Democratic Political Theory

John S Dryzek. Handbook of Political Theory. Editor: Gerald F Gaus & Chandran Kukathas. 2004. Sage Publications Ltd. 

A Plethora of Democracies

Adversarial, aggregative, associative, capitalist, Christian, classical, communicative, communitarian, consensual, consociational, constitutional, contestatory, corporatist, cosmopolitan, delegative, deliberative, developmental, difference, direct, discursive, ecological, economic, electoral, elitist, epistemic, feminist, global, grassroots, green, juridical, industrial, legal, liberal, local, majoritarian, minimalist, parliamentary, participatory, peoples,’ pluralist, populist, presidential, procedural, property-owning, protective, push-button, radical, reflective, representative, social, strong, thin, transnational and unitary are all adjectives that can be, and have been, attached to democracy.

One could write an account of the state of democratic theory by elucidating and juxtaposing the meanings of these 54 adjectives. But life is too short, and pages too limited. Let me begin instead with three observations about this list.

The list is a long one; there is a lot of democracy about, at least in theory, and perhaps in practice.

The categories represented by the adjectives are not mutually exclusive. While there are some obvious binary oppositions (aggregative versus deliberative, participatory versus representative), many combinations are plausible and have their advocates and critics.

The categories represented by these adjectives are not collectively exhaustive. The conversation about democratic development shows no signs of closure. (This is no bad thing; arguably the continuation of this conversation is intrinsic to democracy itself, though only for democracy as conceptualized by some of the categories.)

While covering a lot of territory, democratic theory is not completely unbounded. Contributors to the enterprise all address questions pertaining to the collective construction, distribution, application, and limitation of political authority. These questions define the boundaries of the democratic concourse, the sum of communication about democracy. Within these borders may be found a heartland where practitioners consider what rule by the people and the political equality it implies can mean in contemporary complex societies that also value liberty and efficiency. While it would be nice to be able to specify more precisely a common set of problems that democratic theorists try to resolve, along with a set of standards for what constitute adequate solutions, I believe that is not possible. For such standards emerge in the process of dialogue across theorists (and others), and their content may change with time.

Though historically contingent, these standards have real power, as indicated by the number of dead ducks that have fallen victim to them. The most noteworthy dead duck is democracy’s longstanding authoritarian opponent. In 1989 Robert Dahl could plausibly organize a major statement and defence of democracy using as a foil guardianship, the idea that some elite both knows what is best for society and has the appropriate expertise to implement that programme. In 2004 that would no longer be worth the effort. Serious advocates of guardianship can no longer be found. For example, in the realm of environmental political thought, in the 1970s ecoauthoritarian models were quite popular (for an extreme statement, see Heilbroner, 1974). Come the 1990s, the main flourishing enterprise in ecopolitical thought was green democracy (see for example the essays collected by Mathews, 1996), while advocates of ecological guardianship have almost vanished. Other dead ducks include people’s democracy, workplace democracy, community democracy (in the sense of whole-community mobilization), collectives, Theodore Lowi’s (1969) juridical democracy, and perhaps democratic socialism. Still, there are only a few dead democratic ducks, and they are always outnumbered by new democratic ducklings. Thus with time democratic theory no less than democracy itself becomes more differentiated and complex.

One might despair in the face of ever increasing variety. So, for example, John Dunn suggests that ‘democratic theory is the moral Esperanto of the present nation-state system, the language in which all Nations are truly united, the public cant of the modern world, a dubious currency indeed’ (1979: 2). Democracy does indeed sometimes look a bit like a Christmas tree, a positive symbol to which one can attach any good things one likes. Dahl (1971) prefers to use the term ‘polyarchy’ on the grounds that it provides more in the way of conceptual precision and real-world purchase than ‘democracy.’ Alternatively, one could try to engage in conceptual clarification when it comes to the term ‘democracy’ itself, to cut through the confusion in search of the essential meaning. This is the approach of, for example, Giovanni Sartori, who describes his book on democratic theory as ‘above all, a housecleaning venture, a task of dispelling sloppiness (in argument) and messiness (in conception)’ (1987: xi). Such an approach will not do for two reasons. First, part of what makes democracy interesting in both theory and practice is contestation over its essence. Second, any search for the essential meaning of democracy is undermined by conceptual historians who point to the inevitable historical contingency of key political concepts like democracy (Hanson, 1989), and how democracy’s meaning is itself constitutive of politics at particular times and places.

