Consociational Democracy
Arend Lijphart
World Politics, Vol. 21, No. 2 (Jan., 1969), 207-225.
Stable URL: http://links.jstor.org/sici?sici=0043-8871%28196901%2921%3A2%3C207%3ACD%3E2.0.CO%3B2-E
World Politics is currently published by The Johns Hopkins University Press.
STOR
®
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/jhup.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact support@jstor.org.
http://www.j stor.org/ MonMar 7 05:15:02 2005
CONSOCIATIONAL DEMOCRACY
By AREND LIJPHART*
Types of Western Democratic Systems
IN Gabriel A. Almond's famous typology of political systems, first expounded in 1956, he distinguishes three types of Western democratic systems: Anglo-American political systems (exemplified by Britain and the United States), Continental European political systems (France, Germany, and Italy), and a third category consisting of the Scandinavian and Low Countries. The third type is not given a distinct label and is not described in detail; Almond merely states that the countries belonging to this type "combine some of the features of the Continental European and the Anglo-American" political systems, and "stand somewhere in between the Continental pattern and the Anglo-American."1 Almond's threefold typology has been highly influential in the comparative analysis of democratic politics, although, like any provocative and insightful idea, it has also been criticized. This research note will discuss the concept of "consociational democracy" in a constructive attempt to refine and elaborate Almond's typology of democracies.
The typology derives its theoretical significance from the relationship it establishes between political culture and social structure on the one hand and political stability on the other hand. The Anglo-American systems have a "homogeneous, secular political culture" and a "highly differentiated" role structure, in which governmental agencies, parties, interest groups, and the communication media have specialized functions and are autonomous, although interdependent. In contrast, the Continental European democracies are characterized by a "fragmentation of political culture" with separate "political sub-cultures." Their roles "are embedded in the sub-cultures and tend to constitute separate sub-systems of roles."2 The terms "Anglo-American" and "Continental European" are used for convenience only and do not imply that geo-
* This note represents an intermediate stage of a research project concerning political stability in democratic systems. An earlier and briefer discussion of the concept of consociational democracy, in the context of a critical analysis of the utility of typologies in comparative politics, appeared in the author's "Typologies of Democratic Systems," Comparative Political Studies, 1 (April 1968), 3-44. The author is indebted to the Institute of International Studies, Berkeley, for financial support.
1 Gabriel A. Almond, "Comparative Political Systems," Journal of Politics, xvin (August 1956), 392-93, 405- .
2lbid.y 398-99, 405-07 (italics omitted).
208
WORLD POLITICS
graphical location is an additional criterion distinguishing the two types of democratic systems. This point deserves special emphasis, because some of Almonďs critics have misinterpreted it. For instance, Arthur L. Kalleberg states that the two types "are based on criteria of geographic location and area," and that "Almond does not come out and specify that these are his criteria of classification; we have to infer them from the titles and descriptions he gives of each of his groups of states."3 Actually, Almond does indicate clearly what his criteria are, and he also specifically rejects the criterion of geography or region as irrelevant, because it is not based "on the properties of the political systems."4
Political culture and social structure are empirically related to political stability. The Anglo-American democracies display a high degree of stability and effectiveness. The Continental European systems, on the other hand, tend to be unstable; they are characterized by political immobilism, which is "a consequence of the [fragmented] condition of the political culture." Furthermore, there is the "ever-present threat of what is often called the ťCaesaristic' breakthrough" and even the danger of a lapse into totalitarianism as a result of this immobilism.5
The theoretical basis of Almond's typology is the "overlapping memberships" proposition formulated by the group theorists Arthur F. Bentley and David B. Truman and the very similar "crosscutting cleavages" proposition of Seymour Martin Lipset. These propositions state that the psychological cross-pressures resulting from membership in different groups with diverse interests and outlooks lead to moderate attitudes. These groups may be formally organized groups or merely unorganized, categoric, and, in Truman's terminology, "potential" groups. Cross-pressures operate not only at the mass but also at the elite level: the leaders of social groups with heterogeneous and overlapping memberships will tend to find it necessary to adopt moderate positions. When, on the other hand, a society is divided by sharp cleavages with no or very few overlapping memberships and loyalties— in other words, when the political culture is deeply fragmented—the
3 Kalleberg, "The Logic of Comparison: A Methodological Note on the Comparative Study of Political Systems," World Politics, xix (October 1966), 73-74. Hans Daalder's critical question "Why should France, Germany, and Italy be more 'continental,' than Holland, or Switzerland, or more 'European' than Britain?" seems to be based on a similar erroneous interpretation; see his "Parties, Elites, and Political Developments in Western Europe," in Joseph LaPalombara and Myron Wein er, eds., Political Parties and Political Development (Princeton 1966), 43n.
4 Almond, 392. There is also no reason, therefore, to call the exclusion of Scandinavia and the Low Countries from the "Continental European" systems an "artificial qualifier," as Kalleberg does, 74.
