Political Theory versus Political Science

The divide between democratic theory and the study of political parties can be seen as one part of the broader disconnection between the sub-discipline of political theory (or political philosophy—the terms are taken as interchangeable for present purposes) and political science more generally. Recently, Shapiro has argued that an ongoing process of professionalization and specialization has dissociated political philosophy from the rest of political science, and has separated normative from empirical theory, “with political philosophers declaring a monopoly over the former while abandoning the enterprise of ‘positive’ political theory to other political scientists.” As a consequence, “normative theory … is no longer informed, in the ways that the great theorists of the tradition took it for granted that political theory should be informed, by the state of empirical knowledge of politics.” This separation has not only resulted in an increasing tendency of normative political theory to downgrade concerns over practical feasibility but “has also fed the tendency for empirical political theory to become banal and method driven—detached from the great questions of the day and focused instead on what seems methodologically most tractable.”⁶

Arguably, modern political philosophers have always been more aware of the nature of their activity and the way in which their sub-discipline connects to the rest of political science than their colleagues in other branches of the discipline.⁷

Among political theorists in particular, therefore, the problematic dissociation of political philosophy from the wider activity of political science has not gone unnoticed. Gunnell, in Between Philosophy and Politics, for example, expresses concern with the “alienation” of political theory from substantive political problems.⁸ In Reappraising Political Theory, Ball suggests that political philosophy develop some of the “sensibilities of those among our fellow political enquirers who are conventionally classified as ‘empirical’ political scientists.”⁹ In Deliberative Democracy and Beyond, Dryzek observes a “near-universal tendency among political theorists to treat empirical reality in terms of a few stylized facts, rather than attending seriously to the findings of empirical political science.”¹⁰

Kelly suggests that, if political theory is to continue to command interest and respect within the discipline, “it does more to re-engage with political science,” which it can do “by challenging the terms of much that goes on in political science.” In so doing, political theory could provide a support structure for what political scientists do by conceptual clarification and theory construction, and would act to clarify and challenge the normative dimensions of the theories and questions that political scientists work with.¹¹

At the same time, political scientists should take more notice of the concepts and theories developed by political philosophers. Indeed, Shapiro's cited observations are more than a critique of political theory's navel-gazing. While he condemns the theorists for seeing themselves engaged “in a specialized activity distinct from the rest of political science,” his is as much an assault on triteness and triviality in empirical political science as it is on the narcissism of political theory.¹²

These overviews may seem exaggerated, but we would agree that they are largely accurate. There are many pressures which drive sub-disciplines away from mutual engagement, from the internal organization of academic departments to increasing journal specialization to certain research accounting and funding procedures. Whatever the causes, the diagnosis is clear enough; to move towards tentative prescriptions, we need to address a range of problems that arise from this large-scale disjunction. The specific case of the linking of democratic theory and research into political parties is just one such problem, though (as we hope to show) one of real significance.

Among the various modes of non-communication between political philosophers and political scientists, our concern here centers on the meaning and interpretation of “democracy.” We do not argue that theory and empirical work on a range of democratic institutions have not been mutually engaged—they have, in various ways. But this work, it is fair to say, focuses much more on the “hardware” of democracy rather than the “software” of mediating institutions like political parties. For example, Lijphart's consensus and majoritarian models, Dahl's account of the constitutive features of polyarchy, Guinier's discussion of electoral rules, and Mansbridge's rethinking of representation engage with characteristics and trajectories of a range of background institutions—electoral systems, electoral and other rights, executive accountability, the accountability of individual members of Congress, and so on.¹³

Such authors clearly engage with democratic institutions to help to generate theoretical insights. Notwithstanding such efforts, engagement between work on democratic theory and political parties specifically has remained exceptional.

From that perspective, and with specific reference to work in democratic theory and on political parties, a key challenge for political theorists lies in fostering more critical conceptual engagement among scholars of political parties and the elucidation of the implicit normative assumptions underlying many of their approaches. Democratic theorists, at their best, problematise democracy, drawing out shifting and competing meanings. Some of this work may involve uncomfortable implications for the place of parties in democratic politics, today and in the future. However, when addressing empirical reality and advancing practical solutions for the improvement of the quality of democracy, democratic theorists should be less reluctant to acknowledge the reality (and indeed the potential) of the political party and to incorporate the empirical findings of party scholars.

Party scholars, on the other hand, when addressing normative and empirical issues of democracy, could benefit from developing a greater sensitivity to the broader significance of the place and role of political parties in modern democracy, and a greater awareness of the conceptual and theoretical dimensions underlying democracy. We will pick up a range of such issues in what follows.