Conceptual history will do for political theory in history of ideas mode. Indeed, Russell Hanson’s excellent (1989) history is instructive in that it reminds us that only in the nineteenth century does democracy as a concept cease to be universally reviled, and come to attract positive connotations. In the United States, this comes about largely as a result of the Jacksonian and populist concern for ‘The Democracy’—that is, for the ordinary people against the plutocracy. However, while necessary, conceptual history is insufficient for those with critical and evaluative concerns who want to contribute to the continuing conversation of democratic development. Moreover, conceptual history does not equip us to come to terms with radical variety within an era (such as the present), as opposed to change across eras. Hanson himself confronts present variety only with a lament for the fact that democracy seems to have been emptied of meaning, followed by a plea for the recovery of the class connotations of the term (1989: 85-6).

One way to cope with variety is to isolate and compare the major models of democracy. This is, for example, the approach taken by David Held (1996) in what is the best textbook survey of the field. Held’s models are classical, republican, protective, developmental, direct, competitive elitist, pluralist, legal, participatory, democratic autonomy, and cosmopolitan. This set is very helpful in providing a basic vocabulary and identifying some of the main focal points of democratic theory through the ages. Yet Held only has 11 models: at the outset of this essay I listed 54 adjectives. Not all the adjectives can be squeezed into a particular model. For example, deliberative democrats could be classical, republican, developmental, participatory or cosmopolitan, in Held’s scheme. Some of the adjectives (for example, associative, difference) find no easy home in any of his models. So while Held’s survey is essential reading, it only goes so far in capturing the range of interesting contemporary thinking about democracy.

Just as there are many adjectives to qualify and describe democracy, so there are many axes of contention about what democracy can, ought, and ought not to be. Moreover, different accounts of democracy will dispute the importance of different axes. So how then to proceed, if it is futile to seek a single essence of democracy, if the enumeration of models gets overwhelmed by the complexity of democratic thinking, and if consensus cannot even be found on what constitutes the main lines of contention?

My answer is straightforward. Though democracy comes in many varieties, the dominant current in democratic theory is now a deliberative one. Indeed, it is accurate to say that around 1990 the theory of democracy took a deliberative turn. Thus different accounts of democracy can be appraised in terms of the content, strength, and significance of their relation to the deliberative turn—whether in support, opposition, capture, or qualification.

A second starting point will be the very different view of democracy shared by most of those who study the real world of democracy. Indeed, the depiction of variety when it comes to democracy in theory and practice with which I began will surprise political scientists who study the comparative politics of democracy, as well as more journalistic observers of the recent life and times of democracy. To this latter group, what is striking about contemporary democracy is less its variety than its uniformity. The passing of the Soviet alternative and the contemporaneous withering away of democratic socialism in the West signal to this group the global triumph of an unambiguously liberal, electoralist, elitist, capitalist, and minimalist model of democracy.

This liberal capitalist model captures a great deal of what is important about democracy in the real world. But one of the key tasks of contemporary democratic theory is to engage this model, and not to let it have the easy ride that some of its adherents seek. So again, different accounts of democracy can be appraised in terms of how they connect to this liberal capitalist model. And just as was the case with the deliberative turn, those relationships can involve support, opposition, capture, or qualification. It ought to be especially important to sort out the relationship between deliberative democracy and the minimalist liberal model as deployed by comparativists. However, this turns out to be surprisingly difficult, given the lack of direct engagement between the two approaches to date.

The Deliberative Turn

The deliberative turn in democratic theory occurred in the early 1990s. However, it does have antecedents, reaching back to Aristotle and the Athenian polis, and encompassing conservatives such as Edmund Burke (for whom deliberation connoted mature reflection as opposed to hasty action), as well as liberals such as John Stuart Mill and John Dewey (for a good history, see the introduction to Bohman and Rehg, 1997). There are also continuities in emphasis with participatory democrats such as Carole Pateman (1970) who were dissatisfied with the lack of opportunity for deep democratic experience in contemporary liberal democracies. Benjamin Barber’s (1984) ‘strong democracy’ can be seen in retrospect as a bridge between participatory and deliberative democracy, given his emphasis on ‘strong democratic talk.’ Yet prior to 1990, theorists generally interpreted democracy in terms of the aggregation of the interests or preferences of members of the demos through mechanisms such as voting and representation that produce collective decisions.

With the deliberative turn, the core of democratic legitimacy became instead the right or ability of those subject to a public decision to participate in genuine deliberation (see Manin, 1987; Cohen, 1989; the term ‘deliberative democracy’ was first used by Bessette, 1980). The decision itself has to be justified to these people in terms that, on reflection, they can accept. More formally, deliberation has argumentative, informational, reflective, and social aspects (the last of these referring to the solidarity that can emerge among participants in deliberation). The reflective aspect means that preferences, judgements and views that are taken as fixed in aggregative models are treated as amenable to change in deliberation. Authenticity is therefore a central concern: democratic control should ideally be substantive not symbolic, involving uncoerced communication among competent participants. The importance of the deliberative turn was confirmed in the 1990s by the announcements of the most important liberal theorist John Rawls, and critical theorist Jürgen Habermas, that they were deliberative democrats (Rawls, 1993; 1997: 771-2; Habermas, 1996).

