1
INTRODUCTION
Robert A. Dahl
The Argument
i. Let mc begin with an elementary proposition about politics that I think no one seriously questions. No government receives indefinitely the total support of the people over whom it asserts its jurisdiction. Certainly the government of a large collection of people, such as the government of a country, is never completely supported in all that it does by all the people whom it claims to govern. In no country, in short, does everyone have the same preferences as to the conduct of the government, using the term conduct in its broadest sense. So much is, I believe, obvious.
2. It is an equally well-grounded observation, I believe — though one more often denied, even in highly influential theories of politics — that differences in what people think they want from the government that rules over them tend toward diversity and multiplicity rather than toward bipolarity: to many groupings rather than merely two. Let me call what people think they want the government to be, or to do, their political preferences, or, if you prefer, their political interests. Whenever the barriers to the expression and organization of political preferences are low, one should expect (as the usual tiling) the emergence of a multiplicity of camps, whereas polarization into two internally cohesive and unified camps would be rare.
Although the second proposition may be more debatable than the first, the experiences of the countries described in the chapters of this book lend great weight to it, as I hope to show in a moment.
3. Because people are not in perfect accord as to their political preferences, every political system, if it is to endure, must provide ways for determining which (or whose) political preferences the government responds to. It will be useful to consider two extreme possibilities. At one extreme, a government might respond to the political .preferences of only one person (or perhaps to a tiny and wholly unified minority); it would ignore or override all other preferences. A system of this kind might be
1
i Robot A. Ihhl
• idled :i pure- hegemony. Ai tlie oilier exireme, tlie political preferences n| everyone might be weighted equally, fLod tlie government would respond Always? io ilie preferences of the gre;itesi number. A system of this kind rpighi lie called a pure egalitarian democracy.
4. All political systems in some respects constrain the expression, organization, representation, and satisfaction of political preferences. Given 1 he existence of disagreements as to what the government should do (that is 10 say, given the human condition, if my first assumption is valid), even an egalitarian democracy cannot respond fully to the preferences of
I mill 1 he greatest number and the smaller number who disagree with them. During any given period, therefore, a political system will contain some people who, if there were no barriers or costs to their doing so, would be Opposed to the conduct of the government.
•;. Political systems vary a great deal, however, in the barriers or opportunities they provide for the expression, organization, and representation of political preferences and thus in the opportunities available to potential oppositions.
To cope adequately with these variations among different political systems poses some difficulties. For example, a moment ago I offered two extreme types, hegemony and egalitarian democracy; yet both are purely theoretical types. Although neither exists in pure form, they do hint at a pussiblc continuum, running from one to the other. However, it is more
ii ecu rate for our purposes to think not just of one but of two dimensions. One is the dimension of liberalization or public contestation: the extent to which institutions are openly available, publicly employed, and guaranteed to. at least some members of the political system who wish to contest the conduct of the government. The other is the dimension of participation or inclusiveness: the proportion of the population (or of adults) who are entitled to participate on a more or less equal basis in controlling and contesting the conduct of the government — that is, who arc entitled to participate in the system of public contestation.
In practice, political systems (particularly if we include historical ones) seem to vary over most of the space enclosed by these two hypothetical dimensions, except perhaps for space near the edges. Although the space can be carved up and labeled in many different ways, for the purposes of this essay it seems to me preferable to run the risk of oversimplification by distinguishing no more than three very general categories (see figure i.i).
One: Regimes that impose the most severe limits on the expression, organization, and representation of political preferences and on the opportunities available to opponents of the government. In these systems, individuals are prohibited from expressing public opposition to the incumbent leaders, to their policies and ideology, and to the major social, eco-
liitroJttctioii 3 Figure 1.1 Types of Regimes
Inclusiveness (Participation)
nomic, and political structures. Organized dissent and opposition are prohibited in any form. I shall call such systems hegemonies.
Two: Regimes that impose the fewest restraints on the expression, organization, and representation of political preferences and on the opportunities available to opponents of government. Most individuals are effectively protected in their right to express, privately or publicly, their opposition to the government, to organize, to form parties, and to compete in elections where voting is secret, unintimidated, and honestly counted and where the results of elections are binding according to well-established rules. Ordinarily, the use of violent means is forbidden, and in some cases it is punishable to advocate the use of violence for political purposes. I shall call such systems polyarchies.
