizenship should be committed to the view that citizens of a body politic should be regarded as having equal rights. Once we recognize a right that people should have as citizens of a body politic, we
8 But cf. Darwall (1977, 36) who argued that there are actually two separate concepts of respect involved here, which he named "recognition respect" and "appraisal respect". I tend to think that the former category is just an application of respect to special cases. In any case, the distinction does not affect my argument here.
9 I do not mean "citizenship" in its legal sense; from a moral point of view, it is far from clear that any distinction should be applied to legal citizens as opposed to other categories of residents in a given country.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
321
are committed to holding that, at least in principle, they should have it equally.
Therefore, naturally, the second component of the reply resides in the idea of an individual's right to autonomy. As a matter of fact, it is of course true that people are bound to vary a great deal in the autonomy they actually achieve throughout their lives. Nor do people value their own autonomy to the same degree. Surely, for some people autonomy is much more important than for others. These variations notwithstanding, it is a plausible assumption that people, as citizens of a body politic, should have an equal right to personal autonomy. The right to personal autonomy is the right to a social, legal and political environment that would enable people to have an adequate range of options to choose from, and that would enable them to decide for themselves what kind of life they want to lead, what kind of things they value in life, and how to implement those projects which they regard as valuable. For our present purposes, we should look at the idea of autonomy not from a perfectionist perspective, as a personal achievement, but as a right to the conditions which enable people to lead autonomous lives. Therefore, like any other individual right, the right to have one's autonomy respected is a right people should have to an equal degree. The fact that people differ in their abilities to make use of a given right, or differ in their appreciation of its relative importance, does not normally entail that they do not have an equal right to it.
Thus I suggest that from the perspective of individual rights, people are entitled to an equal respect for their autonomy. Does this entail a right to participate in political decisions to an equal degree? In other words, does equal respect for autonomy entail an equal vote in political decisions? I think that the answer is yes, but only if some further assumptions are added. Surely it would be a violation of the equal respect for a person's autonomy if that person is deliberately disenfranchised or otherwise denied the vote in a given decision procedure. But suppose that a decision procedure is in-egalitarian in a more subtle way. Suppose that a decision procedure is devised which is responsive to the participants' differing capacities to reason. John Stuart Mill (in-)famously suggested something along these lines when he claimed that votes should be allocated unequally, depending on people's level of education (Mill 1991, chap. VIII). Admittedly, this was a crude suggestion, assuming, very problematically, that education is a reliable proxy for political wisdom and deliberative rationality. Suppose, however, that a procedure along Mill's lines could be refined, so as to convince us of its reliability in differentiating people's deliberative capacities with respect to the relevant decision.10 Would a scheme like this violate the principle of equal respect for people's autonomy?
10 See, for example, David Estlund (2002, 175) who argues that strict political equality may be an unattractive ideal because it may damage the quality of political deliberation.
© Blackwell Publishing Ltd 2005.
322
Andrei Marmor
The problem is that at least one plausible affirmative answer to this question might actually impose a severe limit on the desirable scope of democratic decision procedures. Why is that? Because when we think about the moral importance of personal autonomy, it would seem that the main value of personal autonomy is premised on the idea that in matters concerning people's own lives, it is more important for people to be able to choose for themselves than to choose correctly (that is, even if there is a correct answer). But if this is the underlying moral intuition behind the value of personal autonomy, then equal respect for autonomy need not go beyond those realms in which this underlying principle holds true. Namely, it need not extend to those issues in which it is more important to have the right decision than to let people decide for themselves. Put slightly differently, the question is how an autonomy-based principle could give people the right to participate in decisions that affect the lives of other people.11
A very forceful, though somewhat stronger, version of this challenge has been advanced by Arneson. He claims that an equal right to political participation should not be acknowledged precisely because it entails a power to affect other people's lives. His position is "that there is no such basic moral right, because one does not have a basic moral right to exercise significant power over the lives of other people, to direct how they shall live their lives" (Arneson 2004, 46). There are, it seems to me, two responses to Arneson. First, it should be noted that Arneson consistently addresses his argument against the claim that a right to equal political participation could be a basic moral right, arguing that in an inventory of our basic moral rights, one should not find any right which "confers on the right holder the power to direct how another shall live" (ibid., 47). But I have not argued that a right to participate in political decisions, whether equally or not, is a basic right. It certainly is not. To the extent that people should have a right to an equal distribution of the power to make political decisions, it would be a derivative right. It must be derived from the basic right to personal autonomy, combined, as we shall see, with certain assumptions about facts constituting our social lives.12 Whether such a derivation is possible remains to be seen. But again, it is not my contention that a right to political participation, whether equal or not, is a basic moral right of anyone. Furthermore, it is not the case that the only alternative to a basic moral right is an instrumental conception of the relevant good or value. There are many things that we
11 It may be worth keeping in mind that there is a whole tradition in political theory that would resist this formulation of the question since it views political participation as an essential ingredient of personal autonomy (see, for example, Arendt 1958). As Brighouse rightly notes (1996, 136), however, this is a distinctly non-liberal conception of personal autonomy.
12 Perhaps Brighouse's view can be seen as the main target of Arneson's argument, since Brighouse does seem to claim that the principle of equal availability of political influence is a basic moral principle. I do not share this view.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
323
value intrinsically, without necessarily assuming that the good is basic or otherwise foundational.13
There is a second reply to Arneson. Putting aside the question of whether the right to political participation is basic or not, it remains to be asked whether anyone has a right to rule, that is, to make decisions which affect other people's lives. If the answer is yes, then it would be difficult to understand why some people might have such a right and others should not. If, on the other hand, Arneson wants to deny that such a right to rule is ever justified, his position would amount to straightforward anarchism, which, I take it, he does not want to subscribe to. (Nor is it consistent with his general position which justifies political authorities on instrumental grounds.) In other words, once we acknowledge that there are legitimate political authorities, we must also concede that some people have a right to make decisions about other people's lives, and those subject to the legitimate authority would have a duty to follow the authority's prescriptions. The only question that remains is the question of whether or not such a right to rule should be subject to distributive concerns of fairness.
Perhaps Arneson would reply that since the only justification for the exercise of a political authority is an instrumental one, even if an authority is legitimate, it cannot claim a right to rule. I doubt, however, that so much can depend on the question of whether we associate the legitimacy of a political authority with a right to rule. Political authorities exercise power, and if the exercise of this power is deemed legitimate under the circumstances, the authority's subjects are under an obligation (albeit limited, of course) to follow its prescriptions. This is close enough to anything we might want to characterize as a right to direct how another shall live.
These replies to Arneson's argument, however, do not answer our general concern about the possible connections between the right to personal autonomy and the distribution of political power, namely, how to explain the rationale of democratic decision procedures in those realms in which we would normally hold that it is more important to have the right decision than to let people decide for themselves.
Some political philosophers have responded to this challenge by emphasizing the prevalence of value pluralism.14 They point to the fact that we live in societies in which people are deeply divided over their conceptions of the good and the just. Furthermore, these divisions are unavoidable, they claim, in a deep sense: even the most rational people can have reasonable disagreements on those matters which would admit of right and wrong, good or bad,
13 Consider, for example, the idea of friendship. Suppose we hold that friendship is intrinsically valuable. My particular friendship with John, for example, is therefore intrinsically valuable, as it instantiates, non-instrumentally, an intrinsic value. Similarly, there are many rights we hold that are not basic, but they instantiate non-instrumentally, other, more basic rights.