Back to the Future?

Both democratic theorists and party scholars could find inspiration in the great thinkers of the past, as the two particular ‘schools’ have not always been disconnected. Indeed, until the mid-twentieth century it was often precisely through the concept of democracy that normative political theory and empirical political science were closely linked. In the eighteenth century, Hume was the first among the major political philosophers to take up the subject of parties.¹⁴

In his essay Of Parties in General (1742) he observed that parties founded on some real difference of interest (as opposed to affection) and political (as opposed to religious) principles are typical of the modern world: “Parties from principle, especially abstract speculative principle, are known only to modern times and are, perhaps, the most extraordinary and unaccountable phenomenon that has yet appeared in human affairs.”¹⁵ At best, however, Hume accepted parties only as an unpleasant consequence of free government. Edmund Burke was the first to see that see that parties had a positive and necessary use for democracy. In Thoughts on the Cause of the Present Discontent (1770) he made what has subsequently been described as “the first argument in the history of political philosophy for the respectability, not merely the necessity, of parties.”¹⁶

Over the course of the eighteenth and nineteenth centuries, political philosophy gradually came to terms with the continued institutionalization of political parties and their de facto inevitability as permanent elements of the political system (though strong anti-party sentiments continued to permeate many theories of government and democracy). The reality of the political party received a final and decisive push with the advent of mass democracy and the introduction of universal suffrage. As the size of the democratic polity increased with the consolidation of large-scale nation-states, direct links between the state and the individual became increasingly unrealistic, a fact which further legitimized the status of parties as intermediary institutions between individual citizens and the state. More or less contemporaneous with this development was the shift in the dominant meaning of democracy; in the late 1700s Madison dismissed democracy, understood as Athenian-style direct democracy, in favour of republican government. In just over 100 years, Madison's republicanism was well on the path to becoming itself the dominant meaning of democracy.

Until at least the early twentieth century, as the works of major scholars of political parties such as Ostrogorski, Michels, and Weber demonstrate, the empirical study of political parties remained intimately related to democratic theory.¹⁷

To be sure, Ostrogorski and Michels in particular were primarily concerned with democracy within parties, and the undemocratic and oligarchic tendencies of bureaucratic mass parties in particular, and not with democracy as a political system based upon competition between parties. However, their theoretical and empirical analyses of the functioning of political parties are clearly related to the theory of democracy. Ostrogorski's solution to what he saw as the suffocating internal discipline of mass parties, i.e., to create a polity without permanent parties, underlines how he sought to reconcile liberalism with the idea of popular sovereignty, thus addressing a fundamental concern that pervaded much of the nineteenth century. Michels implied that parties should be seen as obstacles to, or the very antithesis of, democracy, at least to the extent that the “iron law of oligarchy” should be seen also to apply to the organization of the state itself. Weberian analysis, on the other hand, suggests that strong and responsible political leadership may be necessary for the healthy functioning of democracy, principally because it serves to prevent the politicization of an unaccountable bureaucracy. On this view, a lack of democracy within parties is not necessarily consequential for democracy at the system level because, as Sartori would later put it, “democracy on a large scale is not the sum of many little democracies.”¹⁸

The Mysterious Absence: Contemporary Democratic Theory and Parties

So, the deep linkage of party (and partisan) interests with philosophical and theoretical reflection on democracy was strong and varied in a range of classic sources. It was in the twentieth century—and arguably in the 1960s and 1970s in particular, as the behavioural revolution in political science advanced while normative political theorizing suffered something of a crisis of identity and relevance—that the study of democratic theory and of political parties separated and diverged. Empirical and institutional work on democracy was still sometimes combined with theory construction and conjecture, as we have seen. But parties specifically tended to be bypassed by democratic theorists, and those scholars specializing in political parties tended to do so quite separately from the concerns of democratic theorists. In this section we briefly offer support for our view that this separation is real. In the next section, we explore the more significant question about why this separation has occurred and been sustained.