Three: Mixed regimes, which in various ways approaclxJb^mojrues or polyarchies. Among the countries of the world, these are the most numerous. I shall sometimes call an intermediate regime that more closely
4 Robert A. Ihihl
ii|i|iii>;ii lu-s ;i hegemony :i n&ar hc&cviony; one chat more closely :ip-|>icp,u lies polyarchy, ;> near polyarchy; aiul one that provides extensive Opportunities lor public contestation t<> a very small elite, a competitive oliiiiirchy.
r>, In any given country, the lower the barriers (the greater the op-poriunities) for expressing, organizing, and representing political preferences, (he greater the number and variety of preferences and interests I Inn will be represented in policy making. The number and variety of preferences and interests represented in policy making are therefore greater if the political regime is a polyarchy than if it is a mixed regime, nml greater under a mixed regime than under a hegemony. Hence the transformation of a hegemony into a mixed regime or polyarchy, or a mixed regime into a polyarchy, would increase the number and variety of preferences and interests represented in policy making.
7. Polarization is an obvious threat to any regime, particularly a polyarchy. But the increase in the number and variety of preferences and interests represented in policy making, as the barriers for expressing, organizing, and representing political preferences are reduced, also helps to create a special risk for each type of regime. By a risk, I mean a sub-si initial probability that a situation will occur which will be seen by those who support the regime as creating a problem for which all solutions entail high costs and which may, if severe, lead to a transformation of the regime.
I propose to take for granted points 1, 3, 4, and 5 of the argument, since these seem to me so obvious or axiomatic or contingently useful that it would be tedious and redundant to discuss them further. Let me then concentrate briefly on points 2, 6, and 7 and in so doing draw upon the later chapters of this volume by way of evidence and illustration.
Polarization and Plurality
That the forces of conflict in society are polarizing, in the sense that they exhibit an irresistible tendency toward a cleavage into two enduring sets of antagonists, is a view that seems to have a powerful attraction. Although it is perhaps best set out in Marxist thought, its reach extends far beyond any particular political philosophy or ideology.
Usually this current of thought stresses the strictly economic or purely socioeconomic axis of conflict; thus societies tend to be bipolar along a socioeconomic line of cleavage. The fact is, however, that the great bulk of experience in the past century — whether in traditional, modernizing, industrial, or postindustrial societies — seems to run counter to this
Introduction 5
hypothesis. It is rare for a country to divide into two camps along socioeconomic lines; it is rare, indeed, for a country to divide into only two camps along any lines. But it is commonplace for countries to display wore than two sets of conflicting interests.
The explanation seems to be that most countries have some conflicts based at least in part on differences in language, religion, race, or ethnic group. Frequently these conflicts cut, wholly or partly, across one another and across differences of status, function, or reward based on economic activity. And they do not disappear or become attenuated with modernization. It is, rather, the conflicts derived from economic differences that often become attenuated, leaving more room, so to speak, for the other conflicts to occur. What is more, different political preferences or interests stimulated by differences in economic function, status, or reward seldom divide a people into only two groups, each with more or less identical preferences. Much more often, it seems, when people respond to economic factors, they divide into more than two groups — frequently, in fact, into a bewildering multiplicity of contesting interests.
What seems to vary most from one country to another is, not so much this tendency toward diversity rather than polarization, but rather the extent to which the societal cloth is cut up into separate pieces by a succession of cuts or, alternatively, is woven with strong strands that hold one part to another. In language more familiar to social scientists (but which reverses the metaphor I have just used) I am of course speaking of the extent to which cleavages "reinforce" one another and hence produce "segmentation" or "fragmentation," or instead "crosscut" one another and thus produce a greater tendency toward cohesion.
The important — and, one would have supposed, by now obvious — point is the comparative rarity of strongly bipolar social conflicts, particularly along an economic axis, and the comparative frequency of conflict involving more than two sets of contestants.
