I PARTY UNITSJoseph A. Schlesinger
II PARTY SYSTEMSHarry Eckstein
The term “political parties” emerged in the nineteenth century with the development of representative institutions and the expansion of the suffrage in Europe and the United States. It designated organizations whose goal was the capture of public office in electoral competition with one or more other parties. Subsequently the term “party” was extended to include political organizations not engaged in electoral competition: minor parties which had no realistic expectations of gaining office through appeals to the electorate, revolutionary organizations seeking to abolish competitive elections, and the governing groups in totalitarian states.
The expansion of the term “party” to include organizations with varying goals has resulted in the scarcity of viable party theory. Apart from the pioneering works of Ostrogorskii (1902) and Michels (1911), there exist mostly descriptive studies of parties in individual countries (e.g., Neumann 1956). The most ambitious attempt to carry party theory forward has been that of Maurice Duverger (1951). But Duverger uses a historically determinist framework which makes the mass membership party the inevitable product of universal suffrage and renders archaic many existing and seemingly durable electoral parties. In contrast, this discussion will focus upon the political organization which actively and effectively engages in the competition for elective office. This emphasis makes it possible to resolve at least one issue which has hampered the development of party theory: the function of the party within the political system.
One consequence of the indiscriminate use of the term “party” has been preoccupation with the party’s functions and goals. The literature on political parties is replete with classifications of parties according to their goals. The most common distinction is that between the mass-based party, which is ideological, doctrinaire, programmatic, or issue-oriented, and the cadre or brokerage party, which is pragmatic and patronage-oriented. Although there is no logical barrier to mass-based parties’ being programmatic or cadre parties’ being doctrinaire, the distinction persists because the question of the function of the party is fundamental: is the party the instrument of its membership, or is it a public agency, primarily responsive to the electorate?
The perception of the party as the instrument of its membership is characteristic mostly of European writings and has led to an emphasis on party structure. The classic example is the work of Michels, who formulated the “iron law of oligarchy” to explain the triumph of the leaders’ ambitions for office over the membership’s revolutionary goals. Michels’ “iron law” is the product of his preoccupation with the German Social Democratic party and his disdainful neglect of the party system in which that organization developed. In Duverger’s work the acceptance of the party as the tool of its members is basic to the conclusion that the entire electorate will inevitably be incorporated into parties, thereby rendering the restricted cadre party archaic. But Duverger’s structural scheme is constructed at the expense of viable cadre parties, notably the American parties and the Radical party of the French Third Republic.
Perception of the party as primarily responsive to the electorate has been mostly a product of the Anglo-Saxon tradition. This perception has sometimes resulted in concern only with the total picture of party competition. Although McKenzie’s study of British parties (1955) does examine structure in relation to the parties’ office goals, Schumpeter (1942) and Downs (1957) ignore the internal workings of the party and concentrate on its place in a competitive system. Most American writing about parties is ambivalent. Although there is acceptance of the office-seeking definition of party, and the importance of party interplay, there is also the persistent longing to make the party’s function service to its membership. As a result, discussions of American parties are often fragmentary and confined to descriptions and criticisms of local machines and formal institutions.
The office-seeking definition of party allows us to recognize that in a competitive system the fundamental issue of the party’s function is a matter of value judgment and a constant source of organizational tension within parties. As complex organizations dealing with matters of importance, major political parties attract participants with a variety of motives. Some are interested in public policy; others enjoy the social life of the party; still others participate because they are ambitious for office or for the rewards that come from association with public officials. While these motives need be neither exclusive nor conflicting, the tension between policy interests and office interests within a political organization is increased within a competitive system. In a competitive system a party maximizes its chances for office by offering policy concessions to marginal voters and parties. What distinguishes the competitive political party from other forms of political organization, therefore, is that the question of electoral tactics is ever present. The competitive party need not always make decisions in tactical terms, and errors in tactics are always possible. But over the long run, if the party is to remain an effective competitor, the goal of office must be the crucial factor in the party’s decisions.
By choosing the office-seeking definition of party, therefore, we are able to distinguish the party from other forms of political organizations. The policy goals of parties change. The problem which has worried observers of parties since Ostrogorskii and Michels is not the rejection of office because of principle but the unseemly pursuit of office at all costs. The weight of the evidence indicates that the goal of office dominates in all parties whose expectations of office are high. In every European country where socialist parties have achieved a realistic chance at office, socialist doctrine has been tempered. However, the extent to which the American Federalists and Whigs were predominantly officeseeking organizations is evidenced by the speed of their disappearance once they lost their chance at office.
For purposes of analysis, the major implication of the office-seeking definition of party is that the party must be viewed in relation to the offices which it seeks to capture. In other words, basic to the understanding of political parties is the awareness of the structure of political opportunities within a given political system: the public offices available; the rules, formal and informal, for their attainment; and the attitudes of politicians and voters toward these offices. A revolutionary organization which rejects existing institutions, a social club, a debating society, a pressure or propaganda group with independent goals, can all be studied as discrete entities. But political parties compete to control a process which they did not establish and which could go on without them. Parties recruit leaders, educate the electorate, and even organize governments; but the state organizes elections. Therefore, everything of interest about a political party—its organization, its leaders, its policies, its income, and its capacity to attract voters—is affected by the structure of political opportunities within a given state.
To assert that party organization reflects the structure of political opportunities is not to deny that parties in turn affect the opportunity structure. Certainly the developments which took place in British party organization at the end of the nineteenth century greatly transformed the process of becoming prime minister. In the United States, parties early captured the Electoral College, which was to choose the president. At the same time, the differences between American and British parties reflect the different methods by which the two countries choose their chief executives. Clearly, parties and institutions affect each other.
It has been fashionable, since the advent of Marx, Weber, and Freud, to reject institutional explanations of the nature and activities of parties, in favor of social, cultural, or psychological explanations (e.g., Truman 1955). Recently, however, some students of parties have concluded that institutions are not epiphenomena but critical variables (e.g., Epstein 1964; Lipset in Ostrogorskii, 1964 edition).
The present discussion treats parties as responses to the ways states structure the opportunities for elective office. This approach is especially useful because the structure of opportunities provides a framework for the comparative analysis of parties and party systems. It enables us to compare the relative standings of parties as office-seeking organizations, in their own countries and across national boundaries.
The basic party unit is the nucleus, or the organization aimed at capturing a single office. The broader structure of the party emerges from the relations among nuclei. Electoral nuclei develop within the constituencies of elective offices. Governmental nuclei form around those offices which popularly elected officials fill from among themselves, as in the selection of legislative leaders or of the executive in parliamentary systems. The nuclei of a party are not necessarily discrete, either in personnel or in other resources. For purposes of analysis, however, we shall examine the basic units first and then the relationships among them.
It is important to note that the development of a nucleus rests upon the expectation that it will be able to capture office—if not immediately, then in the foreseeable future. Seldom are a party’s chances for office evenly divided among all its nuclei. A party will have some safe constituencies, some where it competes, and others where it has no chance at all to win elections. Indeed, a party’s durability will depend in great part upon safe constituencies, which assure some continuity in office regardless of the party’s general electoral fortunes.
At the same time, parties often run candidates in constituencies where they have no hopes of winning office; but we must clearly distinguish these efforts from the party activities. The major British parties contest many hopeless constituencies, in part to provide training and trial runs for candidates who hope to advance to more favorable constituencies, and in part to accommodate the national character of British elections, which means that the campaign in one constituency may well affect the results elsewhere. In the United States during the period of Democratic dominance of the South, the Republican party maintained a network of organizations which drew federal patronage and exerted influence within the national nominating convention. But this network was in effect an appendage of the presidential nucleus. Such efforts do not in themselves constitute party nuclei; they are significant only in relation to the organizations which have a chance to win office.
Cooperation among party nuclei is determined at least minimally by the structure of political opportunities, which gives substance to the party’s goals. The American and the British political systems present two very different opportunity structures and allow us to observe the varying impact of the opportunity structure upon party organization. The American presidential system, which operates within the federal framework and utilizes the popularly elected bicameral legislature, encourages party nuclei to act independently. In the United States thousands of officials are nominated and elected, many on different occasions. Even when and where the potential electorate is the same, different offices can attract substantially different numbers of voters. While there is little pressure for nuclear cooperation to control the independently elected executive, such pressure does exist in the legislature, where positions of leadership can be captured by multinuclear action.
In contrast, the British structure of opportunities imposes tighter relationships among party nuclei. The British parliamentary system provides for a single popularly elected chamber, all of whose members are elected together at a time determined by the dominant party leader. The electoral nuclei in turn define the governmental nucleus. In such a system the structure of political opportunities encourages multinuclear cooperation and party cohesion.
At the same time, within each political system the refinements of the opportunity structure modify its general impact upon party organization. Beyond the number of offices in the opportunity structure, there is the arrangement or hierarchy of offices. In the United States, the federal system provides no long-range career outlets for state officials. For the ambitious state governor there is only the national senate or the presidency and its surrounding administrative offices. Thus, the American federal system provides a restraint upon the independence of state party nuclei as well as pressure for cooperation between the state nuclei and the nuclei for national office. The Canadian federal system, on the other hand, provides for miniature parliaments in its provinces, which make long-range careers possible for provincial leaders. In turn, indigenous provincial parties which have little hope of developing into national organizations are able to flourish. The Social Credit party in Alberta is a good example.
In addition to the arrangement of offices, the opportunity structure imposes upon the party the procedures which it must follow to achieve office. In the United States the provisions for numerous independent elections encourage the independence of party nuclei. But there are also provisions for the sharing of electorates which facilitate multinuclear cooperation. Presidential candidates share the ballot with a host of other candidates for office; governors and senators frequently run for office at the same time; and United States representatives, along with lesser state officials, appeal to the subelectorates of the higher officials. These arrangements make possible the “coat-tail” effect, the possibility that the candidate for one office will influence the vote for other offices; they thereby create pressure for party nuclei to consider the activities and the candidates of other nuclei. On the other hand, in parliamentary systems which use the single-member district and avoid shared electorates, the basis for nuclear independence exists, as, for instance, in the French Third Republic.
Within parliamentary systems, however, the similar behavior and attitudes of different electorates can exert pressure for cooperation among nuclei. Despite the major advances in our knowledge of electoral behavior, the impact of the party upon the voter remains largely unknown. England appears to have a disciplined electorate, receptive to party activity. Such an electorate encourages maximum nuclear cooperation by making possible the electoral destruction of a recalcitrant leader who has been denied his party’s nomination. American and French electorates are rarely willing to behave in this manner. The Converse-Dupeux study (1962) shows that American voters identify with a party more closely than do French voters, but aggregate American election results show that in any given election enough voters will split their ballots to allow victories to both parties within the same constituency.
Undoubtedly the reasons for which voters support parties have important organizational consequences. The reasons are usually complex and include ethnic, religious, social, economic, ideological, and geographical factors in a variety of combinations (Alford 1963). Within a party, shifting voter support can produce factions or clusters of nuclei contending with each other because their electoral bases are divergent. In the French Fourth Republic the rising social and economic status of the Radical voter brought the new Radical leaders into conflict with Radicals of the Third Republic, who wished to retain the party’s prewar doctrinal concern for “the people.” In the United States the geographic support of the South for the Democratic party has increasingly caused conflict with those who support the party for economic and social reasons elsewhere.
Another source of organizational tension within parties is provided by the party system, or the competitive relationship of parties, and by the modifications which the structure of opportunities imposes upon the party system. Party systems vary in the number of parties which have a chance at office. Depending upon the structure of opportunities, party systems vary also in the distribution of the chances for office among party nuclei. More than one party must have a chance at office in order for the party system to be competitive. Yet the chances do not have to be, and often are not, equally distributed, even when there are only two real competitors. Prior to 1932 the American Republican party won most national elections; since 1932 the Democrats have been the dominant national party.
From the standpoint of nuclear cooperation, the uneven distribution of strength among nuclei affects the flow of organizational resources and causes tension within the party. We have assumed that party nuclei will exist only around those offices for which there is the chance of victory. But even for these offices the party’s chances can range from perfect to doubtful. Differences in electoral strength will produce nuclear organizations with different needs and resources. The nuclear organizations with the strongest electoral support are likely to be in the best position to command the other resources of organization, money, and personnel. At the same time, they are under the least pressure to compete with other parties for votes. The nuclei most subject to electoral pressures are those which face the strongest competition. Given the varied electoral needs of its nuclei, a party may well find itself with two or more nuclei or nuclear clusters framing different appeals to the voters.
Of course, a party is subject to additional tensions arising from its needs during the governmental phase. The needs of the governmental nucleus may well conflict with the needs of the electoral nuclei. In government by coalition the fate of competitors who are also allies becomes the concern of the governmental nucleus and imposes further strains upon the electoral nuclei of the dominant party.
The structure of political opportunities determines not only the extent of party organization but also its quality or content. Offices foster nuclear organization, and electoral procedures and practices provoke nuclear cooperation or dissension. Another consequence of the close relationship between party organization and the structure of opportunities is that the limits of party activity are unclear. Imbedded in the electoral and governmental process, the party in its activities is often indistinguishable from voters and pressure groups, on the one hand, and from government, on the other. In the search for office, parties go outside their ranks for leaders; they accept an occasional vote as the basis for party identification; they allow government bureaucrats rather than the party program to define party policy.
In describing the content of party activity, therefore, it is more useful to employ the concept of contribution to the party organization than the concept of membership. The term “membership” connotes an unwarranted clarity in the boundaries between the party and its environment, whereas, as has just been noted, even such critical contributions to the maintenance of the organization as voting, recruitment of candidates, and even candidacy itself may come from individuals not readily identifiable as formal members. Conceiving of party leadership, recruitment, money, and communications as contributions to party organization enables us to ask not only who makes them but also to focus on the flow of contributions from one nucleus to another. Thus, it is the transfer of contributions which creates the multinuclear party.
In party organization, leadership is a contribution of first importance. The fluctuating, ephemeral, and largely voluntary character of most contributions to party activity makes it most useful for someone to move people to participate, to make agreements with other leaders, and to bring together the materials of party combat.
Yet in parties, more than in any other type of formal organization, the official lines of authority are suspect, and there is always implicit the question of who is the “real” leader. This is true in great part because a political party is a leader-producing organization, and the ambitious men attracted to parties find it advantageous at least to seem to be playing a major role. If the “real” party leader is often difficult to locate, it is because many competitors for leadership have a stake in keeping it so.
The contribution of leadership is also difficult to assign because there are two broad categories of party leaders, which in fact may or may not merge. There are the public leaders, men who also represent the party as its candidates for public office, and there are the associational leaders, men whose office is limited to the party organization. Often, but by no means always, there is no sharp distinction in personnel between the two categories; men move from one to the other or hold both public and party office concurrently.