Arguably it was Schumpeter in the 1940s who most influentially reconciled parties with democracy, by defending a minimal conception of democracy in which the circulation of party elites in power through elections became both its essence and justification. In recent years, however, the theory of democracy has by and large become detached from analysis of the character and roles of political parties, despite a few notable exceptions such as Robert Dahl or Giovanni Sartori, both of whom have published normative studies of democracy as well as empirical studies of democratic institutions and political parties. As Sartori observes, however, as scholars of political parties we are “travelling more and more through the ever-growing jungle of party polities without really knowing where we started, let alone where we are heading.”¹⁹

Even here, these authors have often separated out consideration of parties and issues in democratic theory; Democracy and its Critics (Dahl) and The Theory of Democracy Revisited (Sartori), arguably their major works of democratic theory respectively, have little to say about political parties. And key parts of Dahl's work have displayed unresolved tensions between his expression of the democratic ideal and parties and other institutions which he regards as being at the core of modern democracy or polyarchy.²⁰ One can criticize these tensions and still applaud the category-crossing work of Dahl; the tensions themselves point towards the agenda of work whose relative absence is the very thing we are bemoaning here.

Schattschneider's observation that “the political parties created democracy and that modern democracy is unthinkable save in terms of the parties”²¹

has become an oft-cited conventional wisdom among party scholars. Parties have come to be seen as necessary for democracy, even amidst increasing concern that their actual functioning is inadequate for a healthy performance of democracy.²²

Underlying the debates about the contemporary performance of existing parties and party systems, there are fundamental disagreements about the meaning of democracy and the role of political parties within it. What that conception of democracy is, however, remains unclear. Complaints about the decline of party, the growing disengagement from partisan politics, increasing dissatisfaction with and distrust in parties and politicians, the weakening of their representational and governmental roles, and the problems of accountability, responsiveness, and legitimacy, all rest upon, usually implicit, normative assumptions concerning what is valuable about democracy and about how democracy should work.²³

But with relatively few exceptions,²⁴ even when these assumptions are made explicit, they generally are simply stated as self-evident truths, rather than being recognized as contentious choices.²⁵

Just as scholars of political parties tend to disregard the variety in conceptions of democracy, contemporary democratic theorists generally fail to acknowledge the variety in functions and types of party. If not ignored altogether, parties are often at best understood one-dimensionally, as monolithic entities or mysterious black boxes. In particular in varieties of democratic theory belonging to the strands of participatory democracy and civic republicanism (including recently prominent strands of deliberative democracy), political parties are conspicuous by their marginality or absence. Even in theories of democracy that are less averse in principle to the notion of interest representation and the existence of intermediary structures, the status accorded to political parties is quite trivial. Theories of associative democracy, for example, which argue for voluntary citizen associations as the central institutions of governance, have little to say about the role of political parties in democratizing civil society or the state, or the extent to which self-governing voluntary associations should replace or complement the activities of political parties.²⁶

In the same vein, the more recent theories of deliberative democracy, while not necessarily unsympathetic to the notion of representation, define few, if any, of the linkages between “representatives” and “constituents” in terms of party, with parties typically regarded as belonging to the wrong side of the aggregation-deliberation dichotomy.²⁷ More generally, indeed, the implicit understanding of representation in political theory seems to be one in which representation occurs through the presence or actions of individual elected politicians conceived largely in isolation from parties or party structures.

The absence of political parties from these domains of democratic theorizing is quite stark, and is evident in a range of contemporary democratic theory topics and debates. Key parts of the literature on “participatory democracy” in the late 1960s to the early 1980s bypassed parties to look instead at functional and radically decentralized modes of participation.²⁸

Parties are rarely mentioned in a prominent exploration of democracy's “outer and inner edges.”²⁹ Fishkin largely sets parties aside from preferred “deliberative polls” in “the quest for deliberative democracy;”³⁰ Gutmann and Thompson's advocacy of a different brand of deliberative democracy is profoundly focused on principles, and parties do not get a mention.³¹

Parties play a marginal, and to a degree a suspect, role in Young's otherwise extensive examination of the politics of inclusion and democracy,³²

and the same is often true of general texts that seek to engage democratic theory on a wide canvas (e.g., works by Beetham and Saward).³³ A rich array of formal and informal institutions and practices bearing on deliberation are considered in the key essays on deliberative democracy gathered by Fishkin and Laslett,³⁴ but parties barely rate a mention as deliberative, non-deliberative, or potential deliberative bodies or participants; Christiano notes the role of parties as potential deliberative bodies, but they play a strongly subordinate role to principles and more hardware elements of his discussion of “fundamental issues in democratic theory.”³⁵ Similar comments can rightly be made of other noted democratic theory texts, such as those by Dryzek, Macedo, and Benhabib, that engage in particular with the dominant deliberative thread in democratic theory.³⁶ Implicitly, parties are not seen as real or ideal vehicles of democratic deliberation—a key theme to which we return in the final section. Muirhead has commented rightly that “while there are many and varied strains of the deliberative ideal—from abstract philosophical accounts like those of Habermas and Rawls to quite practical treatments like those of Gutmann and Thompson or Fishkin—in general, none give an extended role to party or partisanship'.³⁷ Parties likewise are also bypassed, for the most part, in other leading democratic theory accounts of the ways in which (it is argued) democracy needs to be reinvented or reformed, for example in discussions of green democracy³⁸ and cosmopolitan and transnational democracy.³⁹

