Deep Thoughts: how democratic is any democracy?

The State of Democratic Theory -- Ian Shapiro

THE DEMOCRATIC IDEA is close to nonnegotiable in today's world. Liberation movements insist that they are more democratic than the regimes they seek to replace. Authoritarian rulers seldom reject democracy outright. Instead they argue that their people are not ready for democracy "yet," that their systems are more democratic than they appear, or that the opposition is corrupt and antidemocratic--perhaps the stooge of a foreign power. International financial institutions may be primarily interested in countries' adopting neoliberal market reforms, yet they also feel compelled to call for regular elections and other democratic political reforms. Of course, different people understand different things by democracy, and every democratic order will be thought by some not to be functioning as it should, in the corrupt control of an illicit minority, or otherwise in need of repair. But the very terms of such objections to democracy affirm its obligatory character, since it is the malfunction or corruption of democracy that is being objected to.

Within democratic systems it is accepted that people are free to despise the elected government, but not its right to be the government. Christian fundamentalists may believe they are acting on God's orders, but the fact that they claim to be a "moral majority " indicates that, as far as political legitimacy is concerned, they understand democracy's nonoptional character. Constitutional arrangements sometimes limit democracy's range, particularly in separation-of-powers systems such as the United States. But constitutions generally contain entrenched guarantees of democratic government as well. Moreover, they are themselves revisable at constitutional conventions or via amendment procedures whose legitimacy is popularly authorized. Even liberal constitutionalists like Bruce Ackerman (1993b) agree that critical moments of constitutional founding and change require popular democratic validation if they are to be accepted as legitimate over time.

Nonetheless, democracy's nonnegotiable political status has long stood in contrast to the widespread skepticism about it among political theorists. Generations of scholars following Kenneth Arrow (1951) have questioned the rationality of its inner logic, and many others have been deeply skeptical of its desirability as a political outlook. John Dunn (1979: 26) captured this skepticism well with the observation that although most people think of themselves as democrats, democratic theory oscillates between two variants, "one dismally ideological and the other fairly blatantly utopian." The oscillation Dunn had in mind was between Cold War rhetoric masquerading as theory and arguments for egalitarian and participatory democracy that lacked convincing attention to how they might actually be deployed. Despite its legitimacy in the world, democratic theory did not then seem to be going anywhere interesting or worthwhile.

In the years since Dunn wrote there has been a revival of interest in the study of democracy, fed by the dramatic and unexpected increase in the number of democracies in the world in Africa, Asia, and Latin America--not to mention the countries of the former Soviet Empire. Between 1980 and 2002, some eighty-one countries moved from authoritarianism to democracy, including thirty-three military dictatorships that were replaced by civilian governments (United Nations 2002). Yet if the turn of events has rendered Dunn's opposition anachronistic, it is far from clear that democracy's underlying theoretical difficulties have been addressed satisfactorily. The time seems ripe for a reassessment of the state of democratic theory in light of the actual operation of democratic politics. That is the enterprise attempted here.

Assessments of this sort require yardsticks, two of which suggest themselves. One is normative, implied when we ask how persuasive the theories are that seek to justify democracy as a system of government. The other is explanatory, prompted by asking how successful the theories are that try to account for the dynamics of democratic systems. Normative and explanatory theories of democracy grow out of literatures that proceed, for the most part, on separate tracks, largely uninformed by one another. This is unfortunate, partly because speculation about what ought to be is likely to be more useful when informed by relevant knowledge of what is feasible, and partly because explanatory theory too easily becomes banal and method-driven when isolated from the pressing normative concerns that have fueled worldwide interest in democracy in recent decades. Accordingly, I take an integrative tack, focusing on what we should expect of democracy, and on how those expectations might best be realized in practice.

Sharpening this focus, one inevitably confronts dissensus on both issues. The book is organized around these disagreements. I begin, in chapter 1, with a discussion of the main contending views of democratic purposes, and more specifically the normative claim popular among theorists in what I describe as the aggregative and deliberative traditions that democracy should be geared toward arriving at some notion of the general will that reflects the common good. This is Rousseau's formulation of the problem as set forth in The Social Contract , and aggregative theorists follow his lead in trying to discover the general will by "taking men as they are and laws as they might be" (Rousseau [1762] 1968: 49). They regard preferences as given and concern themselves with how best to tot them up. The aggregative tradition has bequeathed a view of democracy in which competing for the majority's vote is the essence of the exercise, and the challenge for democratic theorists as they conceive it is to come up with the right rules to govern the contest. Deliberative theorists, by contrast, are more Aristotelian in taking a transformative view of human beings (see Aristotle [ca. 330 B.C.] 1977: bk. 2). They concern themselves with the ways in which deliberation can be used to alter preferences so as to facilitate the search for a common good. For them the general will has to be manufactured, not just discovered.