5 Almond, 408.
CONSOCIATIONAL DEMOCRACY
209
pressures toward moderate middle-of-the-road attitudes are absent. Political stability depends on moderation and, therefore, also on overlapping memberships. Truman states this proposition as follows: "In the long run a complex society may experience revolution, degeneration, and decay. If it maintains its stability, however, it may do so in large measure because of the fact of multiple memberships."6 Bentley calls compromise "the very process itself of the criss-cross groups in action."7 And Lipset argues that "the chances for stable democracy are enhanced to the extent that groups and individuals have a number of crosscutting, politically relevant affiliations."8 Sometimes Almond himself explicitly adopts the terminology of these propositions: for instance, he describes the French Fourth Republic as being divided into "three main ideological families or subcultures," which means that the people of France were "exposed to few of the kinds of 'cross-pressures' that moderate [their] rigid political attitudes," while, on the other hand, he characterizes the United States and Britain as having an "overlapping pattern" of membership.9
In his later writings, Almond maintains both the threefold typology of Western democracies and the criteria on which it is based, although the terms that he uses vary considerably. In an article published in 1963, for instance, he distinguishes between "stable democracies" and "immobilist democracies." The latter are characterized by "fragmentation, both in a cultural and structural sense" and by the absence of "consensus on governmental structure and process" (i.e. the Continental European systems). The former group is divided into two sub-classes: one includes Great Britain, the United States, and the Old Commonwealth democracies (i.e. the Anglo-American systems), and the other "the stable multi-party democracies of the European continent—the Scandinavian and Low Countries and Switzerland."10 And in Com-parative Politics: A Developmental Approach, published in 1966, a distinction is drawn between modern democratic systems with "high subsystem autonomy" (the Anglo-American democracies) and those with "limited subsystem autonomy" and fragmentation of political
6 David B. Truman, The Governmental Process: Political Interests and Public Opinion (New York 1951), 508, 511.
7 Arthur F. Bentley, The Process of Government: A Study of Social Pressures, 4th ed. (Evanston 1955), 208.
8 Seymour Martin Lipset, Political Man: The Social Bases of Politics (Garden City i960), 88-89.
9 Almond and G. Bingham Powell, Jr., Comparative Politics: A Developmental Approach (Boston 1966), 122, 263; Almond and Sidney Verba, The Civic Culture: Political Attitudes and Democracy in Five Nations (Princeton 1963), 134.
10 "Political Systems and Political Change," American Behavioral Scientist, vi (June 19^3)» 9-IO-
210
WORLD POLITICS
culture (the Continental European democracies). The third type is not included in this classification.11
In what respects are Switzerland, Scandinavia, and the Low Countries "in between" the Anglo-American and Continental European democracies? Here, too, Almond consistently uses the two criteria of role structure and political culture. A differentiated role structure (or a high degree of subsystem autonomy) is related to the performance of the political aggregation function in a society. The best aggregators are parties in two-party systems like the Anglo-American democracies, but the larger the number and the smaller the size of the parties in a system, the less effectively the aggregation function will be performed; in the Continental European multi-party systems only a minimum of aggregation takes place. The "working multi-party systems" of the Scandinavian and Low Countries differ from the French-Italian "crisis" systems in that some, though not all, of their parties are "broadly aggregative." Almond gives the Scandinavian Socialist parties and the Belgian Catholic and Socialist parties as examples.12 This criterion does not distinguish adequately between the two types of democracies, however: if one calls the Belgian Catholic party broadly aggregative, the Italian Christian Democrats surely also have to be regarded as such. On the other hand, none of the Dutch and Swiss parties can be called broadly aggregative.
Instead of using the extent of aggregation performed by political parties as the operational indicator of the degree of subsystem autonomy, it is more satisfactory to examine the system's role structure directly. Like the Anglo-American countries, the Scandinavian states have a high degree of subsystem autonomy. But one finds a severely limited subsystem autonomy and considerable interpenetration of parties, interest groups, and the media of communication in the Low Countries, Switzerland, and also in Austria. In fact, subsystem autonomy is at least as limited in these countries as in the Continental European systems. According to the criterion of role structure, therefore, one arrives at a dichotomous rather than a threefold typology: the
11 Almond and Powell, 259 (italics omitted).
12 Almond, rapporteur, "A Comparative Study of Interest Groups and the Political Process," American Political Science Review, lii (March 1958), 275-77; Almond, "A Functional Approach to Comparative Politics," in Almond and James S. Coleman, eds., The Politics of the Developing Areas (Princeton i960), 42-43. See also Göran G. Lindahl, "Gabriel A. Almond's funktionella kategorier: En kritik," Statsvetens\aplig Tid~ s\rijt, No. 4 (1967), 263-72; and Constance E. van der Maesen and G. H. Schölten, "De řunctionele benadering van G. A. Almond bij het vergelijken van politieke stelsels," Acta Politica, 1 (1965-66), 220-26.
CONSOCIATIONAL DEMOCRACY
211
Scandinavian states must be grouped with the Anglo-American systems, and the other "in-between" states with the Continental European systems.
The application of the second criterion—political culture—leads to a similar result. Almond writes that the political culture in the Scandinavian and Low Countries is "more homogeneous and fusional of secular and traditional elements" than that in the Continental European systems.13 This is clearly true for the Scandinavian countries, which are, in fact, quite homogeneous and do not differ significantly from the homogeneous Anglo-American systems. But again, the other "in-between" countries are at least as fragmented into political subcultures —the families spirituelles of Belgium and Luxembourg, the zuilen of the Netherlands, and the Lager of Austria—as the Continental European states. Therefore, on the basis of the two criteria of political culture and role structure, the Western democracies can be satisfactorily classified into two broad but clearly bounded categories: (i) the Anglo-American, Old Commonwealth, and Scandinavian states; (2) the other European democracies, including France, Italy, Weimar Germany, the Low Countries, Austria, and Switzerland.