Of course this is a selective sample. Nevertheless readers can judge for themselves our basic claim: contemporary democratic theorists rarely consider the necessity, roles, promise, or democratic character of political parties, and often simply neglect them. Rather, they focus on an array of principles, moral dilemmas, institutional hardware and design, individuals as actors and citizens, “the people,” and new forums and other institutions which may embody democratic ideals. There are exceptions—Christiano to some degree, Phillips and Budge for example⁴⁰

—but they are isolated ones. It is something of an irony that in one of the earliest and most influential articles in the dominant deliberative democracy strand, Joshua Cohen noted that “the question is how we can best approximate the deliberative conception. And it is difficult to see how that is possible in the absence of strong parties, supported with public resources.”⁴¹ Similarly, Bernard Manin insisted that parties make an essential contribution to democracy.⁴² It is fair to say that the suggestions of these early theorists of deliberative democracy have largely been ignored since.

Why Democratic Theorists and Party Scholars Don't Talk to Each Other

With so many topics and concerns in common—not least the fundamental health of democracy—why is there such a divide between democratic theory and the study of political parties? We have hinted at some reasons; in this section we approach the issue explicitly and more systematically. We think that there are at least four key, and interrelated, characteristics of the different literatures' approaches to democracy within which the core reasons can be located; these are referred to as “domains of difference.” These domains are: (1) dominant epistemology approach, addressing fundamental issues of how we know what we know; (2) the level of analysis, dealing with different tendencies to prioritize the focus on certain evident features of polities; (3) definitions of key terms, above all democracy, and (4) positioning of analyses on issues of substance and procedure. No doubt there are different ways of carving up fundamental differences dividing the two literatures concerned. But we hope to show that these four domains are fundamental. Further, there are clear relationships between the four domains. For example, an epistemological approach that prizes normative thinking is more likely to emphasize substance over procedure (or at least to see procedures in terms of principled points of substance). Likewise, scholars of political parties will in part emphasize procedures since their epistemological preferences for empirical approaches induce scepticism about positing essential substance. These different approaches have deep roots and extensive influence on the focus and style of the different literatures. These characteristics, as well as the broader nature of the different approaches, the reasons for them, and their consequences, are schematically represented in table 1.

The first domain of difference concerns the different dominant epistemological approaches of democratic theory and the study of political parties. Although perhaps less directly related to the treatment of democracy per se, these positions have an immediate bearing on it, as well as being related to the other characteristics. Broadly speaking, democratic theory tends to prioritize normative approaches, concentrating on the ideals and fundamental principles of democracy, while the study of political parties tends to prioritize empirical approaches, focusing primarily on the workings of democratic institutions in actual practice. The main reason for this methodological difference is that democratic theory largely emerged out of, and remains grounded in, a philosophical tradition that favours deductive reasoning from abstract axiomatic principles, while the study of political parties is rooted in a later tradition of logical positivism that favours empirical exploration and the verification and falsification of empirically observable evidence.

This means that for the democratic theorist clarifying and debating the democratic ideal remains more important than (and the necessary precondition for) exploring actual empirical examples. The primary force of the argument derives from abstract reasoning and internal coherence rather than a correspondence between theoretical claims and empirical reality (the latter is often more than just a negligible concern, but not often with specific reference to political parties). Empirical examples serve a purpose primarily as selective and isolated illustrations of the argument. To scholars of political parties, on the other hand, empirical observations are vital to theory building, whether inductively or deductively conceived, and contentions and claims only hold true to the extent that they can be supported by empirical evidence. Normative ideas are often taken to be largely irrelevant—it remains implicit that we already know what democracy is, and why it is valuable. From that perspective, arguments about ideal models of democratic deliberation arising from speculative contract theories (as Cohen's does, through Rawls, for example),⁴³

may have points of interest but are unscientific, speculative, and perhaps backward-looking.

