American Exceptionalism

Alan Ware. Handbook of Party Politics. Editor: Richard S Katz & William Crotty. 2006. Sage Publishing.

One of the problematic aspects of the study of party organizations has been how to account for the undoubted differences between American parties and many of the European parties. Of course, taking account of unusual and distinctive institutions is something with which comparative political analysis must deal all the time, but the search for suitable analytic frameworks for party organizations that embrace those in the United States has been hampered by three quite specific factors.

First, there is the impact of what might be termed the ideology of ‘American exceptionalism’ on the American political science community. The Tocquevillian idea that American society was different from other societies, and that consequently its politics was also unique, was a powerful one. Too often the assumption that those exceptional values had a direct effect on organizational forms the parties developed has been held uncritically. Rossiter’s (1960: 37) unsubstantiated claim was typical of a way of thinking; having identified a number of what he claimed were unique aspects of American politics (including local bossism), he noted ‘Nowhere in the world … is there a pattern of politics anything like ours.’

Secondly, among European political scientists there has been much misunderstanding as to how American parties operate. Duverger, for example, understood American parties to be the counterpart of the early cadre parties in Europe—that is, a grouping of notabilities (or elites) who come together to prepare for elections. There are similarities that are evident in many respects, but with one important exception. In Europe the attitude of the notabilities to those they recruited to perform campaigning tasks was that they were not in any real sense participants in the party; in the parties that developed in the United States in the Jacksonian era widespread popular participation in parties was understood as the very cornerstone of democracy.

Thirdly, until the 1980s (and possibly later) relatively little was known about party organizations outside the United States and the larger countries in Western Europe. Although there were a few interesting attempts to broaden the range of parties incorporated into comparative analysis—such as that by Epstein (1964) in the case of Canada—until recently most studies focused on a much restricted range of party organizations in liberal democracies. The result was analytic frameworks that tended to juxtapose the electoral campaign-oriented American organizations with the supposedly policy-oriented, internally democratic, parties of which the European socialist parties were the most outstanding example (Wright, 1971). This reinforced the idea that American parties were, in some sense, different from most parties elsewhere, especially in that ‘the ideological or policy clarification and goal definition function is rejected’ (Wright, 1971: 33). Thus, American parties have often been treated as a type of party that is non-ideological, although Gerring’s (1998) research has demonstrated there has been a pronounced ideological component to American national election campaigns since the 1830s.

Among them, these three factors have contributed to a misunderstanding of the relationship between the American parties and many European parties. Moreover, some political scientists have simply given up on the attempt to incorporate the former into a comparative framework; Panebianco (1988: xv) famously excluded American parties from his analysis by asserting that the factors affecting their emergence and development were different, but without discussing what the difference actually was. This chapter seeks to outline the main differences in party organizational structure in the USA, to consider the various explanations that might be given for those differences, and to explore the problems of incorporating American parties into some of the more popular analytic frameworks.

The Main Differences between American Party Organizations and those of Other Parties

Many of the supposedly distinctive features of American parties are evident elsewhere. The ‘exceptional’ nature of American parties tends to lie in the extent to which some of these features have been developed and also in the particular combination of features prominent in the American parties. Four main differences are especially important in the contemporary United States.

Extensive Legal Control of Parties

From the advent of mass-based parties in the 1830s until the late 1880s there was virtually no legal control. Parties issued their own ballot papers at elections, they controlled their own nomination procedures, they determined the structure of their own organizations, and so on. Beginning with the adoption of the official ballot (known in America as the ‘Australian ballot’) by most states at the end of the 1880s, extensive legal control of party activities commenced. From early in the 20th century most states required major parties to use the direct primary election as the system for nominating candidates for public office, and these parties were also required to have a particular kind of organizational structure, in which the lowest echelons of the party were also directly elected in publicly administered elections.

From a largely unregulated party system, the United States had moved rapidly to one that had the most extensive legal regulation of all liberal democracies. Some political scientists have seen this as being one of the most distinctive aspects of American parties; Epstein (1986: Chapter 6) has argued that parties are conceived in the United States as a kind of public utility, rather than a wholly private form of organization. However, it is important to recognize that the difference is partly a matter of degree; a number of other countries require their parties to operate within a particular legal framework. (In Germany, for instance, a party must be democratic—although the courts rarely venture into the potential minefield as to what counts as democratic.) Moreover, some party-related activities that are not controlled by statute in the United States—an example being the forms of electioneering that are permissible -are regulated elsewhere, with particular practices being banned.

