Why Deliberative Democracy?

Chapter 1

WHAT DELIBERATIVE DEMOCRACY MEANS

To go to war is the most consequential decision a nation can make. Yet most nations, even most democracies, have ceded much of the power to make that decision to their chief executives--to their presidents and prime ministers. Legislators are rarely asked or permitted to issue declarations of war. The decision to go to war, it would seem, is unfriendly territory for pursuing the kind of reasoned argument that characterizes political deliberation.

Yet when President George W. Bush announced that the United States would soon take military action against Saddam Hussein, he and his advisors recognized the need to justify the decision not only to the American people but also to the world community. Beginning in October 2002, the administration found itself engaged in argument with the U.S. Congress and, later, with the United Nations. During the months of preparation for the war, Bush and his colleagues, in many different forums and at many different times, sought to make the case for a preventive war against Iraq.¹ Saddam Hussein, they said, was a threat to the United States because he had or could soon have weapons of mass destruction, and had supported terrorists who might have struck again against the United States. Further, he had tyrannized his own people and destabilized the Middle East.

In Congress and in the United Nations, critics responded, concurring with the judgment that Hussein was a terrible tyrant but challenging the administration on all its arguments in favor of going to war before exhausting thenonmilitary actions that might have controlled the threat. As the debate proceeded, it became clear that almost no one disagreed with the view that the world would be better off if Saddam Hussein no longer ruled in Iraq, but many doubted that he posed an imminent threat, and many questioned whether he actually supported the terrorists who had attacked or were likely to attack the United States.

This debate did not represent the kind of discussion that deliberative democrats hope for, and the deliberation was cut short once U.S. troops began their invasion in March 2003. Defenders and critics of the war seriously questioned one another's motives and deeply suspected that the reasons offered were really rationalizations for partisan politics. The administration, for its part, declined to wait until nonmilitary options had been exhausted, when a greater moral consensus might have been reached. But the remarkable fact is that even under the circumstances of war, and in the face of an alleged imminent threat, the government persisted in attempting to justify its decision, and opponents persevered in responding with reasoned critiques of a preventive war.

The critics are probably right that no amount of deliberation would have prevented the war, and the supporters are probably right that some critics would never have defended going to war even if other nonmilitary sanctions had ultimately failed. Yet the deliberation that did occur laid the foundation for a more sustained and more informative debate after the U.S. military victory than would otherwise have taken place. Because the administration had given reasons (such as the threat of the weapons of mass destruction) for taking action, critics had more basis to continue to dispute the original decision, and to challenge the administration's judgment. The imperfect deliberation that preceded the war prepared the ground for the less imperfect deliberation that followed.

Thus even in a less than friendly environment, deliberative democracy makes an appearance, and with some effect. Both the advocates and the foes of the war acted as if they recognized an obligation to justify their views to their fellow citizens. (That their motives were political or partisan is less important than that their actions were responsive to this obligation.) This problematic episode can help us discern the defining characteristics of deliberative democracy if we attend to both the presence and the absence of those characteristics in the debate about the war.

What Is Deliberative Democracy?

Most fundamentally, deliberative democracy affirms the need to justify decisions made by citizens and their representatives. Both are expected to justify the laws they would impose on one another. In a democracy, leaders should therefore give reasons for their decisions, and respond to the reasons that citizens give in return. But not all issues, all the time, require deliberation. Deliberative democracy makes room for many other forms of decision-making (including bargaining among groups, and secret operations ordered by executives), as long as the use of these forms themselves is justified at some point in a deliberative process. Its first and most important characteristic, then, is its reason-giving requirement.

The reasons that deliberative democracy asks citizens and their representatives to give should appeal to principles that individuals who are trying to find fair terms of cooperation cannot reasonably reject. The reasons are neither merely procedural ("because the majority favors the war") nor purely substantive ("because the war promotes the national interest or world peace"). They are reasons that should be accepted by free and equal persons seeking fair terms of cooperation.

The moral basis for this reason-giving process is common to many conceptions of democracy. Persons should be treated not merely as objects of legislation, as passive subjects to be ruled, but as autonomous agents who take part in the governance of their own society, directly or through their representatives. In deliberative democracy an important way these agents take part is by presenting and responding to reasons, or by demanding that their representatives do so, with the aim of justifying the laws under which they must live together. The reasons are meant both to produce a justifiable decision and to express the value of mutual respect. It is not enough that citizens assert their power through interest-group bargaining, or by voting in elections. No one seriously suggested that the decision to go to war should be determined by logrolling, or that it should be subject to a referendum. Assertions of power and expressions of will, though obviously a key part of democratic politics, still need to be justified by reason. When a primary reason offered by the government for going to war turns out to be false, or worse still deceptive, then not only is the government's justification for the war called into question, so also is its respect for citizens.

A second characteristic of deliberative democracy is that the reasons given in this process should be accessible to all the citizens to whom they are addressed. To justify imposing their will on you, your fellow citizens must give reasons that are comprehensible to you. If you seek to impose your will on them, you owe them no less. This form of reciprocity means that the reasons must be public in two senses. First, the deliberation itself must take place in public, not merely in the privacy of one's mind. In this respect deliberative democracy stands in contrast to Rousseau's conception of democracy, in which individuals reflect on their own on what is right for the society as a whole, and then come to the assembly and vote in accordance with the general will.²

The other sense in which the reasons must be public concerns their content. A deliberative justification does not even get started if those to whom it is addressed cannot understand its essential content. It would not be acceptable, for example, to appeal only to the authority of revelation, whether divine or secular in nature. Most of the arguments for going to war against Iraq appealed to evidence and beliefs that almost anyone could assess. Although President Bush implied that he thought God was on his side, he did not rest his argument on any special instructions from his heavenly ally (who may or may not have joined the coalition of the willing).