Proponents of aggregation and deliberation thus operate with different views of the human condition and the possibilities of collective life; indeed, each group often defends its views as much by pointing to the demerits of the other as by putting forth constructive arguments on its own behalf. Yet both camps share Rousseau's assumption that democracy's task is to express a general will that reflects the common good. For aggregative theorists the alleged impossibility of doing this is said to be proof of democracy's impossibility, whereas for deliberativists the goal is to get people to engage in deliberation so as to forge, and sometimes also implement, policies that serve the common good. I argue that both groups overestimate the importance of the idea of the common good for democracy. Instead, democracy is better thought of as a means of managing power relations so as to minimize domination. My view embodies a view of the common good, to be sure, but it is a stripped-down one: less demanding in its assumptions about collective rationality than either the aggregative or the deliberative view, and more sensitive to considerations about power. Indeed, taking my cue from Machiavelli's ([ca. 1517] 1970: 1.5) intimation at the start of the Discourses , I define the common good in a democracy as that which those with an interest in avoiding domination share. In chapters 2 and 3, I revisit conceptions of politics as deliberation and competition for a majority of votes from this perspective, exploring the conditions under which they can be advantageous once Rousseau's construction of the enterprise has been jettisoned.

If a central task for democracy is to enable people to manage power relations so as to minimize domination, the following questions arise: what is domination? how do we know it when we see it? and how effective can democratic governments aspire to be at reducing it? These questions shape much of the ensuing discussion, but some prefatory comments are in order here. Weber ([1914] 1968: 53) defined domination as "the probability that a command with a specific content will be obeyed by a given group of persons," insisting that "the existence of domination turns only on the actual presence of one person successfully issuing orders to others." I conceive of domination differently, as resulting from the illegitimate exercise of power. In some respects this is a broader conception than Weber's because domination as I conceive it can occur without explicit orders emanating from identifiable agents. Although I do not go all the way with Foucault (1977, 1982) in thinking that domination can be entirely divorced from human agency, I do agree that domination can result from a person's, or a group's, shaping agendas, constraining options, and, in the limiting case, influencing people's preferences and desires. Domination can also occur without the need for explicit commands when one person or group secures the compliance of another as a by-product of their control of resources that are essential for the second person or group, or, in the terminology I will deploy, is in a position to threaten their basic interests.

My conception is narrower than Weber's in that I regard domination as arising only from the illegitimate exercise of power. Compliance is often compelled in armies, firms, sports teams, families, schools, and countless other institutions. Indeed, political theorists from Plato to Foucault have often noted that the ineradicably hierarchical character of much social life makes power relations ubiquitous to human interaction. But this does not mean that domination is. There is a world of difference, for instance, between a teacher's requiring a student to do her homework and his taking advantage of his powerful position to engage in sexual harassment of her. The latter is domination, but the former is not. Hierarchical relations are often legitimate, and, when they are, they do not involve domination on my account. Yet hierarchical relations must concern democratic theorists nonetheless because there is always the possibility that, left unchecked, they can facilitate domination. This is why I have argued elsewhere (Shapiro 1999a) that hierarchies should generally be presumed suspect and be structured so as to minimize the likelihood that they will atrophy into systems of domination.

To be sure, this is often difficult to achieve, partly because governments frequently lack the necessary information and partly because they act with notoriously blunt instruments--creating the danger that the cure can be worse than the disease. Accordingly, there are good reasons for structuring social life, where possible, so that people will discover incentives to democratize things for themselves by creating mechanisms through which those who wield power are held accountable to those over whom it is wielded. Where this fails, I argue, government legitimately intervenes, but one of the more important creative challenges involves calibrating the intervention by reference to the seriousness of the possibility of domination, so as to avoid self-defeating efforts at democratic reform.