Fragmented but Stable Democracies
The second category of the above twofold typology is too broad, however, because it includes both highly stable systems (e.g., Switzerland and Holland) and highly unstable ones (e.g., Weimar Germany and the French Third and Fourth Republics). The political stability of a system can apparently not be predicted solely on the basis of the two variables of political culture and role structure. According to the theory of crosscutting cleavages, one would expect the Low Countries, Switzerland, and Austria, with subcultures divided from each other by mutually reinforcing cleavages, to exhibit great immobilism and instability. But they do not. These deviant cases of fragmented but stable democracies will be called "consociational democracies."14 In general, deviant case analysis can lead to the discovery of additional relevant variables, and in this particular instance, a third variable can account for the stability of the consociational democracies: the behavior of the political elites. The leaders of the rival subcultures may engage
13 "A Functional Approach," 42.
14 Cf. Johannes Althusius' concept of consociatio in his Politico, Methodice Digesta, and the term "consociational" used by David E. Apter, The Political Kingdom in Uganda: A Study in Bureaucratic Nationalism (Princeton 1961), 24-25.
212
WORLD POLITICS
in competitive behavior and thus further aggravate mutual tensions and political instability, but they may also make deliberate efforts to counteract the immobilizing and unstabilizing effects of cultural fragmentation. As a result of such overarching cooperation at the elite level, a country can, as Claude Ake states, "achieve a degree of political stability quite out of proportion to its social homogeneity."15
The clearest examples are the experiences of democratic Austria after the First World War and of pre-democratic Belgium in the early nineteenth century. The fragmented and unstable Austrian First Republic of the interwar years was transformed into the still fragmented but stable Second Republic after the Second World War by means of a con-sociational solution. As Frederick C. Engelmann states, "the central socio-political fact in the life of post-1918 Austria [was that] the Republic had developed under conditions of cleavage so deep as to leave it with a high potential for—and a sporadic actuality of—civil war," The instability caused by the deep cleavage and antagonism between the Catholic and Socialist Lager (subcultures) spelled the end of democracy and the establishment of a dictatorship. The leaders of the rival subcultures were anxious not to repeat the sorry experience of the First Republic, and decided to join in a grand coalition after the Second World War. According to Engelmann, "critics and objective observers agree with Austria's leading politicians in the assessment that the coalition was a response to the civil-war tension of the First Republic."16 Otto Kirchheimer also attributes the consociational pattern of Austria's post-1945 politics (until early 1966) to "the republic's historical record of political frustration and abiding suspicion,"17 Val R. Lorwin describes how the potential instability caused by subcultural cleavage was deliberately avoided at the time of the birth of independent Belgium: the Catholic and Liberal leaders had learned "the great lesson of mutual tolerance from the catastrophic experience of the Brabant Revolution of 1789, when the civil strife of their predecessors had so soon laid the country open to easy Habsburg reconquest. It was a remarkable and self-conscious 'union of the oppositions' that made the
15 Claude Ake, A Theory of Political Integration (Homewood 1967), 113. This possibility exists not only in the fragmented democracies, but also in fragmented predemo-cratic or nondemocratic systems, of course. See also Arend Lijphart, The Politics of Accommodation: Pluralism and Democracy in the Netherlands (Berkeley 1968), 1-15, 197-211.
16 Frederick C. Engelmann, "Haggling for the Equilibrium: The Renegotiation of the Austrian Coalition, 1959," American Political Science Review, lvi  (September 1962),
651'52-.
17 Kirchheimer, "The Waning of Opposition in Parliamentary Regimes," Social Research, xxiv (Summer 1957), 137.
CONSOCIATIONAL DEMOCRACY
213
revolution of 1830, wrote the Constitution of 1831, and headed the government in its critical years."18
The grand coalition cabinet is the most typical and obvious, but not the only possible, consociational solution for a fragmented system. The essential characteristic of consociational democracy is not so much any particular institutional arrangement as the deliberate joint effort by the elites to stabilize the system. Instead of the term "grand coalition" with its rather narrow connotation, one could speak of universal participation, or as Ralf Dahrendorf does, of a "cartel of elites."19 A grand coalition cabinet as in Austria represents the most comprehensive form of the cartel of elites, but one finds a variety of other devices in the other Western consociational democracies and, outside Western Europe, in the consociational politics of Lebanon, Uruguay (until early 1967), and Colombia. Even in Austria, not the cabinet itself but the small extra-constitutional "coalition committee," on which the top Socialist and Catholic leaders were equally represented, made the crucial decisions. In the Swiss system of government, which is a hybrid of the presidential and the parliamentary patterns, all four major parties are represented on the multi-member executive. In Uruguay's (now defunct) governmental system, fashioned after the Swiss model, there was coparúcipación of the two parties on the executive.