For the study of democracy in particular, this means that the concern of the democratic theorist lies primarily with the foundations that constitute the democratic polity, while for the student of political parties it lies with the actual functioning of the institutions within it. This brings us to the second, and related, domain: the level of analysis. Democratic theory and the study of political parties see themselves as operating at different levels of generality. Democratic theory is often grand theory, while the study of political parties operates largely at the middle-range level. While democratic theory focuses primarily on the macro (or system) level and its links to the autonomous individual at the micro level, the study of political parties concentrates mainly on the meso level of the political system. Because democratic theory is essentially concerned with the constitutive features of the democratic state and their moral justification, it tends to focus its attention on the broader system. Moreover, modern democratic theory arises largely from liberal traditions of philosophy. As contractarian sub-traditions of liberal philosophy have been extremely influential in democratic theory, these have biased the latter towards the constitution of states rather than the processes operating within states.⁴⁴

Democratic theory, therefore, often concentrates on questions of constitutionalism and individual rights, which it sees as the important hardware of democracy. Parties are seen as software and as less basic to democracy in temporal, organizational, deliberative, or normative terms. Democratic theorists, indeed, convey real doubt as to whether democracy does “need” parties; partisanship generally is often seem as something to be overcome, an obstacle to democratic reasoning,⁴⁵

and parties can deflect concerns with the common good (note the clear overlaps with themes in classic historical contributions, discussed earlier). While the study of political parties focuses on the intermediary structures between the individual and the state, the higher level of generality of democratic theory tends to deflect its attention from specific institutions, such as political parties, whose operation does not comfortably fit into its broader intellectual orientation or framework. As a consequence of its level of abstraction democratic theory is concerned with the whole, not with parts; it takes a resolutely system-level view of its object. This explains some of its traditional suspicion of political parties, as a party is, by definition, only part of the whole. Further, because parties unite people on the basis of competing ideological principles and opposing interests, they are not by their very nature geared towards more objective perspectives on the whole.⁴⁶

Each of these concerns, in turn, has an impact on how the fundamental concept at hand, democracy, is defined and understood in the two literatures. With regard to the defining terms, therefore, to study democracy means something altogether different for the democratic theorist than it does for the scholar of political parties. This observation brings into focus the third domain of difference. For the democratic theorist, democracy is an abstract and changeable idea as well as a contingent practice. For the scholar of political parties, on the other hand, democracy is largely contextual and exogenous to its object of analysis. In order to understand why this should be the case it is important to underline the sequence in which democratic theory and the study of political parties developed in relation to the establishment of representative government. Democratic theory developed prior to the emergence of modern representative government. Until surprising recently democracy meant direct, face-to-face democracy, a meaning stretching back to its Greek origins over 2,500 years ago. The idea of unmediated, face-to-face politics still hovers in the background as an implicit—and sometimes foregrounded as an explicit—baseline against which to measure degrees of democracy in different contexts.⁴⁷

As a consequence, there is a lingering, detectable apprehension about the broadening of the notion of democracy to the level of the modern nation state and thus (it is normally assumed) consigning any notion of more direct, or strongly participative, democracy to the historical dustbin. As Dahl observes, the large scale of modern nation-states created a conflict “between the theory and practice of representative democracy and earlier conceptions of democratic and republican government that were never wholly lost.”⁴⁸ The skepticism of democratic theory towards large-scale democracy is reflected in a certain “polis envy”⁴⁹ and a persistent suspicion of intermediary structures.⁵⁰

In addition, as we have hinted above, democratic theory largely emerges out of traditions of thinking about politics which stress individualism, on the one hand, and the collective (including the collective interest), on the other. There is little scope for what lies in between that dyad, for mediating institutions and organizations such as political parties. In fact, the existence of political parties was fundamentally incompatible with both the liberal democratic tradition rooted in the resolute individualism of Locke and the radically communitarian and anti-representative democratic tradition inspired by Rousseau, both of which in their different ways are difficult to marry with partisan institutions, which by their very nature both transcend individual interests and refute the existence of some sort of “general will.”⁵¹

The analysis of political parties, by contrast, emerged only after the notion that systems of representative government are democracy had more or less settled. Studies of political parties, therefore, tend to be rather a-historical. They ignore the variety in both historical and contemporary conceptions of democracy and tend to accept elections, and the narrow interpretation of representation associated with them, as benchmark democratic practice. Relatedly, they also tend largely to operate within the uncontested parameters of existing democratic systems without any systematic questioning of their democratic quality. While many empirical studies of political parties and democracy will accept the notion that democracy belongs to the category of “essentially contested concepts,” the literature is not very good at “making democracy strange,” to use the old anthropological way of formulating the first necessary step in analysis. Such approaches can sometimes be seen as committing the “definitional fallacy”⁵²

namely assuming that a system is democratic simply because it fits a type that is commonly regarded as being democratic.⁵³