The difficulties for analysis presented by these overlapping categories pertain especially to the American system. In American parties there is no formal hierarchy of authority or delineation of functions. The committees, chairmen, and conventions which range from the locality to the nation have no consistent authoritative relation to each other or to the parties’ officeholders and nominees. At the same time, the actual conditions for office-holding in the United States, numerous public offices with a high rate of turnover, favor the distinction in leadership. This situation contrasts with that of the parliamentary system, where the availability to the party of “safe” public offices or seats almost always assures the merger of both categories of leaders. Thus, whatever the original source of the party leader’s strength, organization work or officeholding, party and public leadership easily combine and provide the necessary continuity in an organization where most activity is transitory. [SeeLeadership, article onPolitical Aspects.]
Recruitment and nomination
The major task facing the leaders of the nuclear organization is the choice of its candidate for office. Who contributes to this task, and how is it accomplished? There are two aspects to the process. One is the recruitment—and discouragement—of candidates for the nomination. The second is the choice of the nominee from among the active seekers. The latter aspect is more overt and is normally surrounded by rules of procedure which serve to make the nomination authoritative. But the recruitment process is equally critical in defining the choices which can be made within the party.
Although there are a few empirical studies of the recruitment process, those which exist place the burden of the task with the aspirants themselves. There is little evidence from which to conjure the picture of an organization actively seeking candidates. Studies of candidates for lesser offices, such as the state legislatures of the United States (Wahlke et al. 1962), show relatively few who perceive themselves as recruited by party organizations. A high proportion of politicians come from families with active political experience. Party organizations give evidence of actively recruiting candidates for offices where there is little expectation of victory, but when the nomination is of value, the nuclear organization normally responds to choices presented to it by men actively seeking the nomination, who advance along various career lines. [SeePolitical recruitment and careers.]
The opportunity structure, therefore, in defining the paths of advancement, has much to do with the recruitment of party candidates. The loosely ordered American system provides multiple lines of advancement, so that parties have diverse sources for candidates. Parliamentary systems, on the other hand, channel most recruitment through legislative offices.
The contribution of nomination, which follows recruitment, to party organization is equally complex and often obscure. In the United States, reforms such as the convention and the direct primary have clarified some aspects, but some decisive activity still takes place in private. Much of the organizational activity in a nomination consists of discouraging other men’s ambitions and, if the organization is to succeed at the polls, tying their ambitions to those of the candidate. Therefore, the nomination ideally combines a maximum of discouragement of all but one of the ambitious with a minimum of public disgrace for the others.
The discouragement process takes place within a set of rules according to which one man gains the designation “nominee” of the party. In a stabilized party system, capture of this label is important, and the state frequently acts to assure that only candidates selected according to the rules can appear in the general election under that label. This does not prevent officeseekers, however, from organizing their own campaigns to capture the label. As with recruitment, nomination is less a case of an organization’s selecting candidates according to qualifications than it is of providing the framework within which they contest for the nomination.
The actual choice of a candidate may involve numerous individuals, and, although parties normally establish their own procedures for nominations, the state may intervene. In the United States the direct primary imposes upon the party maximum participation in the nominating process. This imposition can produce irrational consequences for the party, for there is no assurance that the primary voter is imbued with the officeseeking drive which works in favor of the most effective electoral candidate. As a result, pre-primary conventions have developed in several American states as a means of controlling the nomination. Party conventions are, in fact, among the most common methods of nomination. They are usually made up of delegates apportioned according to party membership. In Norway the state intervenes in the convention process by paying the expenses of the nominating conventions if they follow prescribed methods (Valen & Katz 1964). Nominations by party caucus or by constituency committee are also common.
Whether or not the state plays a role, the ability to control a nomination is one of the few observable tests of strength within a party. As a result, party leaders usually avoid taking the test unless forced to, or if they do, they seek to support a certain winner. The problem is particularly germane to the delicate relationship between the nuclei for higher offices and lesser offices. For this reason, open intervention by party leaders in the nomination contests for lesser offices is rare. Even in the disciplined British parties, control of constituency nominations by the central office is far less than has been assumed by those who would infer discipline as radiating from the center (Ranney 1965). Although empirical studies of nominating procedures are rare, the evidence seems to be that constituency nuclei are in control in most systems.
Whatever the contribution to nomination, whether it be the discouragement of ambitions or their open defeat, it is dysfunctional if it leaves the nominee too weak to win the general election. Thus, although not all of the actors may be so motivated, all of the candidates who do hold office ambitions will be under a common restraint at least not to oppose, and preferably to support, the nominee. The most effective control the party has is the defeated candidate’s own hope for preferment. It is peculiar to party organization that at one and the same time it encourages men to open conflict and then forces them to curb animosity to achieve a subsequent goal. Much of the two-facedness of political activity as it appears to external observers comes from this inherent characteristic of nuclear organization. That burying the hatchet is not easy is evident from the frequency with which it fails or, if done, is done grudgingly. Still, the constant process of personal advancement through organized conflict either repels or weeds out personalities incapable of sustaining the tension. In his earlier work, Lasswell (1948) emphasized the dominance of power drives in the personalities of political leaders. More recently, he has come to note that such persons do not rise to the top in democracies (Lasswell 1954; see also Lane 1959, pp. 124-128). The pure power seeker, as distinct from the man with office ambitions who sees office as instrumental to other goals, can probably find more satisfactory outlets in areas other than politics. An important aspect of the contribution of nomination is that self-interest must control itself and create cooperation out of what might be the most divisive and corroding of ambitions, the drive for political power. [SeePersonality, Political.]
Once the nuclear organization chooses its candidate, it must seek support for reasons other than its power drives. Although it is known that the voters’ response to parties involves much more than a choice between policy stances, there is a good deal of issue content in electoral decisions, and to succeed, a party must devise a program or platform. The position of the party as seen by both its leaders and the voters is only partially covered by the formal “platform.” Therefore, one must ask who contributes the issues. To a great extent, the process of leader selection decides the policies for which the party will stand. But beyond the party leader there are “leading citizens,” specialists in public affairs, scholars, and journalists on whom the party can call for contributions in devising policy.
The intellectuals’ function, however, is severely limited by the way in which competitive governing parties are forced to devise policy. A party in control of government must make choices, but its range of alternatives is usually narrow. The party is restricted by time, internal differences, the administrative apparatus of government, and other relevant elements of the political system, including foreign relations. A drastic reformulation of public policy, therefore, requires either massive continuous support or the elimination of the competitive rules. Otherwise, the party in government finds its position defined largely by the way it responds to issues which arise during its tenure. Thus, the formulators of the party’s position must consist largely of its elected leaders and its administrative corps. Although parties out of power presumably have greater freedom in defining their stands, they also must respond to the government and to the issues as they arise, and allow public officials the major role.
Parties which govern in coalitions, however, can more easily develop a refined ideological position than can parties which must govern alone. The extent to which a party can be held responsible for government affects its ability to retain or to define a stand independent of the government. In coalitions where the governing responsibility is shared or obscure, parties can govern and retain a doctrine which has little relation to the governing experience.
As in any organization, communication is critical to all aspects of party activity. Very little, however, is known about this contribution. Historically, newspapers were closely associated with the growth of parties, and party newspapers are still common, especially in Europe. But there have also developed independent news media which collect and disseminate information vital to the party. For example, public opinion polls provide information on voter attitudes far more reliable than the reports of party workers; yet more often than not, the polls are conducted and reported by media outside the party’s control. An important party activity, therefore, consists of affecting the information which independent news media report. [SeeCommunication, Political.]
The nuclear organization can use many technical contributions. Periodic contact with the electorate, quite apart from testing their sentiment, is useful in keeping track of the party’s supporters. Every campaign creates a surge of possible tasks for which volunteers must be recruited. Each party affair, rally, or coffee hour involves bringing together as many people as possible to give the appearance that the party enjoys popular support. On election day itself, there are many things which an organization can do to make sure that the committed voter gets to the polls and has his vote tallied. The British parties have career staffs which assign professional agents to constituencies; in this sense the agent is independent of the candidate. Characteristically, in American parties technical contributions are made by small permanent staffs which blossom during the campaign. Such professional staff members exhibit no tendency to develop a sense of direction independent of party leaders. They gain influence in their relations with associational or public leaders of the party, not as a separate bureaucracy.
Since useful services are by no means all voluntary, money is a prime contribution to the nuclear organization. Money represents “instant” organization by helping to satisfy the periodic need to expand activities rapidly. The active nucleus, therefore, puts much effort into gathering money.
Money may come in small amounts from many people. In addition to providing funds, the small contribution also serves to reinforce popular identification with the party. The small contribution may also be used to clarify party membership when it is regularized in the form of dues. This procedure is common in European parties and is not unknown in major American state organizations, e.g., in Wisconsin and Michigan. Fear of the corrupting influence of money has led to the formulation of legal restraints in several countries on the amount and sources of money, as well as on the amounts that can be spent in a campaign. Nevertheless, the consequences of money, either in directing party policy or in affecting elections, are hard to evaluate. [Seepolitical financing.]
It is difficult to assess the impact of any of the connective contributions upon the fate of the political party or the outcome of the electoral process. The effects of the activities of parties are surrounded by an even higher level of uncertainty than those of other organizations. The only test of effectiveness is whether or not a party wins office. But it is seldom clear whether victory or defeat is due to the party’s decisions, the decisions of its opponents, or even to the decision of some foreign government over which the party can have no possible means of control.
Nevertheless, as long as a party remains a viable competitor within the party system, it attracts resources. Any party with a reasonable chance at office will attract those with an interest in office. The greater the expectations of victory, the more attractive the party becomes to individuals and to interest groups concerned with the actions of government. Thus, dominant parties often become holding companies for competing political factions and interest groups; the Indian Congress party, the Israeli Mapai, and the Democratic party in the American South are good examples.
On the other hand, parties whose chances for office are small tend to attract those whose goals are not the goal of office. The disaffected or groups whose limited objectives can be attained through the influence of a minor organization gravitate toward minority parties. Indeed, the major danger to minority parties is their attraction for participants who have a vested interest in preserving a party’s minority status. In order for a political organization to become or to remain a true political party, it must respond or adapt to the structure of political opportunities.
Joseph A. Schlesinger
[See alsoElections; Interest groups; Legislation, article onlegislative behavior; Nonpartisanship; Political machines; Political participation; Representation, article onrepresentational behavior; Voluntary associations; Voting; and the biographies ofKey; Lowell; Michels; Ostrogorskil]
Alford, Robert R. 1963 Party and Society: The Anglo-American Democracies. Chicago: Rand McNally.
Converse, Philip E.; and Dupeux, Georges 1962 Politicization of the Electorate in France and the United States. Public Opinion Quarterly 26:1-23.
Dahl, Robert A. (editor) 1966 Political Oppositions in Western Democracies. New Haven: Yale Univ. Press.
Downs, Anthony 1957 An Economic Theory of Democracy. New York: Harper.
Duverger, Maurice (1951) 1962 Political Parties: Their Organization and Activity in the Modern State. 2d English ed., rev. New York: Wiley; London: Methuen. → First published in French.
Eldersveld, Samuel J. 1964 Political Parties: A Behavioral Analysis. Chicago: Rand McNally.
Engelmann, Frederick C. 1957 A Critique of Recent Writings on Political Parties. Journal of Politics 19: 423-440.
Epstein, Leon D. 1964 A Comparative Study of Canadian Parties. American Political Science Review 58: 46-59.
Key, V. O. Jr. (1942) 1964 Politics, Parties, and Pressure Groups. 5th ed. New York: Crowell.
Lane, Robert E. 1959 Political Life: Why People Get Involved in Politics. Glencoe, III.: Free Press. → A paperback edition was published in 1965.
LaPalombara, Joseph G.; and Wiener, Myron (editors) 1966 Political Parties and Political Development. Princeton Univ. Press.
Lasswell, Harold D. 1948 Power and Personality. New York: Norton.
Lasswell, Harold D. 1954 The Selective Effect of Personality on Political Participation. Pages 197-225 in Richard Christie and Marie Jahoda (editors), Studies in the Scope and Method of The Authoritarian Personality. Glencoe, III.: Free Press.
McDonald, Neil A. (1955) 1961 The Study of Political Parties. New York: Random House.
McKenzie, Robert T. (1955) 1963 British Political Parties: The Distribution of Power Within the Conservative and Labour Parties. 2d ed. London: Heinemann; New York: St. Martins. → A paperback edition was published in 1964 by Praeger.
Michels, Robert (1911) 1959 Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy. New York: Dover. → First published as Zur Soziologie des Parteiwesens in der modemen Demokratie. A paperback edition was published in 1962 by Collier.
Neumann, Sigmund (editor) 1956 Modern Political Parties: Approaches to Comparative Politics. Univ. of Chicago Press.
Ostrogorskii, Moisei I. 1902 Democracy and the Organization of Political Parties. 2 vols. London and New York: Macmillan. → An abridged edition was published in 1964 by Quadrangle Books. See S. M. Lipset’s introduction to the abridged edition.
Ranney, Austin 1965 Pathways to Parliament: Candidate Selection in Britain. Madison: Univ. of Wisconsin Press.
Schattschneider, Elmer E. (1942) 1960 Party Government. New York: Holt.
Schlesinger, Joseph A. 1965 Political Party Organization. Pages 764-801 in James G. March (editor), Handbook of Organizations. Chicago: Rand McNally.
Schumpeter, Joseph A. (1942) 1950 Capitalism, Socialism, and Democracy. 3d ed. New York: Harper; London: Allen & Unwin. → A paperback edition was published by Harper in 1962.
Truman, David B. (1955) 1962 Federalism and the Party System. Pages 115-136 in Arthur W. Macmahon (editor), Federalism: Mature and Emergent. New York: Russell. → A paperback edition was published in 1962 by Harper.
Valen, Henry; and Katz, Daniel 1964 Political Parties in Norway: A Community Study. Oslo: Universitets-forlaget.
Wahlke, John et al. 1962 The Legislative System: Explorations in Legislative Behavior. New York: Wiley.
The study of political parties has two major facets. One involves their characteristics as discrete entities—for example, their various social bases, histories, goals and appeals, formal organizations, and actual power structures. This is the study of party “units,” a subject discussed in the preceding article. The other involves the competitive interaction patterns among party units or, as Duverger puts it, “the forms and modes of their coexistence” (1951, p. 203 in 1954 edition). This is the study of party “systems.”