Admittedly, some of the evidence on both sides of the debate was technical (for example, the reports of the U.N. inspectors). But this is a common occurrence in modern government. Citizens often have to rely on experts. This does not mean that the reasons, or the bases of the reasons, are inaccessible. Citizens are justified in relying on experts if they describe the basis for their conclusions in ways that citizens can understand; and if the citizens have some independent basis for believing the experts to be trustworthy (such as a past record of reliable judgments, or a decision-making structure that contains checks and balances by experts who have reason to exercise critical scrutiny over one another).

To be sure, the Bush administration relied to some extent on secret intelligence to defend its decision. Citizens were not able at the time to assess the validity of this intelligence, and therefore its role in the administration's justification for the decision. In principle, using this kind of evidence does not necessarily violate the requirement of accessibility if good reasons can be given for the secrecy, and if opportunities for challenging the evidence later are provided. As it turned out in this case, the reasons were indeed challenged later, and found to be wanting. Deliberative democracy would of course have been better served if the reasons could have been challenged earlier.

The third characteristic of deliberative democracy is that its process aims at producing a decision that is binding for some period of time. In this respect the deliberative process is not like a talk show or an academic seminar. The participants do not argue for argument's sake; they do not argue even for truth's own sake (although the truthfulness of their arguments is a deliberative virtue because it is a necessary aim in justifying their decision). They intend their discussion to influence a decision the government will make, or a process that will affect how future decisions are made. At some point, the deliberation temporarily ceases, and the leaders make a decision. The president orders troops into battle, the legislature passes the law, or citizens vote for their representatives. Deliberation about the decision to go to war in Iraq went on for a long period of time, longer than most preparations for war. Some believed that it should have gone on longer (to give the U.N. inspectors time to complete their task). But at some point the president had to decide whether to proceed or not. Once he decided, deliberation about the question of whether to go to war ceased.

Yet deliberation about a seemingly similar but significantly different question continued: was the original decision justified? Those who challenged the justification for the war of course did not think they could undo the original decision. They were trying to cast doubt on the competence or judgment of the current administration. They were also trying to influence future decisions--to press for involving the United Nations and other nations in the reconstruction effort, or simply to weaken Bush's prospects for reelection.

This continuation of debate illustrates the fourth characteristic of deliberative democracy--its process is dynamic. Although deliberation aims at a justifiable decision, it does not presuppose that the decision at hand will in fact be justified, let alone that a justification today will suffice for the indefinite future. It keeps open the possibility of a continuing dialogue, one in which citizens can criticize previous decisions and move ahead on the basis of that criticism. Although a decision must stand for some period of time, it is provisional in the sense that it must be open to challenge at some point in the future. This characteristic of deliberative democracy is neglected even by most of its proponents. (We discuss it further below in examining the concept of provisionality.)

Deliberative democrats care as much about what happens after a decision is made as about what happens before. Keeping the decision-making process open in this way--recognizing that its results are provisional--is important for two reasons. First, in politics as in much of practical life, decision-making processes and the human understanding upon which they depend are imperfect. We therefore cannot be sure that the decisions we make today will be correct tomorrow, and even the decisions that appear most sound at the time may appear less justifiable in light of later evidence. Even in the case of those that are irreversible, like the decision to attack Iraq, reappraisals can lead to different choices later than were planned initially. Second, in politics most decisions are not consensual. Those citizens and representatives who disagreed with the original decision are more likely to accept it if they believe they have a chance to reverse or modify it in the future. And they are more likely to be able to do so if they have a chance to keep making arguments.

One important implication of this dynamic feature of deliberative democracy is that the continuing debate it requires should observe what we call the principle of the economy of moral disagreement. In giving reasons for their decisions, citizens and their representatives should try to find justifications that minimize their differences with their opponents. Deliberative democrats do not expect deliberation always or even usually to yield agreement. How citizens deal with the disagreement that is endemic in political life should therefore be a central question in any democracy. Practicing the economy of moral disagreement promotes the value of mutual respect (which is at the core of deliberative democracy). By economizing on their disagreements, citizens and their representatives can continue to work together to find common ground, if not on the policies that produced the disagreement, then on related policies about which they stand a greater chance of finding agreement. Cooperation on the reconstruction of Iraq does not require that the parties at home and abroad agree about the correctness of the original decision to go to war. Questioning the patriotism of critics of the war, or opposing the defense expenditures that are necessary to support the troops, does not promote an economy of moral disagreement.

Combining these four characteristics, we can define deliberative democracy as a form of government in which free and equal citizens (and their representatives), justify decisions in a process in which they give one another reasons that are mutually acceptable and generally accessible, with the aim of reaching conclusions that are binding in the present on all citizens but open to challenge in the future.³ This definition obviously leaves open a number of questions. We can further refine its meaning and defend its claims by considering to what extent deliberative democracy is democratic; what purposes it serves; why it is better than the alternatives; what kinds of deliberative democracy are justifiable; and how its critics can be answered.

How Democratic Is Deliberation?

In its origins, deliberative politics has an ambivalent relation to modern democracy. Its roots can be traced to fifth-century Athens. According to Pericles, political leaders then saw discussion not as a "stumbling-block in the way of action" but as an "indispensable preliminary to any wise action at all."⁴ Aristotle was the first major theorist to defend the value of a process in which citizens publicly discuss and justify their laws to one another.⁵ He argued that ordinary citizens debating and deciding together can reach a better decision than can experts acting alone. But the Athenian democracy of Pericles and Aristotle was quite different from ours. Only a small portion of the residents counted as citizens; many were slaves. The deliberation took place in an assembly open to all citizens, not in a legislature or in the campaigns that characterize democratic practice in our time. And though Aristotle saw the virtues of deliberation by the many, he preferred aristocracy, wherein the deliberators would be more competent, and the deliberation more refined.