Governments can help structure the power dimensions of human interaction so as to ameliorate domination in many walks of life, but since they wield power willy-nilly, they are also potential agents of domination themselves. Indeed, for much of the twentieth century governments were arguably the most fearsome sources of domination in the world, and they continue to be so in many places. The appalling excesses of fascism and communism may have led some political theorists to adopt too govern-mentalist a view of politics, missing the ways in which governments underwrite domination throughout civil society, however implicitly, and undervaluing the ways in which they can ameliorate it. Yet there is no doubt that governments continue to be major exercisers of power in the world, and that a central task for democratic theorists is to devise ways of making them more accountable to those over whom their power is exercised--servants of the people rather than their masters.

If the task for democratic theorists is to devise better ways for governments to render the exercise of power legitimate, then democratic theory should be informed by considerable attention to the nature of power. Unfortunately, at the same time as the democracy literature has been oddly innocent of the research on power, the power literature been preoccupied with epistemological issues to the virtual exclusion of their implications for the theory and practice of democracy. I try to redress this mutual blindness in my discussions of deliberation and electoral competition, by showing how insights from the power literature can help us specify the conditions under which these practices can operate to minimize domination.

In chapter 2, I argue that although deliberation can sometimes be inimical to undermining domination, there are settings in which it can be helpful. Because people cannot really be forced to deliberate or to pursue any particular goal when they do deliberate, the challenge for democratic institutional designers is to structure the incentives so that people will want to deploy deliberation to minimize domination in the course of their collective endeavors. The best way for government to try to foster this is to strengthen the rights of appeal, delay, and in extreme cases even veto--but only of those who are vulnerable to the power of others because they have basic interests at stake in a given setting. Strengthening these rights may lead to the reduction of domination through deliberation, and, even when it fails to achieve this result, it nonetheless makes sense from the perspective of a stripped-down conception of the common good geared to reducing domination. Strengthening the hand of the vulnerable in such settings is desirable, even if they end up using it to bargain or negotiate rather than to deliberate.

In chapter 3, I turn to the literature on majoritarian competitive democracy. The classic power-centered analysis of it is Joseph Schumpeter's Capitalism, Socialism, and Democracy , published in 1942, and I organize my discussion around an examination of the debates between Schumpeter and his critics. Like many liberal constitutionalists, Schumpeter was keenly aware of the potential for the legitimate exercise of power to atrophy into illegitimate forms of domination. The liberal constitutionalist impulse is to try to wall power off by limiting the sphere of collective action. This is not an approach, I argue, that is attractive or even coherent when viewed through the prism of the power literature. More plausible is the Schumpeterian impulse to control power by making it the object of electoral competition. In this, I argue, he delivers more effectively than anything that can be found in The Federalist on the Madisonian aspiration to ensure that "ambition will be made to counteract ambition" (Hamilton, Madison, and Jay [1788] 1966: 160).

Schumpeter's critics fall into two groups: those who think his competitive democracy desirable but insufficient and those who think it undesirable. I argue that, when compared with the going alternatives, wholesale rejections of Schumpeterianism are unpersuasive. Those who are hostile to Schumpeterian democracy usually value agreement and consensus more than competition, whether because of beliefs about what deliberation can do, or convictions that unanimity is inherently desirable, or because competition is thought to lead to destabilizing conflict. I argue that all three rationales for preferring consensus-based to competitive conceptions of democracy are wrongheaded. We do better, instead, to conceive of bipartisan agreement in antitrust terms as collusion in restraint of democracy. This is not to say that Schumpeterian democracy is without flaws, but I argue that the more fruitful path to addressing them involves exploring ways to make democracy more genuinely competitive than it is, to expand its reach beyond governmentalist institutions, and to supplement it with complementary institutional devices. I discuss a variety of ways in which this can be done, better to structure power relations so as to minimize domination. Inter alia, this involves supplementing the idea of nondomination with a principle of affected interest, geared to ensuring that there is genuine competition by decision-makers for the votes of those who are actually affected by their decisions.