The Absence of a Dues-Paying Membership

In many countries the link between the would-be participant in a party and that party became primarily one of formal membership. That is, the party formally enrolled its activists, often after a trial period in which they were associated with the party; on joining the party, the member would be liable to pay dues annually to the party, and he or she could be removed from the party as a result of activities held to be incompatible with membership. Members had rights to be involved in certain procedures -such as the nomination of candidates—that non-members did not. The original exemplar of the membership model had been the German Social Democratic party, though the general model was adopted by most socialist parties. In turn, and as Duverger (1959) had observed, other parties in Europe started to make use of the device of the fee-paying member, even though they did not always grant extensive formal powers to their members.

By contrast, in the United States, neither major party developed this kind of structure. Before the introduction of legal regulation, the question of who could become involved in a party’s activities was not a matter that was policed rigorously, although those known to have supported the opposing party recently might be excluded from the relevant meetings -by force, if necessary. After the introduction of devices like the direct primary, it became difficult to identify party ‘traitors’ and hence exclude them from activities such as candidate selection—even in states that opted for so-called ‘closed primaries.’ In practice, therefore, today the American parties have ‘members’ who differ from the dues-paying members of most parties, in that they are entirely self-selected. However, as with legal regulation, the absence of a dues-paying membership is not unique to American parties; for example, the two largest Canadian parties, the Liberals and the Progressive Conservatives, for most of the 20th century did not enrol members in this way.

Highly Decentralized Party Structures

Even when they started out as decentralized structures, as in France, with power residing among locally based leaders, over time most parties centralized a number of key activities. National parties might still leave control over certain functions to local units—for example, the selection of parliamentary candidates in the case of the British Conservatives—but many functions came to be controlled at the centre. In the United States, there was no such centralization. Not only did the power of nomination continue to rest at the local level, but those who were elected to public office could remain largely independent of their fellow party representatives, providing they retained the loyalty of their local voters. Furthermore, over time power within the nomination process has become even more decentralized. In the 19th century it was the county-wide party organizations that were normally the most important actors in that process. With the advent of television campaigning, individual candidates could assume that role; after the 1960s increasingly at all levels of office candidates’ own organizations became the main structures deployed both in nomination contests and in general elections.

While this feature does mark a difference with many other parties, it is important not to exaggerate that difference. First, there are some significant centralizing tendencies in American parties. For example, the ability of major politicians to raise large sums of money that they can then disperse to likely challengers from their party creates informal links of obligation that tie otherwise independent political actors to each other. Consequently, crude versions of the ‘business firm’ model of parties tend to overemphasize the autonomy American politicians have in office. Not only must they keep the local interests in their own electoral coalition content, they also operate in an environment where major opponents in their party might be able to help resource a primary challenge to them; for instance, the behaviour of Republican moderates in the House over the Clinton impeachment becomes inexplicable unless this factor is recognized. (For a version of the model that does not make this kind of mistake, see Schlesinger, 1985.) Secondly, not all the other democracies displayed an early tendency towards centralization in their party structures. France, for example, did not; it was not until the incentives provided by the switch to a semi-presidential system in the Fifth Republic that the essentially local nature of French parties was transformed. Thirdly, personal campaign organizations are not unique to American parties. They are also found in those electoral systems that tend to pit different members of a party against each other—for example, the single transferable vote (as used in Ireland) and the single non-transferable vote (as used in Japan before the mid-1990s).

The Non-Programmatic Nature of Party Competition

American parties do not develop policy programmes within their organizations, nor do they campaign on the basis of such a programme at elections. The policy platforms that are published before an election are general in nature, do not constitute a kind of promise to the electorate, and the party’s candidates are not bound to support it. Consequently, these platforms receive little attention in the campaign. Because they do not develop policy, the parties lack the kinds of research units within their organizations that might develop policy programmes.

However, this different approach to campaigning does not mean that American parties are non-ideological—a charge frequently raised against them. Gerring’s (1998) systematic analysis of presidential election campaigns shows that, once mass-oriented parties emerged, there have been clear differences between the two main parties in the ideological appeals they make to voters. Those appeals have changed over time, but ever since the 1830s there has been electoral competition based on ideology between the parties. Moreover, the non-programmatic style is one that is not unique to the United States; it is common in those presidential elections where, at some point in the electoral process, the winning candidate has got to secure the votes of more than half of those voting. This requirement tends to discourage policy programmes—partly because, at least formally, legislation is not the responsibility of the elected chief executive, and partly because the size of coalition needed to win a presidential election may be so great that it becomes difficult to construct detailed winning programmes.

What are the Causes of the Differences?

While most commentators agree on the differences between American party organizations and organizations elsewhere, there is considerable disagreement as to what the causes of those differences are. That disagreement includes the question of whether there is one main cause of ‘exceptionalism,’ or several unrelated causes of that apparent phenomenon.