In the early modern period, deliberation was more explicitly contrasted with democracy. When the term "deliberative" was first used to refer to political discussion (evidently as early as 1489), it referred to discussion within a small and exclusive group of political leaders. By the eighteenth century, deliberation was part of a defense of political representation that pointedly resisted appeals to popular opinion. Edmund Burke's "Speech to the electors of Bristol," which declared that "Parliament is a deliberative assembly," is famously a defense of a trustee conception of representation that today seems more aristocratic than democratic.⁶ Neither did the founders of the new American nation embrace a fully democratic form of deliberation. The authors of the Federalist Papers certainly sought institutions that would promote deliberation. But although in the view of one commentator their constitutional design "combined deliberation and democracy,"⁷ the degree of democracy they tolerated remained very much limited in scope and membership.

The most prominent nineteenth-century advocate of "government by discussion"--John Stuart Mill--is rightly considered one of the sources of deliberative democracy. But he too continued to prefer that this discussion be led by the better educated.⁸ It was not until the early part of the twentieth century that deliberation came to be decisively joined to democracy. In the writings of John Dewey, Alf Ross, and A. D. Lindsay we finally find unequivocal declarations of the need for political discussion in a polity recognizably democratic in the modern sense. These theorists not only included widespread deliberation as part of democracy, but saw it as a necessary condition of this form of government. Lindsay regarded discussion as "the essential of democracy."⁹

More than any other theorist, Jürgen Habermas is responsible for reviving the idea of deliberation in our time, and giving it a more thoroughly democratic foundation. His deliberative politics is firmly grounded in the idea of popular sovereignty.¹⁰ The fundamental source of legitimacy is the collective judgment of the people. This is to be found not in the expression of an unmediated popular will, but in a disciplined set of practices defined by the deliberative ideal. Some critics, however, complain that his conception does not adequately protect liberal values, such as freedom of religion or human rights. His proceduralism, the critics suggest, realizes democracy at the expense of liberalism. They believe that a theory of justice like that of John Rawls provides a more secure foundation for these values without denying the legitimate claims of democracy.

We note later that Habermas and Rawls are not so far apart as this contrast suggests. But here the point to keep in mind is that the democratic element in deliberative democracy should turn not on how purely procedural the conception is but on how fully inclusive the process is. While deliberation is now happily married to democracy--and Habermas deserves much of the credit for making the match--the bond that holds the partners together is not pure proceduralism. What makes deliberative democracy democratic is an expansive definition of who is included in the process of deliberation--an inclusive answer to the questions of who has the right (and effective opportunity) to deliberate or choose the deliberators, and to whom do the deliberators owe their justifications. In this respect, the traditional tests of democratic inclusion, applied to deliberation itself, constitute the primary criterion of the extent to which deliberation is democratic. (It must be said, however, that this defense of deliberative democracy does not suffice to show that it has overcome its aristocratic origins. One of the recurring objections, which we take up later, is that deliberative democracy is exclusive in various ways, excluding some people not by legal or formal restrictions as early deliberative politics did, but by informal norms defining what counts as proper deliberation.)

What Purposes Does Deliberative Democracy Serve?

The general aim of deliberative democracy is to provide the most justifiable conception for dealing with moral disagreement in politics. In pursuing this aim, deliberative democracy serves four related purposes. The first is to promote the legitimacy of collective decisions. This aim is a response to one of the sources of moral disagreement--scarcity of resources. Citizens would not have to argue about how best to distribute health care or who should receive organ transplants if these goods and services were unlimited. In the face of scarcity, deliberation can help those who do not get what they want, or even what they need, to come to accept the legitimacy of a collective decision.

The hard choices that public officials have to make should be more acceptable, even to those who receive less than they deserve, if everyone's claims have been considered on the merits, rather than on the basis of the party's bargaining power. Even with regard to decisions with which many disagree, most of us take one attitude toward those that are adopted after careful consideration of the relevant conflicting moral claims, and quite a different attitude toward those that are adopted merely by virtue of the relative strength of competing political interests.

The second purpose of deliberation is to encourage public-spirited perspectives on public issues. This aim responds to another source of moral disagreement--limited generosity. Few people are inclined to be wholly altruistic when they are arguing about contentious issues of public policy, such as defense spending or health-care priorities. Deliberation in well-constituted forums responds to this limited generosity by encouraging participants to take a broader perspective on questions of common interest.

To be sure, politicians are not automatically transformed from representatives of special interests into trustees of the public interest as a result of talking to one another. The background conditions in which the deliberation takes place are critical. Deliberation is more likely to succeed to the extent that the deliberators are well informed, have relatively equal resources, and take seriously their opponents' views. But even when the background conditions are unfavorable (as they often are), citizens are more likely to take a broader view of issues in a process in which moral reasons are traded than in a process in which political power is the only currency.

The third purpose of deliberation is to promote mutually respectful processes of decision-making. It responds to an often neglected source of moral disagreement--incompatible moral values. Even fully altruistic individuals trying to decide on the morally best standards for governing a society of abundance would not be able to reconcile some moral conflicts beyond a reasonable doubt. They would still confront, for example, the problem of abortion, which pits the value of life against the value of liberty. Even issues of national security can pose questions about which people can reasonably disagree--under what conditions is a nation justified in starting a war, on its own, against another nation?

Deliberation cannot make incompatible values compatible, but it can help participants recognize the moral merit in their opponents' claims when those claims have merit. It can also help deliberators distinguish those disagreements that arise from genuinely incompatible values from those that can be more resolvable than they first appear. And it can support other practices of mutual respect, such as the economy of moral disagreement described earlier.