14 Most recently, for example, Waldron 1999a.
© Blackwell Publishing Ltd 2005.
324
Andrei Marmor
just and unjust, etc. Note that the purpose of this move is not to doubt that there are sound political decisions, or that there are issues on which it is more important to have the correct decision than to let people decide for themselves. The argument from value pluralism is based on the premise that there is something wrong in imposing an authoritative ruling on people who may reasonably disagree with it (see Estlund 2000). Thus, in a way, the requirement of an equal vote as an application of equal respect for peoples' autonomy becomes something like a fall-back position. Since we cannot rule out the possibility of rational disagreement about conceptions of justice and the good, we must assume that people should have an equal right to participate in the decision process, even if, objectively speaking, it may be an instance where it is more important to have the correct decision than to let people choose for themselves.
There are two problems with this argument, however, and both of them concern its potential scope. First, the argument from value pluralism must be reconciled with the fact that even in the political domain, not everything is potentially subject to reasonable, principled, disagreement. A great many functions of legal and political decisions in a modern society concern the resolution of such issues as collective action problems, cost-benefit analysis, the creation of public goods, and other tasks which have very little to do with principled disagreement. And if there is no principled disagreement lurking in the background of such political decisions, it is difficult to see how respect for people's autonomy entails a requirement of equal right to participate in the making of the decision (see Marmor 2002).
It is true that sometimes it is controversial whether an issue involves, say, only a solution of a collective action problem or disputed matters of principle. The boundaries between those issues about which people can reasonably disagree and those about which they cannot are sometimes blurred, and may well be subject to reasonable disagreement. Perhaps this would extend the scope of the argument, but not enough. We should be careful not to generalize from some cases to all.
Secondly, the argument from value pluralism entails the problematic conclusion that the more homogenous a society is, the less democracy it requires. Pluralism is a fact in certain societies, but not in all of them. Those "circumstances of politics," as Waldron called them,15 whereby different segments of the population are deeply divided over conceptions of the good and the just, simply do not obtain in all contemporary societies. Would we want to say that in homogenous societies, where people share a conception of the good and evaluative schemes, there is no need for democracy? (That is, at least not from the vantage point of fairness.) Perhaps some political theorists who endorse the argument from pluralism would be willing to bite
15 Waldron (1999a, 102, 144, 159-60) defines the "circumstances of politics" as a combination of two facts: value pluralism and the practical need to have authoritative decisions.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
325
the bullet at this point, simply admitting that democracy is not universally valuable; it may not be required in all contemporary societies. Others have argued that the lack of pluralism in a given society is in itself something wrong that needs to be changed (see, for example, Rawls 1993, 35-8). Both of these views raise many difficult questions that I cannot deal with here. Instead, I will argue that it is not mainly the background of pluralism that explains the need for equal participation in politics, but, more importantly, the background of scarcity and competition over those social resources which, ultimately, constitute the conditions of autonomy. As Thomas Chris-tiano aptly put it, it is mainly a conflict of interests and not a conflict of judgments that grounds the need for democracy (see Christiano 1996, 53).16 It is true that the distinction between these two is not very sharp: A conflict of interest tend to translate into conflicts of judgment, and often conflicts of judgment lead to conflicts of interest. Nevertheless, the distinction is a very important one, for two reasons. First, a conflict of interest is present even if the parties to the conflict do not grasp it as such. Conflict of interest does not necessarily depend on its self-awareness. More importantly, by focusing on conflicts of interest we can show that the need for democracy is not confined to pluralistic societies, since conflicts of interest can be pervasive even in homogenous societies where people share, more or less, the same conception of the good and the just.
It may be objected, however, that a conflict of interests, in itself, is irrelevant to the justification of democracy because the latter must show that it is important to take people's judgments into account in making decisions. Otherwise, why consult them at all? This is true, of course. The importance of focusing on conflicts of interests, however, consists in the thesis that we should take conflicts of judgment seriously precisely because they tend to manifest genuine conflicts of interests. In other words, the question is which type of conflict we regard as basic. The view advanced by Christiano, I take it, is that we have reason to take conflicts of judgments seriously mainly because, and to the extent that, they manifest conflicts of interest. Let me try to explain, in some detail, why I think that he is right about this.
One crucially important aspect of political authorities concerns the fact that they deal with problems relating to scarcity and competition over resources. Scarcity is not confined to material resources; it pervades all those social factors which affect the conditions of personal autonomy. In a complex society, individuals have very limited abilities to control their lives and shape the environment in which they live. Most of the options we can choose from, and our actual ability to make those choices which constitute our autonomy, are determined by forces outside our control. Some of these limits are determined by natural causes, but many others are determined by
16 Generally speaking, the argument that follows reaches similar conclusions to Christiano's approach.
© Blackwell Publishing Ltd 2005.
326
Andrei Marmor
economic and social factors. In one way or another, all political decisions affect these conditions and continuously reset the boundaries of social opportunities. Even decisions which have nothing to do with material resources would normally affect people's condition of autonomy: They would make certain forms of life relatively more easy or difficult to pursue. In other words, an individual's conditions of autonomy are determined by a very complex set of factors prevailing in the society in which she lives. Some of these factors consist of material resources, but many others do not. One's ability to choose and pursue a certain form of life is also determined by such factors as social confirmation, self esteem, political opportunities, means of expression, culture, religion, and what not.
Generally speaking, the kinds of values, rights, and ideals which are protected by legal-political decisions, and the extent of their protection, greatly affect people's conditions of autonomy; they determine the forms of life which could be pursued in that community, and how easy or difficult it would be to pursue them. This essentially competitive feature of political decisions lends considerable support to the claim that political power, that is, the power to make authoritative political decisions, could be regarded as a kind of "primary good" which should be subject to concerns of fairness and distributive justice.17 Thus, assuming, as we did, that the ideal of equal citizenship carries with it the ideal of an equal right to have one's autonomy respected, it would be natural to conclude that the fair distribution of political power should be, at least prima facie, an egalitarian one.
To be sure, it is not assumed here that the value of personal autonomy increases indefinitely in direct proportion to the number of choices available. The availability of an adequate number and diversity of choices is sufficient in order to fulfill the ideal of personal autonomy. However, the argument does assume that for most people living in a complex society, the conditions of autonomy present a problem of scarcity and competition over economic and social resources. In a world of limited resources, the political affirmation of one form of life normally comes at the expense of another. Furthermore, it should be kept in mind that scarcity of resources creates competition even when the relevant parties share the same conception of the good or the just. The present argument need not assume that there must be any principled disagreement in the background of such competition over the limited resources which constitute people's conditions of autonomy.
It may be argued, however, that the scope of this argument is more limited than suggested above, since there are many types of societal goods created or modified by legal-political decisions which are not essentially competi-
17 I do not intend to put too much weight on the idea that political power is a "primary good", partly because the latter lends itself to several possible interpretations, and I cannot go into this here. For my purposes it is sufficient to maintain that political power is the kind of good which should be subject to fair distribution.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
327
tive. The creation of public goods, so this argument runs, is not subject to competition since those are the kinds of good which people can enjoy without subtracting from the potential enjoyment of others. Public goods, once created, can be enjoyed by all in a non-exclusive and non-excludable manner. Therefore, this argument concludes, even if there are conflicts of interest and competition over resources in the background of many political decisions, this cannot be generalized to all cases; the creation of public goods, which forms a major part of the role of political authorities, should be excluded from the argument from scarcity.