Scope and definition
Party units and party systems are so closely related that one cannot really deal adequately with either without reference to the other. Interactions among parties obviously are affected by the nature of the interacting units, while the interactions in turn have repercussions on the units themselves. These relations between units and systems will, of necessity, be alluded to frequently in this article. However, only party systems will be treated systematically here, party units being referred to only in passing, literally by allusion. This is not simply a matter of arbitrarily limiting the scope of the discussion, for, to an extent, party systems both can and must be treated independently of the units that constitute them. In no sense are they merely sums of their parts. The interactions of party units invariably have attributes not fully stated in the description of the units, and while they are obviously much dependent on individual party characteristics, they also reflect other factors, not least the broader settings—society, political system, general structure of political competition—within which parties exist.
Party units and party systems constitute separable subjects also in two other senses. First, defining party systems as competitive interaction patterns among parties implies that political systems could conceivably have parties without having party systems. Manifest cases in point are those monolithic parties that serve regimes as instruments of control, coordination, and permeation, rather than being structures of genuine political competition. Second, a party might stand outside of the party system of a polity even where such a system exists. A party will do so if it does not competitively interact with other parties in a substantial way—for example, if it is a very minor party exercising very little political influence, or if it is essentially a pressure group (or revolutionary organization) that goes through the motions of party competition only to publicize or disguise itself or to get advantages that party status often confers (such as free railway travel and free radio or television time).
One could, of course, make participation in a party system a condition of including an organization in the universe of parties in the first place, regardless of what the organization calls itself. One probably should do so if the concept of party is to delimit a coherent theoretical subject—that is, a set of cases similar enough to one another, yet different enough from others, to permit the formulation of informative generalizations that pertain to all of the cases and some that pertain to them alone. “False” parties, needless to say, must be studied but under rubrics other than that of parties. They may form special theoretical universes of their own, and they may belong to larger universes to which genuine parties, as well as many other structures, also belong—for example, the very large world of structures for seeking and maintaining political power or the still larger universe of human organizations.
To delimit the concept of party systems precisely, it is necessary to concentrate on, and to be more specific about, the patterns of competition that characterize the interactions of their units. Political competition takes, after all, a great many forms. The protagonists in a civil war are certainly engaged in a kind of political competition; so are lobbyists maneuvering for minor advantages, legislators attempting to constitute majorities or to attain sensitive positions in the law-making process, bureaucrats promoting pet projects, and mobs demonstrating or rioting in the streets. What sort of political competition, then, is peculiar to party systems? Manifestly, it is electoral competition: overt participation in open, formalized, genuine elections. Not only is this a process that cannot possibly be confused with the activities of other political participants; not only does focusing upon it promise to yield a homogeneous and highly differentiated universe of organizations for theoretical study; but the association of party competition with elections happens also to be implicit in conventional usage, at least since the time when the parliamentary combinations of representative oligarchy, under pressure of democratization, first developed extraparliamentary machines to provide them with new sources of power.
When one speaks of party systems, then, one speaks at bottom of interactions in a particular kind of political process. That process, however, is closely tied to certain political structures and functions. The structures are representative governments (although a party system may well grow out of organizations operating in other contexts). The functions are those important ones that elections perform in representative systems: political recruitment and relating the recruitment of leaders to the articulation and aggregation of political goals—in Neumann’s words, “organizing the chaotic public will” (1956, p. 397) in order to shape choices that elections can resolve. The aggregation performed by party systems should not, however, be conceived solely as the internal process of resolving conflicts between particular men and measures that always precedes electoral competition; the whole history of parties, which gradually turns their labels into “images,” is involved as well.
Aggregation is necessary not only for the sake of electors but also to provide a foundation for authority in representative, especially parliamentary, regimes. Effective authority in a democracy must always rest on “solid masses of steady votes” in the representative assembly, and these must perforce be provided by organizations that form a persistent legislative will out of the myriad possible shifting constellations that might arise in representative assemblies. But while in both cases the formation of collective will is involved, the two functions of aggregation—to represent bodies of opinions and to support authority—may very well conflict. The more faithfully divergent opinions are represented, the more likely it is that authority will be insufficiently supported; the reverse is true as well: the maximum support of authority would come from a party system so overarchingly aggregative as to offer no choice at all. In the final analysis, then, aggregation involves the search for an optimum between the conflicting values of unity and divergence, the one to support power, the other to provide choice.
However vital the functions that party systems perform, and however closely they may be tied to the process of electoral competition, it is nevertheless inadvisable to make these functions central to a delimitation of party systems. This is for two reasons. One is that many substantially different structures, in the same and different political systems, recruit leaders and aggregate preferences. In representative systems, for example, aggregation of preferences is carried out not only within the legislative process but also by any moderately complex pressure group and, increasingly nowadays, by bureaucrats, who consult among themselves and with others both in and out of government. The same holds true for recruitment. The second reason is that recruitment and aggregation are certainly not the only functions of party systems. The units that constitute them serve a host of other ends. Parties continuously provide political education and inculcate political skills. Above all they do so internally to their militants, by such means as discussions in branches, study courses (most typical of Marxist parties), conferences and summer schools (as in Britain), the publication and dissemination of research reports and handbooks, and training schools (like those of the Neo-Destour in Tunisia); most obviously, they provide, in their internal structure, facilities for the acquisition of political skills of every kind. In many countries they serve also as frameworks for their members’ social life, sometimes indeed all-embracing frameworks. This is the case mainly in some non-Western countries, where parties often furnish a badly needed cement for men suddenly cut off from their traditional ties and often perform, especially in urban areas, some of the functions discharged elsewhere by traditional structures. Sometimes this involves rather surprising activities; Hodgkin cites the example of a section of an African party which combines the activities of “emancipating young women from family influences, assisting the process of matrimonial selection, providing on a contributory basis marriage and maternity benefits (including perfume and layettes for the newborn), preserving the Dioula tribal spirit, and running an orchestra” (1962, pp. 144-145). Such activities are characteristic of sham parties that serve as adjuncts to dictatorial regimes, but one can also find them in less dubious cases; Scandinavian parties, for instance, often sponsor boy-scout groups and summer camps, provide adult education courses, and—less peculiarly—publish newspapers and journals. In some cases, the party system even directly supplies governmental services, although this is unusual and occurs mainly in new or newly independent countries where governmental frameworks have not yet hardened. More commonly, party systems perform the functions of directly controlling the executive, most often in colonial systems that permit party competition but sometimes also in highly developed countries like Britain, where informal legislative committees probably do more to call the executive to account than formal parliamentary institutions. Finally, party systems, quite apart from their constituent units, may serve perhaps the most basic function that must be performed in any sound polity: that of giving the polity an identity above the separate realities of its constituent units. The regular interplay of competing electoral forces may in some incohesive new nations be the only central point to which loyalties that rise above parochial attachments may be tied.
To summarize: The subject of “party systems” is concerned with the interaction patterns among significant and genuine electoral organizations in representative governments—governments in which such systems serve pre-eminently (whether well or badly) the functions of providing a basis for effective authority and for defining choices that can be resolved by electoral processes.
The study of party systems
Until very recently studies of party systems almost invariably consisted simply of discussions of the units constituting them and dealt chiefly with particular countries, while little attention was paid to general system characteristics. This applies particularly to the early textbook and monographic literature on parties. But it even applies to comparative and theoretical works like those of Weber, Ostrogorskii, and Michels, which, possibly excepting some highly value-laden passages in Ostrogorskii (1902), deal entirely with party units. Today, despite some small advances toward the special theoretical study of party systems, it is still this sort of literature that provides the main underpinning for the subject. No descriptive concepts for characterizing party systems have been settled upon (although various ones are used), and no generalizations about such systems have been adequately tested (although many have been proposed). Hence, no settled body of ordered knowledge can be reported in this article, and much space must be devoted to quarrels, gropings, uncertainties, and tentative suggestions.
The first works that accord party-system characteristics any sort of explicit treatment appeared during the late interwar and early postwar years. During this period scholars became increasingly aware of the links between patterns of party competition and the performance of representative governments. Representative government in the interwar years was in a state of crisis, which often seemed to stem from the party systems, particularly the fragmentation and ideological intensity of party competition in certain countries. Critical fire was directed at representative systems because the facts of party life—bosses, oligarchies, cliques, interest-mongering, unscrupulous electioneering, insensitivity to opinion or real problems—seemed to make a mockery of the ideals of liberal government. Early writings on party systems, consequently, concentrated heavily on remedies and prophylaxis: on how party systems could be made more democratic, more responsible, less fragmented, and less dogmatic, so that antidemocratic criticism might be disarmed and crises averted. Typical of this genre of writings is F. A. Hermens’ Democracy or Anarchy? (1941), a book which argues with great vigor that proportional representation produces party systems inimical to successful democracy and recommends the single-member simple-majority system as a cure for practically all democratic ills. An even more famous specimen is the special report, Toward a More Responsible Two-party System, produced by a committee on political parties of the American Political Science Association (1950), which makes an intimidatingly large series of interrelated proposals designed to produce parties “able to bring forth programs to which they commit themselves” and possessing “sufficient internal cohesion to carry out these programs,” parties able to provide both integrated government and an integrated opposition, parties more able to resist pressure and to formulate the public interest, parties responsive to public will, and party leaders responsive to party members.
This premature but useful emphasis on therapy was followed, in the 1950s, by some first efforts to systematize our knowledge and understanding of party systems: their anatomy and physiology, so to speak; how they respond to conditioning factors; and how in turn they affect their larger settings— all matters that therapy presupposes to be known.
The main pioneering work in the systematic empirical study of party systems is Maurice Duverger’s Political Parties (1951). Duverger devotes about half of that work to party units (his term is party structure) and half to party systems, making him, to my knowledge, the first to raise the distinction between units and system to a highly explicit level. In both sections of the book he tries to construct general descriptive categories for characterizing and classifying cases, as well as empirical generalizations about the cases and types.
Duverger’s book does not pretend to furnish a finished theory; how could it as a first exploration? On the contrary, it is hedged at every turn with caveats and disclaimers. His theories, according to Duverger himself, are “vague, conjectural and approximate” and intended merely to raise issues and provide tools for further inquiry. Few even of Duverger’s more mordant critics seem to have had greater doubts about his work than Duverger himself; almost everything he says is presented as merely suggestive, indecisive, tentative, superficial, vague, preliminary, incomplete, and imprecise— the adjectives again being his.
Unfortunately, few other scholars have so far tried to go beyond Duverger’s first attempt. A symposium volume, Modern Political Parties (Neumann 1956), did provide some well-summarized data on many national party systems, as well as a few well-formulated questions and concepts in its introduction and conclusion; but it too remained very much on the level of preliminary concept formation and very tentative hypothesizing. Nor do works like Hodgkin’s African Political Parties (1962) or the many textbook sketches of party systems in non-Western areas and countries go much further. They add a great deal to our data, but, as to theory, they either present ad hoc interpretations or apply uncritically Duverger’s formulations.
However, the main lines of inquiry to be pursued in the general study of party systems have by now emerged. First and foremost, one needs adequate descriptive concepts for characterizing, distinguishing, and classifying party systems; without such concepts, theorizing in other forms can hardly proceed. Second, there is the problem of finding the determinants of the various forms of party systems. Third, there is the problem of the functional impacts of different party systems on the larger political and social systems in which they exist.
Variations in party systems
Number of interacting units
In thinking of differences in the patterns of interaction among party units, one variable has always come first to mind: the number of interacting units, or, in other words, the extent to which party systems are consolidated or fragmented. In fact, this is not just the foremost but the sole dimension on which most existing studies distinguish party systems. Much the most frequently used typology of party systems has been a purely numerical division into three classes: oneparty systems, two-party systems, and multiparty systems; with, occasionally, one-party systems subdivided, particularly to distinguish dictatorial from representative structures. Practically all generalizations about the determinants and functional impacts of party systems are couched in terms of this typology. Examples of such generalizations are propositions to the effect that the process of nation making, or a commitment to rapid social and economic development, leads to and/or requires one-party systems, that the single-ballot simple-majority system of elections leads to two-party systems, and that multiparty systems make parliamentary governments unstable and ineffective.
No doubt, the number of parties is a basic and obvious dimension along which party systems vary. But despite appearances, it is not a dimension that lends itself to making distinctions clearly and easily. In fact, the numerical typology has seemed to act as a positive barrier to the discovery of trustworthy hypotheses. Writings on party systems are very rich in empirical generalizations yet extraordinarily poor in any that have withstood rigorous testing; so often does reality seem to confound propositions based on the traditional typology, that one has reason to think that the fault lies in the typology itself.
The fundamental difficulty with the established numerical typology involves its application to concrete cases: the typology cannot sensibly be taken to mean literally what it says, and it is difficult to use in a nonliteral sense. Take, to begin with, the concept of a one-party system. Strictly speaking, there can be no such thing. If party systems involve interactions among party units in the process of electoral competition, then the idea of a one-party system is logically absurd, for one cannot have a competition or an interaction with only one actor. It follows that if the concept is to be used, there must be specific conditions under which the existence of all parties but one should be ignored for purposes of generalization, even in genuinely competitive systems, or under which a single party without competitors is nevertheless considered to be genuinely engaged in electoral competition. This in fact is the procedure in most writings on one-party systems. Duverger, for example, includes in the category—like many other writers—the American South as well as Nazi Germany, Fascist Italy, Portugal under Salazar, and Turkey from 1923 to 1950 (1951, p. 275 in 1954 edition). Coleman has listed as one-party systems in Africa (as of 1959) northern and eastern Nigeria, Ghana, Somalia, and the Federation of Rhodesia and Nyasaland, despite the fact that in each case more than one party actually contested elections and managed to win representation. Blanksten similarly regards Mexico as a one-party system (Almond & Coleman 1960).
These writings raise the essential problem of finding sensible but unconventional criteria for including and excluding parties in a party system. In much of the literature, however, such criteria are not stated at all, and parties are ignored or taken into account practically at will, while works that do state their criteria usually arrive at them on an ad hoc, arbitrary basis. The result is that the universe of one-party systems depicted in the literature, sometimes even in a single work, is extremely heterogeneous, including such very diverse phenomena as monolithic dictatorial parties that suppress all opposition (for example, totalitarian “parties”), parties that do not use a high degree of coercion but nevertheless have a literal monopoly (for example, the True Whig party of Liberia or the African Democratic Rally in the Ivory Coast, Niger, and the Sudanese Republic), “unified” nationalist and postnationalist movements that consist of many, sometimes not very cooperative, groups (for example, the Burmese Anti-Fascist People’s Freedom League until 1958), and parties that simply greatly overshadow their competitors (like the Democrats in much of the American South, or the Northern People’s Congress in Nigeria, or the Congress party in India). That such heterogeneous cases do not yield valid generalizations is hardly surprising.