Inevitably, citizens and officials make some mistakes when they take collective actions. The fourth purpose of deliberation is to help correct these mistakes. This aim is a response to the fourth source of disagreement, incomplete understanding. A well-constituted deliberative forum provides an opportunity for advancing both individual and collective understanding. Through the give-and-take of argument, participants can learn from each other, come to recognize their individual and collective misapprehensions, and develop new views and policies that can more successfully withstand critical scrutiny. When citizens bargain and negotiate, they may learn how better to get what they want. But when they deliberate, they can expand their knowledge, including both their self-understanding and their collective understanding of what will best serve their fellow citizens.

It is all too easy to assume that we already know what constitutes the best resolution of a moral conflict, and do not need to deliberate with our fellow citizens. To presume that we know what the right resolution is before we hear from others who will also be affected by our decisions is not only arrogant but also unjustified in light of the complexity of the issues and interests that are so often at stake. If we refuse to give deliberation a chance, not only do we forsake the possibility of arriving at a genuine moral compromise but we also give up the most defensible ground we could have for maintaining an uncompromising position: that we have fairly tested our views against those of others.

Tugging on the coattails of Thomas Jefferson, a little boy (in a New Yorker cartoon) once asked: "If you take those truths to be self-evident, then why do you keep on harping on them so much?" The answer from a deliberative perspective is that such claims deserve their status as self-evident truths for the purposes of collective action only if they can withstand challenge in a public forum. Jefferson himself argued for open deliberative forums, indeed even periodic constitutional conventions, in which citizens could contest conventional wisdom.¹¹ An implication of taking the problem of incomplete understanding seriously is that the results of the deliberative process should be regarded as provisional. Some results are rightly regarded as more settled than others. We do not have to reargue the question of slavery every generation. But the justification for regarding such results as settled is that they have met the deliberative challenge in the past, and there is no reason to believe that they could not do so today.

Why Is Deliberative Democracy Better Than Aggregative Democracy?

To appreciate the value of deliberative democracy, we need to consider the alternatives. Obviously, there are many conceptions of democracy, and many moral theories that support these conceptions. To begin, we should distinguish first- and second-order theories.¹² First-order theories seek to resolve moral disagreement by demonstrating that alternative theories and principles should be rejected. The aim of each is to be the lone theory capable of resolving moral disagreement. The most familiar theories of justice--utilitarianism, libertarianism, liberal egalitarianism, communitarianism--are first-order theories in this sense. Each theory claims to resolve moral conflict, but does so in ways that require rejecting the principles of its rivals. In contrast, deliberative democracy is best understood as a second-order theory. Second-order theories are about other theories in the sense that they provide ways of dealing with the claims of conflicting first-order theories. They make room for continuing moral conflict that first-order theories purport to eliminate. They can be held consistently without rejecting a wide range of moral principles expressed by first-order theories. Deliberative democracy's leading rivals among second-order theories are what are known as aggregative conceptions of democracy.¹³

The deliberative conception, as we have indicated, considers the reasons that citizens and their representatives give for their expressed preferences. It asks for justifications. The aggregative conception, by contrast, takes the preferences as given (though some versions would correct preferences based on misinformation). It requires no justification for the preferences themselves, but seeks only to combine them in various ways that are efficient and fair. Some preferences may be discounted or even rejected, but only because they do not produce an optimal result, not because they are not justified by reasons.

The best way to reveal the essential differences between these conceptions is to examine their responses to the basic problem of democratic politics that both of them purport to address--how to make legitimate decisions for the society as a whole in the face of fundamental disagreement. The core of the problem is not merely that people disagree, but that some of the disagreement is reasonable.¹⁴ It is built into the circumstances of social and political life. When citizens disagree about such issues as the morality of abortion, capital punishment, starting a preventive war, or funding health care, deliberation does not produce agreement, and perhaps even should not.¹⁵ Let us assume, then, that there are some disagreements that at any particular time cannot be resolved by deliberation. Yet governments must make decisions. How should they decide?

Aggregative theories offer two seemingly different but closely related methods. The first is a form of majoritarianism: put the question to the people and let them vote (or let them record their preferences in public opinion surveys).¹⁶ The most common version of this method is to let the representatives of the people make the decision, again by majority vote, or some similar rule, in the legislature. The representatives themselves are chosen in elections, which are viewed as "competitive struggle[s] for the people's vote."¹⁷ The electoral process is modeled on the analogy of the market. Like producers, politicians and parties formulate their positions and devise their strategies in response to the demands of voters who, like consumers, express their preferences by choosing among competing products (the candidates and their parties). Whatever debate takes place in the campaign serves a function more like that of advertising (informing the voters about the comparative advantages of the candidates) than like that of argument (seeking to change minds by giving reasons).

The second aggregative method gives less deference to the votes and opinions of citizens: officials take note of the expressed preferences but put them through an analytic filter--such as cost-benefit analysis--which is intended to produce optimal outcomes. In some versions of this process, preferences based on misinformation or faulty heuristics can be corrected, and sets of preferences that produce irrational results (such as cyclical majorities) can be modified. This method originates in classical utilitarianism and owes its contemporary pedigree to welfare economics. But it is not necessarily democratic. Giving voters the final word is not the most rational way to produce policies and laws that maximize welfare. Experts may be more competent at finding laws and policies that serve that end. But proponents of this method usually welcome democratic procedures such as elections, because they recognize that experts and the politicians who appoint them cannot always be trusted to pursue the public interest.

What these methods have in common--and what defines aggregative conceptions--is that they take the expressed preferences as the privileged or primary material for democratic decision-making. Preferences as such do not need to be justified, and aggregative conceptions pay little or no attention to the reasons that citizens or their representatives give or fail to give. They regard reasons as significant only insofar as the reasons help predict or correct preferences. (The reasons might, for example, enable politicians to anticipate future preferences, or they might help analysts identify preferences that are based on misinformation.) Aggregative theorists thus believe that the collective outcomes produced by their various methods need no further justification beyond the rationale for the method itself. The majoritarian or utilitarian assumptions underlying the method provide its justification. Reasons can be given for the outcomes, but they are to be found not in the preferences but in the rationale for the method of combining the preferences.