There is a grain of truth in this objection, but much less than meets the eye. We need to see the complete picture in order to get a better sense of proportion. Let us ask ourselves what are the kinds of goods which can be created or modified by political decisions. To begin with, legal-political decisions often affect the allocation or re-allocation of private goods. The law frequently transfers goods from one person to another and, more basically, creates the legal and social structures which define the conditions and boundaries of private ownership and possession. Surely, these are the kinds of decisions which are needed precisely because there is a scarcity of resources and potential competition over their distribution. In addition to this obvious impact of politics on the allocation of private goods, however, there is a whole range of societal or communal goods which can be created and modified by various forms of political decision. Let me suggest that there are three main types of such communal goods: collective, public, and common goods.18
Collective goods are defined as such by the means of their production; namely, they are the kind of goods which require some form of collective action to produce and sustain. Such goods, for example, as national defense, a clean environment, safety in transportation, and many others, all require some collective action to create and sustain. Needless to say, collective action is always subject to competition because it always raises issues about the distribution of burdens and avoidance of free riding.19 Public goods, on the other hand, need not require collective action to produce (although often they do), because what marks them as "public" is the character of their consumption. Once a public good exists, it is of such a nature that everybody can consume it without subtracting from the potential consumption of others. The consumption of public goods is not exclusive and not excludable, and therefore, allegedly, not subject to competition. As the examples of a clean environment or national defense show, many collective goods are
18 For a more complete account of these distinctions, see Marmor 2001b, 213-25. Cf. Christiano 1996, 59-61.
19 With one exception: The solution of a pure coordination problem does not involve competition. Pure coordination problems, however, are normally resolved without authoritative involvement (see Marmor 2001a, 44-5).
© Blackwell Publishing Ltd 2005.
328
Andrei Marmor
also public goods, and vice versa. Clean air, for example, is something everybody can enjoy without thus subtracting from the enjoyment of others. Finally, there are common goods, which are those goods that can only be consumed in their communal manifestations. Common goods are of such a nature that they cannot be enjoyed in isolation from a whole community which takes part in the consumption of that good, because the communal aspect of the good is an essential ingredient of what makes it a good. For example, such goods as culture, solidarity, nationalism or national pride, or as Waldron once suggested, the conviviality of a party (see Waldron 1993, 355), can only be consumable goods in the context of their communal manifestations. There must be a party which is convivial, or a group which exercises solidarity, for any particular individual to be able to take part in the enjoyment of such goods.
The competitive aspect of a good either can concern its mode of production, or its mode of consumption, or both. As we have already seen, both the allocation of private goods and the creation of collective goods are essentially competitive and thus support the argument from scarcity. But what about public goods and common goods? Are they not free of competition? Let me begin with the latter. Common goods often, but certainly not always, require a concerted action to produce. What marks such goods as essentially competitive, however, is mostly their consumption end. The consumption of common goods is competitive in two main ways: first, common goods are typically of such a nature that they involve exclusion. The solidarity of a group, for instance, can only be entertained in connection with some conception of membership in that group, thus by necessity excluding others who are not members. Similarly, the conviviality of a party can only be enjoyed by those who participate in it, thus excluding those who do not participate. In other words, since common goods are goods only in their communal manifestations, such goods are closely tied to a notion of membership in the relevant community; and membership is always a matter of including some and excluding others. This is one main aspect of common goods in which they are essentially competitive: They would always give rise to potential competition over inclusion in, or exclusion from, the relevant community.
Secondly, though not less importantly, the consumption of common goods is made possible, as such, only because there is a whole group of people who share its appreciation and enjoy it as a good of a certain kind. One can hardly value the solidarity of one's community unless others in that community share those values which constitute solidarity and regard their solidarity as a value. Solidarity, national pride, and many aspects of our culture are things we share, as a group, and we can only share them if we share at least some of the values which are associated with those goods. Therefore, the consumption of common goods is partly a matter of sharing certain conceptions of the good with other members of one's community. As
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy 329
such, common goods are as potentially competitive as any other variety of the conception of the good.
Finally, and this is the core of the objection under consideration here, it may be argued that at least the consumption of public goods is not necessarily subject to competition. As I have mentioned earlier, the production of many types of public goods requires collective action, so at least in that respect, we cannot rule out competitive elements even in the sphere of public goods. Furthermore, we should keep in mind that the production of any type of good, including public goods, involves opportunity costs which can be very significant. This is yet another way in which the production of public goods is potentially competitive. But I think that even at the consumption end, public goods tend to be more competitive than is generally assumed. To begin with, the non-excludability of public goods is a matter of degree. A beautiful public park may be enjoyed by many without the enjoyment of one subtracting from the enjoyment of another. But only up to a point, of course; once too many people begin to use the park it may well become less enjoyable. There are very few goods which are perfectly non-excludable in their consumption. Perhaps the beauty of a work of art, or the achievement of an artistic genius, may be examples of near-perfect non-excludability, although the possibility of exclusion can never be ruled out even in such innocuous examples. More important examples of public goods, however, are definitely more competitive than this. Consider, for instance, the suggestion that freedom of speech is a public good. A society which protects free speech, it is often argued, creates a public good that can be enjoyed by all, without thus subtracting from the consumption of this good by others. But this is actually false. An extensive protection of freedom of speech might protect such expressions which make speech itself much more difficult for some than for others. If freedom of speech protects the continuous hate-speech of one group against another, the people who belong to the target group, which could be a vulnerable minority, may find the effectiveness or perhaps even the possibility of their free speech seriously impeded. And similar considerations pertain to many other public goods. It is true that their consumption is not necessarily exclusive, but in practice they do tend to involve many competitive elements. The number of public goods which are perfectly non-exclusive and non-excludable, involving no conflict of interest between persons who might be affected, would shrink under closer scrutiny to a very insignificant number.
I believe that this, admittedly contingent, element of socio-economic scarcity which lurks in the background of political decisions, may help us to bridge the gap between the ideal of equal respect for people's autonomy and the principle of equal participation in the political decision process. The competitive aspect of politics lends considerable support to the conclusion that an equal respect for people's right to personal autonomy should be implemented by a right to an equal participation in the political decision
© Blackwell Publishing Ltd 2005.
330
Andrei Marmor
process. To be sure, this is not a conceptual link; but then, few arguments in political theory should be expected to be entirely abstract. In a very different, much less competitive world, people might need different principles of fairness and different political institutions. That should come as no surprise.
Arguably, however, there is still a missing link in the chain. It can be objected that the argument from autonomy merely supports the conclusion that the conditions of autonomy ought to be distributed fairly, but it does not follow that the political power to make decisions about such matters must also be distributed fairly. This is a forceful objection, and it has two possible aspects. First, it might be an issue of institutional design: Can we think of any alternative political structure that could guarantee a fair distribution of people's conditions of autonomy, without also providing them an equal opportunity to shape the outcomes of such political decisions? I do not think so, but of course, I cannot be sure. If there is such an alternative political structure, then we would have to admit that at least from the vantage point of fairness, there is a great deal to be said in its favor.
Secondly, and more importantly, however, there is a question of principle here. Suppose that I am wrong about the institutional issue, and a non-democratic political structure could be envisaged so that it would achieve a fair or equal distribution of people's conditions of autonomy. Could it also manifest respect for people's autonomy? Respect for a person's right to personal autonomy is a respect we owe to her freedom of choice. How can we demonstrate such respect if we do not ask the person to express her choice in any meaningful way? To be sure, there are many goods which can be distributed fairly without the need to have the subjects of the distributive principle express their views about the proper distribution or influence of its outcomes. But this cannot be the case when the relevant distributive concern is essentially tied to people's right to choose the kinds of goods they would want to have in the first place. Only an equal right to participate in the decision process can instantiate to a full extent the need to respect people's choices, and enable them to have a fair chance to shape the underlying conditions that would enable them to exercise their choices (see Brighouse 1996, 125). Needless to say, the nature of the equal right to participate in political decisions is problematic, and the question of what kind of equality it would entail needs careful elaboration. The following sections take up some of these difficulties.