The fourth domain of difference is that of substance versus procedure. Whereas democratic theory tends to emphasize the substantive aspects of democracy, the study of political parties emphasizes the procedural aspects of democracy. To do political philosophy is to write about morality, about issues pertaining to rightness and wrongness, which lends itself to theorizing about substance, not procedures.⁵⁴

Democratic theory tends to use a currency of fundamental principles such as freedom, equality, and justice, which have a timeless and moral quality. Detailed empirical examples may be explored, but these tend to be selected in the light of, and strongly driven by, principled ideals with deeper roots. The study of political parties, on the other hand, focuses on how systems operate and concentrates on how these principles are deployed in a core domain of actual democratic practice. This in turn means that democratic theory is often influenced by broader political philosophy tendencies to focus on the substance of policy rather than on the political process, since policies can, arguably, be right or wrong, moral or not, in principle. Policies lend themselves to substantive normative analysis, allowing political philosophers to stipulate what states should do (or refrain from doing) in policy terms, selectively ignoring the procedural aspects of democracy that form the ground on which political parties operate, thereby almost deliberately detaching itself from ordinary politics.

Why Democratic Theorists and Party Scholars Should Talk to Each Other, and How They Could

There are many good reasons for rectifying the contemporary lack of communication between democratic theorists and scholars of political parties. Political parties occupy an ambiguous position in modern democracies, which is in part a product of the tension between their de facto inevitability as key institutions of modern democracy and their increasing inability to perform the representative functions that legitimized their emergence and that are widely seen as essential to the quality of democracy. The centrality of political parties can be demonstrated empirically by the fact that they are firmly rooted in the established western democracies, and have rapidly acquired relevance in the more recently established democracies which have emerged out of what Huntington has called the “third wave,”⁵⁵

to the point that it is difficult even to conceive of a contemporary democratic polity without political parties. However, the failure of the empirical study of parties to take questions of democratic theory to heart, and to identify the relationship between normative and institutional prescriptions (difficult as it may be to unite the two paradigms) is a cause for serious concern, particularly given the important challenges faced by modern democracy.

Modern democracy, as Hardin puts it either “cannot entail massive citizen participation or it is irrelevant to actual practice in modern politics.”⁵⁶

Modern democracy thus depends on mediation, and therefore requires mediators. Democratic theorists need to take on board much more fully the mediating software—above all and despite their many challenges, parties—at the meso level of polities. It is here that pristine conceptions and principles are modified and acted out. In other words, it is crucial to incorporate these mediating structures and linkage mechanisms into democratic theory, and it seems equally evident that it is precisely through the agency of party that state and society can be interlinked. However inadequately they may do it, parties gather and bundle disparate preferences and interests and attempt to articulate collective visions out of varied and diverse particularities. Democratic theorists often propose new mediating structures—deliberative bodies, participative budgeting, and so on—but leap too blithely over the existing, if troubled, core mediating devices, namely parties.

Both theories of liberal democracy and deliberative democracy offer some potential to extend and to reconnect in this respect, as neither is in principle averse to the notions of representation and mediation. However, further progress would require a deeper appreciation and analysis of the place and functions of parties as the principal mediators in the democratic system. Even arguments for new mediating or representative institutions that might afford a deeper realization of democratic norms in practice need to deal with the issue of whether they can or should supplement or replace party-based processes.

What are some of the key areas in which this renewed engagement should occur? Consider a snapshot of selected core concerns of democratic theorists and party scholars respectively:

Each of the items in the left column of table 2 can be seen as democratic theory demands on democracy, and a set of areas in which democracy might be extended or deepened. They reflect the widespread conviction in democratic theory that the process of democratization is always unfinished, and that democracy is always an imperfectly realized ideal. Each of the items in the right column, deriving largely from the study of parties and party systems, can be seen as either constraints on the supply of democracy or vehicles through which the demands might in part be met. Both columns also include suggestions how both fields might take on board and engage more directly with each others' concerns.

With regard to A, equality, democratic theorists could gain by considering how, and to what extent, parties, in present or transformed configurations, can act as vehicles for the democratic ideal of equality. Discussions of equality lie at the heart of justifications of democracy,⁵⁷

but these are often individualized and decontextualized discussions. The capacities of parties to promote equality in their internal structures, their incentives to foster procedural equality in governance institutions, and their policy programs for addressing wider social inequalities affecting citizens' sense of political efficacy and participation rates, are all relatively neglected but critical parts of such discussions. Party scholars could gain from attempting to pinpoint the extent to which different parties and party system configurations realize the democratic ideal of equality in practice. This might take the form of renewed, theory-driven attention to the extent of openness and inclusion of party members in internal structures. It may further involve investigation of the commitments of parties to democracy and the support of the democratic process; contra Downs,⁵⁸ do parties have an incentive to deepen or bolster system-wide democratic practice, or are they in fact vote-winning machines for whom “democracy” is one more piece of handy rhetoric?