Aggregative conceptions have important advantages. First, they produce determinate outcomes, at least in principle.¹⁸ The result of an election or the conclusion of a cost-benefit analysis yields definite decisions. This is no small advantage in dealing with the problem of disagreement, especially in disputes that are not resolvable on reasonable terms. Deliberative democrats recognize of course that decisions must be made--even when the reason-giving process is incomplete. On any conception of democracy, elections must be held, and in elections citizens express their will without giving reasons. But deliberative democrats tend to emphasize the provisionality of political outcomes more than their finality.

A second advantage of aggregative conceptions is that they rely on relatively uncontroversial procedures to resolve disagreement. They also provide ways of reaching decisions that can be said to express the views of most citizens, and may even be regarded as fair under the circumstances. The methods of aggregative democrats are not morally neutral, as they sometimes claim, but the methods do not entail positions on most substantive issues, and do not pass moral judgment on the individual preferences that citizens express, however base or noble they may be. The most common methods take preferences as given, and therefore can be said to be less paternalistic. Even the methods that correct the preferences still seek to respect what citizens or voters actually desire, or would desire if they were better informed--not what they should desire if they were more public spirited or if they were more inclined to respect the principle of reciprocity.

Despite these substantial advantages, the aggregative conception is seriously flawed, and cannot serve as a principled basis for democratic decision-making. By taking existing or minimally corrected preferences as given, as the base line for collective decisions, the aggregative conception fundamentally accepts and may even reinforce existing distributions of power in society. These distributions may or may not be fair, but aggregative conceptions do not offer any principles by which we can decide. Even more important, they do not provide any process by which citizens' views about those distributions might be changed.

A second fundamental problem of aggregative conceptions is that they do not provide any way for citizens to challenge the methods of aggregation themselves. The "preference" for a different method of decision-making--the argument for a deliberative process, for example--cannot be treated in the same fashion as other preferences are, and simply factored into a cost-benefit analysis.¹⁹ The argument rejects in critical respects the assumptions of such an analysis. Furthermore, aggregative methods do not welcome all kinds of primary preferences equally. Those that can be readily translated into economic categories fit much better than those that express values that are incommensurable. Sometimes governments have to put a price on life and health, but they have to recognize that the value of life and health is not completely captured by their price, even in decision-making about public policy. Even on their own terms, then, aggregative methods do not always answer the critical question decision-makers must ask: should a government give priority, for example, to treating conditions that are not life-threatening but cause large numbers of people considerable discomfort, or to treating conditions that are life-threatening but affect only small numbers of people?

Consider the problem the state of Oregon faced in the early 1990s: how to allocate the state's limited resources for health care for residents enrolled in Medicaid. To set priorities for its publicly funded health care under Medicaid, the Oregon Health Services Commission created a list of several hundred conditions and treatments, ranked mainly on the basis of cost-benefit calculations. Treatments lower on the list were regarded as less cost-beneficial than those higher on the list, and therefore less likely to receive funding. In essence, the Commission followed the second kind of method recommended by aggregative democrats. The ranking did not correspond to popular opinion about what the most serious diseases are: some life-threatening conditions ranked lower because their treatment was relatively expensive or affected relatively small numbers of people. But the ranking was a good-faith attempt to maximize the welfare of the largest number of citizens, given the limited resources the state had at its disposal.

The list of priorities provoked a public outcry. Health policy that followed these priorities might maximize the welfare of most citizens, but the rankings departed so far from what most citizens thought was right or fair that no state official could continue to justify the policy. Capping a tooth ranked much higher than an appendectomy, for example. The Commission might have reverted to the first method recommended by aggregative democrats--conducting a survey or referendum and taking the results as final. But the Commission realized that public opinion on this complex set of issues was inchoate, and would depend on how the questions were phrased. Instead, the Commission wisely enlisted the aid of the methods of deliberative democracy.

The Commission undertook an elaborate process of consultation. It sponsored community meetings at which participants were "asked to think and express themselves in the first person plural . . . as members of a statewide community for whom health care has a shared value." Deliberation went through stages, as leaders presented their proposals, citizens responded, leaders revised, citizens reacted. This is what we call the reiteration of deliberation. It is an illustration of the dynamic character of deliberation. Eventually, the Commission presented a revised list, one that most observers deemed an improvement over the original plan.

Yet the Oregon experience should remind us that deliberative democracy is not a perfect way to deal with the problem of moral disagreement. (We defer until a later section discussion of the general criticisms of deliberative democracy, and consider here only two objections that are most relevant to the comparison with aggregative conceptions.) First, deliberative democracy does not provide a natural way to come to a definite conclusion short of consensus, which is not to be expected in most cases of decision-making. Deliberative politics almost always has to be supplemented by other decision procedures--in the Oregon case by a recommendation of a commission and a vote by the legislature. The community groups provided helpful input, which informed the further deliberations by both the Commission and the legislature, but in the end the disagreement that remained had to be resolved by a majority vote in the legislature. Deliberation must end in a decision, but deliberative democracy does not itself specify a single procedure for reaching a final decision. It must rely on other procedures, most notably voting, which in themselves are not deliberative.

Second, the deliberative conception relies on explicitly moral principles rather than the seemingly neutral ones of aggregative conceptions. Reciprocity is an explicitly moral principle. Deliberation therefore invokes substantive moral claims that may be independent of the preferences citizens put forward. In the Oregon episode, the most serious flaw in the proposed policy was not the ranking of treatments per se, but the unfairness of rationing under these circumstances. Because only citizens below certain income levels were eligible for any support at all, the rationing necessary at the relatively low level of funding available would cause some poor citizens to lose out to other poor citizens. Some of the participants in the deliberative process recognized that this was unfair, but to express that recognition they had to appeal at least implicitly to a principle of justice that not everyone accepted. Furthermore, in order to eliminate this unfairness they had to call for an increase in the total budget for health care--an option that went beyond the agreed-upon agenda of the community meetings.