Much more than a straightforward count of organizations calling themselves parties is necessary to make sense of the category “one-party systems”; and even excluding all but genuine electoral organizations does not quite overcome the difficulty. The same is true of two-party systems. The category is not logically absurd, but two-party systems in the literal sense are extremely rare, if indeed they have ever existed. Not even Great Britain, usually considered the paragon of bipartism, really fills the bill. For one thing, minor parties have always operated alongside major ones in Britain—parties like the Liberals of today, the early Labour party, the Irish Nationalists, the Liberal Unionists, National Labour, the Independent Labour party, the Communists, the Fascists. Furthermore, in the 77 years between 1885 and 1962, the era of modern parties and elections, Britain has had only 43 years of straightforward one-party majority government. It can be argued as well that the major British parties function cohesively only at election time, when much less is decided than is generally believed, and that at other times they are simply combinations of factions, despite their discipline in parliamentary voting, much in the manner of the major American parties. That at any rate seems a tenable position in regard to certain party functions, for example, the formulation of interests. Talking about two-party systems, therefore, also requires one to choose what units to include in and exclude from a party system, and such choices, if made without sensible and explicit criteria, will once again lump together a very heterogeneous set of cases, difficult or impossible to generalize about. Take but one example: the familiar proposition that bipartism promotes stable government will almost certainly founder if tested against both British and postwar Italian experience. The reason is that in Italy the opposing forces (until the recent, still inconclusive “breakthrough to the left”) were constituted very differently from the British. One force, the Christian Democrats, has consisted of factions representing a far wider spectrum of far more antagonistic positions than either British party, while the other, the Communist-Socialist alliance (formal until 1952, thereafter informal and to some extent uncertain—but little more so than the alliance of forces in the Christian Democratic party), has been substantially an antidemocratic force, the existence of which, one may suspect, is all that has bound together its opponents. (There are, of course, other parties in Italy, but all extremely weak; in the 1958 election no other party got as much as 7 per cent of the vote.)
The concept of a multiparty system also poses a logical difficulty. Again, if taken literally, it is a classification that does not classify, for virtually all party systems fall under it. Even if we departed somewhat from literal meaning (for example, by discounting very minor or vastly overshadowed forces), the concept would still describe a staggering variety of cases. On one extreme would be countries like Australia with its three major parties (the Australian Labour, Liberal, and Country parties) or Canada with its two large and two smaller parties (the Liberals, Progressive Conservatives, the New Democratic party, and Social Credit); on the other extreme would be countries like Indonesia which has had an astonishingly large number of electioneering groups, more than forty having contested one election. The classic multiparty systems, so often held up to our dismay, like Weimar Germany (which generally had about nine or ten significant electoral forces) or the French Fourth Republic (which had six or seven), would only fall around the middle of the spectrum. Moreover, the category would comprise some systems in which single parties can win and have at times won outright majorities (for example, Sweden and Norway); systems in which single parties consistently come close to that point (Denmark and Norway); systems in which no party even approaches a majority (the Weimar Republic, the Fourth Republic); systems in which the distribution of party strength is fairly even and systems in which it is remarkably skewed; systems in which effective party alliances are easy to form and maintain and systems in which they are hard to form and harder to preserve; systems rent by dogmatic ideological conflicts and systems divided by less intense, sometimes picayune, policy disputes or by regional, religious, ethnic, linguistic, or socioeconomic divisions; systems in which party cooperation is virtually nonexistent and systems in which parties collaborate on the parliamentary level almost as much as distinct party factions in some so-called two-party systems collaborate in elections. Under these circumstances, again, a lack of tested generalizations is just what one would expect.
The upshot is that if the categories one-party, two-party, and multiparty systems are taken at face value they yield a classificatory scheme that simply does not classify. Taken literally, all but a very few idiosyncratic cases fall into the third category, since the first is logically untenable and the second is empirically nearly empty. And if the categories are not taken literally, criteria for counting and discounting units must be specified, justified, and consistently applied, which is nowhere done and is very difficult to do. In addition, however, a further difficulty should be mentioned: The propensity for failure of generalizations about party systems that are couched in the traditional typology may be due to the fact that many factors independent of numbers shape the impacts of such systems and reflect the conditions that impinge on them; among these factors the number of party units may be only a minor source of difficulty.
Constructing an adequate set of concepts for characterizing and classifying party systems thus clearly requires (a) technical precision about establishing the number of competing party units so that the numerical dimension can be precisely used and (b) proper attention to other dimensions of variability.
Distributions of strength
The two requirements just mentioned can be met to a considerable extent by distinguishing among party strengths. In regard to the number of party units in party systems, measures of strength can help one identify the number of genuinely competitive units, those strong enough to play a role worth noting in electoral competition and in any combinative processes subsequent to elections. If it is sensible to characterize a game by the number of players, then surely it is also sensible to count only those players who really make a difference. Beyond that, persistent (that is, patterned) differences in distributions of strength among party units merit consideration as a separate dimension in characterizing and classifying party systems.
Party strength refers in the first instance, and most obviously, to quantities: the number of votes or representative offices that a party is able to obtain. One speaks of a two-party system in Great Britain chiefly because of the extreme quantitative weakness of all but two parties in that country and of a one-party system in northern Nigeria because of the overwhelming strength of the ruling Northern People’s Congress, not because it has no competitors at all. Duverger, in this connection, classifies parties into four types: (1) parties with a majority bent, that is, capable of commanding an absolute parliamentary majority; (2) major parties, which cannot normally command absolute majorities but can govern with some external support; (3) medium parties, which can participate in government only by playing a subordinate role in coalitions and which cannot get an opposition to coalesce around them; and (4) minor parties, which are so small as to be incapable of playing any significant role at all in government or opposition (1951, pp. 283-290 in 1954 edition). One might add the concept of the dominant party—which Duverger uses in another connection and different sense—to denote a very strong party not confronted by any significant opposition.
Party strength, however, is not a matter of quantities pure and simple; note that Duverger’s criteria for classifying parties according to their strength are not explicitly numerical but are concerned with the actual or potential role of parties in government and opposition. While that role is certainly determined substantially by the size of their representation and electoral support, it may also be affected by other factors. Some of these are themselves quantitative. It is necessary, for example, to consider the size of a party’s competitors. One must also consider the distribution of numerical strength among the larger parties; in some cases that distribution might allow small independent parties persistently, not just under unusual circumstances, to call the tune in the formation of governments and in decision making. In addition, a small political force could be important because of its sheer capacity for entering coalitions of many kinds in rather fragmented party systems (that is to say, because of its opportunistic desire for offices) or because its location in the political spectrum makes it somehow indispensable in coalition making, regardless of whether its leaders are opportunists or men of principle. The Weimar Republic’s Center party thus played a role in government far greater than its size might lead one to suppose (its vote varied between 14 and 19 per cent) precisely because it was truly a center party in a very wide spectrum of parties and, by Weimar standards, a party conspicuously lacking dogmatic principles or monolithic class support. The case of the Radical Socialists in France is even more blatant. Never a massive force in sheer numbers they were nevertheless the dominant party in the Third Republic, the payoff value of its representation (as measured by access to offices) being greatly enhanced by the quantitative weakness of, and deep divisions among, other parties, by its center position, and by its leaders’ quite extraordinary appetite for office. The Radicals usually controlled important positions, especially the Ministry of the Interior and the premiership itself, and this regardless of electoral fluctuations.
Qualitative factors of this sort can also make a party less of a force than its size might indicate. Above all, a party may be weakened by lack of internal cohesion. Indeed, the nominal unity of parties that are in fact very disunited is often best disregarded in the categorization of party systems. It is, therefore, possible to speak without violation of logic of a one-party system even where no nominally separate organizations fight elections. The term then denotes that genuine competition takes place among persistent factions within a formally unified party rather than among formally autonomous parties. Usually, but not always, this is what is meant by a parti unifié as against a parti unique,terms first developed to distinguish between African party systems that are monolithic and those that display considerable intraparty pluralism (Schachter 1961, p. 306). Similarly, a nominal two-party system can consist of such weak structures that it might best be classified with certain multiparty systems, a case sometimes argued in regard to the American party system—implicitly by V. O. Key and explicitly by James McG. Burns. This is particularly the case where factions belonging to the nominal parties frequently collaborate across formal party lines, rather than merely representing, as in the British case, distinct intraparty tendencies that rarely, if ever, coalesce with kindred factions in other parties. On the other hand, certain formally independent parties can be counted as a single force because of their persistent and close alliance; such is the case of the communists and Nenni Socialists in Italy before the Hungarian uprising (the parties then were “two-in-one and onein-two,” according to Nenni).
It is very difficult to specify precise, general, and easily applicable indicators of the capacity for, and strategic position in the process of, coalition making, of the degree of interparty identification, and of internal party cohesion. While some useful measures exist (for example, the amount of cross-party voting in legislatures or the frequency of significant participation in coalitions), these qualitative factors must necessarily be gauged to a large extent by thorough acquaintance with particular cases; they are more matters of informed judgment than of exact measurement. For that reason they may be used to modify the results obtained from quantitative assessments of party strength, but the latter must necessarily be basic, if only because they produce the least equivocal results.
It may be useful at this point to illustrate how quantitative indicators of party strength are and could be used to determine the number of units in a party system—although this exercise should be prefaced by saying that not enough has as yet been done along these lines to permit more than suggestions. In existing writings on party systems, two measures of party strength are used, usually implicitly, for this purpose. One involves the absolute size of party units, that is, parties with less than a certain percentage of votes or seats are disregarded or parties with more than a certain percentage are considered to constitute single-party systems. The other involves relative sizes or ratios of forces, parties being disregarded if they fall far below the size of the principal party or parties. Both measures, although they may lead to different results, clearly help establish the number of genuinely competitive units in a party system, as do such less frequently used measures as the extent of division of control over certain offices over time or the rate of alternation in office of various parties.
On consideration, it would seem advisable to combine measures of absolute and relative size to determine the number of competitive party units. Suppose, for example, one classified as a one-party system any country in which a single party, over a number of elections, gets 60 per cent or more of the votes or seats. In that case one will end up placing in the same class systems which clearly are greatly dissimilar, for example, systems in which the remaining 40 per cent go to a second party, which has at least a chance of becoming the governing party (precisely the situation in Dahomey as of 1959—before the cessation of party politics—and approximated in South Africa and Southern Rhodesia in 1958), and systems in which the 40 per cent are divided among many inconsequential parties, none of which has the slightest chance of governing (as in northern Rhodesia and the Cameroons in 1959). Such a questionable result can be avoided if one takes into account the ratio of forces between parties, but that criterion alone also seems unsatisfactory. For example, classifying as a one-party system any in which the ratio between the largest and next largest party is more than 2:1 would make likely an even more questionable result: classifying as a one-party system one in which no party can govern by itself, because none gets even a bare majority, but in which one party is markedly stronger than the others (a condition approximated in the Weimar Republic until 1924 and again from 1928 to 1930). Combining the two criteria—for example, defining as a oneparty system one in which a single party over time gets more than, say, 60 per cent of votes and seats and in which the ratio between the largest and next largest party is more than 2:1—avoids both difficulties. Similar formulas could be worked out easily for party systems having any number of units, although distinctions beyond four or five units may not be important to make and may be adequately covered under some more general rubric. (For example, a two-party system might be defined as one in which two parties get about 75 per cent or more of votes and seats, and in which the ratio between the first and second party is less than 2:1 and that between the second and third party more than 2:1—the 75 per cent minimum, rather than a lower one, being advisable because parties which poll less than 35 per cent of the votes are not likely to have a “majority bent/’
The obvious advantage of such formulas is that they can make the concept of a one-party system logically tenable and the concept of a two-party system empirically relevant, while introducing some basic distinctions into the very large, nearly all-encompassing and very heterogeneous universe of multiparty systems—distinctions that particularly draw lines between very fragmented systems and those that display a considerable amalgamation of political forces. They thus solve the most basic problems that arise in a classification of party systems by number of constituent units. But, to repeat, the results obtained by such formulas must still be modified by assessment of qualitative factors that affect party strength within the broad categories used, and such assessments may even lead to shifting a party system from one category to another.
Usually one will want to make further quantitative distinctions to get more detail about the comparative strength of parties within the various categories, particularly when strength is treated as a separate dimension and not as a tool in the count of party units. For some theoretical purposes, for example, a distinction may have to be made between one-party systems that barely satisfy the minimum quantitative requirements and those in which a single party virtually has a monopoly of votes and seats—no less than between dominant parties that are highly cohesive and those that are extremely pluralistic. In addition to the distinction between cohesion and pluralism in the units of two-party systems, one may have to distinguish between two-party systems that persistently and markedly favor a single party (like the Bonn Federal Republic so far) and those which, like Britain, consist of two rather evenly matched forces. So also three-, four-, or five-party systems may be evenly balanced or markedly skewed, and this difference may matter a great deal. The Scandinavian party systems, for example, are certainly pluralistic, but so skewed that they have not prevented effective authority in parliamentary government. The same might be said today of India.
To characterize a pattern of competition one needs to know not only the number of competitive units and their relative strengths but also how the units compete. This aspect of party systems has many facets, most of which are difficult to discuss in language at once general and precise. There is, for example, the large question of what sorts of appeals parties commonly use—whether, among other things, their appeals tend to be emotionally intense or sober, intellectually dogmatic or pragmatic, a matter of policies or slogans, programs or platforms. There is the very large question of tactics—whether, for example, the parties want mainly to make new converts or to mobilize old supporters, try to exaggerate divisions or to minimize their importance, operate independently or in more or less close, more or less national, more or less open alliances. There is the related question of what sort of supporters the parties can muster and toward whom precisely they direct appeals: for example, whether the parties in a system tend to have distinct social bases or targets or whether they are rather similar in these respects (as American parties are similar in their heterogeneous supporters and targets, and the Indian and some African party systems are similar in recruiting militants from a single social segment, the urban middle class). There is the question of the more general similarity or dissimilarity of parties in a system: whether the parties fundamentally resemble each other, so that there is a rather simple pattern of competition between them (for example, in that all are “communal” parties or “national” parties, all “interest” parties or “aggregative” parties, all “ideological” or “pragmatic” parties, all “mass” parties or “caucus” parties), or whether they belong to fundamentally different species and thus use fundamentally different practices in more complex patterns of competition (as is the case, for example, in many transitional societies before stable regimes, competitive or noncompetitive, are established—for example, in the early years of Indonesia, where one found simultaneously religious and secular parties, broadly aggregative and narrow interest parties, national and ethnic parties, organizationally strong and weak parties, ideological and pragmatic parties).
Despite the rich technical language available for characterizing individual parties, there is no general and precise vocabulary for dealing with variations of this sort on the party-system level; and one must rely, to a large extent, on unsystematic descriptions of particular party systems. However, at least a start has been made toward the construction of appropriate concepts, perhaps the most notable attempt being Helmut Unkelbach’s use of the concept “party system integration” (parteipolitische Integration), which can subsume many aspects of the different modes of party competition (1956, see especially pp. 36-41, 51-95).