Given the sheer number of democratic theorists who now sail under the deliberative flag, as well as the historically different schools of thought from which they come (conservatism, liberalism, and critical theory), there really ought to be substantial variety among deliberative democrats. But what is now striking is less the variety than the uniformity. For better or for worse, deliberative democracy has mostly been assimilated to liberal constitutionalism, which can be defined in terms of the reconciliation of interests established prior to political interaction under a neutral set of rules and rights. This rapid assimilation, celebrated by James Bohman (1998) as ‘The coming of age of deliberative democracy,’ ought to be surprising. For in its beginnings the deliberative turn involved a critical challenge to established liberal ways of thinking about democracy—especially to the idea that interests and preferences are only aggregated and reconciled, not transformed, in political interaction.

The assimilation happened in three ways (see Dryzek, 2000: 10-17). First, a commitment to deliberative principles can be used to justify some (but not all) of the rights long cherished by liberals. This argument applies most straightforwardly to freedom of expression and association, but also, with a bit more stretching, to freedom of religion and political equality (Cohen, 1996). Second, liberal constitutions can be interpreted as devices that promote deliberation. This is, for example, how Bessette (1980; 1994) interprets the United States Constitution, especially what it says about Congress. Other theorists emphasize deliberation in courts rather than legislatures (for example, Rawls, 1993: 231). Third, constitution-making can be seen as the quintessentially deliberative process, guided by impartial public reason, even if normal politics under the constitution is aggregative, strategic, and driven by partial interests. Deliberative theorists who emphasize constitution-making include Bruce Ackerman (1991), David Estlund (1993), and Rawls (1993) (who also wants deliberation guided by public reason to cover matters of ‘basic justice’). Amy Gutmann and Dennis Thompson (1996) are particularly comprehensive in deploying all three of these linkages.

The main shortcoming of liberal constitutionalist deliberative democracy is that it says little about extra-constitutional agents of distortion, such as the dominant position of business in liberal states, oppressive discourses (in Foucault’s sense), and imperatives dictated to states by either security concerns or the transnational political economy and its institutions. In this light, it is perhaps surprising that Habermas, who once highlighted such forces, has now turned his back upon them and taken his critical theory of deliberation quite close to liberalism. In Between Facts and Norms (1996) he advances a ‘two-track’ model of deliberative democracy. One track is rooted in the public sphere, the other in the legislature. Influence formulated through deliberation in the informal public sphere is converted to communicative power—especially via elections and thence to administrative power through legislation, whose precepts are followed to the letter by government bureaucracies. Constitutions are necessary to detail these elements and, in particular, to specify the rights necessary to enable the public sphere to flourish. The public sphere itself is seen in less insurgent terms than was the case for an earlier Habermas (for example, 1989), and there is no recognition of any need to democratize the economy, the administrative state, or the legal system, all of which receive easy legitimacy.

However invigorating this assimilation of deliberative democracy might be for liberalism, it may be bad news for democracy. Some deliberative liberals are not especially democratic. Notably, Rawls in the end wants to entrust deliberation to experts in public reason such as Supreme Court justices, who only need to deliberate in the personal as opposed to the interactive sense of the word (see Goodin, 2000, for an explicit defence of personal as opposed to interactive deliberation). This is justified because Rawls believes public reason is unitary, accessible in the same terms by any reflective individual, and thus best accessible by the most qualified individual(s). Even setting aside such non-democratic deliberation, thorough assimilation to liberal constitutionalism blunts any critical edge deliberative democracy might have had, and diverts attention away from extra-constitutional agents of both distortion and democratic influence, as well as novel sites of democratic innovation. Liberal constitutionalism emphasizes what is undeniably an important location—the state—but one that is arguably increasingly constrained by economic forces that determine the content of public policy.

Critics of Deliberation

Deliberative democracy has three prominent sets of critics, who otherwise have absolutely nothing in common: social choice theorists, difference democrats, and sceptical egalitarians.