Neither of these disadvantages is fatal to the case for deliberative democracy, however. Indeed, the problems each identifies can be turned to the advantage of deliberative democracy. The fact that deliberative democracy does not in itself define a unique method for bringing deliberation to a justified conclusion (short of a moral consensus) means that it acknowledges that no single method can justify whatever results from its implementation. No decision-making method, for example, should be able to justify a war of aggression. Deliberative democracy can accommodate many different kinds of decision-making procedures to reach final decisions, including voting and executive order, provided they are justified in a deliberative forum. More important, the open-ended nature of deliberation enables citizens or legislators to challenge earlier decisions, including decisions about the procedures for making decisions. Deliberative democracy's provisionality checks the excesses of conventional democracy's finality.

The appeal to justice and other potentially controversial principles may intensify the disagreement that exists, but it can also lead to new ways of dealing with it, ways that would otherwise have been neglected. In the Oregon case, the deliberative process forced officials and citizens to confront a serious problem of injustice that they had previously evaded--the unfairness of a harsh rationing scheme that affected only poor citizens. As a result, the basic unfairness in the policy was somewhat lessened in a way that neither most of the critics of the plan nor its proponents expected. When the legislators finally saw what treatments on the list would have to be curtailed or eliminated under the projected budget, they managed to find more resources, and increased the total budget for health care for the poor. Apart from the results, the year-long deliberations helped citizens, legislators, and health-care professionals come to a better understanding of their own values--those they shared and those they did not. They were able, as they went forward, to confront the difficult decisions they had to make about health-care policy, to work together in a more cooperative "first person plural" spirit. The continuing process also exposed another glaring defect of the original process--the absence of the poor citizens who would be most affected by the policy. It became clearer that in the future they should be more adequately represented in the process.

In the face of disagreement, deliberative democracy tells citizens and their representatives to continue to reason together. If the disagreement is resolvable on reciprocal terms, deliberation is more likely than aggregation to produce agreement. If it is not so resolvable, deliberation is more likely than aggregation to produce justifiable agreement in the future, and to promote mutual respect when no agreement is possible. By engaging in deliberation, citizens acknowledge the possibility that they may change their preferences. The preferences that they assert now may not be the preferences they find they wish to express later. The very nature of the deliberative process of justification sends a signal that its participants are willing to enter into a dialogue in which the reasons given, and the reasons responded to, have the capacity to change minds.

On many disagreements, especially reasonable ones, people will not change their minds, no matter how respectfully they deliberate with their opponents. If citizens persist in defending the position with which they began, what difference does it make if they come to regard their opponents' positions as morally reasonable? This thicker kind of respect encourages citizens to consider their opponents' positions on the merits, rather than to try to explain them as products of unfavorable conditions, such as impaired judgment, misguided motives, or cultural influences. Such an attitude is more conducive to appreciating that even benevolent and intelligent but fallible people are likely to disagree on morally difficult matters such as military intervention and heath-care policy--as well as abortion, capital punishment, affirmative action, and many other overtly moral issues. Moreover, considering positions on their merits generally builds a stronger basis for respect for persons than explaining positions as a product of unfavorable conditions. Certainly, some disagreements are the result of such conditions, and when a position can be shown to be justifiable mainly from a perspective that depends on such conditions, mutual respect (of both persons and positions) does not prevent, and may require, that the critics of the position point out its defective origins. But in the absence of a specific showing of this kind, the presumption of respectful deliberation is that positions should be challenged on their merits.

What Kind of Deliberative Democracy?

Deliberative democrats have to deal with another kind of disagreement--not among citizens but among themselves. They disagree about the value, status, aims, and scope of deliberation, and their disagreements yield different versions of the theory of deliberative democracy. Some of these differences, we suggest, can be reconciled, and some cannot. In either case, recognizing the differences can help clarify the nature of both the theory and the practice of deliberative democracy.

Instrumental or Expressive?

Deliberative democrats disagree about whether deliberation has only instrumental value, as a means of arriving at good policies, or whether it also has expressive value, as a manifestation of mutual respect among citizens. On the instrumental view (sometimes called the epistemic view), deliberating about political issues has no value in itself. It is valuable only to the extent that it enables citizens to make the most justifiable political decisions.²⁰ On the expressive view of deliberation, significant value resides in the act of justifying laws and public policies to the people who are bound by them.²¹ By deliberating with one another, decision-makers manifest mutual respect toward their fellow citizens.

When the Oregon Commission consulted with community members about alternative proposals for funding health-care services, citizens could reasonably expect that Commission members would arrive at better outcomes than when they decided, without public deliberation, to rank capping teeth above treating acute appendicitis. The same citizens could also reasonably believe that the Commission's deliberation promoted a value basic to any democratic government--the expression of mutual respect between decision-makers and their fellow citizens. By their willingness to exchange views before rendering a binding decision, the commission members treated their fellow Oregonians as subjects, not merely objects, of decision-making. Had the Commission acted without deliberation, the value of this expression of mutual respect would have been lost, however correct or just the policy might have been.

These two views of the values that deliberative democracy is supposed to promote are not incompatible. Indeed, any adequate theory must recognize both. If deliberation tended to produce worse decisions than other processes in the long run, then it would not serve the expressive purpose. A process that generally produced bad outcomes would hardly express mutual respect. Citizens might participate on equal terms, but with results that few would see as worthy. The value would at best be like the faint satisfaction that players feel on a team that constantly loses its games. The instrumental view reminds us that because the stakes of political decision-making are high, and deliberation is a time-consuming activity, a deliberative process should contribute to fulfilling the central political function of making good decisions and laws.