To Unkelbach, a party system is integrated when it involves a low level of conflict or enmity (Gegnerschaft) or a relatively small “distance” between parties; it is unintegrated when the opposite holds. The term may describe dynamic processes as well as stable states: party systems “disintegrate” when Gegnerschaft is activated in them and become integrated when it is toned down. And it can be used to describe particular parties or complexes of parties within a party system as well as the system in general. Unkelbach mentions, for example, two special forms of party integration which obviously involve subsystems of party systems: two-party systems in which each is highly aggregative and successful in avoiding centrifugal tendencies, and groups of parties that are both close to one another and capable of commanding parliamentary majorities between them. He also distinguishes between genuine integration, which is based upon a consensus sufficient to allow different parties or factions to carry out a common program, and false integration, that is, mere electoral collaboration.
The crucial consideration throughout is the political “distance” between parties or factions, and this concept summarizes many important aspects of the modes of party competition. It is gauged (unfortunately, as yet not measured) on the basis of five general criteria: (1) differences in fundamental views or orientations (weltanschauliche Grundlagen); (2) differences in concrete goals or in the rank order of goals pursued; (3) differences in the means of striving for concrete goals; (4) differences in the assessment and valuation of political personalities; and (5) differences in the social composition of the parties or factions, particularly their class basis and the extent to which they aggregate diverse forces. The results obtained on these bases can, however, be modified by two related conditions: the existence of a higher consensus, which creates an emphasis on the requirements of the common welfare and thus lessens political distance, and the predominance of elements of conflict over elements of harmony, which increases the distance by creating “artificial” divisions. In parliamentary regimes another criterion, one that does not involve electoral activity, can be used to supplement, or to indicate, the others: the character of collaboration in coalitions. The collaboration may be frictionless or involve only minor differences, indicating political proximity or even kinship; or it may involve great differences or prevent the maintenance of coalitions, indicating various degrees of enmity. (Collaboration in opposition might indicate the same things, as might the collaboration of legislative factions in presidential regimes.)
Assuming the possibility of reasonably precise assessment, the concept of party-system integration holds great promise. It can help to determine accurately the number and relative strengths of truly competitive units in a party system; for example, whether a single nominal unit, like the Malayan Alliance, should be counted as several real units, or whether several nominal units form a single real one, like the Italian communists and Nenni Socialists before 1956 and the two main Austrian parties, which contest elections but usually collaborate in government through the proporz system. More important, the concept can be used to make crucial distinctions between party systems consisting of similar numbers of units. As an example, contrast a two-party system like the British, in which the distance between the parties is slight, with one like the Italian (assuming it qualifies as a two-party system), in which the distance could hardly be greater. Such systems must surely result from different conditions and have different consequences. Furthermore, it might well be found that the conditions and consequences of party-system integration or lack of integration are everywhere the same, regardless of the number or strengths of units that are involved. Beyond all this, the concept gets summarily at many aspects of party competition and could provide answers for important questions about the impacts of party systems on the larger political system.
The dimensions discussed so far—the number, relative strengths, and integration of party units—treat party systems entirely without reference to their contexts (with the exception of one of the subdimensions of party-system integration, social composition, which inherently requires reference to something beyond the systems themselves). But party systems have contextual no less than immanent characteristics. Not only are they conditioned by (and in turn condition) the larger political and social systems and some of their other subsystems, but they also participate with other structures in various political processes. The ways in which they do so are part and parcel of their structural characteristics and may, like their integration, be placed on a single, but divisible, dimension: “salience.”
In general, party-system salience refers to the functional significance of party systems in political processes, in other words, to the extent that they are in a position to contribute to, or undermine, the operation of political systems or their parts. This involves at least two closely related considerations: the significance of party systems relative to other structures in over-all processes of political competition (including the extent to which other structures are subsumed to the units of party systems) and the range of the political (and social) functions the party systems actually perform.
Political competition has one or more of three objects: to recruit decision makers, to influence the making of decisions, and to affect the specific ways in which decisions are applied and values are allocated. In regard to each of these, party competition may range from being all-important to being immaterial relative to other sorts of competitive processes. A familiar example is the difference in party-system salience between spoils systems and those in which the appointive and allocative powers of elected officeholders are highly circumscribed. In this case it is largely formal-legal rules that define the salience of party systems, but many other factors may have similar effects. No rules, for example, underlie the presently growing relative significance of pressure groups in most Western countries; rather this tendency reflects, among other factors, the expansion of governmental activities; the increase in comprehensive memberships and effective organization of the groups; the fossilization of many parties (Lowi 1963); and the great consensus on the broad, aggregative policy alternatives that electoral competition usually defines. Nor do rules underlie the relative lack of salience in leadership recruitment in the case of fragmented party systems that do not yield clear electoral results and where the recruitment function is largely performed jointly by parliamentary groups and individuals, who may operate without close reference to party lines; Italy in the age of trasformismo and France during most of the modern era are cases in point. On the other hand, constitutional rules greatly contributed to the relative insignificance of the party system in Imperial Germany, where the executive was not a creature of parliament and was accountable to it only in very restricted ways, although Bismarck’s skill in manipulating the party system was certainly a factor in his personal success.
The salience of party systems relative to other structures of political competition reflects, but is not quite the same thing as, the range of political functions to which party systems contribute. It is not the same thing because party systems contributing to a great variety of functions may not contribute to them very much; yet the range of functions is itself a partial measure of their salience. Perhaps the main distinctions to be made on this subdimension are between party systems that restrict activity to the more manifest functions of such systems (recruitment and the aggregation of preferences), those that emphasize more latent functions (such as the creation of a sense of national identity, the awakening of participant attitudes toward the political system, and the provision of political education and training), and those that combine both types of functions. The last category undoubtedly is much the largest. However, the earliest Western parties, those of the age of representative oligarchy, Duverger argues, belonged uniformly to the first group, while the broader party systems formed in response to democratization or in anticipation of suffrage expansion at first fell mainly into the second group, as do many contemporary non-Western parties.
Party-system salience is an important dimension not just for purposes of description but also because variations in salience may be expected to have important consequences for theories about the determinants and functional impacts of party systems. The more salient party systems are bound to have more considerable functional repercussions than the less salient ones, and such systems are also likely to be more sensitive to external conditioning factors. They may be expected to be the crucial nerve centers, buttresses, or impediments of representative government. But if party systems matter little compared to other structures, and this is realized, the tendency of social forces will be to work upon the other structures. Hence any generalization about the other dimensions of party systems will be vulnerable to the extent that the weight of such systems in political processes is ignored.
Up to this point only static variables have been discussed—that is to say, variables that can be applied at any point in time as well as to describe persistent patterns over time. Undoubtedly these variables are best applied to highly persistent patterns, since very transient structures resist rigorous analysis. But patterns that cover long durations can rarely be discussed adequately in static terms alone. They are always likely to change, to develop; some party systems, indeed, are so unstable that a classification of them at any moment in time will be virtually useless. While changes in party systems can largely be described by characterizing them statically at successive points in time, inherently dynamic categories that describe general, frequently occurring characteristics of the process of change itself are also useful.
A distinction should be made at the outset between dynamics within party systems and the dynamics of party systems. For the first, Duverger has developed a very useful typology that deals with “success patterns” and “displacement patterns” within party systems that change in content but not in form. Duverger distinguishes four basic types of intrasystem dynamics. One is alternation—the regular movement from office to opposition and opposition to office which one finds in some, but not all, two-party systems, and which may matter to the construction of generalizations about such systems quite as much as system integration. The second he calls stable distribution—the absence of serious variations in party strength over a long period (as in Holland, Switzerland, and Belgium between the wars). The third is leftism—the gradual disappearance of parties on the right and the birth of parties on the left, or the gradual weakening of the right and increasing strength of the left, or the displacement of moderate leftist by more radical leftist parties, without any change in the formal characteristics of the system. Finally, there is domination—either the continuous dominance of a party (like that of the Radicals in the French Third Republic) or alternation in which the swing of the pendulum in one direction is consistently much longer than in the other. Additionally, Duverger distinguishes between “normal evolution” and “sudden mutations” within a party system, the latter referring to very abrupt shifts in the political balance, as in the rapid rise of Nazism in Germany after 1930 and of Rexism in Belgium in 1936.
The basic distinction to be made in characterizing the dynamics of party systems is of course between stable and mutable systems—those that, despite intrasystem developments, maintain their formal characteristics over time and those that do not. The latter category, however, obviously requires some subcategories. For example, continuous and rather prolonged tendencies (like the gradual fragmentation and weakening of the dominant parties in India and in Pakistan before 1958) should be distinguished from sudden discontinuities, like those that result from coups and revolutions (for example, the rapid destruction of the old German and Italian party systems by the Nazis and Fascists and their subsequent reappearance in radically altered form). One must also distinguish fluidity in party systems from patterned trends. Fluidity involves seemingly random, or at any rate highly ephemeral and often contradictory, movements in party life—the constant appearance and disappearance of parties, the continuous ebb and flow of consolidation and fragmentation, that characterized the limited, oligarchic representative systems of nineteenth-century Europe, with their parties of notables and caucuses, and that seems equally typical of some new and transitional nations today, especially those in which personalism is the characteristic form of party leadership. In contradistinction are patterned trends toward consolidation or fragmentation (decreases or increases in the number of competitive units) or toward the increased integration or disintegration of party competition.
The origins of party systems are also an aspect of party-system dynamics, for origins are always a part of developmental patterns. For example, party systems that originate in the followings of aristocrats or other political notables in oligarchic representative systems may be contrasted with those that begin with mass organizations. Similar distinctions can be made along many other lines. One can contrast the party systems of new states like the Philippines, where the colonial power fostered parties before independence, with those in which the beginnings of independence and party life coincide; and one can contrast party systems that originate in the followings of dynamic personal leaders, in a broad but unmaintained nationalist or revolutionary consensus, in ethnic or other “communities,” or in interest-group formations that turn from semi-political to fully political activity. In making such contrasts, however, we are no longer merely characterizing party systems, we begin to deal with the determinants of their varieties; for the origins of party systems, while an aspect of their dynamics, are of interest primarily because they may condition subsequent development.
Determinants of party systems
To account for the characteristics of party systems, three broad aspects of their contexts can be explored. One is the general political system or certain of its substructures: party systems might be treated mainly as responsive to larger or other aspects of politics. The second is social structure and culture: party systems might be treated primarily as structures that politically crystallize sociocultural forces. The third is their own histories: party systems with a past might be conditioned largely by that past, that is, might be self-propelling along predictable paths once set in motion, or even self-maintaining, regardless of external forces. Each of these approaches obviously comprehends many possibilities that must be disentangled, and while in some respects they may be mutually exclusive, in others they can be treated as complementary and combined.
Among the hypotheses that link party systems to other aspects of political structure, indeed among all existing hypotheses about party-system determinants, one looms much larger than the rest, in terms of both the frequency and rigor with which it has been argued. This is the theory that party systems are essentially products of electoral arrangements, of the formal rules governing their competitions.
The most compelling reasons for the prevalence of this theory are that the manifest objective of party competition is to maximize a group’s representation and influence in government and that different electoral systems, like different rules of games, require varying strategies in pursuit of that objective. Above all, the formal mechanics of electoral systems seem to have a bearing on combination strategies—on deciding whether groups with similar interests and opinions can maximize their influence by operating independently in elections, or through different kinds of alliances, or by permanently subsuming themselves to larger aggregative groups. Election rules thus influence the number, sizes, and integration of units and dynamic tendencies in party systems. Hypotheses about the effects of electoral systems are also popular because they involve formal-legal political structure, long the special concern of political scientists, and because they are particularly suitable to studies that have a remedial intent. Electoral systems, after all, are easy to identify and manipulate, while social and historical forces are more elusive since they are generally givens and thus defy control. [SeeElections, article onelectoral systems.]
Despite their frequency and surface plausibility, however, theories about the relations between electoral rules and party systems, closely examined, leave one with some feelings of unease. Invariably, they have more logical than empirical content. Moreover, in the course of their development, they have become constantly more complex (reversing the usual course of scientific inquiry) and increasingly qualified by other factors, both tendencies resulting from attempts to neutralize cogent criticisms without yielding fundamental ground. These points, which at least make one wonder, are illustrated by the three most comprehensive works so far written on the subject.
The first systematic exposition of the relations between electoral and party systems was that of Hermens (1941). The argument of that work, somewhat simplified, was that proportional representation undermines democracy by the effect it has on party systems: it tends to fragment such systems into large numbers of independent units and, in the sense we have used the term here, to disintegrate them; in contrast, the majority system of elections consolidates and integrates the parties and thus promotes effective government. The case for this theory was in the first instance made logically, and that part of it carried much conviction. Then empirical materials were produced to support the logic; at this point, however, the argument became much less compelling, since the data were clearly selected to fit the case and contradictory cases could easily be found. These cases in fact soon became the basis of a large anti-Hermens literature.
Duverger’s Political Parties (1951) was the next major work on the subject (see also Duverger 1950). It resembles Hermens’ work both in its basic arguments and the force of its logic but differs from it in several respects. In the first place, Duverger’s argument is more complex because it concerns three general types of electoral arrangements, not two: proportional representation, the simple-majority single-ballot system, and the majority system with two ballots (as used in France during most of the Third Republic). To the last Duverger attributes results similar in many respects to those of proportional representation. The basic implication is obvious: a major modification of the majority system can produce results different from those that Hermens attributed to it. More important, Duverger conscientiously adduces empirical exceptions to his arguments, sometimes indeed as many exceptions as cases in point; for example, the hypothesis that the simple-majority single-ballot system leads to a two-party system with alternation in power (which, according to Duverger, comes as close as it is possible to come to a sociological “law”) is illustrated by Britain, Uruguay, the United States, and Turkey (since 1950), but is contradicted by the “exceptions” of Belgium before 1894, Sweden before 1911, Denmark before 1920, and contemporary Canada. The deviant cases are then explained away by various “special conditions,” implying, again in opposition to Hermens, that there is more than electoral logic to the conditioning of party systems, even if electoral logic retains pride of place. However, to call the other relevant factors “special conditions” and the cases they govern “deviant cases” is to assume that the rule “electoral systems determine party systems” has already been established; but Duverger nowhere shows just how this has been done. His manifest object is to complicate the Hermens thesis somewhat in order to accommodate the more obvious contradictory cases and to disarm empirical criticisms of the more complicated theory itself by invoking saving conditions wherever it seems endangered by less obvious cases.