Anti-Party Political Values

Many American political scientists subscribe to the Tocquevillian view that the origins of the distinctive party structures lie in unique political values, though few now articulate the more general argument associated with an earlier generation that it was a whole set of exceptional political values and practices that the United States acquired (Rossiter, 1960). Rather it is a particular argument that is now often invoked to account for most of the observed differences—namely that in the United States there has always been at least an ambivalence, and at worst outright hostility, to political parties. It is argued that anti-partyism can account directly for the introduction of a compulsory legal framework imposed upon the major parties; that in turn meant that membership-based parties were neither possible nor necessary, and they were under less pressure to become programmatic. Even the decentralized state of the parties might be understood as a response to the unpopularity of parties—keeping them decentralized (and hence more low-profile) made it less likely that they would stir up yet more antagonism among mass publics.

In the last two decades, when survey evidence indicates that parties certainly did become less popular than they had been, the argument from anti-partyism has been much used in accounting for the distinctiveness of party organizations (see, for example, Katz and Kolodny 1994: 26-7; Pomper, 1992: 132-4; Wattenberg, 1991: 32-4). The argument has an obvious appeal but there are three main objections to it that suggest it is of limited utility.

First, in spite of frequent assertions that anti-partyism is stronger in American politics (or, perhaps, permeates politics there more deeply), there is no available cross-national survey data of either political elites or mass publics to demonstrate that this is true. There are also plenty of examples of periods in European countries when anti-partyism was strong, so that it is an explanation that is far from self-evident. Secondly, even if it could be demonstrated that, for much of the 20th century, anti-party sentiments were more prevalent in America than elsewhere, it remains the case that American party structures became established in the 1830s, and that for more than 60 years parties were at the centre of social and political life (McCormick, 1986; Silbey 1991). Arguments that somehow earlier deep-rooted anti-partyism (of a kind usually associated with Madison) somehow just got submerged in this period, only to resurface later, are unconvincing—unless a firm link can be established between different periods of anti-partyism, and so far it has not been. Thirdly, although it is often argued that anti-partyism was responsible for the legal regulation of party nominating procedures, and the subsequent adoption of the direct primary, the evidence actually points in a different direction. As with the earlier adoption of the Australian ballot, it was party politicians, rather than anti-party reformers, who were at the forefront of moves for legislation. Their motivation was usually the desire to modernize practices that no longer worked well, and such reforms could not be effected through party rules, with the law being the only way of doing so (Ware, 2002).

The Decentralization of the American State

Party organizations tend to develop structures that reflect the structure of governmental institutions they are attempting to control. As Finegold (1995: 29) observed of the impact of American local government reform at the end of the 19th century: ‘The structure of local party organization reflected the structure of local government. Governmental consolidation encouraged party consolidation, and governmental fragmentation encouraged party fragmentation.’ More generally, the point could be made that in a state that had the separation of powers, federalism, and, after the Jacksonian era, a very large number of directly elected public offices, a decentralized party structure was inevitable. However, by itself, this kind of argument is insufficient—it cannot account for the absence of a phenomenon evident in some Canadian provinces. There, one of the effects of the pressures of federalism has been the formal separation of the federal and the provincial parties, and, in some cases, also the successful formation of parties that compete at the provincial level but not at the federal level. With only a few minor exceptions, in the United States the same parties have engaged in competition for all levels of office.

Consequently, an argument from the impact of governmental fragmentation has to be used in conjunction with an argument from the development of a national political career structure in America (Watts, 1987). Two features of this are relevant in this context. Unlike Canadian provincial politicians, state politicians often run for federal office (and, less frequently, the opposite career move also occurs). However, while at all levels of office individual politicians have their own base of electoral support, the US Congress and most state legislatures organize their business around parties; it is parties that make committee assignments, for example. Although some political careers have been sustained by politicians who chose not to align themselves with either of the major parties, there are strong incentives not to do this. The result is a near-monopoly enjoyed by the two parties at all levels of partisan contest.