Social choice theory does of course long predate the deliberative turn, but for reasons I will explain shortly, its contemporary significance may now lie mostly in its relationship to deliberative democracy. The social choice account of democracy takes its bearings from Kenneth Arrow’s (1951) demonstration of the impossibility of any collective choice mechanism, such as a voting system, simultaneously achieving an innocuous set of conditions (unanimity, non-dictatorship, transitivity, unrestricted domain of preferences, and independence of irrelevant alternatives). William Riker (1982) radicalized the social choice critique of democracy by observing that different voting systems and rules avoid the Arrow problem only by introducing an element of arbitrariness into collective choice. Given that different mechanisms will produce different results from identical distributions of preferences, there is no such thing as a popular will independent of the mechanism that ascertains it. This is especially true if, as Riker believes, there is no particular reason to prefer any mechanism (e.g. majority rule, or approval voting, or consensus) over any other. Democracy is then emptied of meaning.

Social choice theory has developed for half a century alongside a rational choice account of politics, though the two enterprises are actually logically distinct. At a time of democratic advance in the real world, their main thrust, with a few exceptions, has been in exactly the opposite direction. The public choice field that they constitute is home to many demonstrations of the arbitrariness, instability, perversity, and inefficiency of democratic politics. Beyond Riker’s exposé of the vacuum at the heart of democracy, public choice theorists have argued that:

  • In political systems of any size, voting is irrational.
  • Majority rule entails the Pareto-suboptimal exploitation of minorities.
  • Self-interested elected representatives at best create programmes that benefit their own constituents at the expense of the public interest, at worst deliberately design programmes badly such that their own intercession is required to deliver benefits.
  • Public spending levels are mostly a consequence of self-interested bureaucrats maximizing budgets. Bureaucrats can conspire with special interest groups and their supportive politicians to divert public resources for their own benefit.
  • More generally, ‘distributional coalitions’ such as labour unions and employers secure laws and policies to protect their own privileges at the expense of economic efficiency.
  • Democratic politics is intrinsically irresponsible because all actors seek benefits for themselves while imposing costs upon others; the result is a negative-sum game where total costs outweigh total benefits.

Now, not all public choice analysts argue all of these points; the most unremittingly anti-state aspects are found only in Virginia-style public choice (see Mitchell and Simmons, 1994, for a statement). The more highly mathematical rational choice treatments of politics found in (say) the pages of the American Political Science Review have a less overtly political agenda. Yet it remains true that when such analyses do have implications for democratic politics, the news is usually bad. Thus can Russell Hardin conclude that public choice analyses have ‘largely helped to expose flaws—grievous, foundational flaws—in democratic thought and practice’ (1993: 170).

However, if, to use the title of Hardin’s survey, the conflict is ‘Public choice versus democracy,’ then by the early 2000s democracy was winning. The claims of rational choice as explanatory theory have been severely dented within political science (Green and Shapiro, 1994). Social choice theory in its Rochester-style anti-democratic manifestation has been destroyed by Mackie (2003). Gerry Mackie shows that every real-world example of a voting cycle (A beats B beats C beats A) adduced by William Riker or his followers to illustrate the potential for arbitrariness, instability, and manipulation in collective choice is actually inconsistent with the historical evidence.

What, then, remains of public choice as democratic theory? The answer is that it provides a set of warnings about what democratic politics could be like if political actors behaved in Homo economicus fashion, and if no mechanisms existed to curb these behavioural proclivities and their consequences. Deliberative democracy provides both a communicative paradigm of personhood and mechanisms to bring Homo economicus and his interactions under control (a non-deliberative alternative can be found in Shepsle’s 1979 idea of structure-induced equilibrium).

Now, social choice theorists can still try to pour cold water over deliberation because it is easy to demonstrate that the very conditions of free access, equality, and unrestricted communication conducive to authentic deliberation are exactly the conditions conducive to instability, arbitrariness, and so strategic manipulation (van Mill, 1996; see also Grofman, 1993: 1578; Knight and Johnson, 1994). Deliberative democrats can reply that there are mechanisms intrinsic to deliberation that act to structure preferences in ways that solve social choice problems (Dryzek and List, 2003). For example, deliberation can disaggregate a dimension on which preferences are non-single-peaked (one major cause of cycles across three or more alternatives that are at the root of the kind of instability Riker identifies) into several dimensions on each of which single-peakedness prevails (Miller, 1992). To the extent this deliberative reply succeeds, then the social choice critique undermines only an aggregative account of democracy in which all actors behave strategically, and can actually be deployed to show why deliberation is necessary.