If Schumpeterian democracy needs supplementation, questions arise: what should it consist of, and who should do the supplementing? Debates about these issues are taken up next. Because there are no perfect decision rules, a purely procedural scheme like competitive majority rule can produce self-defeating results. Most obviously, majorities can use their power to undermine democratic freedoms by abolishing opposition and undermining future political competition. The evidence suggests that the likelihood of this occurrence is often exaggerated, but the possibility exists and democratic procedures can in any case have perverse consequences in a host of subtler ways. Yet the difficulty with those who would question the procedural results of majority rule is that they do not--and in all likelihood cannot--agree on a "substantive" standard by reference to which those results can be measured. In this situation I argue that the best solution is to agree with those who propose a middle-ground approach, in which courts or other second-guessing institutions should play a reactive, escape-valve, role in limiting the perverse consequences of democratic procedures when they produce results that foster domination. Defending this view leads me into a discussion of alternative ways for courts to behave in democratic systems. I argue that their legitimacy appropriately varies with the degree to which they act in democracy-sustaining ways, and I supply a variety of examples of what this means in practice in the American context.

This is followed, in chapter 4, by a consideration of the literatures on democratic transitions and consolidation. The scholarship in this area prompts the thought that the state of democratic theory is a bit like the state of Wyoming: large, windy, and mainly empty. It reveals that we know something about some of the necessary conditions for some democratic transitions, but that there are numerous possible paths to democracy, and that we should not be looking for a single general theory--certainly not a predictive one. The scholarship also reveals that we know something about the economic preconditions for viable democracy, but, notwithstanding the confident assertions of numerous commentators, we are mainly in the dark about the cultural and institutional factors that influence democracy's viability. Prudence suggests that it is nonetheless wise to try to inculcate support for democracy among those who operate it, though it is far from clear just how important this is, or, indeed, how best to accomplish it.

It is often claimed that certain types of societies, to wit, "deeply divided" ones, are inherently incapable of democracy. I explore this claim in some depth, finding it unsupported by either evidence or convincing theoretical argument. Too often, arguments about divided societies operate as rationalizations for resisting democratization and sustaining domination. Nonetheless, I argue for an incremental approach, given the dearth of reliable knowledge about the adaptability of politicized identities to the requirements of competitive democracy, and I survey the different ways in which reform of electoral systems might be pressed into the service of this goal. This raises normative questions about the circumstances under which group aspirations merit deference in democratic polities. I take these questions up in the final section of chapter 4, analyzing them from the interest-based perspective geared to limiting domination developed in chapters 2 and 3.

Some potential sources of domination come from within the political system; others come from outside it. Perhaps the most important one of these derives from economic vulnerability to the power of others. This subject is of particular interest from the standpoint of democratic theory because many nineteenth-century thinkers--both defenders and critics of democracy--believed that imposing majority rule with a universal franchise on democracy in highly inegalitarian societies would lead the poor to soak the rich. This intuition was formalized in the twentieth-century political science literature via the median voter theorem, which predicts that in highly inegalitarian settings the median voter will vote for downward redistribution. In chapter 5, I take up the puzzle presented by history's having demurred. This is reflected in the reality that in the United States today the bottom fifth of the population either lives in poverty or is in imminent danger of so doing. Not only is there no systematic relationship between expanding the franchise and downward redistribution of income or wealth, but democracies often engage in regressive redistribution--sometimes with strong bipartisan support.

Whatever justice-related reasons one might have for concern about this, I argue that we have democratic reasons as well--given that so large a segment of the population is vulnerable to domination. I examine the relationship between democracy and distribution first from the supply side, where I explore institutional and other factors that impede the kinds of political competition that would be expected to lead to more redistribution to the bottom quintile of the population. Then I turn to the demand side, looking to issues concerning beliefs, information, ideologies, framing effects, and the psychology of interpersonal comparisons in accounting for the relative dearth of demand for downward redistribution. I also consider the counterintuitive hypothesis that the more unequal a society, the less likely it is that there will be effective demand for downward redistribution, for reasons having to do with the physical separation between rich and poor, the opening up of what I describe as empathy gulfs between them, and structural incentives for the rich and the middle classes to marginalize the poor. Like many other areas in the study of democratic politics, this is one in which a great deal of research remains to be done. But, in light of what we do know and can reasonably surmise, I advance a variety of suggestions for viable reforms that would make the American political system more responsive to the interests of the poor, undermining their vulnerability to domination. By that token at least, these reforms would move us a little closer to a genuine democracy than can the system we have at present.