The Development of Mass Political Organizations before Urbanization

This is an argument about the effect of changes of scale in the polity—changes that could account for the pressure to adopt an extensive legal framework for the parties. That in turn would account for the failure of membership-based parties to emerge. American parties began mobilizing mass electorates (almost exclusively white and male) in the 1830s in a country that consisted mainly of small towns and rural areas. In such a society formal rules of participation were often unnecessary, and the proceedings of parties at all levels were largely informal—relying on tradition and convention to resolve disputes. This was a face-to-face society in which order could be maintained in the party without recourse to formal rules, and many state parties did not have written rules. With population growth electoral districts became larger, and with mass immigration it was no longer the case that this was mainly a polity in which participants in a party knew each other. Consequently, abuses of political procedures increased, so that many party elites had an incentive to attempt to reform their parties. The classic example of this is the party ballot. Intra-party disputes about nominations could lead to rival candidates for local offices being run at general elections, with alternative ballots being issued to voters. This produced relatively high levels of ticket splitting, a practice that party elites within a county usually wished to minimize. With the introduction of the Australian ballot at the end of the 1880s, a reform supported by many of those elites, ticket splitting was reduced in the following decade (Reynolds and McCormick, 1986).

This third argument holds that one key difference between the United States and other liberal democracies is that mass party structures elsewhere had to deal with problems associated with larger scale at their inception. Thus one of the consequences of democratization preceding urbanization in America was that large parties, mobilizing mass electorates, could develop without formal internal rules and procedures. In conjunction with the decentralized form of the state, these parties could operate effectively for a while without the kind of structures that parties in other countries would need later when they started mass electoral mobilization. The American parties had to address the problems associated with scale eventually, but by that time there were already structures containing political actors who had a stake in maintaining certain aspects of the status quo. That restricted the form that organizational change could then take—a point that is compatible with Panebianco’s (1988) most general argument about the constraints on long-term change in party organizations.

The Impact of Presidential Government

A recent, and important, explanation for the failure of American parties to develop a policy-making role is that it was the result of how the presidency developed under Franklin Roosevelt. This is an argument most closely associated with Milkis (1993: 5), who argued that:

Roosevelt aimed at building a more progressive form of government within the presidency, rather than through a more permanent link between the executive and legislature. This required extending the personal and nonpartisan responsibility of the president to the detriment of collective and partisan responsibility.

The expansion of federal government activity in the 1930s could have been done in a way compatible with a more programmatic role for parties, but it was not. Not all political scientists agree with Milkis that Roosevelt actually intended to weaken parties. Coleman, for example, argues that Roosevelt had a desire for ‘more cohesive parties and for institutional entrenchment of the New Deal, not simply a system of presidential aggrandizement.’ However, he too accepts that the effect of the presidency developing in the way that it did was to prevent the development of a policy formation role for parties: ‘Institutionalizing New Deal reforms in the state would not build a “party state” but would lead instead to a diminished stature for party over time.’ (Coleman 1996: 59).

The main query to be raised about this kind of argument is whether a debasing of the role of party in policy-making is something that is always likely to happen in a presidential system. Not only do presidential contests typically turn less on battles of rival programmes than do parliamentary elections, because of the need to construct a large electoral coalition, but once in office presidents often have the resources to develop policy semi-independently of party colleagues in the legislature. Thus, the differences that might be being explained by this factor are more differences between the role of parties in presidential and parliamentary systems, rather than differences that are peculiar to the United States.

The Impact of Competition from other Forms of Party Structure

An argument developed by Duverger (1959) was that superior forms of party organization would drive out inferior ones—the latter being those that were less successful at mobilizing mass electorates. Thus, according to him, mass parties relying on formally enrolled members had tended to replace cadre parties organized around a small caucus—an argument rejected by Epstein (1967), on the grounds that looser forms of party organization were more flexible, and more appropriate to new forms of election campaigning. But can the absence of competition from membership-based parties account for the failure of the major American parties to develop this, or other forms, of party structure?

The answer is probably ‘no.’ The alleged efficacy of membership-based parties lay in their ability both to reach a large number of potential voters, and to acquire resources to enable themselves to carry out the activities associated with that. However, the American parties had devised means of doing this in the 1830s without formal membership. On the one hand, elections were held frequently, and the parties turned the activities associated with this regular vote mobilization into a form of participatory recreation. This style of campaigning started to decline towards the end of the 19th century (see McGerr, 1986), but in the meantime it had contributed to a massive vote mobilization for several decades. On other hand, the use of the spoils system gave parties access to resources through control of public office; they simply did not need the income that membership dues would have provided.

How do American Parties Fit into Some Models of Party Structures?

One of the consequences of the influence of the three factors identified in the opening section of this essay is that frameworks for analysing party structures have often been devised into which the American parties do not easily fit. Certainly, this is true of three of the most famous frameworks.