While social choice critics of democracy fear the unmanageable diversity that deliberation can encourage, difference democrats criticize deliberation for exactly the opposite reason: that it represses diversity. To greater or lesser degrees difference democrats take their bearings from the postmodern theory of identity and difference, in which the essence of democracy is seen in terms of the creative encounter of those with disparate identities (for example, Connolly, 1991). Just as for the deliberative democrats, the core of democracy is therefore seen as communication. However, difference democrats problematize communication, and criticize the allegedly neutral forms of communication emphasized by deliberative democrats for their cultural biases. Notably, Iris Young (2000) argues (in a US context) that any main or exclusive emphasis on rational argument further disadvantages minorities who are not well versed in its niceties. Her ‘communicative democracy’ would feature greeting, rhetoric, and storytelling (or testimony, or narrative) as well as argument—forms of communication she believes are more accessible to disadvantaged minorities. This resonates with Young’s earlier (1990) advocacy of guaranteed representation and veto power over policies that affect them for disadvantaged groups.

Deliberative democrats who are not under the sway of an exclusive Rawlsian belief in unitary public reason or an overly narrow Habermasian account of communicative rationality could reply that there is nothing in deliberation that excludes these alternative forms of communication (though they ought to balk at any suggestion of veto power). However, Young’s trio should not be accepted uncritically, as she eventually recognizes (2000: 77-80). Instead, they need to be held up to the tests of non-coercion, capacity to induce reflection, and ability to link the particular with the general (Dryzek, 2000: 68-71).

A third group of critics of deliberation, those I style sceptical egalitarians, defend more traditional accounts of democracy against the deliberative turn. In Shapiro’s pithy (1999) terms, ‘enough about deliberation, politics is about interest and power.’ In this light, those interested in improving the quality of democracy should seek the equalization of power; here, issues of democracy become linked to distributive justice. Such sceptics can point to the rather embarrassing fact that deliberation cannot be a complete theory of democracy because its advocates do not specify how collective decisions get made (Saward, 2000). If so, then deliberative democrats might have to retreat to more familiar aggregative mechanisms, and the deliberative/ aggregative dichotomy is proven false, for then democracy is necessarily aggregative, and votes have to be taken (Przeworski, 1998: 140-2). Goodin (2000) points out that deliberation is an activity that can never realistically involve more than a handful of people. Saward (2000) believes that such considerations mean that egalitarians should therefore oppose deliberation’s aristocratic leanings that would exclude those with non-deliberative preferences; far better, in this light, to extend democracy in more direct fashion (for example, by greater use of referenda).

Deliberative democrats can reply to the sceptics who charge that deliberation can only be an elite activity in several ways here. In Fishkin’s (1995) deliberative opinion polls, participants for a deliberative forum are selected at random from the population, and complete a questionnaire at the end of the process. Citizens’s juries too are recruited by random selection, but conclude with a policy recommendation crafted and agreed upon by the jurors rather than a questionnaire (Smith and Wales, 2000). Fishkin argues that a deliberative poll represents what public opinion would be if everyone could deliberate; the same might be said for citizens’s juries.

Alternatively, deliberative democrats could allow that deliberation can coexist with a variety of mechanisms for reaching binding decisions, be they voting in referenda, elections, or the legislature, the decisions of courts, consensus among stakeholders in an issue, or even administrative fiat. More radically, they might think about ways in which the deliberative contestation of discourses in the public sphere can generate collective outcomes not only in its indirect influence on public policy, but also via cultural change and paragovernmental action (Dryzek, 2000).

Before leaving the deliberative turn and its critics, one further argument that might weigh against deliberation should be noted. If democracy involves aggregation (however much it is downplayed by deliberative democrats), that can be across judgements and not just across preferences as emphasized in social choice theory. Such judgements can involve disagreement over (say) what is in the common good. This epistemic way of thinking about democracy is associated with Rousseau, according to whom the general will can be ascertained by voting. Bernard Grofman and Scott Feld (1988) argue that if indeed there is such a thing as the common good, though people differ in their judgements about which option will best serve it, then Condorcet’s jury theorem applies. This theorem demonstrates that if each citizen has a better than even chance of being correct in his/her judgement, then the larger the number of voters, the greater the chance of the majority choosing the correct option. The jury theorem therefore justifies the rationality of majoritarian democracy, at least in a republican context of a search for the common good, though only if each citizen reaches and exercises independent judgement. So there should be no factions (which reduce the effective number of voters) and, it might seem, no communication. These, at least, were Rousseau’s own views: deliberation should only be a matter of internal reflection, not communication. However, as Robert Goodin (2002: 125) and others point out, discussion is fine so long as people then subsequently exercise their own independent judgements when voting. Goodin then questions an epistemic democracy rooted in Rousseau and the jury theorem by pointing out that in a dynamic context, their implication is that minorities should rationally and immediately cease their opposition when a majority votes against them. Persistent opposition therefore makes sense only when values differ, but not when only factual judgements vary (2002: 144). If only factual judgements are at issue, an epistemic approach threatens to wipe out the contestatory aspect of democracy.