Unkelbach’s work (1956), which also contains many of the same basic arguments, is a distinct advance over Duverger at least in one respect: many deviations from the general rules are themselves due, in Unkelbach’s argument, to variations in electoral systems, not to “special” conditions. The book shows, with mathematical logic, how even minor changes in an electoral system can dictate important adaptations in party strategy and thus in party systems. Hence, Unkelbach does not deal with only three types of electoral rules. Instead he adduces separate arguments for a huge number of them, even when they seem to have much in common. For example, within the general category of “majority systems” he distinguishes the effects of the relative majority system (Britain), the absolute majority system (Imperial Germany), the New York City variant of the relative majority system, the majority system with ballotage (Third Republic), the majority system with the alternative vote (Australia), the “two-step” relative majority system (Uruguay), and the various kinds of majority systems in multimember constituencies. But this procedure, while making it less necessary to introduce nonelectoral factors into the analysis and while being logically more convincing in its very nature, creates an important difficulty of its own. It ends up making a separate case for almost every electoral system in every country and thus makes it virtually impossible to test the logical arguments by empirical generalization. And even so, Unkelbach is compelled to make his findings fit the data by systematically invoking three additional determinants: other aspects of political structure, underlying social forces (a large category indeed), and the “free” actions of decisive personalities. These, moreover, are not treated as merely minor influences that can to some extent modify the tendencies emanating from electoral systems: they can reverse them and must therefore be, in some instances at least, the fundamental determinants of party systems.
The whole literature linking electoral and party systems thus confronts one with a paradox: it has great logical force and has acquired increasingly greater, if not simpler, force of logic over the years, but it often lacks empirical fit, no matter how complicated the logic has been made. The explanation for this discrepancy between logic and experience is that the conditions assumed in logical arguments are not always satisfied, perhaps even rarely satisfied, in actual cases. These assumed conditions are very much like those posited in economic theories: that parties try to maximize their influence and that they are infinitely malleable institutions that can readily adjust their behavior to changes in the conditions of competition. But both are doubtful assumptions when indiscriminately applied, particularly the latter. Because of this, theories like Duverger’s and Unkelbach’s may be more useful as models against which to gauge the actual rationality, power drive, and malleability of units in party systems than as theories that account for variations between such systems.
To illustrate, take a case that would seem highly deviant to all the electoral-systems theories: a country that uses the British electoral system but has nothing like a two-party system with alternation in power, namely, India. The Indian party system is somewhat ambiguously poised between the oneparty dominant and polyparty categories. There is one very large aggregative party, Congress, and a great number of very small, unaggregative, communal, and ideological parties, like Jan Sangh, Mahasabha, Ram Rajya Parishad, and the Communists. Congress may resemble the kinds of parties we normally find in two-party systems but the rest certainly do not, and the system as a whole lacks integration; nor has there been any tendency toward party consolidation and integration in the opposition—if anything, the reverse.
Weiner’s study of the Indian party system (1957) mentions four factors that help explain this situation. First, Indian parties typically are not just electoral organizations but loci of their members’ social life, “total organizations” that are often substitutes for family, caste, village, status group— even job. Such organizations, frequently found in non-Western nations and not unknown in the West itself, clearly mean too much to their adherents to be very malleable in the light of coldly rational electoral calculations. Second, the Indian parties lack power drive, since the general Indian culture particularly undervalues the political realm. Third, many opposition parties fail to make calculations that would improve their electoral chances simply because they are not committed to working within the parliamentary system (something also familiar in the West and, even more, in other areas). Finally, there is a general lack of interest organizations. This seems to militate against party consolidation and integration because a group submerged in a larger organization for electoral purposes does not have available alternative channels—pressure-group politics—for pursuing its special interests, at least not to the extent it would in many Western nations where the value sacrifices involved in electoral consolidation are correspondingly much lower.
These and other observations we have made suggest a generalization about the relations between electoral rules and party systems. Electoral rules govern the forms of party systems along the lines of maximizing models only under specifiable conditions, the absence of which may lead to results very different from, and even contradictory to, those predicted by the models. These conditions are (1) that the logic of the rules be understood by the actors; (2) that there be commitment by the actors to representative rule; (3) that they have a strong appetite for power as an end and that significant power positions be attainable through electoral competition; (4) that the actors’ political intensity be low (that is, their political style be pragmatic, not doctrinaire); (5) that the units through which they act be relatively specific, not highly diffuse, in social purpose; and (6) that there be available in the political system useful alternative channels for pursuing interests.
Three points follow from these conditions. Since the requirements for unfettered maximizing actions in party competition are numerous, such actions are likely to be rare, and hypotheses confined to the relations of electoral rules and party systems are likely to be weak. Second, one must pay at least as much attention to other aspects of the political contexts in which party systems operate as to electoral rules; these include other kinds of formal rules and, above all, aspects of political culture and the wider structure of political competition in a society. [SeePolitical culture.] And third, since the latter are ineluctably tied to sociocultural forces, only exceptional party systems may be explicable within a framework of purely political considerations—which raises the possibility that social and cultural conditions may themselves suffice to account for their characteristics.
It has been argued (by Downs 1957) that party systems may be conceived of simply as showing the distribution of politically relevant preferences in society; for example, the more modal preferences there are, the larger the number of parties. This position may be useful as a premise for abstract model-building, but it is not even a tenable truism. Parties, after all, may, and often do, “aggregate” to such an extent that they comprehend internally various modal preferences—think of the Christian Democrats in Italy, Congress in India, or either American party. Also, modal preference groups have available many channels other than parties through which to act politically, particularly pressure groups and revolutionary organizations; for that reason alone “modal preference” need not equal “party,” although it may in special cases.
Party systems are obviously susceptible to the conditioning of social structure and culture. Tribally fragmented societies are not likely to produce consolidated or integrated party systems and have rarely done so, although party itself might sometimes be an antidote to social fragmentation. The doctrinaire and schismatic tradition of Islam seemed to make the preservation of an aggregative dominant party far more difficult in Pakistan than was the case in India (Almond & Coleman 1960, p. 196). Sometimes, indeed, the salient divisions of social structure are virtually mirrored in those of the party system, as in the Belgian party split between Catholics and freethinkers, a division corresponding to the overriding national split between those who speak Flemish and the Walloons who speak French (Neumann 1956), or in the division in South Africa between English- and Afrikaansspeaking people, or in the ethnic bases of the Nigerian parties.
Many such cases can be cited, but the relations between party systems, social structure, and culture are far from simple or readily apparent. One difficulty is to determine which among the many aspects of social structure and culture most impinge upon party systems, for such systems, on the evidence, may or may not reflect almost any aspect of social life. In one case, Scandinavia, the most relevant factor, if there is one, would seem to be the long past division of society into estates; in Belgium, coinciding religious and linguistic divisions; in America, regional divisions; in Italy, perhaps a split between the sexes, as much as anything else; and in many of the new nations, demographic factors and generational conflict.
More important is the fact that even the pro-foundest sociocultural cleavages often are not reflected, or are only imperfectly reflected, in party systems, no matter what they are. Some sociocultural splits never enter the political realm at all. Even if they do, they may find political outlets other than parties. They may have to find such alternative outlets if parties are, as they well may be, unresponsive to sociocultural divisions or changes in such divisions, particularly since it is not easy to found new competitive units where the old are well entrenched. And, just as many sociocultural conflicts are integrated in the units of party systems, so there are party splits that simply do not correspond to sociocultural ones. Party itself may be the main thing that divides some societies, especially where the appetite for office and spoils is great; to what sociocultural divisions, after all, did the party and factional divisions of transformist Italy correspond?
All this means that, like electoral exigencies, sociocultural conditions give only a very partial account of party systems, in that they account only for limited numbers of cases and/or account only for some aspects of the cases. A combination of the two approaches would clearly yield better results than either alone, for electoral strategy is clearly one factor that bears upon how parties reflect sociocultural forces, while sociocultural forces in turn can obviously distort and limit a purely maximizing approach to elections.
The past of party systems
There is, however, yet a third way to account for the characteristics of party systems, which seems to work well in numerous cases. Party systems, at any point in time except their origin, are largely creatures of their own histories; once in being, they become self-moving, even in the sense of self-perpetuating. The argument implies that party systems are highly responsive to external conditioning only at their formation, that subsequently they either retain their essential characteristics or change only very slowly, unless totally destroyed in political upheavals. In other words, their development is largely inherent in their original shape. Party systems may simply become incorporated into the habit background of society (for example, acquire extensive followings of nonfloating voters). Like so many organizations, parties, especially modern, bureaucratized, mass parties, may spawn groups with strong vested interests in keeping the systems substantially as they are. Or the party system may itself create strategic imperatives that condition its development: for example, if a consolidated system emerges early, it may in all cases be more politic for small groups to seek to work within it rather than independently.
One of the most conspicuous cases illustrating this theory is the remarkable similarity between the party systems of Imperial Germany and the Weimar Republic, despite a change in the electoral system, and the great change in German party life that followed the destruction of the old organizations during the Nazi period. Even electoral-systems theorists implicitly provide cases in point. An example is the now common argument that while proportional representation, contrary to earlier beliefs, might not fragment or disintegrate party systems, it cannot lead to their consolidation and integration—which may only be an instance of the inability of electoral systems to affect party systems at all.
Party origins are used particularly often to account for one-party dominant systems—systems that for obvious reasons have never been related by anyone to electoral machinery. The most common hypothesis is that one-party dominance comes into being where party systems are formed in struggles for national independence, since such struggles channel political activity toward a single, widely shared, intensely held objective and reduce domestic conflict by focusing political Gegnerschaft on the clash with the external forces (Hodgkin 1962, pp. 22-23). Such dominance may then persist through the very machinery, symbols, and habits it creates, even where the political system remains in some sense competitive. It may even persist when unity begins to disintegrate, since the process of disintegration may give rise only to many small, uncompetitive opposition fragments. This familiar process may be related to extreme ethnic, regional, religious, or ideological divisions; it could, however, also result from the fact that the very strength of the dominant party makes the consolidation of opposition useless (under any electoral system) and thus provides no benefits to balance against the unavoidable costs of factional combination.
Origins may also play a significant role in shaping other kinds of party systems. It has been argued, for example, that fragmented party systems tend to come into being (and then to persist) where the beginnings of wide political participation coincide with the simultaneous existence of deeply divisive conflicts over fundamental issues, such as issues of national identity, constitutional order, and basic internal and international policy. Contrasts are often drawn in this connection between Britain—where issues were settled in the following sequence: national identity, religious policy, constitutional order, basic socioeconomic policy, and the question of Britain’s international function—and France and Germany, where all such issues arose at once, just when the party systems were being formed. Contrasts have also been drawn between party systems that originate in competitive and noncompetitive systems, as full-fledged participants or nonparticipants within competitive systems, within or outside of parliaments, in “preparty” formations or directly as party formations. To mention some examples: Parties that originate as clandestine groups in noncompetitive systems seem generally to retain a highly uncompromising style and highly disciplined organization, becoming either instruments of noncompetitive rule (in the manner of the Moroccan Istiqlal or the Tunisian Neo-Destour) or disintegrated units in genuine party systems. Parties that do not early in their histories play a fully participant role in government—like the German Imperial or prewar Japanese parties, which contested elections but had no chance to run affairs—seem to develop highly doctrinal, “irresponsible,” and oppositional styles that lead to disintegrated party systems. Parties that originate in parliamentary combinations, like the old British caucus parties, seem to put much more emphasis on electoral struggle than other goals and ties and thus to respond more to the exigencies of electoral systems than parties formed outside of parliaments. And parties originating, like so many contemporary African parties, in other associations—trades and farmers’ unions, youth associations, literary societies, study circles, tribal associations, religious sects, etc.—also seem highly resistant to electoral logic, tending to be “total organizations” in one-party dominant or disintegrated party systems.
In theories of party systems, the idea of inertia must play a serious role. Yet changes in party systems do occur, and not only after great political upheavals or always in ways that seem immanent in earlier stages. This, coupled with previous arguments, implies that just as party systems cannot be characterized on a single dimension, so their characteristics cannot be accounted for by one set of determinants. Numerous conditions manifestly are involved.
The present task of study is, however, not merely to add up determinants to yield plausible accounts of specific cases—nothing is easier—but to discover the general circumstances under which each of the conditions discussed above may be of special significance. For example, it may well be, as already suggested, that sociocultural factors are of particular significance during the genesis and youth of party systems—much as environmental stimuli most affect human personalities early in life; that thereafter party systems become constantly less sensitive to such forces and more selfdetermining; but that, under conditions earlier specified, they will replace the pulls and pushes of broad social forces with responsiveness to electoral rules. It may also be that through such a process, party systems become gradually less salient in political competition compared to structures more sensitive to social forces and, in some cases perhaps, become a positive hindrance to effective representative rule, regardless of the degree of consolidation, distribution of strength, or amount of integration in the systems (cf. Lowi 1963, pp. 573-575).
The functional impacts of party systems
The last sentence touches upon the third set of problems that arise in the study of party systems, namely, problems about their functional impacts. By this is meant broadly how party systems affect the contexts that affect the systems: how party systems contribute to the viability and effective working of other structures, including political systems as a whole, and to the achievement of specific goals. Questions concerning the functional impacts of party systems are inseparable from questions about their salience. But since it can be assumed that generally only high salience can give party systems great functional significance, study can be concentrated on the other characteristics of the systems, or salience can be treated as itself a general aspect of the functions as well as the forms of party systems.
In a field generally in infancy, the study of functional impacts is at present much the least developed. We lack not merely tested theories but any theories at all, or even explicitly stated problems about which to theorize. Possible party and party-system functions are usually just inventoried, and even that only partially, without attention to the kinds of party systems that perform them adequately or the conditions under which they do so (Lowi 1963, pp. 571-573).
Thus far, the functional impacts of party systems have been studied mainly in connection with one problem that arose naturally in the therapeutic literature on the subject: What kinds of party systems support or undermine stable and effective representative rule? The answer most often advanced has, equally naturally, been couched in terms of the classical typology of party systems: two-party systems supposedly make democracies stable and effective, multiparty systems do not. In view of what has been said here, the terminology used in this proposition itself casts doubt upon it. But even if one ignores that point, the hypothesis still leaves much to be desired both in logic and empirical fit. Logically, one can construct a situation in which a polarization of the party system might have devastating consequences—for example, a polarization of moderate and extremist forces that would make electoral competition tantamount to civil war and concentrate all the energies of the moderates on the process of keeping rather than using power. Where political conflict has great potential for violence a wide dispersion of political forces might be much preferable, even if not ideally desirable. In addition, a highly consolidated party system unsupplemented by alternative channels of political competition—a hyperaggregative system—could lead to massive alienation from representative government. It would be equally easy to construct logical models of working multiparty systems, not least because such systems have actually existed and now exist—in Norway, in Denmark, in Iceland, in prewar Czechoslovakia, and in postwar Holland. All this quite apart from the question of what sort of party system optimizes the quite separate values of representing opinions and supporting authority, while performing well the more manifest party-system functions discussed at the outset.