In the Colombian and Lebanese presidential systems, such a sharing of the top executive post is not possible because the presidency is held by one person. The alternative solution provided by the Lebanese National Pact of 1943 is that the President of the Republic must be a Maronite and the President of the Council a Sunni, thus guaranteeing representation to the country's two major religious groups. In Colombia, the Liberal and Conservative parties agreed in 1958 to join in a consociational arrangement in order to deliver the country from its recurrent civil wars and dictatorships. The agreement stipulated that the presidency would be alternated for four-year terms between the two parties and that there would be equal representation (paridad) on all lower levels of government. In the Low Countries, the cabinets are usually broadly based coalitions, but not all major subcultures are permanently represented. The typical consociational devices in these democracies are the advisory councils and committees, which, in spite
18 Lorwin, "Constitutionalism and Controlled Violence in the Modern State: The Case of Belgium" (paper presented at the annual meeting of the American Historical Association, San Francisco, 1965), 4 (italics added). For a description of the establishment of consociational democracy in the Netherlands, see Lijphart, The Politics of Accommodation, 103-12.
19 Dahrendorf, Society and Democracy in Germany (Garden City 1967), 276.
214
WORLD POLITICS
of their very limited formal powers, often have decisive influence. These councils and committees may be permanent organs, such as the powerful Social and Economic Council of the Netherlands—a perfect example of a cartel of economic elites—or ad hoc bodies, such as the cartels of top party leaders that negotiated the "school pacts" in Holland in 1917 and in Belgium in 1958.
The desire to avoid political competition may be so strong that the cartel of elites may decide to extend the consociational principle to the electoral level in order to prevent the passions aroused by elections from upsetting the carefully constructed, and possibly fragile, system of cooperation. This may apply to a single election or to a number of successive elections. The paridad and alternáciou principles in Colombia entail a controlled democracy for a period of sixteen years, during which the efHcacy of the right to vote is severely restricted. Another example is the Dutch parliamentary election of 1917, in which all of the parties agreed not to contest the seats held by incumbents in order to safeguard the passage of a set of crucial constitutional amendments; these amendments, negotiated by cartels of top party leaders, contained the terms of the settlement of the sensitive issues of universal suffrage and state aid to church schools. A parallel agreement on the suffrage was adopted in Belgium in 1919 without holding the constitutionally prescribed election at all.
Consociational democracy violates the principle of majority rule, but it does not deviate very much from normative democratic theory. Most democratic constitutions prescribe majority rule for the normal transaction of business when the stakes are not too high, but extraordinary majorities or several successive majorities for the most important decisions, such as changes in the constitution. In fragmented systems, many other decisions in addition to constituent ones are perceived as involving high stakes, and therefore require more than simple majority rule. Similarly, majority rule does not suffice in times of grave crisis in even the most homogeneous and consensual of democracies. Great Britain and Sweden, both highly homogeneous countries, resorted to grand coalition cabinets during the Second World War. Julius Nyerere draws the correct lesson from the experience of the Western democracies, in which, he observes, "it is an accepted practice in times of emergency for opposition parties to sink their differences and join together in forming a national government."20 And just as the forma-
20 Nyerere, "One-Party Rule," in Paul E. Sigmund, Jr., ed., The Ideologies of the Developing Nations (New York 1963), 199.
CONSOCIATIONAL DEMOCRACY
215
tion of a national unity government is the appropriate response to an external emergency, so the formation of a grand coalition cabinet or an alternative form of elite cartel is the appropriate response to the internal crisis of fragmentation into hostile subcultures.
Furthermore, the concept of consociational democracy is also in agreement with the empirical "size principle," formulated by William H. Riker. This principle, based on game-theoretic assumptions, states: "In social situations similar to ^-person, zero-sum games with side-payments [private agreements about the division of the payoff], participants create coalitions just as large as they believe will ensure winning and no larger." The tendency will be toward a "minimum winning coalition," which in a democracy will be a coalition with bare majority support—but only under the conditions specified in the size principle. The most important condition is the zero-sum assumption: "only the direct conflicts among participants are included and common advantages are ignored."21 Common advantages will be completely ignored only in two diametrically opposite kinds of situations: (i) when the participants in the "game" do not perceive any common advantages, and when, consequently, they are likely to engage in unlimited warfare; and (2) when they are in such firm agreement on their common advantages that they can take them for granted. In the latter case, politics literally becomes a game. In other words, the zero-sum condition and the size principle apply only to societies with completely homogeneous political cultures and to societies with completely fragmented cultures. To the extent that political cultures deviate from these two extreme conditions, pressures will exist to fashion coalitions and other forms of cooperation that are more inclusive than the bare "minimum winning coalition" and that may be all-inclusive grand coalitions.
Almond aptly uses the metaphor of the game in characterizing the Anglo-American systems: "Because the political culture tends to be homogeneous and pragmatic, [the political process] takes on some of the atmosphere of a game. A game is a good game when the outcome is in doubt and when the stakes are not too high. When the stakes are too high, the tone changes from excitement to anxiety."22 Political contests in severely fragmented societies are indeed not likely to be "good games." But the anxieties and hostilities attending the political process may be countered by removing its competitive features as much
21 William H. Riker, The Theory o j Political Coalitions (New Haven 1962), 29, 32-33.
22 Almond, "Comparative Political Systems," 398-99.
216
WORLD POLITICS
as possible. In consociational democracies, politics is treated not as a game but as a serious business.