But if we were to regard deliberation as only instrumental, we would fail to recognize the moral significance of the political fact that the decisions of government bind people other than the decision-makers themselves. Political officials cannot rightly decide an issue simply by claiming that they know that their preferred policies are right for their fellow citizens. They need to seek the views of those citizens who have to live with the results of the policies. When binding decisions are routinely made without deliberation, the government not only conveys disrespect for citizens, but also exposes its lack of adequate justification for imposing the decision on them. Furthermore, there is a practical reason for officials to recognize the expressive value of deliberation: they can thereby increase the likelihood not only of discovering but also of implementing good public policy. If citizens perceive that their views are not being respected, they may seek to block otherwise good policies.

If political deliberation tends to produce better decisions in the long run, and if political decision-makers in a democracy owe justifications to those who are bound by their decisions, then the instrumental and expressive rationales for deliberation can be mutually supportive. By deliberating with their fellow citizens, decision-makers can arrive at better, more adequately justified decisions and, in the process, express mutual respect among free and equal citizens.

The instrumental and expressive values cannot of course be reconciled in practice in every particular case. A deliberative process that otherwise expresses mutual respect can nonetheless produce an unjust outcome. And a nondeliberative process can produce a more nearly just result in some cases. Yet deliberative democracy, as we shall see, has the capacity both to criticize unjust outcomes and to recognize its own limits. In this way it tends, over time, to reconcile its own instrumental and expressive values.

Procedural or Substantive?

Another, closely related conflict that has divided deliberative democrats can also be resolved more readily than has usually been assumed. This conflict concerns the status of the principles of the theory: should they be procedural or substantive? Pure proceduralism holds that the principles should apply only to the process of making political decisions in government or civil society.²² Thus the principles should not prescribe the substance of the laws, but only the procedures by which laws (such as equal suffrage) are made and the conditions necessary for the procedures to work fairly (such as free political speech). Democratic theory, the proceduralists hold, should not incorporate substantive principles such as individual liberty or equal opportunity because such constraints are not necessary to ensure a fair democratic process. Pure proceduralists do not deny that substantive principles such as freedom of religion, nondiscrimination, or basic health care are important, but they insist on keeping these principles out of their democratic theory.

Deliberative theorists who favor a more substantive conception deny that procedural principles are sufficient. They point out that procedures (such as majority rule) can produce unjust outcomes (such as discrimination against minorities). Unjust outcomes, they assume, should not be justifiable on any adequate democratic theory. A theory that allows for the possibility that such outcomes are justified should be especially objectionable to deliberative democrats. A fundamental aim of deliberative democracy is to offer reasons that can be accepted by free and equal persons seeking fair terms of cooperation. Such reasons could rarely justify unjust outcomes. The idea of free and equal personhood itself provides substantive moral content for principles that would reject an unjust decision even if it had been reached by procedurally just means.

The reasons most often offered in defense of both substantive and procedural principles are associated with liberalism, broadly speaking. They reflect what it means to respect individuals as free and equal citizens. Those rights that are fundamental to human agency, dignity, or integrity (freedom of religion, racial nondiscrimination, and so on) need to be secured, along with rights related to the procedural aspects of democracy (such as the right to vote). Appreciating that majoritarian procedures can support aggressive wars, racial or religious discrimination, and other patently unjust policies, the principles of deliberative democracy, the substantive theorists insist, must go beyond process.²³

A purely procedural conception of deliberative democracy, on its face, shares with aggregative theories the advantage of minimalism. Once the right procedures are in place, whatever emerges from them is right. It follows that if majority rule is right, then so are its results. But few pure proceduralists defend pure majority rule. Much like aggregative theorists, pure proceduralists usually support a more complex procedural conception, and they also offer reasons other than minimalism in its defense. To oppose the inclusion of substantive principles, they invoke a particular view of moral and political authority. Who has the authority to legislate in a democracy? Democratic citizens, not democratic theorists, pure proceduralists answer. Citizens or their representatives, within broad procedural limits, should be as free as possible to determine the content of laws. The substantive principles that some theorists would include in their conceptions of deliberative democracy in effect preempt the moral and political authority of citizens. Racial and religious discrimination and aggressive wars are usually wrong, but the questions of whether a particular law or decision should be so described, or whether the wrong decision should be overridden by more compelling considerations, should be left to citizens and their accountable representatives, not to theorists and their substantive principles. A deliberative theory that includes substantive principles, so the argument goes, improperly constrains democratic decision-making, including the process of deliberation itself.

Substantive theorists reply that the principles they propose are no less fundamental and no more contestable than the principles on which proceduralists rely. Procedural principles have substantive content, too. If majority rule is better than minority rule, it must be for moral reasons. These reasons refer to such values as free and equal personhood, the same values that support substantive principles. How procedural principles should be interpreted and how they should be applied are often controversial, and reasonably so. Procedural theories therefore cannot occupy a privileged place relative to substantive theories. Procedural and substantive principles alike require democratic deliberation, at least with respect to how they should be interpreted and applied. Both threaten to usurp legitimate democratic authority if they are put forward, without benefit of democratic deliberation, as morally and politically authoritative.

It follows that if the moral and political authority of free and equal citizens is to be safeguarded, then neither procedural nor substantive principles of deliberative democracy can claim priority. Both need to be treated as morally and politically provisional (in ways that we explain more fully below). Procedural and substantive principles should both be systematically open to revision in an ongoing process of moral and political deliberation. If the principles are understood in this way, the usual objections against including substantive principles lose their force. The provisional status of all principles, procedural and substantive alike, thus constitutes a distinctive strength of deliberative theory, and at the same time offers deliberative democrats an effective way of uniting procedural and substantive principles into a coherent theory.