In the functional assessment of party systems, it is precisely such balances between different functional impacts that must be struck, for party systems impinge upon much more than the stability of parliamentary institutions. Indeed, in many contemporary societies other functional questions, as yet unstudied, even unformulated, loom much larger. For instance: What types of party systems, under what conditions, conduce to or hinder the formation of new political identities, especially broad national identities? What systems, under what circumstances, help create feelings of political community, that is, ties within as well as ties to new political structures? What party systems are suitable to political mobilization, that is, to engendering participant attitudes and behavior among political “parochials”? What party systems tend to adapt well to new political demands and forces? Do any party-system characteristics particularly help or hinder the rapid “modernization” of societies, in politics or in a wider sense? What party systems, under what conditions, encourage the proliferation and salience of other structures of political competition, and with what general effects on politics and society? What party systems produce what sorts of special political skills, help overcome personal disorientation and insecurity in periods of rapid social change, and bring out, accentuate, or effectively sublimate destructive conflicts?
We know little or nothing about such questions; they can only be proposed as items on a large agenda. But the questions themselves indicate some of the most poignant difficulties, even dilemmas, of representative rule, as well as the centrality of party systems to such rule. Party systems operate at the most sensitive points of representative government, where special interests, aggregated conflicts, national authority, and national purpose conjoin. Hence their great multifunctionality. But can any party system satisfy equally the various functions to which it may be crucial? It seems unlikely. Even the basic ends of representing opinions and supporting authority may be contradictory, not just separate; similarly, party systems that activate politically large masses may be precisely those that accentuate conflict rather than cohesion, segmental loyalties rather than national identity. Equally serious, party systems appropriate to certain ends may be least likely to exist where they are most urgently needed. For instance, where national identity and community must be created, party systems are likely to be most fragmented and unintegrated precisely because of the lack of larger social ties; where broad political mobilization is still merely a goal, party systems are most likely to consist of units based on restricted elites; where social transformation is an overriding end, party systems may tend to become foci of social recalcitrance.
Few problems of comparative politics, then, are more crucial than those of discovering what kinds of party systems optimize satisfaction of the various ends that they may have to serve and how, and to what extent the development of such systems may be promoted. Before these problems can be coped with, however, the whole universe of party systems must be better charted, and the forces that bear upon them, and which they in turn exert, must be better understood. My object here, given a subject conspicuously understudied, has been to provide an indication of how these ends might be reached.
[See alsoElections, article onElectoral Systems; Government; Political Sociology; Politics, Comparative; Representation, article onRepresentational Systems; Systems Analysis, article OnPolitical Systems.]
Almond, Gabriel A.; and Coleman, James S. (editors) 1960 The Politics of the Developing Areas. Princeton Univ. Press.
American Political Science Association, Committee on Political Parties 1950 Toward a More Responsible Two-party System: A Report. New York: Holt.
Downs, Anthony 1957 An Economic Theory of Democracy. New York: Harper.
Duverger, Maurice 1950 L’influence des systemes electoraux sur la vie politique. Paris: Colin.
Duverger, Maurice (1951) 1962 Political Parties: Their Organization and Activity in the Modern State. 2d English ed., rev. New York: Wiley; London: Methuen. → First published in French.
Engelmann, Frederick C. 1957 A Critique of Recent Writings on Political Parties. Journal of Politics 19: 423-440.
Hermens, Ferdinand A. 1941 Democracy or Anarchy? A Study of Proportional Representation. Univ. of Notre Dame Press.
Hodgkin, Thomas 1962 African Political Parties: An Introductory Guide. Harmondsworth (England): Penguin.
Key, Valdimer O. (1942) 1964 Politics, Parties, and Pressure Groups. 5th ed. New York: Crowell.
Key, Valdimer O. 1949 Southern Politics in State and Nation. New York: Knopf. → A paperback edition was published in 1962 by Random House.
Lavau, Georges 1953 Partis politiques et réalités sociales: Contribution a une etude realiste des partis politiques. Paris: Colin.
Leiserson, Avery 1958 Parties and Politics: An Institutional and Behavioral Approach. New York: Knopf.
Lewis, W. Arthur 1965 Beyond African Dictatorships: The Crisis of the One-party State. Encounter 25, no. 2:3-18.
Lowi, Theodore 1963 Toward Functionalism in Political Science: The Case of Innovation in Party Systems. American Political Science Review 57:570-583.
Mcdonald, Neil A. 1955 The Study of Political Parties. Garden City, N.Y.: Doubleday.
Neumann, Sigmund 1954 Toward a Theory of Political Parties. World Politics 6:549-563.
Neumann, Sigmund (editor) 1956 Modern Political Parties: Approaches to Comparative Politics. Univ. of Chicago Press.
Ostrogorskii, Moisei I. 1902 Democracy and the Organization of Political Parties. 2 vols. London and New York: Macmillan. → An abridged edition was published in 1964 by Quadrangle Books.
Schachter, Ruth 1961 Single-party Systems in West Africa. American Political Science Review 55:294-307.
Stokes, Donald E. 1963 Spatial Models of Party Competition. American Political Science Review 57:368-377.
Unkelbach, Helmut 1956 Grundlagen der Wahlsystematik: Stabilitdtsbedingungen der parlamentarischen Demokratie. Gottingen (Germany): Vandenhoeck.
Weiner, Myron 1957 Party Politics in India: The Development of a Multi-party System. Princeton Univ. Press.
"Parties, Political." International Encyclopedia of the Social Sciences. . Encyclopedia.com. (May 22, 2018). http://www.encyclopedia.com/social-sciences/applied-and-social-sciences-magazines/parties-political
"Parties, Political." International Encyclopedia of the Social Sciences. . Retrieved May 22, 2018 from Encyclopedia.com: http://www.encyclopedia.com/social-sciences/applied-and-social-sciences-magazines/parties-political
Modern Language Association
The Chicago Manual of Style
American Psychological Association
Political parties are organizations subscribing to an ideology or formed around a special interest, with the aim of attaining power within government. They participate in elections, select candidates for public office, mobilize voters, raise funds, articulate political positions, coordinate policy making, develop campaign strategies, and generate symbols of party identification and loyalty. Parties are rooted in political, religious, sectional, ethnic, racial, and/or economic class interests. Parties are frequently coalitions of groups espousing disparate interests. Persons who support a party’s candidates, espouse its policies, and work to advance its political objectives are partisans. An opposition party does not challenge the legitimacy of the government, but only its policies.
European political parties formed during the seventeenth and eighteenth centuries to advise the monarchy. The model of a political party can be traced to Great Britain as the Tory and Whig parties fought for control of Parliament. Alexander Hamilton (1755–1804) founded the first U.S. political party, the Federalists, in 1792 to support his fiscal and political policies.
Although a competitive party system is considered to be an essential prerequisite for political freedom today, seventeenth- and eighteenth-century Anglo-American thinkers were wary of political parties, faulting them for serving special interests rather than the public good. Parties often were associated with treason and conspiracy. George Washington (1732–1799), the first president of the United States, in his “Farewell Address” deplored “the baneful effects of the Spirit of Party” (Washington 1796, p. 226). John Adams (1734–1826), the second president, held that “a division of the republic into two great parties … is to be dreaded as the greatest political evil under our Constitution” (Adams 1851, p. 508). Political leaders who serve only the interests of their parties rather than the common good were condemned as corrupt. The Federalists and Jeffersonian Republicans did not regard each other as legitimate opposition parties, but as threats to the republic that should be eliminated. When the Federalist Party collapsed in 1814, the political consensus that ensued—the “Era of Good Feeling”—was cited as evidence that the U.S. system had succeeded.
Others wrote more sympathetically and prophetically about the positive role that parties played in political life. Edmund Burke (1729–1797) saw party competition as a necessary good. Famously defining a party as “a body of men united for promoting by their joint endeavors the national interest upon particular principle in which they are all agreed,” Burke declared that “[p]arty divisions whether operating for good or evil are things inseparable from free government” (Burke 1925, p. 229). One of the authors of the U.S. Constitution, James Madison (1751–1836), following the English philosopher David Hume (1711–1776), held in Federalist Paper number 10 that parties, or “factions,” could not “be removed” because they are rooted in man’s natural propensity to differ. The “mischief” of factions that cause, though, could be curbed by fostering a multiplicity of “factions and parties” that would render unlikely “that a majority of the whole will have a common motive to invade the rights of citizens.” If they do, sufficient impediments, such as bad roads and poor communications, would make it difficult for them “to act in unison” (Madison 1787, pp. 55–61).
Although there is no mention of political parties in the U.S. Constitution, political parties are a logical outcome of the constitutional system. The First Amendment of the Constitution guarantees freedom of speech and assembly, both necessary conditions for the emergence of voluntary political organizations. The Constitution’s delegation of legislative power to elected representatives encouraged the formation of political parties. Parties, in turn, transformed and democratized the constitutional system. Thomas Jefferson (1743–1826) is credited with having created the first popular political party—the Democratic-Republicans later known as the Democratic Party. For his election to the presidency in 1828, Andrew Jackson (1767–1845) transformed this party into the first national, massed-based party.
Because power is shared between the central government and the individual states, the U.S. federal system of government impedes the development of the type of unified, cohesive parties found in, for example, the European parliamentary systems. A parliamentary system demands greater party discipline because the majority party in Parliament forms the government. Except when the Democrat and Republican parties gather every four years at their national conventions to nominate candidates for the offices of president and vice president and to write their platforms, they are not in any meaningful way national parties. Rather, each is really a coalition of state parties that are themselves confederations of semiautonomous local governmental parties. Even presidents or candidates for the presidency have only feeble control over state and local party members and elected officials. Under the U.S system, a divided government is possible, when one or both houses of Congress and the presidency are held by different parties.
The ideological spectrum of political parties typically runs from left to right. Right-wing political parties espouse conservative or reactionary views, whereas left-wing parties are associated with progressive or radical policies. The Conservative Party of Canada and the United Kingdom and the Republican Party of the United States are right-wing parties, and the Labour Party of the United Kingdom, the Liberal Party of Canada, and the Democratic Party of the United States are generally considered to be left-wing parties. The British National Front Party and the Front National of France are examples of far right-wing parties, and the Green, Communist, and Socialist parties are all on the extreme left.
Political party systems vary across the world. Nonparty states, such as Saudia Arabia and the United Arab Emirates, have no political parties at all. Single-party states, such as the People’s Republic of China, North Korea, and Cuba, allow only one-party rule. The constitution of the former Soviet Union officially established the primacy of the Communist Party. The prohibition against the formation of opposition parties is evidence of the absence of liberty. In contrast, dominant-party states do allow other parties, but one party, such as the People’s Action Party of Singapore and African National Congress Party of South Africa, typically wins most if not all elections. In the U.S. South until the 1970s, the Democratic Party won nearly all the general elections.
Two-party states, such as the United States, have two dominant national parties that vie for power. The system for electing the president in the United States entails an indirect election in which the electorate votes for a slate of electors who cast their votes for the president and vice president in an electoral college. This system impedes the formation of third parties: Because the slate of electors for the electoral college are elected in winner-take-all state elections, minor party candidates rarely win sufficient electoral college votes to have an impact on a national election. The election of legislative representatives in winner-take-all district elections further discourages the development of third parties. The principle that the single-member district plurality voting system results in a two-party system has been called Durverger’s law, after the French sociologist, Maurice Duverger (b. 1917), who formulated it. The two parties tend to resemble each other because each party gravitates toward the middle in an effort to capture the independent vote.
Multiparty states, such as the United Kingdom, Israel, and Canada, have a number of parties that compete for power. A multiparty system commonly exists in states with a parliamentary form of government. If no party wins a majority of seats in a parliamentary election, a coalition government is formed between two or more parties. The majority of the members of Parliament vote for a leader who serves as the head of the government. Proportional Representation, a scheme of voting used in several European states, encourages further proliferation of parties. Under this system, legislative seats are allocated according to the percentage of popular votes that the party received in the most recent election.
Modern political scientists regard political parties as beneficial avenues through which political interests and opinions can be channeled. “Democracy,” political scientist E. E. Schattschneider (1892–1971) concluded, “is unthinkable save in terms of parties” (Schattschneider 1942, p. 124).
Other thinkers, though, were more cautious in their assessment of the democratic influence of political parties on government. Most significantly, the Italian anarcho-syndicalist sociologist Robert Michels (1876–1936) proposed the iron law of oligarchy which stipulated that no matter how democratic an organization may be initially, it will eventually develop into an oligarchy. All large organizations, to attain greater efficiency and decision-making coordination, tend to concentrate power into the hands of a few. As parties grow in size and complexity, they become more hierarchically organized. True democracy, given the premises of Michels’s argument, is practically and theoretically impossible.
The noted American sociologist C. Wright Mills (1916–1962) observed that the top of the American political system was becoming more unified and powerful. He described in his most famous book, The Power Elite (1956), the interconnecting relationship between corporate, military, and government leaders. His controversial argument was that there was a growing power gap between a class of people Mills called “the power elite” and the increasingly manipulated and controlled masses.
Not all theorists agreed with the pessimistic implications of these arguments. The popular definition of democracy as “government by the people,” Schattschneider contended, exaggerates the power of the public. Rather, a more realistic definition that emphasizes organization and leadership rather than spontaneous grass-roots politics is needed. “Democracy is a competitive political system,” he observed, “in which competing leaders and organizations define the alternatives of public policy in such a way that the public can participate in the decision-making process” (Schattschneider 1960, p. 141).
Some of the factors that weaken the power and influence of political parties include declining patronage, the direct primary, the role of the media, the proliferation of nonpartisan political action committees, and the growing importance of the Internet as a tool for raising campaign funds and disseminating information about candidates and issues. The direct primary diminishes the ability of party leaders to select preferred candidates to run for political office. Patronage induces voters to support the party’s candidates by handing out jobs, contracts, or promotions for political reasons rather than merit. As easily accessible information about candidates and issues becomes available to voters through the mass media and the Internet, voters become less dependent on parties. Better informed voters tend to be less deferential to parties and more likely to split their votes among candidates from two or more parties.
Many analysts of the U.S. political party system detect a continuing trend toward party dealignment. A growing number of voters are declining to affiliate with any political party, preferring instead to identify themselves as independents. Since 1988 the plurality of independent voters has steadily increased. In 2004 39 percent of the American voters identified themselves as independents. Polls suggest that a large portion of the American public is allergic to partisan politics. They want candidates who promise to rise above partisan bickering and party loyalty to work on enacting legislation that will effectively solve the nation’s most pressing social and economic problems.