Cadre and Mass Parties

In their original form in the mid-19th century, American parties could be said to exhibit features of both cadre and mass parties. As Scarrow (1996: 19) observes, ‘For Duverger a true mass party is identified by its aspirations to enrol a wide segment of supporters, and to offer them year round opportunities for participation.’ However, this is precisely what American parties did. The ethos of the political system then was that political participation was desirable in a polity, and that it should be transmitted through the parties. Because so many offices became elected during the 1830s, and because terms of office were usually short, most communities had at least one set of elections each year. In that era American parties were engaged in mass activity much of the time. In that sense they were not elitist institutions. However, they were elitist in a different sense—in that there were few formal decision-making procedures by which organizational leaders were selected. Leaders emerged from the set of interpersonal relations within the party, and for that reason it was appropriate for Duverger to understand American parties as a kind of cadre party. However, Duverger also believed that mass parties were distinguished by their ‘formal enrolment procedures,’ and most definitely the American parties lacked such procedures.

After the 1890s the major American parties tended to lose some of their ‘mass’ characteristics. Participatory activities during election campaigns were replaced by ones in which party supporters were more akin to spectators (McGerr, 1986). During the Progressive era the enthusiasm for as extensive a system of elective office as possible waned, and over the next century tenure of office also tended to increase. The result was that activity in the parties became much more intermittent than it had been in the 19th century. Moreover, with the move towards more candidate-centred styles of campaigning, it became possible to see the parties as being a kind of cadre party in which the main political elites were the candidates themselves (Ware, 1985: 14-15). Yet, with respect to their origins, they do not fit easily into either the cadre or the mass category.

Mass Parties and Catch-all Parties

Yet another feature of Duverger’s concept of the mass party poses difficulties when examining American parties. The mass party was the ‘representative of pre-defined sectors of society’ (Katz and Mair, 1995: 7). Yet this was precisely the role of American parties in the 19th century—they mobilized specific social groups. What Thelen (1986: 23) says of Missouri was true of most other states: ‘Parties became the political arm of ethnic, religious and sectional cultures. Each new group from Europe joined the party opposite to the one that had attracted its most bitter enemy from their homeland.’ However, there were three differences between the social bases of parties in the United States and those found in much of Europe. First, class was not a source of party mobilization in the two major parties, and there were no large class-based mass parties. Secondly, the decentralized nature of the American parties meant that the same social group might be mobilized by one party in one city and by the other party in a different city, so that nationally the parties’ social bases resembled much more a mosaic than clear lines of division. Thirdly, one of the effects of presidentialism had been to provide a disincentive for the formation of more than two major parties, so that social groups tended to combine with other groups in a given party, rather than having a party of their own to promote their interest.

During the 20th century the connections between specific social groups and particular parties became weaker, thereby furthering the popular impression that American parties sought votes from wherever they could obtain them. Thus, when Kirchheimer (1966) introduced the idea of the ‘catch-all’ party that was replacing the mass party in Europe, it was misleadingly seen as an aspect of the ‘Americanization of European politics’ (Katz and Mair, 1995: 8). What was correct about this view was that since the end of the 19th century American parties had had to pursue more of a catch-all strategy as social group links started to weaken. What was misleading was that the American experience had actually been remarkably similar to the European—except that the change to a more ‘catch-all’ approach to election campaigning had occurred about six decades earlier. Until the 1890s American parties had largely won elections by mobilizing the party faithful, rather than by engaging in ‘catch-all’ strategies.

Catch-all and Cartel Parties

One of the most recent models of parties, the cartel party, sees parties as becoming part of the state. The ‘parties still compete, but they do so in the knowledge that they share with their competitors a mutual interest in collective organizational survival and, in some cases, even the limited incentive to compete has actually been replaced by a positive incentive not to compete’ (Katz and Mair, 1995: 19-20). Such parties draw many of the resources they need from the state directly (in the case of financial subventions) or indirectly (in the case of access to publicly controlled television channels). However, if this is a new form of party model, dating perhaps from 1970, then it must be admitted that in the case of the United States some of the features of the cartel party were more evident in the 19th century than they are now. Certainly, the parties then used control of government to generate the contracts and jobs that were the lifeblood of the party. Although they competed fiercely against each other for spoils, when their mutual interests were threatened they worked together to keep out other parties. For example, following the introduction of the Australian ballot one of the most persistent problems facing third parties was actually to get on to the official ballot; in many states rules were designed to make it difficult for these parties to mount a challenge. Moreover, legislatures were usually organized in such a way as to disadvantage representatives of other parties.

This is not to say that the cartel party is not a useful concept, nor to deny that in many respects there is growing evidence of recent co-option of the state by parties—even by parties in the United States. However, there is a case for arguing that some aspects of cartelization were to be found in America in an earlier era, and this exposes further the need for party models to be more sensitive to the particular development of American parties. That development resembles many aspects of party development elsewhere, but there are also important differences—and those differences are not always the ones to which commentators draw attention.