Factors Conducive to Consociational Democracy
Consociational democracy means government by elite cartel designed to turn a democracy with a fragmented political culture into a stable democracy. Efforts at consociationalism are not necessarily successful, of course: consociational designs failed in Cyprus and Nigeria, and Uruguay abandoned its Swiss-style consociational system. Successful consociational democracy requires: (i) That the elites have the ability to accommodate the divergent interests and demands of the subcultures. (2) This requires that they have the ability to transcend cleavages and to join in a common effort with the elites of rival subcultures. (3) This in turn depends on their commitment to the maintenance of the system and to the improvement of its cohesion and stability. (4) Finally, all of the above requirements are based on the assumption that the elites understand the perils of political fragmentation. These four requirements are logically implied by the concept of consociational democracy as defined in this paper. Under what conditions are they likely to be fulfilled ? An examination of the successful consociational democracies in the Low Countries, Switzerland, Austria, and Lebanon suggests a number of conditions favorable to the establishment and the persistence of this type of democracy. These have to do with inter-subcultural relations at the elite level, inter-subcultural relations at the mass level, and elite-mass relations within each of the subcultures.
RELATIONS AMONG THE ELITES OF THE SUBCULTURES
It is easier to assess the probability of continued success of an already established consociational democracy than to predict the chance of success that a fragmented system would have if it were to attempt consociationalism. In an existing consociational democracy, an investigation of the institutional arrangements and the operational code of inter-elite accommodation can throw light on the question of how thorough a commitment to cooperation they represent and how effective they have been in solving the problems caused by fragmentation. The length of time a consociational democracy has been in operation is also a factor of importance. As inter-elite cooperation becomes habitual and does not represent a deliberate departure from competitive responses to political challenges, consociational norms become more
CONSOCIATIONAL DEMOCRACY
217
firmly established. And, as Gerhard Lehmbruch states, these norms may become an important part of "the political socialization of elites and thus acquire a strong degree of persistence through time."23
There are three factors that appear to be strongly conducive to the establishment or maintenance of cooperation among elites in a fragmented system. The most striking of these is the existence of external threats to the country. In all of the consociational democracies, the cartel of elites was either initiated or greatly strengthened during periods of international crisis, especially the First and Second World Wars. During the First World War, the comprehensive settlement of the conflict among Holland's political subcultures firmly established the pattern of consociational democracy. "Unionism"—i.e., Catholic-Liberal grand coalitions—began during Belgium's struggle for independence in the early nineteenth century, but lapsed when the country appeared to be out of danger. As a result of the First World War, unionism was resumed and the Socialist leaders were soon admitted to the governing cartel. The Second World War marked the beginning of consociational democracy in Lebanon: the National Pact—the Islamo-Christian accord that provided the basis for consociational government for the country—was concluded in 1943. In Switzerland, consociational democracy developed more gradually, but reached its culmination with the admission of the Socialists to the grand coalition of the Federal Council, also in 1943. The Austrian grand coalition was formed soon after the Second World War, when the country was occupied by the allied forces. In all cases, the external threats impressed on the elites the need for internal unity and cooperation. External threats can also strengthen the ties among the subcultures at the mass level and the ties between leaders and followers within the subcultures.
A second factor favorable to consociational democracy, in the sense that it helps the elites to recognize the necessity of cooperation, is a multiple balance of power among the subcultures instead of either a dual balance of power or a clear hegemony by one subculture. When one group is in the majority, its leaders may attempt to dominate rather than cooperate with the rival minority. Similarly, in a society with two evenly matched subcultures, the leaders of both may hope to achieve their aims by domination rather than cooperation, if they ex-
23 Lehmbruch, "A Non-Competitive Pattern of Conflict Management in Liberal Democracies: The Case of Switzerland, Austria and Lebanon" (paper presented at the Seventh World Congress of the International Political Science Association, Brussels, 1967), 6. See also Lehmbruch, Proporz democratic Politisches System und politische Kultur in der Schweiz und in Österreich (Tübingen 1967).
218
WORLD POLITICS
pect to win a majority at the polls. Robert Dahl argues that for this reason it is doubtful that the consociational arrangement in Colombia will last, because "the temptation to shift from coalition to competition is bound to be very great."24 When political parties in a fragmented society are the organized manifestations of political subcultures, a multiparty system is more conducive to consociational democracy and therefore to stability than a two-party system. This proposition is at odds with the generally high esteem accorded to two-party systems. In an already homogeneous system, two-party systems may be more effective, but a moderate multiparty system, in which no party is close to a majority, appears preferable in a consociational democracy. The Netherlands, Switzerland, and Lebanon have the advantage that their subcultures are all minority groups. In the Austrian two-party system, consociational politics did work, but with considerable strain. Lehm-bruch states: "Austrian political parties are strongly integrated social communities . . . and the bipolar structure of the coalition reinforced their antagonisms."25 The internal balance of power in Belgium has complicated the country's consociational politics in two ways. The Catholic, Socialist, and Liberal subcultures are minorities, but the Catholics are close to majority status. The Catholic party actually won a legislative majority in 1950, and attempted to settle the sensitive royal question by majority rule. This led to a short civil war, followed by a return to consociational government. Moreover, the Belgian situation is complicated as a result of the linguistic cleavage, which cuts across the three spiritual families. The linguistic balance of power is a dual balance in which the Walloons fear the numerical majority of the Flemings, while the Flemings resent the economic and social superiority of the Walloons.