For better or for worse, the deliberative approach sets the agenda for contemporary democratic theory. However, it should be clear that there remains plenty to argue about, both among those who share the deliberative orientation, and those who reject it. Yet there are scholars of democracy who remain untouched by the deliberative approach, and to these I now turn.

Liberal Minimalism and its Alternatives

The model of democracy most popular among comparative politics scholars, especially those in the burgeoning field of democratic transition and consolidation, expects far less from democracy than do the deliberative democrats. This model is essentially that proposed long ago by Schumpeter (1942): democracy is no more than competition among elites for popular approval that confers the right to rule. In the 1950s this idea became the foundation for ‘empirical’ theories of democracy happy with the generally apathetic role of the ignorant and potentially authoritarian masses (Berelson, 1952; Sartori, 1962). Such competitive elitist models have long been discredited among democratic theorists—not least those such as Dahl (1989) who had earlier believed in them as both accurate descriptions of United States politics and desirable states of affairs. Yet they live on among transitologists and consolidologists, who see the hallmark of a consolidated democracy as a set of well-behaved parties representing material interests engaged in electoral competition regulated by constitutional rules (see, for example, Di Palma, 1990; Huntington, 1991; Mueller, 1996; Schedler, 1998). The deliberative democrat’s concern with authenticity is nowhere to be seen. Active citizens play no role in such models. There is no outlet for citizen engagement with politics beyond regular elections where the mostly uninformed, uninterested and apathetic masses can register preferences across a limited range of candidates or parties.

What can explain the popularity of this minimalist, electoralist model? Partly it is a matter of the undeniable analytic purchase the model provides for those who study the real world of democracy. To such scholars, the contested character of democracy in political theory is a nuisance when it comes to devising empirical indicators for the comparison of different countries and the tracking of democratic transition and consolidation in particular countries. Acceptance of the minimalist model makes life much easier. It can be applied, for example, in Huntington’s (1991: 267) famous two-election test for consolidated democracy, which requires a freely elected government to cede power in a subsequent electoral defeat. Or it can underwrite a temporal scale for assessing the degree to which democracy is consolidated; Lijphart (1984: 38) suggests 30 to 35 years.

Perhaps a more important reason for the popularity of liberal minimalism is its consistency with developments that see capitalist marketization and democratization marching together. Since the mid 1970s, the adoption of liberal democratic systems by ever more countries has gone hand-in-hand with the global expansion of capitalism. It has of course long been noted that there is a correlation between capitalism and liberal democracy. Exactly why such a correlation exists is a matter of dispute. To Lipset (1959) it was a matter of capitalism producing a middle class that had all the right democratic virtues of toleration and moderation. To Rueschemeyer, Stephens and Stephens (1992) the answer lies instead in the fact that capitalism produces a working class with a vested interest in redistribution that is promoted by effective universal franchise. To Adam Przeworski et al. (2000) it is not that capitalism causes democracy, for countries can become democratic at any level of development. However, capitalism produces wealth which in turn provides protection against the overthrow of democracy.

While capitalism facilitates the development or stability of a minimalist liberal democracy, it impedes any strengthening of democracy beyond this minimum. As Lindblom (1982) among others notes, the capitalist market context automatically punishes governments that pursue policies that undermine the confidence of actual or potential investors by causing disinvestment and capital flight. Thus when it comes to public policy, democracy can only operate in what Lindblom calls an ‘unimprisoned’ zone. The corollary is that too much state democracy means dangerous indeterminacy in public policy (Dryzek, 1996). Democracy may no longer mean, as Plato defined it in the Republic, ‘a state in which the poor, gaining the upper hand, kill some and banish others, and then divide the offices among the remaining citizens, usually by lot.’ But there is a lingering possibility that too much democracy might undermine the inequalities on which effective wealth creation rests. The minimalist model therefore seems uniquely suited to the contemporary liberal capitalist political economy.

This combination of capitalism and liberal minimalist democracy received perhaps its most positive gloss (and a dash of Hegel) in the triumphalism of Francis Fukuyama’s (1989; 1992) ‘end of history.’ Fukuyama’s thesis lost plausibility in the ensuing decade, but only in terms of the persistence (or renewal) of challenges such as religious fundamentalisms, ethnic nationalism, and Confucian capitalism. But the basic idea that democracy is globally dominant and that the liberal capitalist model of democracy has few if any plausible challengers that merit the title ‘democracy’ is still the dominant view among transitologists. Life with this model, and without the kinds of critical questions that democratic theorists are apt to raise, is certainly less complicated for the transitologist. Obviously happy about this state of affairs, Sartori wants to be done with the critics: ‘the winner is an entirely liberal democracy, not only popularly elected government, but also, and indivisibly, constitutional government; that is, the hitherto much belittled “formal democracy” that controls the exercise of power’ (1991: 437).

