Political Parties and Deliberative Democracy?

James Johnson. Handbook of Party Politics. Editor: Richard S Katz & William Crotty. Sage Publications, 2006.

The study of political parties is related to the study of democratic theory in complex, ongoing ways. Over the past two decades democratic theorists have devoted considerable attention to what is called ‘deliberative democracy’ (Bohman, 1998; Freeman, 2000). This literature largely neglects the topic of political parties. Empirical and analytical research on political parties has reciprocated and passes over the topic of deliberation largely without comment. In this essay, then, I will not review a large body of research. That is not possible since such a literature does not exist. Instead, I first suggest that the neglect of parties by advocates of deliberation is somewhat surprising. I then identify the source of that neglect. Finally, I sketch one way that we might reconnect the study of democratic deliberation and political parties.

Democracy and Deliberation

Democracy is among the institutional arrangements that people have adopted to address what can be called the ‘circumstances of politics’ (Waldron, 1999; Weale, 1999). For any population these circumstances are constituted by inescapable diversity across multiple, overlapping dimensions, including material interests, moral commitments, and cultural attachments. Such diversity means, in turn, that disagreement is an unavoidable condition of politics. This is, in part, because the individuals and groups who constitute some relevant population have interests, commitments, and attachments that are not only diverse, but irreducibly so. There simply is no neutral metric that will accommodate competing demands without remainder. But the inevitability of disagreement also partly reflects the fact that, precisely as members of a relevant population, those individuals and groups are, as it were, stuck with one another. In short, their lives are highly, irrevocably interdependent. Thus, despite their diversity and the disagreement to which it gives rise, they require some means of coordinating their ongoing social and economic interaction.

At the most general level, democracy means ‘rule by the people’. More specifically, democracy consists in an institutional arrangement for making binding political decisions in ways that are responsive to the views of the public. Any such arrangement will include formal or official decision-making forums, even if it additionally requires such features as a sustaining set of civil and political rights, supporting institutions (such as schools), and an extensive environment of political organizations and secondary associations. It perhaps is easiest to see where political parties fit into this definition, namely as political organizations that connect citizens to government by coordinating citizens for electoral purposes. Beyond that, however, the analytical and empirical literature on parties is divided regarding the more precise roles that parties play in democratic politics. Indeed, even the fundamental question of whether parties make government more or less responsive to the views of the electorate remains unsettled (Stokes, 1999).

Advocates of deliberation insist that political argument or debate conducted under conditions of freedom and equality is a crucial component of any democratic decision-making process. They commonly insist that, so understood, deliberation aims at the formation of political judgements or preferences. In this sense it stands in contrast to voting, which seeks solely to aggregate pre-existing preferences. Deliberation thus is best seen as part of a process of forming majorities which is a crucial aspect of democratic politics (Spitz, 1984). In this respect it would seem that advocates of democratic deliberation might find political parties a natural focus of inquiry. Unfortunately they rarely so much as mention parties.

This state of affairs is somewhat surprising. Historically, theorists such as Edmund Burke or John Stuart Mill ‘who stress the role of deliberation … in politics’ also ‘justify the existence of parties’ (Manin, 1987: 368). What is more striking is that two of the now classic essays in the contemporary literature explicitly accord political parties a central role in democratic deliberation. Cohen (1989: 31-2) insists that independent, publicly funded political parties could contribute to democratic deliberation in two ways. First, because parties provide organizational resources, they might help offset the material inequalities that render the outcomes of deliberative processes suspect. Second, because parties, unlike interest groups, need to address a broad range of issues, they could help keep the focus of political debate on matters of general concern rather than on local or issue-specific matters. Manin (1987: 356-7) insists that political parties are an important means of overcoming the necessarily ‘bounded’ nature of deliberative processes. Since not all possibilities can be examined, parties operate to focus discussion and debate on some subset of the possible ways to resolve any political disagreement. After this apparently promising start, however, the subsequent literature on deliberative democracy has almost nothing to say about political parties. One exception is Christiano (1996: 222-4, 244-8) who argues, like Cohen, that parties work to focus on general issues and away from candidate-centered politics and, like Manin, that they operate to focus attention on particular sorts of response to political problems. And he stresses that parties play a useful role in the political division of labor by structuring discussion of public issues in ways that are accessible to non-specialist voters. He depicts political parties as actors in electoral campaigns which he interprets, in turn, as a ‘process of competitive debate’ aimed at persuading voters. But among recent advocates of deliberation Christiano is a clear exception.

Why the Neglect of Parties?

One might conjecture that the reason why theorists of deliberative democracy neglect parties is that their concerns are primarily normative while the literature on parties is primarily empirical. But recent surveys make clear that analysis of parties is suffused with normative concerns (Pomper, 1992; Stokes, 1999). Conversely, theories of democratic deliberation raise sets of analytical and empirical problems that cannot plausibly be set aside to concentrate solely on normative concerns (Johnson, 1998). One might, alternatively, argue that the neglect of parties reflects the abstract level at which treatments of deliberation are pitched. But this suggestion, too, is unpersuasive. Even briefs for democratic deliberation that focus on problems of institutions (Ferejohn, 2000; Guttmann and Thompson, 2004) or that purportedly analyse ‘actual deliberation in non-ideal conditions’, do not treat parties in any sustained way (Guttmann and Thompson, 1996: 39; 2004; Macedo, 1999).