Consociational democracy presupposes not only a willingness on the part of elites to cooperate but also a capability to solve the political problems of their countries. Fragmented societies have a tendency to im-mobilism, which consociational politics is designed to avoid. Nevertheless, decision-making that entails accommodation among all subcultures is a difficult process, and consociational democracies are always threatened by a degree of immobilism. Consequently, a third favorable factor to inter-elite cooperation is a relatively low total load on the decision-making apparatus. The stability of Lebanon is partly due to its productive economy and the social equilibrium it has maintained so
24 Dahl, Political Oppositions in Western Democracies (New Haven 1966), 337.
25 Lehmbruch, 8.
CONSOCIATIONAL DEMOCRACY
219
far, but it may not be able to continue its successful consociational politics when the burdens on the system increase. Michael C. Hudson argues that the Lebanese political system is "attuned to incessant adjustment among primordial groups rather than policy planning and execution." As a result, its "apparent stability ... is deceptively precarious: social mobilization appears to be overloading the circuits of the Lebanese political system."26 In general, the size factor is important in this respect: the political burdens that large states have to shoulder tend to be disproportionately heavier than those of small countries. Ernest S. Griffith argues that "democracy is more likely to survive, other things being equal, in small states. Such states are more manageable. . . ."27 In particular, small states are more likely to escape the onerous burdens entailed by an active foreign policy. Lehmbruch states that the Swiss, Austrian, and Lebanese cases "show that the preservation of the inner equilibrium presupposes a reduction of external demands to the political system." And he even goes so far as to conclude that the type of politics found in these three countries "seems to work in small states only."28
INTER-SUBCULTURAL RELATIONS AT THE MASS LEVEL
The political cultures of the countries belonging to Almond's Continental European type and to the consociational type are all fragmented, but the consociational countries have even clearer boundaries among their subcultures. Such distinct lines of cleavage appear to be conducive to consociational democracy and political stability. The explanation is that subcultures with widely divergent outlooks and interests may coexist without necessarily being in conflict; conflict arises only when they are in contact with each other. As Quincy Wright states: "Ideologies accepted by different groups within a society may be inconsistent without creating tension; but if ... the groups with inconsistent ideologies are in close contact . . . the tension will be great."29 David Easton also endorses the thesis that good social fences may make good political neighbors, when he suggests a kind of voluntary apartheid policy as the best solution for a divided society: "Greater success
28 Hudson, "A Case of Political Underdevelopment," Journal of Politics, xxix (November 1967), 836.
27 Griffith, "Cultural Prerequisites to a Successfully Functioning Democracy," American Political Science Review, l (March 1956), 102.
28 Lehmbruch, 9.
29 Wright, 'The Nature of Conflict," Western Political Quarterly, iv (June 1951), 196.
220
WORLD POLITICS
may be attained through steps that conduce to the development of a deeper sense of mutual awareness and responsiveness among encapsulated cultural units." This is "the major hope of avoiding stress."30 And Sidney Verba follows the same line of reasoning when he argues that political and economic modernization in Africa is bringing "differing subcultures into contact with each other and hence into conflict."31
This argument appears to be a direct refutation of the overlapping-memberships proposition, but by adding two amendments to this proposition the discrepancy can be resolved. In the first place, the basic explanatory element in the concept of consociational democracy is that political elites may take joint actions to counter the effects of cultural fragmentation. This means that the overlapping-memberships propositions may become a self-denying hypothesis under certain conditions. Secondly, the view that any severe discontinuity in overlapping patterns of membership and allegiance is a danger to political stability needs to be restated in more refined form. A distinction has to be made between essentially homogeneous political cultures, where increased contacts are likely to lead to an increase in mutual understanding and further homogenization, and essentially heterogeneous cultures, where close contacts are likely to lead to strain and hostility. This is the distinction that Walker Connor makes when he argues that "increased contacts help to dissolve regional cultural distinctions within a state such as the United States. Yet, if one is dealing not with minor variations of the same culture, but with two quite distinct and self-differentiating cultures, are not increased contacts between the two apt to increase antagonisms?"32 This proposition can be refined further by stating both the degree of homogeneity and the extent of mutual contacts in terms of continua rather than dichotomies. In order to safeguard political stability, the volume and intensity of contacts must not exceed the commensurate degree of homogeneity. Karl W. Deutsch states that stability depends on a "balance between transaction and integration" because "the number of opportunities for possible violent conflict will increase with the volume and range of mutual transactions."33 Hence, it may be desirable to keep transactions among antagonistic subcultures
30Easton, A Systems Analysis of Political Life (New York 1965), 250-51 (italics added). See also G. H. Schölten, "Het vergelijken van federates met behulp van systeem-analyse," Acta Politica, 11 (1966-67), 51-68.
31 Verba, "Some Dilemmas in Comparative Research," World Politics, xx (October 1967), 126 (italics added).
32 Connor, "Self-Determination: The New Phase," World Politics, xx (October 1967), 49-50.
33 Deutsch, Political Community at the International Level (Garden City 1954), 39.
CONSOCIATIONAL DEMOCRACY
221
in a divided society—or, similarly, among different nationalities in a multinational state—to a minimum.