2.
So far, we have seen that equal respect for people's autonomy needs to be implemented by acknowledging a right to an equal participation in the decision process. Therefore, from now on, I will assume that in the political
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
331
domain people have a right to an egalitarian distribution of political power20 and I will try to suggest a few clarifications about what kind of a right to equality it is. First, however, we need to mention a familiar but important distinction. A political process which results in an authoritative resolution comprises two main stages: deliberation and decision. Political power is measured by one's ability to influence both of these stages of the political process. Deliberation begins with setting the agenda. For any authoritative result, it matters a great deal what the agenda for decision is, and how it is determined and defined. People's ability to influence the setting of the agenda is a first and crucial step in the exercise of political power. Once the agenda is set, a process of deliberation begins, whereby people attempt to influence the result in various ways, by proposing rational (or irrational) arguments, bargains, enticements, and what not. A considerable aspect of political power is determined by people's ability to influence this process of deliberation.21 At some point, however, deliberation has to come to an end and an authoritative decision must be reached. Needless to say, it matters a great deal who makes the decision and what kind of decision-procedure is followed.
Now suppose that democracy is a commitment to a principle of fairness which should apply at both of these stages of the political process,22 and suppose that this is some principle of equality. It often goes unnoticed that such a principle of equality need not be the same kind of equality with respect to these two main stages of the political process, namely, deliberation and decision. An equal distribution of political power with respect to the deliberation process is basically a principle of equal opportunity. Whereas the kind of equality which would be required to implement fairness at the
20 The notion of "political power" should be understood here simply to mean the power to make authoritative public decisions. It should be noted, however, that the argument of the previous section is limited to politics at the basic structure of society, so to speak; it is not intended to apply, without qualifications, to many "micro" levels of decision making, such as a faculty meeting or a condo association.
21 One could claim that agenda-setting and deliberation are two separate stages in the political process. In many respects surely that is correct, but for the purposes of the argument under consideration, it makes no difference. I will regard both as aspects of the deliberation stage of democratic procedures.
22 It may be objected that the very need for deliberation cannot be explained on grounds of fairness. Only the objective of reaching sound political decisions can rationalize the importance of the deliberation stage of democratic decision procedures. I think that Estlund makes this objection (2000, 111). See also Richardson 1999, 359. This is not quite true, however. Once we assume that people should have an equal opportunity to influence political results, the deliberation stage of democracy cannot be exempt from requirements of fairness; after all, part of what it means to have political power is determined by one's ability to influence deliberation. Furthermore, it should be kept in mind that alternative ways of addressing a political problem, and different structures of decision making, are almost always biased in favor of particular outcomes. Perhaps Estlund would be right in this objection if one assumed that fairness is the only consideration supporting democracy; but we have made no such claim.
© Blackwell Publishing Ltd 2005.
332
Andrei Marmor
stage of the authoritative decision is a principle of an equal share of the power to make the decision, namely, majority vote.
Dworkin's distinction between "impact" and "influence" may be helpful in explaining the first point. According to Dworkin, the distinction between political impact and influence is as follows: "Someone's impact in politics is the difference he can make, just on his own, by voting for or choosing one decision rather than another. Someone's influence, on the other hand, is the difference he can make not just on his own but also by leading or inducing others to believe or vote or choose as he does" (Dworkin 2000,191). Dworkin is right to conclude that at least with respect to the democratic deliberation process, if equality matters, it is equality of influence, and not impact, that should count. After all, the whole point of the exercise of political power at the stage of deliberation is to have the maximum influence on the emerging decision, and that surely means one's ability to lead or induce others to vote or choose as he does. But, Dworkin argues, equality of influence is not an attractive political ideal. Basically, his argument is as follows: There is a substantial variety of factors which determine people's differing abilities to influence political decisions, such as money, social status, natural skills, willingness to devote time and resources, and so on. Even on the face of it, it would seem to be morally groundless to strive to eliminate the differences between people on all these different dimensions. Perhaps the influence of money should be eliminated because wealth is not the kind of difference that should affect people's relative political influence. But other differences are not illegitimate: Presumably, a person who is willing to devote more time, energy, and personal resources to political participation is entitled to reap the benefit of her effort by having greater influence on the result. In other words, Dworkin claims that people cannot complain about inequality of influence unless they can trace the relative lack of influence to a source that is itself illegitimate (ibid., 199). From this Dworkin concludes that "equality of influence is incompatible, even in principle, with other attractive aspects of an egalitarian society" (ibid., 198).
Even if we accept the premises of this argument, as I think we should, it does not follow that equality of influence is not the principle of fairness which is required at the deliberation stage of the political process. The natural conclusion that follows is that the relevant principle should be the principle of equality of opportunity of political influence (see Estlund 2002, 176; and Brighouse 1996). That is so, because equality of opportunity is always relative to a set of factors which we consider as independently illegitimate in determining social outcomes. The distinction between potentially relevant and irrelevant differentiating factors is one which is presupposed by any principle of equality of opportunity. Suppose, for example, that we hold an egalitarian conception with respect to the opportunity of school children to receive adequate education. Surely, such an ideal must be driven by the assumption that among the factors which can affect
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
333
children's educational prospects, there are some which ought to be eliminated because they are illegitimate sources of differentiation. However, there must also be factors, such as, perhaps, natural intelligence, or talent, that we regard as legitimate sources of influence on the educational outcome. Otherwise, if nothing can legitimately differentiate the outcomes, equality of opportunity necessarily collapses into equality of outcome. Dworkin's argument, therefore, cannot be regarded as an objection to the principle of equality of opportunity of political influence at the deliberation stage. Roughly, the idea is that people should have an equal opportunity to influence the setting of the political agenda and once the agenda is set, they should have an equal opportunity to influence the results of the deliberation.
Now, this is an essential part of what democracy is all about. The commitment of democratic regimes to equality of opportunity of political influence is manifest in a wide range of principles and institutions which we normally regard as essential to the proper functioning of a democracy. These principles and institutions include freedom of association, equality of access to political institutions, an extensive protection of freedom of speech, regulation of political parties, lobbying groups, etc., campaign finance restrictions, and so on. All these, and doubtless many other principles and institutions, aim to guarantee the equality of opportunity of political influence at the stage of deliberation. Needless to say, the details of these principles are controversial. For example, people differ in their views about political campaign finance restrictions, some seeking more, and others less, regulation. Some of these differences of opinion are about matters of fact (e.g., what kind of restrictions are feasible, more effective, and so on), but others are about matters of principle and fairness. This is natural: As with any other principle of equality of opportunity, people tend to have different views about the distinction between legitimate and illegitimate differentiating factors. Although we tend to agree, for example, that people's willingness to devote time and energy to political participation is a legitimate factor that may be allowed to affect political deliberation, we tend to disagree about the legitimacy of the factor of wealth. But again, once we regard the relevant consideration of fairness at the stage of political deliberation as one of a principle of equality of opportunity, these controversies are precisely those which are to be expected.
3.