The more critical stances that democratic theorists are inclined to take would highlight the limitations on democracy that this global dominance of minimalist liberal democracy plus capitalism entails. But any such critical response is easily countered if it remains devoid of ideas about how such dominance might realistically be challenged (without retreating to ungrounded idealism). Part of the response might involve the strengthening and democratization of international institutions in response to the migration of political power from the state to the transnational political economy. This is, for example, the approach taken by Held and his fellow advocates of a cosmopolitan democracy that would involve a more inclusive United Nations Security Council, a strengthened UN General Assembly, cross-national referenda, and international economic, military and judicial authorities accountable to regional and global parliamentary bodies (Held, 1995; Archibugi, Held and Köhler, 1998). Alternatively, if state democracy can only be minimalist, theorists might explore non-state locations for the pursuit of democracy. Such locations might involve public spheres in both domestic and transnational civil society that remain distant from state power though still oriented to public affairs (Cohen and Arato, 1992; Fraser, 1992; Dryzek, 1996: 46-53), and home to social movements. Community-based grassroots democracy, collectives, and workplace democracy would also fit here, but today seem to have fallen on hard times. Feminist proposals for democratization of areas of life traditionally considered private, such as the household, remain perhaps more promising (Rowbotham, 1986), but also more tangential, at least in the sense that they do not confront the state and its enmeshment in the transnational capitalist political economy head-on.

A rare normative defence of the minimalist model is provided by Przeworski (1999) who argues that the model at least puts an end to large-scale political violence once those defeated accept that they have a realistic chance to return and win another day (proponents of consensual democracy such as Lijphart, 1999, could respond that power-sharing not majority rule is the best defence against violence in a divided society). Given that the structure of interests in a complex society means that competing interests can never be reconciled, their provisional resolution in electoral competition is about the best we can ever do. Riker’s (1982) attempted social-choice-theoretic defence of minimalism is that though voting is meaningless, periodic elections at least provide an opportunity for the removal of tyrannical, incompetent, or corrupt leadership. But Riker’s defence fails because his own analysis shows that there is no will of the voters independent of the mechanism that is supposed to measure it—and this has to include the will to dismiss tyrants or incompetents (Coleman and Ferejohn, 1986: 22).

Democratic Theory and Practice

Acceptance of the minimalist model would render most democratic theory unnecessary. But minimalism fails in its own terms, for the following reasons. Democratic theorists are well placed to highlight these failures and move the conversation beyond them.

First, minimalism can allow forms of democracy that are very thin indeed, to the extent they barely merit the description ‘democratic.’ For example, what Guillermo O’sDonnell (1994) calls delegative democracy passes the minimalist test. Under delegative democracy, found especially in Latin America but also in the post-communist world, leaders submit themselves to regular elections, but otherwise govern without accountability, without any sense that election promises need to be remembered and without constitutional constraint (except of course the one specifying regular free elections). Delegative democracy completely misses what Philip Pettit (1999) calls the contestatory as opposed to electoral aspect of democracy. In light of this aspect, the guarantee of freedom (defined as non-domination) is the ability of citizens to contest the content of collective decisions under fair terms, be it via access to courts, legislatures, or administrative review.

Second, minimalism is insensitive to the variety of forms that democracy can take in practice as well as theory, leading to misinterpretation of events and developments, and so undermining the analytical purchase that is one of minimalism’s main justifications. For example, under sway of a liberal model of democracy, Juan Linz and Alfred Stepan (1996) fear for democracy in post-communist Poland and the Czech Republic because of the legacy of the kind of politics that characterized their oppositional civil societies in the Soviet era: ‘Ethical civil society represents “truth” but political society in a consolidated democracy normally represents “interests”’ (1996: 272). However, the kind of politics they criticize is consistent not only with the deeper republican history of these two countries (reaching back to the eighteenth century in Poland), but also with contemporary civic republican political theory (Sandel, 1996). In this light, the practices and discourses bemoaned by Linz and Stepan actually provide resources for those interested in consolidating and deepening democracy (Dryzek and Holmes, 2002: chs 14 and 15).