ELITE-MASS RELATIONS  WITHIN THE  SUBCULTURES
Distinct lines of cleavage among the subcultures are also conducive to consociational democracy because they are likely to be concomitant with a high degree of internal political cohesion of the subcultures. This is vital to the success of consociational democracy. The elites have to cooperate and compromise with each other without losing the allegiance and support of their own rank and file. When the subcultures are cohesive political blocs, such support is more likely to be forthcoming. As Hans Daalder states, what is important is not only "the extent to which party leaders are more tolerant than their followers" but also the extent to which they "are yet able to carry them along."34
A second way in which distinct cleavages have a favorable effect on elite-mass relations in a consociational democracy is that they make it more likely that the parties and interest groups will be the organized representatives of the political subcultures. If this is the case, the political parties may not be the best aggregators, but there is at least an adequate articulation of the interests of the subcultures. Aggregation of the clearly articulated interests can then be performed by the cartel of elites. In Belgium, the three principal parties represent the Catholic, Socialist, and Liberal spiritual families, but the linguistic cleavage does not coincide with the cleavages dividing the spiritual families, and all three parties have both Flemings and Walloons among their followers, Lorwin describes the situation as follows: "The sentimental and practical interests of the two linguistic communities are not effectively organized, and the geographical regions have no administrative or formal political existence. There are no recognized representatives qualified to formulate demands, to negotiate, and to fulfill commitments."35 The religious and class issues have been effectively articulated by the political parties and have by and large been resolved, but the linguistic issue has not been clearly articulated and remains intractable. In Switzerland, the parties also represent the religious-ideological groups rather than the linguistic communities, but much of the country's decentralized political life takes place at the cantonal level, and most of the cantons are linguistically homogeneous.
A final factor which favors consociational democracy is widespread
34 Daalder, 69.
85 Lorwin, "Belgium: Religion, Class, and Language in National Politics," in Dahl, ed., Political Oppositions in Western Democracies, 174.
222
WORLD POLITICS
approval of the principle of government by elite cartel. This is a very obvious factor, but it is of considerable importance and deserves to be mentioned briefly. For example, Switzerland has a long and strong tradition of grand coalition executives, and this has immeasurably strengthened Swiss consociational democracy. On the other hand, the grand coalition in Austria was under constant attack by critics who alleged that the absence of a British-style opposition made Austrian politics "undemocratic." This attests to the strength of the British system as a normative model even in fragmented political systems, where the model is inappropriate and undermines the attempt to achieve political stability by consociational means.
Centripetal and Centrifugal Democracies
An examination of the other two types of the threefold typology of democracies in the light of the distinguishing characteristics of consociational democracy can contribute to the clarification and refinement of all three types and their prerequisites. In order to avoid any unintended geographical connotation, we shall refer to the homogeneous and stable democracies as the centripetal (instead of the Anglo-American) democracies, and to the fragmented and unstable ones as the centrifugal (instead of the Continental European) democracies. The centrifugal democracies include the French Third and Fourth Republics, Italy, Weimar Germany, the Austrian First Republic, and the short-lived Spanish Republic of the early 1930's. The major examples of centripetal democracy are Great Britain, the Old Commonwealth countries, the United States, Ireland, the Scandinavian states, and the postwar Bonn Republic in Germany.
The French Fourth Republic is often regarded as the outstanding example of unstable, ineffective, and immobilist democracy, but the explanation of its political instability in terms of cultural fragmentation has been criticized on two grounds. In the first place, Eric A. Nord-linger rejects the argument that the "ideological inundation of French politics" and its "fragmented party system" were responsible for its chronic instability; he states that this explanation conveniently overlooks "the way in which the game of politics is actually played in France. Although ideologism pervades the parties' electoral and propaganda efforts, this public ideological posturing of French politicians does not prevent them from playing out their game of compromise in the Assembly and its couloirs. In fact, the political class thinks of compromise as a positive principle of action, with parliamentary activity
CONSOCIATIONAL DEMOCRACY
223
largely revolving around nonideological squabbles. . . ."36 The elites of the center parties that supported the Republic fulfilled to some extent all of the logical prerequisites for consociational democracy except the most important one: they lacked the ability to forge effective and lasting solutions to pressing political problems. They indeed played a nonideological game, but, as Nathan Leites observes, with a "well-developed capacity for avoiding their responsibility."37 In other words, they were nonideological, but not constructively pragmatic. To turn a centrifugal into a consociational democracy, true statesmanship is required. Moreover, it is incorrect to assume that, because the elites were not divided by irreconcilable ideological differences, mass politics was not ideologically fragmented either.38
The second criticism of the cultural fragmentation thesis alleges, on the basis of independent evidence, that not only at the elite level but also at the mass level, ideology played a negligible role in France. Philip E. Converse and Georges Dupeux demonstrate that the French electorate was not highly politicized and felt little allegiance to the political parties.39 But the lack of stable partisan attachments does not necessarily indicate that the political culture was not fragmented. Duncan MacRae argues persuasively that political divisions did extend to the electorate as a whole in spite of the apparent "lack of involvement of the average voter." Even though political allegiances were diffuse, there were "relatively fixed and non-overlapping social groupings" to which "separate leaders and separate media of communication had access."40 The combination of fragmentation into subcultures and low politiciza-tion can in turn be explained by the negative French attitude toward authority. Stanley Hoffmann speaks of "potential insurrection against authority," and Michel Crozier observes that this attitude makes it "impossible for an individual of the group to become its leader."41