Equality at the stage of the decision procedure is a different matter altogether. Here we need a principle of a fair distribution of the actual power to make the decision. An equal distribution of the power to make a decision would seem to entail the familiar principle of one vote to one person, that is, a straightforward majority decision procedure. This is not so simple, however, for a number of reasons. First, the idea of an equal distribution of
© Blackwell Publishing Ltd 2005.
334
Andrei Marmor
political power does not necessarily entail a system of majority vote. If the whole point of a majority vote is to provide each one of the participants with an equal share of the power to determine the result, then any fair lottery system would seem to be an adequate way of implementing this egalitarian principle (see, for example, Waldron 1999b, 160-1; Estlund 2000, 119-20). Maybe so. But we have good reasons, based on the considerations we have mentioned so far, to reject the lottery system. A lottery decision procedure would undermine the rationality of the deliberation stage of democracy. Let us recall that the considerations of fairness at the deliberation stage were premised on the idea that people should have an equal opportunity to influence the political decision. It makes no sense to maintain that people should have a meaningful opportunity to influence a political result when the result is entirely determined by chance. The rationality of the principles of fairness at the stage of deliberation requires that people's input into the deliberation process have a potential influence on the output. This can only be achieved if there is some causal connection between the deliberation and the decision. Any type of a lottery system severs such a causal connection, and thus renders the deliberation process entirely superfluous. In other words, democratic theory assumes that deliberation is, at least potentially, a rational process that can lead to changes in people's views which will culminate in the final vote.
The idea that an equal distribution of political power at the decision stage requires the "one person, one vote" model is obviously a simplification. Let me consider several complications, beginning with the question of the appropriate "majority." It is natural to assume that an equal distribution of political power requires the 1 /n formula for determining the voting procedure, where n simply represents the number of voters. This would seem to entail that anything which deviates from a simple majority model amounts to a non-egalitarian distribution of power and is therefore, at least prima facie, unwarranted. In most familiar democracies, however, supermajority requirements are not uncommon. Are these procedures undemocratic? Is it always wrong to deviate from the simple majority model?
One rationale which is often suggested for supermajorities ties its justification to the relative importance of the decision. Many people find it intuitively problematic, almost frightful, that crucial political decisions, profoundly important to the future of their nation, can be decided by simple majorities. There is a strong temptation to believe that the more important a decision is, the more far-reaching its long term effects, the larger the majority which should be required for its adoption. The intuition here is clear enough; it relies on the assumption that there is a strong element of contingency, almost arbitrariness, in the particular majority which happens to prevail at any given set of circumstances, and people feel that such arbitrary factors should not determine crucial political decisions. Given how easily political processes are subject to manipulation, there is certainly a grain of
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
335
truth in this kind of concern. But as a matter of principle, it is not warranted. If anything, the contrary is true. From the perspective of fairness, the more important a decision is, the more scrupulous we should be with respect to the equal distribution of the power to make that decision. A supermajority requirement always disrupts the equal distribution of political power because it is necessarily biased in favor of the status quo ante; it always reduces the relative political power of those who favor a social change. As a matter of principle, then, the relative importance of the decision, by itself, cannot justify a supermajority decision procedure.
It may well be argued, of course, that there are certain decisions in the public domain which should be immune from change. Perhaps there are certain rights or principles which should be protected from the vagaries of political decisions and removed from the ordinary democratic decision process. This is the main idea behind the constitutional protection of rights, principles and certain political structures.23 The question of whether there are, indeed, matters of political morality which should be protected from standard democratic decision procedures is a very complicated one, and I must leave its discussion for another occasion (on this issue, see Marmor 2005, chap. 9).
Other cases of a supermajority requirement, however, are not necessarily at odds with the equal distribution of political power. Consider, for example, the case of a relatively large persistent minority: suppose that country A is comprised of two main social groups, let us call them the greens and the reds. The greens form about 55% of the population, and the reds 45%. Now suppose that these two social groups are in a deep and longstanding conflict with each other, whereby the greens and the reds tend to vote on most issues according to their group interests which are typically at odds with the interests of the other group. Under these circumstances the reds are extremely unlikely to have their interests protected by a simple majority. In a very simplified way we can say that instead of having their political objectives materialized in about 45% of the cases, the reds are actually going to lose in 100% of the cases. Under these circumstances, it seems that a super-majority decision procedure may actually facilitate a much more egalitarian distribution of political power, thereby giving the reds a fair chance of having their votes affect the political outcomes.24 In other words, we some-
23 Constitutional entrenchment is always a form of supermajority decision procedure. Constitutions vary a great deal in the details of the procedures and in the degree of entrenchment. Basically, the more rigid a constitution is, the greater the supermajority which is required to introduce a constitutional change.
24 To be sure, I do not want to suggest that a supermajority decision procedure is always the best way to handle the problems of persistent minorities. Particularly when the minority forms a smaller proportion of the population, other means might be required, such as the constitutional protection of the minority's rights.
© Blackwell Publishing Ltd 2005.
336
Andrei Marmor
times need a supermajority decision procedure as a corrective measure when certain social forces make it very unlikely that a simple majority voting rule will implement an equal distribution of political power.25
To be sure, a supermajority requirement is not the only potential deviation from a strict majoritarian principle. Many other familiar political procedures and institutions raise similar worries. For example, it is probably impossible to defend the American system of senatorial representation among the states (namely, an equal number of representatives regardless of the population size) as a proper implementation of a majority-rule principle. Similar concerns apply to various forms of non-proportional representation systems, various "creative" designs of voting districts, and many other familiar features of contemporary democracies. I do not intend to argue that any democratic institution which deviates from the majority rule principle is therefore unjustified or illegitimate. Democratic regimes strive to implement many values and moral political concerns, some of which are bound to conflict with each other. It would be naive to presume that any single principle can justify such complex institutions. Each case must be judged on its particular merits.
Another familiar worry about majority voting concerns the fact that such a decision procedure is not designed to take into account the voters' intensity of preferences. This is usually understood to be troublesome for those theorists who strive to justify democracy on utilitarian grounds. This understanding of democracy is premised on the idea that the good government is the one which maximizes the overall satisfaction of preferences. Within the utilitarian model, however, the issue of the intensity of preferences becomes a serious problem: If each voter gets to cast one vote which is meant to express her preference, the intensity of such preferences cannot be taken into account, and this would seriously corrupt the utilitarian calculus.26 But even if we reject the utilitarian model, the question of whether the intensity of preferences should be taken into account remains, though from a different angle. Majority decision procedures are agnostic about the reasons for which people cast their vote. Intensity of preferences is typically determined by reasons; one's reasons for a given vote determine, among other things, how much one should care about the issue, and how important it actually is for
25 There are other cases in which a supermajority can be justified. For example, sometimes the rationale of a supermajority requirement stems from a concern about very low voter turnout. In particular, constituencies which face relatively frequent initiatives and referenda tend to have very low turnout at these elections, which makes it problematic to leave crucially important decisions turn on a fraction of the votes. In such cases the requirement of a supermajority seems justified (see Garrett 2001; for a rationale of supermajority requirements in legislative bodies, see Levmore 1992).
26 This should not be confused with the problems created by the ranking of preferences that lead to the famous cycling problems extensively discussed in the Public Choice literature. See Arrow 1970, and Mueller 1989, 384-99.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
337
that person to have the result come out as he would have it. Doesn't fairness require that such considerations be taken into account?