Such practices and attitudes might also include the civic attributes fostered in associational life that are, according to Robert Putnam (1993; 2000), the key to ‘making democracy work.’ For Putnam, a widely shared civic orientation that is not reducible to private material interest is necessary to defend state democracy against amoral clientelism (as in southern Italy) or rampant individualism (as in the United States in recent decades).

A variety of democratic systems observable in contemporary nation-states passes the minimalist test: libertarian and social democratic, elitist and pluralist, presidential and parliamentary, nationalist and cosmopolitan, dense and weak civil societies. Uncritical application of liberal minimalism fails to pick up on the importance of these variations. Of course, the minimalist might reply that the variations are unimportant; but he or she should be required to demonstrate this fact rather than merely assert it, and the theorist can at least identify the dimensions along which a response is required.

A third reason why minimalism is inadequate is that it is untrue to democratization as understood by many political actors in transitional systems. The more idealistic, such as Vaclav Havel, President of Czechoslovakia and then the Czech Republic, see continued experimentation with and dialogue about forms of democracy at the centre of the democratic project (Lienesch, 1992: 1012; on the idea of democracy as an open-ended project, see also Downs, 1987: 146). While Havel’s idealism may put him in a minority, minimalism fails to do justice to the variety of conceptions that political elites and ordinary people in these societies bring to bear when it comes to their expectations of and hopes for democracy (for evidence for 13 post-communist countries, see Dryzek and Holmes, 2002).

These sorts of considerations might suggest that those who study the real world of democracy (and especially democratic transitions) ought to listen more to democratic theorists. But the converse is also true: democratic theorists should attend more to real-world constraints and possibilities that empirical social science can help to illuminate. In common with many areas of political theory, democratic theory can sometimes lapse into a self-referential dialogue in which connections to real-world events, constraints, and possibilities are lost (Gunnell, 1986). As Jeffrey Isaac (1995) points out, the fall of the Berlin Wall went largely unnoticed by political theory. Whether or not this is a satisfactory state of affairs depends in the end on one’s conception of the value and role of political theory. Yet normative democratic theory at least generally looks as though it is developing prescriptions that it has some interest in being followed, or at least attempted, in the real world. It is then problematic when theorists propose schemes that stand little chance of being implemented in the world as it is. For example, Young (1992), picking up on the model of associative democracy proposed by Joshua Cohen and Joel Rogers (1992), wants the first task of the state to be the organization of oppressed minorities into forces capable of exercising real power. This is not a kind of state whose existence it is at all plausible to postulate, especially given the many constraints and imperatives to which real states are subject. The theorist’s last line of defence here might be that such abstractions are necessary in order to maintain a critical distance, to present counter-factual ideals that expose the shortcomings of real-world situations in particularly stark form. But does everyone need to do that? Might not contextually sensitive critique be more productive? And is not a closed, self-referential discourse of democratic theory reflexively undemocratic in cutting itself off from those who struggle to promote, defend, develop, and deepen democracy? (A thoroughly reflexive approach to democratic theory would begin with popular conceptions of democracy; see Dryzek and Berejikian, 1993.)

An additional reason why democratic theorists should attend more to democratic practice is that sometimes problems that concern democratic theorists may actually find solutions in political practice. For example, David Schlosberg (1999) argues that the problem of engagement across deep difference identified by postmodern theorists as the key democratic challenge has been successfully negotiated in the political practice of the environmental justice movement in the United States.

I have argued that the two main poles in contemporary thinking about democracy are the deliberative approach and liberal minimalism. Some schools of thought have engaged both poles. So social choice theory can provide both a defence of minimalism and a critique of deliberation, though it can also be deployed to reach almost the opposite position. Civic republicanism has many synergies with deliberation, and can also be used to criticize liberal minimalism’s application to real-world political systems. However, there has yet to be any direct engagement between deliberative democrats and the liberal minimalism of the comparative scholars of democratic transition and consolidation. The two approaches have only connected via intermediaries. This state of affairs might be indicative of the gap between democratic theory and democratic practice, though as I have argued, the minimalists have not got the practice right. And the fact that there are intermediaries shows that the gap can be bridged.

Conclusion

One measure of democratic theory’s success will be the extent to which it can loosen the grip of liberal minimalism on those who study the comparative politics of democracy and democratization. A second more demanding measure would be found in the degree to which it can contribute to the global conversation about democratic development, in established liberal democracies and the transnational arena, no less than transitional societies and new democracies. These are of course both external tests: the internal conversation currently flourishes without them being passed, though there are sporadic exceptions (for example, in connection with deliberative opinion polls and citizens’ juries). The fact that the internal conversation is full of vigour is cause for self-congratulation. But democratic political theory, precisely because it is democratic political theory, cannot get off so lightly in this respect as most other areas of political theory.