The source of neglect resides more plausibly in the way advocates of deliberation frame their enterprise and specifically in the fact that they consistently defend the relative normative attractiveness of ‘deliberative’ practices and institutions by setting them in opposition to ‘aggregative’ ones. Thus, in an early essay Cohen (1989: 29) insists that his ‘deliberative conception’ of democracy ‘construes politics as aiming in part at the formation of preferences and convictions, not just at their articulation and aggregation’. And much more recently, two prominent theorists ask rhetorically ‘Why is deliberative democracy better than aggregative democracy?’ (Guttmann and Thompson, 2004: 13). The problem should be clear. Political parties typically are understood as ways of coordinating citizens for electoral purposes (Pomper, 1992). They thus fall on the aggregative side of this divide and so beyond what most advocates of deliberation take as their purview. Unsurprisingly, the few advocates of deliberation who do discuss political parties in a serious manner resist the aggregation-deliberation dichotomy (Christiano, 1996).

Framing the discussion of democratic politics in this way—in terms of deliberative versus aggregative conceptions—is doubly unhelpful. First, it distorts the history of political thought in which defenses of representative government standardly insist that it both relies on elections to select government officials and demands that binding political decisions must survive prior public discussion and debate (Manin, 1997). In short, modern democratic institutions do not break down easily along the deliberative-aggregative divide. Second, the deliberative-aggregative dichotomy distracts attention from theoretical reasons we have for suspecting that neither mechanism alone affords a sufficient basis for arriving at political decisions within a democratic framework. Advocates of deliberation offer no reason to suspect that, given the circumstances of politics as I describe them above, political discussion, debate or argument will generate substantive consensus on even minor policy matters. Nor have they offered any reason to think that such substantive agreement is uniformly desirable in a complex, pluralist society (Knight and Johnson, 1994). Conversely, social choice theorists notoriously argue that any voting mechanism that meets even a relatively minimal set of normative criteria can generate cyclical or unstable collective choices.

Reconnecting Deliberation and Parties

Consider an alternative approach. Instead of setting deliberation and aggregation in opposition to one another, we might examine the way they interact in democratic politics. The first step here would be to recall that if the preference rankings of individual voters are structured in particular ways then voting mechanisms need not generate the collective instability that social choice theorists identify. The most regularly discussed of these preference structures is ‘single-peakedness’, but there are several others that are sufficient to avoid collective irrationality (Sen, 1966). If voters have single-peaked preferences they essentially share a common understanding of the issue space that sets the parameters on any substantive political disagreement they might have. Put otherwise, while they might continue to disagree substantively voters with single-peaked preferences agree in a second-order way. The second step would be to recognize that political debate and argument can induce just such a shared understanding. This point was intimated by prominent social choice theorists (Arrow, 1963: 85; Riker, 1988: 122, 128), was spelled out explicitly in relatively early discussions of the relationship between deliberation and aggregation (Miller, 1992; Knight and Johnson, 1994), and has been revived even more recently in the same context (Dryzek and List, 2003).

The payoff that advocates of deliberation might derive from these insights is considerable. First, there is some empirical evidence that deliberative practices do indeed help to structure preferences in just this way (List et al., 2000). Thus the approach just sketched will afford them an empirically plausible mechanism for explaining how deliberation ‘works’ (Johnson, 1998). Second, the claim that deliberation establishes second-order agreement deflates the complaints of postmodern critics who presume that deliberation necessarily aims at unwarranted consensus (Mouffe, 2000). The reason is simple. This view does not require that deliberation induces consensus. In the first place, a shared understanding about the dimensions of conflict does not eliminate substantive disagreement over how best to resolve the conflict. What is common to members of the relevant population is the structure of their preference orderings, not the content of their preferences. But, perhaps more importantly, even second-order agreement regarding what is at stake in a given conflict need not be unanimous. Indeed, aggregation mechanisms can generate collectively rational outcomes if as few as 70-75 percent of the relevant population have single-peaked preferences (Niemi, 1969). There is, in other words, room for considerable disagreement among the population over both substantive matters and second-order understandings. It is, therefore, perhaps more appropriate to claim that deliberation structures disagreement rather than to insist that it induces agreement. Finally, since advocates of deliberation are not committed to direct democracy (Cohen, 1989), the focus on the interaction of deliberation and aggregation locates their work more firmly in the tradition of democratic theorists as various as James Madison, John Stuart Mill, and John Dewey all of whom defended representative government as including both aggregative and deliberative aspects.

Here we return to the question that motivates this entry. Where do political parties enter into all this? In a rare discussion of deliberation from the perspective of one who studies political parties, Ian Budge (2000: 206) sees party competition as imposing a left-right dimension on electoral issues and hence as an alternative to deliberation. In so doing Budge, tacitly at least, embraces the aggregation-deliberation dichotomy that frames most briefs for deliberative democracy. Moreover, his view of the effects of party activity is not universally accepted (Stokes, 1999). Yet here we might well elaborate on Christiano—who in turn follows Mill (1991: 315), for whom public debate fulfills the ‘function of antagonism’ in politics—and interpret party competition as a vehicle for rather than an alternative to public persuasion and debate. Then, if we understand the primary effect of deliberation as establishing second-order agreement on the dimensions of conflict, a clear confluence emerges between the case for deliberative democracy and the analysis of political parties. This is a proposition that bears scrutiny.