The answer is both yes and no. From the perspective of the egalitarian distribution of political power, intensity of preferences should not matter. In a way, the whole point of a majority decision procedure is to ensure that at the end of the day, when a decision has to be reached and votes cast, each person should have an equal share of the power to make the decision. This simply requires counting the votes. However, we should keep in mind that this egalitarian principle is premised on the value of personal autonomy. Although the right to personal autonomy should be distributed equally, any fair political decision procedure must take account of the fact that political decisions affect people's lives and social opportunities differently; therefore, some people may have good reasons to care more about the decision than others. A certain increase in the tax rate, for instance, might be much more burdensome for some people who happen to be in the minority, than its advantage to those who happen to form the majority. Thus it would seem to be unfair if we disregard the reasons for voting preferences altogether. But democracies can actually cope with this problem by facilitating the means of reasoning and bargaining. First, intensities of preferences can make some difference at the deliberation stage of the political process, which allows certain, though limited, mechanisms for reasoning, bargaining, and compromise. Much more importantly, however, an essential part of the rationale of representative democracies is to facilitate this bargaining aspect of the decision procedures and allow for intensities of preferences to count. Political parties will find it much easier, at the parliamentary level, to engage in bargaining, compromise, and log rolling which would give effect to the intensities of the preferences they represent (see Buchanan and Tullock 1967, and Cooter 2000).
In other words, considerations of fairness about the distribution of political power must be mixed. On the one hand, we must give each person an equal share of the power to make the decision, and this is, basically, a matter of counting votes. On the other hand, considerations of fairness also require that due weight be given to the fact that political decisions matter differently to different groups of people. This means that any fair decision procedure should make room for fair bargaining and compromise. Representative democracy is particularly well suited for accommodating both of these concerns.27
Acknowledging the need for bargaining raises another puzzle: Why don't we take one further step and allow vote-trading? If a particular decision is much more important for some people than others, why do we not allow
27 This is not meant to be a general argument against direct democracy. The latter involves many complicated issues that cannot be explored here. The point in the text does suggest, however, that even from the limited perspective of fairness, it should not be assumed that the more direct a democracy is, the better.
© Blackwell Publishing Ltd 2005.
338
Andrei Marmor
those who would be willing to pay for the votes of others to purchase them? From the vantage point of fairness, it is not obvious that the right to vote should be regarded as inalienable, as it is, in all contemporary democracies. The principle of fairness, in itself, would not explain the inalienability of the right to vote: The principle of an equal distribution of political power requires that each person be given a right to vote. It does not explain why people cannot trade those rights under any circumstances. Let me consider a tempting but wrong answer first. It might be tempting to think that the right to vote is inalienable because voting expresses one's consent to be governed. This argument is premised on the idea that the legitimacy of a government resides in the consent of the governed, and without the subjects' consent, the authority is illegitimate. The exercise of the right to vote is then taken to embody this element of consent, and its inalienability is regarded essential for maintaining the legitimacy of the government vis a vis each and every one of its putative subjects.28
This line of thought, however, is doubly flawed. First, it relies on the mistaken assumption that the legitimacy of an authority depends on the consent of its putative subjects. Nothing in the conditions for the legitimacy of a practical authority warrants such a thesis. Neither the soundness of the authoritative decision, that is, the NJT, nor the considerations of fairness we have mentioned above, rely on the idea of consent (see, for example, Raz 1981). But suppose that I am wrong about this, and that consent is rightly regarded as a condition of legitimacy. Would it render the right to vote inalienable in the requisite sense? It is very difficult to see how. Theorists who tend to regard voting as the expression of the consent to be governed immediately acknowledge that those who deliberately abstain from voting must also be taken to have expressed their consent to be governed, that is, they are taken to have acquiesced in whatever result turns out at the relevant ballot. But this makes it impossible to explain why selling one's vote voluntarily and under fair bargaining conditions would violate the consent element of the vote: If deliberate abstention amounts to acquiescence, why not allow voluntary transfer as well? Contractual rights, like property rights, are transferable.29
Perhaps we should look for the rationale of the prohibition on vote-trading in more mundane places. A familiar problem with vote-trading consists in the fact that it is particularly vulnerable to exploitation. The temptation to trade a non-tangible asset like the vote for immediate material benefits would be extremely strong among poor people: After all, they know that the marginal impact of their personal vote amongst millions of others is almost nil.
28 Perhaps this is the sentiment which is manifest in those democracies which make it legally obligatory for their citizens to vote in elections (e.g., Australia, Brazil).
29 It is worth keeping in mind that only the trade of vote for money (or other tangible goods) is prohibited; political vote-trades, in various forms, are part and parcel of political bargaining.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy 339
Since only the rich, or the very rich, would be able to afford to buy votes in sufficient quantities to make a political difference, a vote-trading system would dramatically increase the effect of wealth on politics, entrenching, de facto, a very unequal distribution of political power. Thus the prohibition of vote-trading is basically a safeguard against the slide from democracy to wealth-aristocracy.30 Admittedly, this is basically a "slippery slope" argument. But it is worth keeping in mind that the slope is very steep here.
Finally, we should bear in mind that the fairness of majority decision procedures crucially depends on the demarcation of the constituency. When we say that power must be distributed equally, the question always arises: Equally amongst whom? In the complex world we live in, constituencies can not be regarded as some kind of natural political entities, given in advance of political interests and decision procedures. When democratic decision procedures pervade such an enormous range of public decision levels, from condo associations through town-meetings, up to national and even international levels, the demarcation of constituencies becomes a ubiquitous problem, and there is little hope for generalizations and abstract answers.31 There is, however, one important point that should be stressed here. The argument for an equal distribution of political power was grounded on the idea of equal respect for people's personal autonomy. I have argued that in the context of political authority, this principle entails a majoritarian decision procedure. But the argument is limited to the basic structure of the political domain. In many other, let us say "micro" contexts, such as a faculty meeting or a condo association, the rationale of a majority decision procedure might be quite different. I would venture to suggest that in most of these contexts, the requirement of a majority vote is grounded on some principle of equal respect, but not necessarily respect for the participants' autonomy. In the context of a faculty meeting, for instance, it is presumably the equal respect for the participants' professional judgment that we should appeal to if we want to justify a majority decision procedure. At other times, it may be a principle of equal respect for people's stake in the outcome of the decision which could warrant an egalitarian distribution of power. In any case, it should not be assumed that the justification of democracy and the principles of fairness it instantiates necessarily apply, without modifications, to all levels of decision making, even if those decisions ought to be conducted according to some majority-rule principle.
30 One may be tempted to think that while a system of vote trading increases the inequality of political power, it may increase the equality of material resources, transferring money from the rich to the poor. This is just unrealistic since votes are ludicrously cheap to buy. I have been told that in the happy days of former Mayor Daley in Chicago, votes were bought for a few cans of beer. Vote trading does not seem to be a good system for the redistribution of wealth (see Hasen 2000). Furthermore, as Levmore argues, a system of vote trading would be plagued with problems of collective action (see Levmore 2000).
31 For some discussions of the constituency demarcation problem, see, for example, Raz and Margalit 1995, chapter 6.
© Blackwell Publishing Ltd 2005.
340
Andrei Marmor
4.
So far we have regarded the fairness of an authoritative decision procedure and the soundness of its result as separate issues. Sometimes, however, soundness and fairness are closely entangled, and at other times they may turn out to be in conflict. For the rest of this essay I would like to explore some of these relations between fairness and soundness of democratic procedures.