We return to B, deliberation, shortly, for reasons we shall make clear.

With regard to C, participation, decentralizing tendencies within democratic theory—oriented towards getting power closer to the people—might engage much more fully the local-national nexus in party systems and structures, and ask, for example, to what extent decentralized parties might act as a surprisingly subversive vehicle for disrupting the resolutely national (and therefore highly indirect and often distant) focus of democratic politics. Democratic theorists might also reflect on the role of parties in linking the local with the national, or on empowering the people through the agency of party in the democratization of decision-making structures and governing processes at the level of the nation-state. And indeed, since cosmopolitans and greens within democratic theory have been concerned to explore democratization in the international arena, links between parties in different countries, as well as the organization of international parties, ought to be explored. Party scholars might take less for granted the national basis of democracy and ask, for example, to what extent the dominance of national party organizations underpins the persistence of a minimal democratic politics and undermines a more diverse, local and even direct or participatory democratic politics. They might also further explore how the increased use of mechanisms of direct democracy, both inside and outside party channels, for decision-making and candidate selection impact on democratic ideals for deepening democracy.

For D, inclusion, democratic theorists' concerns about facilitating the representation of new and marginalized interests in democracies could focus less on general institutional fixes (such as quotas for women) or exotic imports and creations (citizens juries, for example), and more on the democratic potential of provocative ideas like the “open party”—a party in which its grassroots members, and perhaps also supporters more broadly, are responsible for setting political and policy goals. In other words, they could engage more fully in a discussion on how a deeper realization of democratic inclusiveness might be achieved through a modification or transformation of the existing representative institutions rather than ignoring the party-based process or tacitly assuming that it is best replaced by other structures. Students of parties, on the other hand, could fruitfully consider the extent to which parties fail to prioritize the inclusion of concerns of the marginalized in their search for electoral pre-eminence and government office, and how in turn they may unwittingly act as brakes on the deepening of democracy. Do parties lack a democracy-supporting incentive to build long-term capacity as opposed to maximizing short-term gain? On the level of the party system, they could examine the extent to which the formation of “cartel parties” has encouraged the exclusion or effective marginalization of certain interests and thus undermined a more inclusive democratic politics, and explore possibilities to counteract current trends towards less inclusiveness.⁵⁹

In all of these areas, and others, each camp has much to learn—or perhaps more to the point, much to be provoked by—in the different orientation of the other one. So, democratic theorists need to engage more, for example, with the constraints and opportunities which the realities of parties and party systems embody with respect to the realization of core democratic ideals. And party scholars need to engage more with the demandingness, restlessness, and changeability of those ideals, as well as with difficult issues of how far empirical realities offer a sufficient realization of democratic ideals.

Clearly the four areas are closely related. Participation is not the same as inclusion but the demand for one can often be a demand for the other. Inclusion might involve issues about citizenship, raising questions about the boundaries of the political community and who is to be included in the equality concern. Deliberation is not necessarily a basis of, or a producer of, equality, but issues of equal access to deliberative forums are crucial to the claims of deliberative theorists. We lay the issues out here as we do for clarity's sake.

Of course, it is one thing to call for more engagement, and point towards how it might be done. It is tougher actually to do the work. So let us try to put this partly right by spelling out for illustrative purposes what it might mean to follow our own injunction for our second democratic demand B, deliberation. We do not have space to be much more than indicative in so doing.

Scholars of political parties have largely ignored the “deliberative turn” in democratic theory. And theorists of deliberative democracy have largely ignored political parties. Even when they recognize the limited contribution deliberation can make to democratic politics and governance, democratic theorists still do not discuss parties.⁶⁰

This may be because theorists see parties as part of the problem of political disaffection and voter ignorance, and not part of a potential solution. Or because the deliberative ideal has roots in thinking oriented to the general interest, roots which influence their preference for certain alternative devices—citizens' juries and deliberative polls for example—and steer them away from suspect partisan institutions like parties (and thus, according to Budge's critique, taking “the politics out of politics.”)⁶¹