The most important function of the national political parties today is to raise funds for candidates and wage “get-out-the vote” campaigns.
SEE ALSO Burke, Edmund; Campaigning; Centrism; Cleavages; Constitutions; Dahl, Robert Alan; Dealignment; Democracy; Elections; Electoral Systems; Federalism; Hamilton, Alexander; Hume, David; Ideology; Interest Groups and Interests; Left and Right; Left Wing; Madison, James; Michels, Robert; Mills, C. Wright; Oligarchy, Iron Law of; One-Party States; Pluralism; Political Science; Politics; Power Elite; Representation; Right Wing; Schattschneider, E. E.; State, The; Washington, George
Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Political Parties in America. Chicago: University of Chicago Press.
Burke, Edmund. 1925. Works, ed. C. F. Adams. London: Oxford University Press.
Duverger, Maurice. 1964. Political Parties: Their Organization and Activity in the Modern State. Trans. Barbara and Robert North. New York: Wiley and Sons.
Hofstadter, Richard. 1969. The Idea of a Party System: The Rise of Legitimate Opposition in the United States, 1780-1840. Berkeley: University of California Press.
Michels, Robert.  1915. Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracies. Trans. Eden Paul and Cedar Paul. New York: Free Press.
Mills, C. Wright. 1956. The Power Elite. New York: Oxford University Press.
Patterson, Kelly D. 1996. Political Parties and the Maintenance of Liberal Democracy. New York: Columbia University Press.
Schattschneider, E. E. 1942. Party Government. New York: Holt. Schattschneider, E. E. 1960. The Semisovereign People: A Realist’s View of Democracy in America. New York: Holt, Rinehart, and Winston.
Washington, George. 1931–1941. The Writings of George Washington. Vol. 35. Washington, D.C.: U.S. Government Printing Office.
White, John Kenneth, and Daniel M. Shea. 2004. New Party Politics: From Jefferson and Hamilton to the Information Age. 2nd ed. Belmont, CA: Wadsworth.
W. Wesley McDonald
"Political Parties." International Encyclopedia of the Social Sciences. . Encyclopedia.com. (May 22, 2018). http://www.encyclopedia.com/social-sciences/applied-and-social-sciences-magazines/political-parties
"Political Parties." International Encyclopedia of the Social Sciences. . Retrieved May 22, 2018 from Encyclopedia.com: http://www.encyclopedia.com/social-sciences/applied-and-social-sciences-magazines/political-parties
Modern Language Association
The Chicago Manual of Style
American Psychological Association
POLITICAL PARTIES, along with other political organizations (such as Political Action Committees, or PACs) have the ability to increase the political effectiveness of individuals by bringing them into an aggregate. The importance and distinctiveness of parties as political
organizations spring from their domination of electoral politics. Candidates are identified solely by party affiliation on the ballots and, although many candidates in the United States now use television advertisements that omit this information, the party label is still the principal cue for the voter at the polls.
The relationship of the parties to mass electoral politics is apparent in their evolution. Originally, political parties were legislative caucuses and elite nominating organizations. They assumed their modern form with the expansion of the male suffrage in the first half of the nineteenth century, when the parties first gained broad support in the electorate and a network of constituency-based local parties. Thus, they became instruments for organizing and representing the expanding electorates. By the latter half of the century, in fact, they had become instruments by which the new masses of voters wrested control of cities from old patrician and economic elites.
Origin and Development of American Political Parties
The origin and development of the American political parties stand entirely apart from the U.S. Constitution. Nowhere in it are they mentioned or even anticipated. Throughout American history they have been instruments of the democratization of the Constitution as well as a result of that process. The parties and their system of loyalty transformed the electoral college and the entire process of electing an American president into something approaching a majoritarian decision.
Scholars have identified five party systems that arose out of what political scientist V. O. Key called "critical elections," or periods of political realignment. Political participation rose suddenly, after a decline, and key components of a national coalition rearranged themselves. For example, the realignment of 1896 ushered in a generation of Republican dominance of politics in all areas except the South. Newly arrived immigrants affiliated for the first time with the GOP ("Grand Old Party," as the Republican Party is known), because the Republicans addressed themselves to the interests of urban workers by sponsoring protectionism. The Democrats, under William Jennings Bryan, addressed themselves to agrarian issues at the expense of urban interests.
The first party system, of Federalists and Democratic Republicans, took on the character of a mass-based party system after the election of 1800 which featured a dramatic rise in turnout that eventually included the majority of adult white males in most states. This "Revolution of 1800," as well as the first party system collapsed in the 1820s; the Jacksonian, or second party system, followed the critical election of 1828, with two parties that were competitive in every state of the Union, the Whigs and the Democrats. The third party system followed the 1860 election and continued for a generation after the Civil War. In this period, Democrats and Republicans achieved the highest rates of turnout ever recorded in American elections. This party system was driven by deep sectional, religious, ethnic, and racial antagonism. The election of 1896 promoted Republican dominance of American politics until 1932. With the exception of the Wilson years in the White House (1913–1921), the Democrats spent most of this time deeply divided between the agrarian and the urban wings of the party. The fifth party system, following the election of Franklin D. Roosevelt in 1932, was the model by which Key developed his notion of critical elections. Republicans and Democrats, while retaining some of their sectional and religious differences, primarily appealed to different voters on the basis of their socioeconomic classes rather than their regional or religious backgrounds. Thus for the first time, African Americans, who had always voted for the party of Lincoln, now found their economic interests better addressed by the New Deal Democrats and they altered their party loyalties accordingly. The fifth party system endured until the 1960s, when the Democratic coalition, forged in the New Deal, of labor, immigrants, small farmers, Catholics, Jews, African Americans, and white southerners began to break down. In the chaotic politics of the 1960s, southern whites, Catholics, and labor felt increasingly alienated from what they saw as the "cultural" politics of the 1970s: inclusive politics emphasizing the interests of women, African Americans, and Latinos. During this period, these groups became a swing coalition altering their partisan preferences from election to election, becoming Democrats for Nixon, Democrats for Carter, Reagan Democrats, and Clinton Democrats. The result of this large swing component in the electorate has been what some political scientists call "de-alignment," or the detachment of voter identification from consistent party loyalty.
American parties are also marked by a distinctive, three-part character that sets them apart from other political organizations. They are composed of an identifiable set of committees and activists (the party organization), a group of public officeholders and would-be officeholders (the party in the government), and a large contingent of loyalists who consider themselves to be members of the party (the party in the electorate). Ordinary usage recognizes any of the three sectors as the party, and, as in parties elsewhere, American party organizations and parties in government have contested for supremacy in the party and for control of its symbols and decisions. It is peculiar to the American party system that party organizations have rarely subjected the party's officeholders to even the mildest forms of direction or sanctions.
Special Characteristics of American Parties
While the development of American parties was similar to that of parties in other Western democracies and for most of the same reasons, the American party system has always had special characteristics. In form it has long been marked by considerable decentralization, by nonbureaucratic, skeletal organizations, and by the persistence of only two competitive parties. That is to say, the American parties have always been loose confederations of state and
local party organizations. Never have they developed the strong national executives or committees that parties elsewhere have. Nor have they developed the membership organizations common in the twentieth century in other countries. Largely without formal memberships or career bureaucracies, they have been staffed at most levels—except perhaps within the classic urban machine—by only a few party functionaries investing only limited time and energy in the business of the party. Along with the British parties and few others, the American parties have remained two in number. The parliamentary systems in English-speaking democracies have opted for a first-past the-post system of elections, except in Ireland. This kind of system creates strong incentives for a two-party system to maximize the vote.
Related to these formal organizational characteristics has been the parties' chief functional trait: the pragmatic, almost issueless majoritarianism through which they piece together electoral majorities through strategies of compromise and accommodation. They have been much less involved in the business of doctrine or ideology than similar parties elsewhere. Platforms have revealed only modest differences between the two parties. Periodically one finds movements and candidates within the parties who have been intensely programmatic, but, until the presidency of Ronald Reagan, their records of success, even when they have captured their party's presidential nominations, have not been good.
The American parties have found their major role as nominators and electors of candidates for public office. They waxed in the nineteenth century in their ability to confer the party label on candidates, first in party caucuses and then in the more widely consultative conventions that Jacksonian Democracy favored. Especially during the prevailing one-partyism of so much of American politics at the end of the nineteenth century, the excesses of party power in those nominations led to the advent of the direct primary in the years between 1900 and World War I. By the 1970s, party control of nominations was limited to some degree by primary laws in every state. In several states, to be sure, the primary law left some nominations to party conventions; in others in which it did not, the parties devised ways (especially in preprimary endorsements) of affecting the primary-election outcomes. The quadrennial national conventions at which the parties choose their presidential candidates remain an important but increasingly vestigial remnant of the party's once unchallenged control of nominations.
Control of nominations has shifted away from the party organization to the party in government; the same is true of the control of election campaigns. The vigorous political organizations of the late nineteenth century and early twentieth century controlled, even monopolized, the major election resources. Its army of workers publicized the candidates, raised their own campaign funds, and recruited their own workers. And they have been able to find sources of campaign experience other than the party organization—the opinion pollsters, the political public relations firms, the mass media. Just as the primary ended the party organization's monopoly of nominations, the rise of the new campaign expertise threatens its control of the election campaign.
Nonetheless, most American officeholders reach office on the ticket of one of the major American parties. American presidents and governors are party leaders, and in the early part of the twentieth century all state legislatures (except the nonpartisan legislature of Nebraska) were organized along party lines. Beginning in the 1990s, moreover, Congress and, to a lesser extent, the state legislatures began to adopt a more assertive form of partisanship. Beginning with the Republican Party's Contract with America in the 1990s, the House and Senate leadership put into effect an ideologically oriented public policy. In the Contract with America, in budget negotiations with the White House, and with the impeachment of President Bill Clinton, "party votes" roll calls—in which the majority of one party opposed a majority of the other—began to increase, after a decline that had lasted for almost all of the twentieth century.
American presidents, on the other hand, have found it necessary to be less partisan than their congressional colleagues. Building majority support for a president's program, to be successful, has in recent years required votes from the opposition party. Ronald Reagan achieved his legislative successes with the help of the Boll Weevils, the conservative Democrats from the Deep South and Texas who supported his tax-cutting policy. Bill Clinton relied on moderate Republicans not only for assistance in getting his budget bills passed, but also for his very survival in the Senate trial after his impeachment. Despite the president's less partisan approach, it is the president's record that most reflects on the party. The power of the political party has been joined to the power of the presidency. As coalitions led by the executive, American parties find their governing role conditioned above all by the American separation of powers. That role contrasts sharply with the role of the cohesive parties that support cabinets in the parliaments of most other Western democracies.
Broder, David. The Party's Over: The Failure of Politics in America. New York: Harper, 1972.
Chambers, William Nisbet, and Walter Dean Burnham. American Party Systems: Stages of Political Development. New York: Oxford University Press, 1975.
Formisano, Ronald P. The Birth of Mass Political Parties: Michigan, 1827–1861. Princeton: Princeton University Press, 1971.
Key, V.O. Politics, Parties, and Pressure Groups. 5th ed. New York: Crowell, 1965.
Ladd, Everett Carl. American Political Parties: Social Change and Political Response. New York: Norton, 1970.
Sorauf, Frank J. Political Parties in the American System. Boston: Little, Brown, 1964.
Shafer, Byron E. The End of Realignment?: Interpreting American Electoral Eras. Madison: University of Wisconsin Press, 1991.
"Political Parties." Dictionary of American History. . Encyclopedia.com. (May 22, 2018). http://www.encyclopedia.com/history/dictionaries-thesauruses-pictures-and-press-releases/political-parties
"Political Parties." Dictionary of American History. . Retrieved May 22, 2018 from Encyclopedia.com: http://www.encyclopedia.com/history/dictionaries-thesauruses-pictures-and-press-releases/political-parties
Modern Language Association
The Chicago Manual of Style
American Psychological Association
Party systems also take a variety of different forms, from the multi-party system at one extreme, to the one-party monopolistic state at the other. Multi-party systems (often with two principal parties) are strong in liberal democratic societies like Britain, the United States, France, and Germany, while the dominance of one political party is especially evident in African countries such as Kenya and Zimbabwe. It has been suggested that the type of party system is linked to the stage of development of a society, but local historical and political factors are probably more important in influencing the type of system that emerges.
Political sociologists have focused on political parties as organizations and studied their organizational dynamics. Issues of interest include the socio-economic background of leaders, activists, and supporters; the socio-political ideologies espoused by parties; the distribution of power between the different collectivities embraced by the party organization; and the techniques for mobilizing support. A major pioneering study of political parties was conducted by the German sociologist Robert Michels. In his study of organizational power, he noted the oligarchical tendencies of party leaders and officials who come to dominate the party, as it becomes increasingly bureaucratic. Their beliefs and attitudes, directed towards their own personal goals, are invariably less radical than those of rank-and-file members. Furthermore, where organizational procedures are used to stifle popular aspirations, radical objectives are inhibited. However, evidence from research elsewhere suggests that the oligarchical tendencies of party leaders should not be overstated, especially in accounts of the institutionalization of political parties.
Political scientists have also explored the role of parties in the political process and the extent to which different political regimes may be described as open or closed. The liberal view is that political parties, along with pressure groups and other interest groups, engage in competition for power as the representatives of different socio-economic groups in society. As a result of open competition, power in pluralist political systems is non-cumulative and shared. This benign view of the role of political parties in liberal democracies has been the subject of much criticism. It has been argued that certain groups dominate the political decision-making process–most obviously those who dominate in the economic realm. Furthermore, while observable party politics is worthy of study, the more subtle forms of power (such as agenda-setting) should not be ignored. Thus, while liberals emphasize the important role of political parties in representative democracies, neo-Marxists play down their significance. In capitalist societies, it is argued, since the dominant economic power is also the ruling class, parliamentary politics is illusory, and simply an ideological strategy which diverts attention away from the real sources of political power.
Many have argued that this Marxist view of political parties and power is at least as unsophisticated as the liberal alternative. True, power may be concentrated, but it is possible for the views of ordinary people to influence political outcomes. In this respect, political parties are not inconsequential, and play an important role in the political sphere of advanced capitalist societies.
"political parties." A Dictionary of Sociology. . Encyclopedia.com. (May 22, 2018). http://www.encyclopedia.com/social-sciences/dictionaries-thesauruses-pictures-and-press-releases/political-parties
"political parties." A Dictionary of Sociology. . Retrieved May 22, 2018 from Encyclopedia.com: http://www.encyclopedia.com/social-sciences/dictionaries-thesauruses-pictures-and-press-releases/political-parties