Perhaps the closest link between the fairness of a democratic decision procedure and the soundness of the result obtains in those relatively rare cases in which the democratic decision procedure instantiates a form of pure procedural justice. As Rawls defined it, "pure procedural justice obtains when there is no independent criterion for the right result: Instead there is a correct or fair procedure such that the outcome is likewise correct or fair, whatever it is, provided that the procedure has been properly followed" (Rawls 1971, 86). Now it is true that public authoritative decisions are rarely of this nature. Normally, we should be able to know what the criteria for the right decision are, regardless of the decision procedure. However, there are some cases in which something like pure procedural justice obtains. Consider, for example, a case where we have to choose a certain public official and there are several equally competent candidates (or, incommensurably on a par with each other). Assuming that those candidates do not represent different ideological constituencies, and that the election is purely a question of personal competence, this may well be a case in which any decision which is reached by a fair procedure is rendered correct and fair, and only because it results from such a fair decision procedure. Note, however, that in such cases a democratic decision procedure is only one option; any other fair procedure, like a fair lottery, would instantiate pure procedural justice and render the result equally acceptable. But again, such cases are relatively rare, and I do not want to argue that pure procedural justice plays a major role in democratic decision procedures. Nevertheless, the model is important, as I will explain in a moment.
Second, we should consider those cases in which a public decision is good only because most people actually want it. These are the cases that Dworkin labeled choice-sensitive issues: "Choice-sensitive issues are those whose correct solution [. . .] depends essentially on the character and distribution of preferences within the political community" (Dworkin 2000, 204).32 Roughly, then, choice-sensitive public decisions are those which become the correct decisions just in virtue of being supported by the majority. Consider, for example, a need to choose between allocating certain public funds to the construction of a swimming pool or a football stadium. Presumably, the
32 The missing part of this quote includes the words "as a matter of justice". I omitted these words advisedly, since it implies an unnecessary restriction. See also Raz 1989, 778.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
341
right and fair decision would be the one which is supported by the majority. But why? There are two possible explanations for this intuitive stance. One would rely on the utilitarian conception which favors the maximization of the overall satisfaction of preferences. Assuming that we leave the intensity of preferences constant (or that we just ignore this complication), it would be natural to conclude that a majority decision procedure is likely to lead to the best results: If more people prefer the swimming pool over the football stadium, building the former would satisfy more preferences than the alternative. One familiar problem with the utilitarian model is that it treats all public decisions in this manner, as if the satisfaction of subjective preferences is all that counts. This is a crude mistake, as it ignores the fact that most preferences we have are preferences which are based on reasons, and those reasons can be right or wrong, sound or unsound.33
But there is a better, non-utilitarian explanation for the rationale of choice-sensitive decisions. Some choices are basically a matter of taste, either because they are not supported by any particular reasons, or else, because the reasons which support them are entirely agent-relative. The essential nature of a choice which reflects taste, as opposed to other evaluative choices, consists in the self-regarding nature of its underlying reasons. Such choices are not based on reasons which ought to apply to others. Perhaps I prefer to have a swimming pool in my neighborhood, but this is not a preference that I can rationally regard as one that others should have as well. If they prefer something else, so be it. When decisions have to be reached on matters concerning preferences of taste, nobody can claim to have a better other-regarding reason than anyone else supporting any particular decision. Therefore, fairness of the decision procedure and soundness of the result are very closely linked here: If we cannot have all our taste-preferences satisfied, and none is better supported by reasons than any other, at least we should have the result which gets most people what they want (see Raz 1989, 778).34
The scope of choice-sensitive decisions in politics, understood as concerning matters of taste, is very limited. Political decisions are far more complex, and even if some choices are based on preferences of taste, they are often inextricably entangled with other, substantive issues of justice, culture, efficiency, and so on. Thus in most political contexts, the fairness of the decision procedure, and the soundness of the result, are separate issues, quite independent of each other. This brings us back to the question of
33 On the dependence of preferences on reasons I have elaborated in Marmor 2001a, chap. 8.
34 It is arguable that in such cases, when a decision has to be reached on matters of taste, a lottery decision procedure would be just as fair as voting. I think that this is basically correct. As a practical matter, however, there would often be a problem in isolating choice sensitive decisions from others. Often such decisions are closely entangled with other decisions which ought to be reason-based and open for rational deliberation.
© Blackwell Publishing Ltd 2005.
342
Andrei Marmor
legitimacy, and the reasons for following democratic authorities. As we have noted earlier, the primary reasons for following an authoritative decision consists in the NJT, that is, in the assumption that the subjects are likely better to comply with the reasons which apply to them by following the authoritative decision, than by trying to figure out, or act on, those reasons by themselves. Now, the relations between the fairness of the decision procedure and the NJT raises two important questions: First, are there reasons to believe that democratic decision procedures are likely to yield sound results? Second, would we have reasons to follow democratic decisions only because they are fair, even if the decisions do not comply with the NJT? An attempt to answer the first question would need to rely on a great deal of empirical evidence, and I cannot hope to answer such a question here.35 I would like to conclude, however, with a few sketchy comments on the second question.
Thus, the question is this: Would people have reasons to follow democratic decisions even if they are not sound, that is, even if the decision does not comply with the NJT? This is a very broad and complicated issue, so let me try to narrow it down since I do not want to discuss the entire range of arguments which have been offered to justify political obligation in democratic regimes. I will assume here, following Joseph Raz's conception of authority, that people's primary reasons for following an authoritative decision consist in the NJT. As we have noted from the outset, however, the fairness of the decision procedure does not seem to form any part of the reasons for following an authoritative directive. But this is not quite accurate, and it is time now to qualify this statement. I will consider two such qualifications.
First, in addition to the primary reasons for obeying an authority, captured by the NJT, people often have secondary or auxiliary reasons as well. The main example of such auxiliary reasons obtains where obedience to a mistaken decision is instrumentally conducive to supporting an otherwise legitimate and well functioning authority (see Marmor 2001a, 102-3). The question remains whether the fairness of the authoritative decision proce-
35 For example, many of those who are inclined to justify democracy on Utilitarian grounds think that the famous Condorcet "jury theorem" provides a very promising line of thought: Roughly, this theorem holds that if the following two conditions obtain, the probability of getting the correct result increases with the number of voters; the conditions are, first, that each and every voter is more likely than not to get the decision right, and second, that votes are cast independently of each other. The main problem with the application of the Condorcet theorem to democratic decision procedures is that these two conditions rarely obtain. Generally, even under ideal conditions of deliberation, we have no assurance that all voters, or even that most of them, have a greater than 0.5 probability of reaching the right decision when they cast their vote. Nor is it generally true that in a democratic process, voters cast their votes independently of each other: Voters often tend to follow those whom they regard as more knowledgeable than they are; then we have bargaining, strategic behavior, and other familiar factors, all of which undermine the second condition of the Condorcet theorem (see, for example, Dahl 1989,141-2; Waldron 1999a, 140; Levmore 1999).
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
343
dure is sufficient to ground such auxiliary reasons for obedience. It is not difficult to see that under certain circumstances, the answer would be affirmative.
When we have a reasonably fair and well functioning democracy, the duty to obey mistaken decisions can be derived from the duty to support just institutions (Rawls 1971, 354). When the democratic institutions operate in a fair and just manner, people would normally have good reasons to support the democratic institutions, even if such a support occasionally involves the need to obey mistaken decisions. Note, however, that this is a very limited conclusion, based on instrumental considerations, and therefore it does not entail a general obligation to obey the law. Not every act of obedience to the law is instrumentally valuable for supporting the democratic regime (see Raz 1979, 241). For instance, the publicity element is a crucial factor: If the authority's decision is mistaken, and one can disobey the decision without anyone else knowing about it, it is difficult to see how the duty to support just institutions would tell against such an act of disobedience. So this line of thought entails a very modest conclusion: The subjects of a democratic authority would have a reason or maybe an obligation to obey a mistaken authoritative decision only if such an obedience is, as matter of fact, instrumentally valuable as a means of support for the proper functioning of the democratic regime, assuming that the regime is by and large legitimate.