But what if both party scholars and democratic theorists took up Cohen's early challenge to explore parties as vehicles for deliberative democracy? Deliberation, theorists argue, can produce more informed preferences and enhance legitimacy, although the ultimate purpose of deliberation—the education and development of citizens, the forging of a broad consensus, etc.—has a real impact on the sites and mechanisms of deliberative democracy. Consider first parties themselves as sites of deliberation. We need to accept, of course, that parties are partisan, but that need not mean narrowly partisan. Parties have to sell a general interest vision to voters. And, especially in an era where relatively fixed class roots of party ideologies are loosening, there is arguably more scope for deliberation over alternative party visions, programmes, and policies. Parties could be umbrellas for positive deliberation over alternative proposals, all still within the broad ideological “frame” of the party. There may be scope to increase party memberships (in steep decline in many countries) if genuine opportunities to deliberate over policy are seen to be available.⁶²

Carefully managed, rebranding a party as committed to internal openness and transparency through deliberative processes may carry strategic benefits—there is no need to deny that parties are driven, at least in part, by vote-seeking objectives. Internal deliberative forums need not be non-partisan; deliberation that engages partisan stakeholders can also be effective.⁶³

Analysts could also look at decentralist and consultative mechanisms already present in party organization, exploring and perhaps deepening democratic deliberation on the back of them (with potential benefits in terms of identification, legitimation, and even strategic advantage). Such mechanisms could include bottom-up policy-making, degrees of local party autonomy, party primary elections and other participative modes of candidate selection, and party conferences. Further, there is little reason why more-favoured deliberative mechanisms, such as deliberative polls, could not be deployed within party debates.

Next, consider parties as potential promoters of societal deliberation. Party competition clearly fosters societal deliberation, through the media and otherwise—observers might analyze the quality and extent of this deliberation, and how parties do (and might) contribute to both. Could parties take part in policy and candidate debates more often and more prominently—perhaps even to their strategic advantage? Could they, in this light, sponsor citizen's juries or consensus conferences, rather than the more narrow focus groups that are often more favoured by parties? Even more broadly, what of the different incentive structures inherent in different electoral systems for promoting inter-party and diffuse societal deliberation? Perhaps the shift to cartel parties, ironically enough, could foster deliberation, since the latter is sometimes easier to engender in more elite contexts.⁶⁴

Observers might pay closer attention to deliberation between parties in legislatures. To what extent are parties vehicles for deliberation as well as representation in that context? And how can the potential complementary nature of these two roles be enhanced? Further, at a time when environmental and other issues are leading us to question the very idea of domestic policy, deliberation's capacity to work across borders and contexts may be enhanced by parties engaging for example with sister parties in other countries through policy deliberation. More generally, the value of party-fostered diffuse and episodic societal deliberation, not confined to one specific mechanism or device, has been under explored.

In these brief comments there is a range of hypotheses, born of looking past the mutual disregard that is the core topic of this article. Refining and exploring these hypotheses would require new orientations for both democratic theorists and party scholars. The former would need, for example, to engage with parties as unavoidable, if flawed, vehicles of policy aggregation and political representation; to recognize deliberation's limits; to work with institutional constraints; and to recognize the grey zone between particular and general interests. The latter would need, for example, to accept that democracy is inevitably a normative notion and take on board the power of the deliberative ideal; to recognize the potential of deliberative devices as supplementing and potentially improving party legitimacy and procedures; and to recognize the importance of critical questioning of party contributions to the quality of democracy, and deliberation. By such routes, new ways for scholars to address the widespread disaffection with politics in general, and parties and politicians in particular, may be opened up. And that would be significant for the relevance and utility of political science.

Conclusion

If it is true, as Bryce argued at the beginning of the twentieth century, that modern representative democracy cannot function without political parties, current developments indicating that parties are losing their capacity to act as agents of representation have far-reaching implications for the nature of democracy.⁶⁵

In order to address the fundamental tension between the centrality of parties and their marginalization in an area quintessential to any modern democracy, both normative theories of democracy and empirical studies of political parties are of vital importance. In particular when faced with the challenges of consolidating more recently established democracies, resolving problems arising from the changing nature of parties in the established democracies, or addressing the democratic deficit of the European Union and other transnational organizations, it is imperative that empirical observations can take guidance from normative theories and that normative claims are grounded in empirical reality. For that reason, the literatures on political parties and democratic theory should engage more frequently, substantially, and imaginatively with one another. In this article, we have outlined several of the reasons why the two strands might have parted ways and sketched various ways in which they could possibly re-engage with one another. It is only when they do that we can try to make sense of the place of parties in contemporary democracy and, indeed, of the nature of modern democracy itself and its potential futures.