The contrast between procedural and substantive theories ofdemocracy is sometimes thought to be reflected in a disagreement between Jürgen Habermas and John Rawls. Habermas is said to favor democratic deliberation over individual rights, and Rawls, rights over deliberation. But on the more careful interpretations of their theories, neither Habermas nor Rawls defends a purely procedural or purely substantive conception of democracy. As Habermas writes: ". . . private and public autonomy mutually presuppose each other in such a way that neither human rights nor popular sovereignty can claim primacy over its counterparts."²⁴ The democracy that Rawls defends is also fundamentally committed to securing both substantive and procedural principles.²⁵ The convergence between Habermas and Rawls suggests that the most compelling theories of deliberative democracy combine both substantive and procedural principles. They also both recognize that all democratic principles require substantive defense.

Consensual or Pluralist?

The disagreement among deliberative democrats who seek consensus and those who accept pluralism is more intractable than the disputes we have so far considered. Should deliberation aim at achieving consensus through realizing a common good or through seeking the fairest terms of living with a recalcitrant pluralism?

Deliberative democrats who identify with the republican tradition or with communitarianism in political theory typically seek a comprehensive or thick common good, one that goes beyond agreement on basic principles, whether procedural or substantive.²⁶ They do not assume that we will reach this goal, but they believe that it is nevertheless a worthy aim. It fulfills the deepest moral promise of deliberative democracy--a form of cooperation that all citizens could accept despite their deep differences of identity. Other deliberative democrats, drawing on the liberal tradition, argue that it is not always desirable to seek a comprehensive common good rather than to try to live respectfully with moral disagreements. One reason, they point out, is that some of these disagreements are inherent in the human condition. They arise because of our incomplete and incompatible moral and empirical understandings.²⁷

Virtually all deliberative democrats can agree that a primary aim of deliberation is to justify decisions and laws that citizens and their representatives impose on one another. In this sense, deliberative democrats share a consensus that deliberation aims at least at a thin conception of the common good. Finding fair terms of cooperation among free and equal persons is a common good for both individuals and society as a whole.

This agreement among deliberative democrats breaks down when we ask whether the common good can or should be comprehensive. Consensus democrats recognize that a comprehensive common good is an ideal and will not often, if ever, be achieved, but they regard the failure to achieve it as a sign of defects that can and should be remedied, whether they lie in the capacities of citizens and their representatives or in the practices and institutions of the polity. Consensus democrats also tend to resist the idea, which pluralists willingly embrace, that a great deal of political disagreement is built into the conditions of collective life, and that to eliminate it entirely would be undesirable. Consensus democrats criticize pluralists for settling for too thin a conception of the common good. Agreement on fair terms of cooperation, they argue, does not create a community in which citizens find common ground at the deepest level of their social identities. It does not even require citizens to engage deeply with one another over their deepest moral differences. They must set these differences aside, the pluralists are assumed to say, in order to find fair terms of cooperation and to arrive at just decisions and laws. The pluralists' common good therefore does not serve even deliberation itself well. A thin pluralist conception of the common good produces passive citizens, the consensus democrats argue, who too readily settle for the role of consumers of material commodities, rather than producers of public goods.

Pluralists reply that a democracy that seeks a comprehensive good threatens to become tyrannical. If moral differences are as deep and pervasive as pluralists believe, they can be eliminated in politics only by repression. As long as people in power and those to whom they are accountable are neither omniscient nor angelic, and as long as they reasonably disagree about how to rank incompatible values, deliberation should aim at achieving a noncomprehensive common good, and at finding good ways of living with ongoing moral disagreements. If deliberators aimed primarily at a comprehensive common good, they would be tempted to tolerate less diversity than the disharmonious moral universe demands.

Although pluralists agree that deliberation should strive to justify as much agreement as possible, they also seek ways of living well with those disagreements that cannot or should not be eliminated at any given time. This is the deep and irreconcilable difference between democrats who accept pluralism as part of the human condition and those who see it always as a serious po litical problem to be overcome by deliberation. Some disagreements--for example, a call to exclude blacks, Jews, or homosexuals from various associations--cry out for a democracy to confirm its commitment to the principles of nondiscrimination and equal opportunity in their core form. But other disagreements should not be resolved. We call these deliberative disagreements: they involve conflicts not between views that are clearly right and clearly wrong, but between views none of which can be reasonably rejected.²⁸ In the face of such disagreements, deliberative democrats should practice the economy of moral disagreement described earlier.

But economizing on moral disagreement does not eliminate it. Consider the debate on homosexual unions.²⁹ In the spirit of moral economizing, the state would seek a compromise: it would grant legal recognition to both homosexual and heterosexual unions, giving the same legal rights to partners of both kinds of union. This recognition would respect the principles of nondiscrimination and civic equality. At the same time, the state would not require religious associations to recognize either homosexual or heterosexual unions. Such tolerance would respect freedom of religious association as well as the right to argue, whether on a religious basis or not, that marriage should be a union of only men and women and that homosexual acts are sinful. Some citizens would want the law to require that all associations not discriminate. Others would continue to defend the freedom of private associations to discriminate, although they themselves might not oppose homosexual unions. And still others would insist that the state should legally recognize homosexual unions as marriages in every respect including by name.

Because reasonable differences will persist, democratic governments and their citizens should learn from the way they are expressed and dealt with. By their nature, reasonable differences contain partial understandings. Each alone is likely to be mistaken if taken comprehensively, all together are likely to be incoherent if taken completely, but all together are likely to be instructive if taken partially. A democracy can govern effectively and prosper morally if its citizens seek to clarify and narrow their deliberative disagreements without giving up their core moral commitments. This is the pluralist hope. It is, in our view, both more charitable and more realistic than the pursuit of the comprehensive common good that consensus democrats favor.