Second, in addition to such auxiliary reasons for political obligation, there may also be cases in which it is initially more important for the subjects of an authority to have a fair decision procedure than a sound result. In fact, there is one very limited type of case in which this is evidently so: When the authoritative decision instantiates a form of pure procedural justice, it is clearly the case that the fairness of the procedure is more important than the soundness of the result, simply because, ex hypothesis, we do not have a procedure-independent criterion for the soundness of the result. Now, despite the fact that pure procedural justice has a very limited application to politics, it does suggest a more general point. The fairness of the decision procedure in cases of pure procedural justice is decisive of the legitimacy of the result because we just do not have any independent criterion for the desired result. But the question of whether we do or do not have a procedure-independent criterion for the soundness of the result is not necessarily a dichotomous issue; even if there are such criteria, sometimes they are not knowable, or extremely difficult to ascertain, or not supported by enough available evidence. In other words, we can regard pure procedural justice as a limiting case, with some other cases closer or farther removed from it to various degrees. This would suggest that if the criteria for the soundness of an authoritative result are in serious doubt, it may become more important to have a fair decision procedure than a sound result.
A note of caution might be appropriate here. I am not suggesting that the fact of controversy, or pluralism, renders certain aspects of democratic
© Blackwell Publishing Ltd 2005.
344
Andrei Marmor
politics close enough to the model of pure procedural justice. The fact that people disagree on the criteria for a sound political result does not necessarily render the existence and content of such criteria in serious doubt. Controversy, by itself, does not entail warranted skepticism. However, to the extent that skepticism about criteria for sound political results is objectively warranted in particular cases, the fairness of the decision procedure may become decisive in determining the reasons for following an authoritative decision.
Let me sum up: We have seen two types of cases in which the reasons for following an authority's decision are primarily determined by the fairness of the decision procedure rather than the soundness of the result. One type concerns cases in which the duty to obey the authority's decision derives from the duty to support just institutions. Another type concerns cases in which the democratic decision procedure comes very close to a model of pure procedural justice. Admittedly, both of these cases are rather limited in scope. Ultimately, the justification of democratic authority must also rest on the soundness of the results. No political authority can be regarded as legitimate unless it meets the condition of the normal justification thesis. Fairness, I have tried to argue here, is very important. But fairness is not enough. If we are to be content with democracy, we must also assume that, by and large, democratic decision procedures are likely to result in sound decisions.
University of Southern California The School of Law & The School of Philosophy Los Angeles, CA 90089-0071, USA E-mail: amarmor@law.usc.edu
References
Arendt, H. 1958. The Human Condition. Chicago, IL: University of Chicago Press.
Arneson, R. 2004. Democracy is not Intrinsically Just. In Justice and Democracy. Eds. K. Dowding, R. E. Goodin, C. Pateman. Cambridge: Cambridge University Press.
Arrow, K. 1970. Social Choice and Individual Values. 2nd ed. New Haven, CN: Yale University Press.
Brighouse, H. 1996. Egalitarianism and Equal Availability of Political Influence. The
Journal of Political Philosophy 4: 118-41. Buchanan, J. M., and G. Tullock. 1962. The Calculus of Consent: Logical Foundations of
Constitutional Democracy. Ann Arbor, MI: University of Michigan Press. Christiano, T. 1996. The Rule of the Many. Boulder, CO: Westview. Cohen, J. 1999. Deliberation and Democratic Legitimacy. In Deliberative Democracy.
Eds. J. Bohman and W. Rehg. Cambridge, MA: MIT. Cooter, R. 2000. The Strategic Constitution. Princeton, NJ: Princeton University Press. Dahl, R. A. 1989. Democracy and Its Critics. New Haven, CN: Yale University Press. Darwall, S. 1977. Two Kinds of Respect. Ethics 88: 36-49. Dworkin, R. 2000. Sovereign Virtue. Harvard: Harvard University Press.
© Blackwell Publishing Ltd 2005.
Authority, Equality and Democracy
345
Estlund, D. 1999. Beyond Fairness and Deliberation: The Epistemic Dimension of Democratic Authority. In Deliberative Democracy. Eds. J. Bohman and W. Rehg. Cambridge, MA: MIT.
Estlund, D. 2000. Jeremy Waldron on Law and Disagreement. Philosophical Studies 99: 111-28.
Estlund, D. 2002. Political Quality. In Democracy. Ed. D. Estlund. Oxford: Blackwell. Garrett, E. 2001. Issues in Implementing Referendums in Israel: A Comparative Study
in Direct Democracy. Chicago Journal of International Law 2: 159-82. Hasen, R. L. 2000. Vote Buying. California Law Review 88: 1323-71. Hershovitz, S. 2003. Legitimacy, Democracy, and Razian Authority. Legal Theory 9:
201-20.
Levmore, S. 1992. Bicameralism: When Are Two Decisions Better than One? International Review of Law and Economics 12: 145-62.
Levmore, S. 1999. Voting Paradoxes and Interest Groups. Journal of Legal Studies 28: 259-81.
Levmore, S. 2000. Voting with Intensity. Stanford Law Review 53: 111-61.
Marmor, A. 2001a. Positive Law and Objective Values. Oxford: Oxford University Press.
Marmor, A. 2001b. Do We Have a Right to Common Goods? The Canadian Journal of
Law and Jurisprudence 14: 213-25. Marmor, A. 2002. Book Review of Law and Disagreement. Ethics 112: 410-14. Marmor, A. 2004. The Intrinsic Value of Economic Equality. In Rights, Culture and the
Law: Essays after Joseph Raz. Eds. L. H. Meyer, S. L. Paulson, and T. W. Pogge.
Oxford: Oxford University Press. Marmor, A. 2005. Interpretation and Legal Theory. Revised Second Edition. Oxford: Hart. Mill, J. S. 1991. Considerations on Representative Government. Buffalo, NY: Prometheus.
(1st ed. 1861.)
Mueller, D. 1989. Public Choice II. Cambridge: Cambridge University Press.
Rawls, J. 1971. A Theory of Justice. Oxford: Oxford University Press.
Rawls, J. 1993. Political Liberalism, New York, NY: Columbia University Press.
Raz, J. 1979. The Authority of Law. Oxford: Clarendon.
Raz, J. 1981. Authority and Consent. Virginia Law Review 67: 103-31.
Raz, J. 1986. The Morality of Freedom. Oxford: Oxford University Press.
Raz, J. 1989. Liberalism, Skepticism, and Democracy. Iowa Law Review 74: 761-86.
Raz, J., and A. Margalit. 1995. The Right to Self Determination. In Ethics in the Public
Domain. Ed. J. Raz. Oxford: Clarendon. Richardson, H. 1999. Democratic Intentions. In Deliberative Democracy. Eds. J.
Bohman and W. Rehg. Cambridge, MA: MIT. Waldron, J. 1993. Liberal Rights. Cambridge: Cambridge University Press. Waldron, J. 1999a. Law and Disagreement. Oxford: Oxford University Press. Waldron, J. 1999b. The Dignity of Legislation. Cambridge: Cambridge University Press.
© Blackwell Publishing Ltd 2005.