36 Nordlinger, "Democratic Stability and Instability: The French Case," World Politics, xviii (October 1965), 143.
37 Leites, On the Game of Politics in France (Stanford 1959), 2.
38 Nor does the reverse assumption hold true. Giovanni Sartori relates the instability of Italian democracy to "poor leadership, both in the sense that the political elites lack the ability for problem-solving and that they do not provide a generalized leadership." This weakness of leadership, he continues, "is easily explained by the fragmentation of the party system and its ideological rigidity." ("European Political Parties: The Case of Polarized Pluralism," in LaPalombara and Weiner, eds., Political Parties and Political Development, 163.) The example of the consociational democracies shows that this is not a sufficient explanation.
39 Converse and Dupeux, "Politicization of the Electorate in France and the United States," Public Opinion Quarterly, xxvi (Spring 1962), 1-23.
40 MacRae, Parliament, Parties, and Society in France: igqo-igtf (New York 1967),
333-
41 Hoffmann and others, In Search of France (Cambridge 1963), 8 (italics omitted); Crozier, The Bureaucratic Phenomenon (Chicago 1964), 220.
224
WORLD POLITICS
Strong cohesion within the subcultures was mentioned earlier as a factor conducive to consociational democracy; the lack of it in France can explain both that the French people were fragmented but at the same time not politically involved, and that the political elites did not have the advantage of strong support from the rank and file for constructive cooperation.
On the other hand, the example of France also serves to make clear that the lack of problem-solving ability as a cause of political instability must not be overstated. After all, as Maurice Duverger points out, in spite of all of the Fourth Republic's flaws and weaknesses, it "would have continued to exist if it had not been for the Algerian war."42 The critical factor was the too-heavy burden of an essentially external problem on the political system. Similarly, the fragmented Weimar Republic might have survived, too, if it had not been for the unusually difficult problems it was faced with.
Germany's experience with democracy also appears to throw some doubt on our threefold typology and the theory on which it is based. Weimar Germany was a centrifugal democracy but the Bonn Republic can be grouped with the centripetal democracies. In explaining this extraordinary shift, we have to keep in mind that cultural fragmentation must be measured on a continuum rather than as a dichotomy, as we have done so far. The degree of homogeneity of a political culture can change, although great changes at a rapid pace can normally not be expected. Three reasons can plausibly account for the change from the fragmented political culture of the unstable Weimar Republic to the much more homogeneous culture of the Bonn Republic: (i) the traumatic experiences of totalitarianism, war, defeat, and occupation;
(2) "conscious manipulative change of fundamental political attitudes," which, as Verba states, added up to a "remaking of political culture";43
(3) the loss of the eastern territories, which meant that, as Lipset argues, "the greater homogeneity of western Germany now became a national homogeneity."44
The degree of competitive or cooperative behavior by elites must also be seen as a continuum. Among the consociational democracies, some are more consociational than others; and many centripetal democracies have some consociational features. The phenomenon of war-
42 Duverger, "The Development of Democracy in France," in Henry W. Ehrmann, ed., Democracy in a Changing Society (New York 1964), 77.
43 Verba, "Germany: The Remaking of Political Culture," in Lucian W. Pye and Verba, eds., Political Culture and Political Development (Princeton 1965), 133.
44 Lipset, The First New Nation: The United States in Historical and Comparative Perspective (New York 1963), 292.
CONSOCIATIONAL DEMOCRACY
225
time grand coalition cabinets has already been mentioned. The temporary Christian Democratic-Socialist grand coalition under Chancellor Kiesinger falls in the same category. In fact, the stability of the centripetal democracies depends not only on their essentially homogeneous political cultures but also on consociational devices, to the extent that a certain degree of heterogeneity exists. The alternation of English-speaking and French-speaking leaders of the Liberal party in Canada may be compared with the Colombian device of alternación. In the United States, where, as Dahl states, "the South has for nearly two centuries formed a distinctive regional subculture,"45 cultural fragmentation led to secession and civil war. After the Civil War, a consociational arrangement developed that gave to the South a high degree of autonomy and to the Southern leaders—by such means as chairmanships of key Congressional committees and the filibuster—a crucial position in federal decision-making. This example also shows that, while consociational solutions may increase political cohesion, they also have a definite tendency to lead to a certain degree of immobilism.
Even in Denmark, which is among the most homogeneous of the centripetal democracies, one finds considerable consociationalism. This does not appear in grand coalition cabinets—in fact, Denmark is known for its long periods of government by minority cabinets—but in the far-reaching search for compromise in the legislature. The rule of the game prescribes that the top leaders of all four major parties do their utmost to reach a consensus. This is glidningspoliti^ which Gerald R. McDaniel translates as the "politics of smoothness"46—an apt characterization of consociational politics.
45 Dahl, 358.
46 McDaniel, The Danish Unicameral Parliament (unpubl. Ph.D. diss., University of California, Berkeley 1963), iv.