Thus in describing the social basis for opposition in the single-party states of tropical Africa, Foltz refers to the impact of a half-dozen kinds of differences. Language and ethnic group reinforce each other to make tribal differences perhaps the dominant source of conflict; yet differences in region, mode of livelihood, religion, caste, modernity, urban-rural residence, and education reduce the cohesiveness of the tribe and enable politicians to build intertribal coalitions. As elsewhere in the world — India, Canada, Belgium, or Spain, for instance — the most serious conflicts are likely to arise where regional differences coincide with some of the other differences, most notably in language and ethnic groups. It is this kind of combination, in fact, that seems much more likely to lead to polarization or segmentation than any purely socioeconomic difference alone has ever
Robert A. Dahl
succeeded in doing. Throughout the world, the overt or l;ircnr forces of iiiitioiinlistn continue to be the most powerful sources of political antagonisms.
■Segmentation and pluralism arc central themes in Indian society, as Kiitluiri shows in his essay. In India there has not yet been any clear poliiri/alion of conflict, cither between the dominant Congress party and its opponents or among the incredible variety of castes, subcastes, and <:thnic:, linguistic, and regional groups. Although the differences between the preponderant Hindu population and the Muslim minority are an ever present source of danger, these do not form a principal line of cleavage in political conflicts; most political conflicts occur among Hindus, not between Hindus and Muslims. As to socioeconomic cleavages, the peasantry is of course numerically overwhelming, but it is not a unified force; and conflict between the urban working class and employers is simply one cleavage among many. As Kothari says, "India is still far from becoming a mass society; its pluralism also is of a different kind from t hat in Western democracies. It is less a confrontation between aggregated subsystems and more a coexistence between historically autonomous diversities and identities. Hence the great variety and diffusion of oppositions and their lack of clear-cut boundaries."
Tropical Africa and India are, of course, areas where traditional society anil its differences exert an extraordinarily powerful pull, and it would not be unreasonable to argue that as they undergo the changes that come with modernization, traditional societies would be most prone to a diversity of cleavages. Latin America is hardly a region of traditionalism in quite the same sense as India or Africa, though it is an area where the forces Of modernization are nearly everywhere bringing about rapid changes in previous patterns of social, economic, and political life. As Dix points out, in the early decades of this century economic growth and urbanization in a number of Latin American countries, stimulating the expansion of new social sectors — "the industrialists, the middle sectors, and urban and mining proletariats" — helped make traditional political coalitions less effectual and increased the political importance of "categorical groupings based on occupation and economic interest." Yet these bases of conflict and coalition have "far from overwhelmed the past. Modern institutions and orientations often coexist with others that are traditional, sometimes in surprisingly compatible ways." Wealthy industrial entrepreneurs do not see eye to eye with upper-class landowners. The middle classes are at best a nominal category, not cohesive social classes; they are divided in their attitudes toward public policy, in ideology, in their commitment to Catholicism or anticlericalism, and in their personal and party loyalties. The lower, or "popular," classes, to use Dix's term, are divided by more than an urban-rural cleavage; in the countryside, traditional agriculture
Introduction
vies with modern agriculture, and in the city the skilled worker or the worker employed in a modern industry feels little or no solidarity with the unskilled, with the newly arriving peasant from the countryside, or with the "marginals living in the jerry-built slums that surround many of the large urban centers." As in many Catholic societies, women are more likely to support the church than men, while anticlericals are predominantly male. And of course the generation gap (or gaps) is (or are) us marked in Latin America as anywhere else.
Despite palpable evidence of this kind in Latin America as elsewhere in the world, many political activists and intellectuals continue to interpret politics as a struggle between classes. Certainly socioeconomic differences a re sufficiently important as bases of cleavage and conflict in Latin America (as elsewhere) to give considerable credibility to their interpretations. The fact is, however, that other lines of cleavage, and the plurality of cleavages based on socioeconomic differences, so complicate political life in these countries that when political movements base their strategies on a pure class interpretation, they suffer grievous political defeat. Despite the presence of poverty, revolutionary rhetoric, and severe conflict, the Marxist parties of Latin America have failed to acquire a leading role except in Chile, Cuba, and Guatemala, where, in Dix's words, they were "tails to the kites of non-Communist nationalistic revolutions." In the rare cases where a Communist party has acquired a substantial electoral following, as in Chile, or where it is relatively small and hopes to expand its mass following in order to compete effectively in elections, as in Venezuela, party leaders seem to have recognized the failure of the classic strategy based on hopes of polarization. Like Communist leaders in Italy and France, they now seek alliances that would lead to a broad and-heterogeneous collection of interests capable of winning an electoral majority.
Many people who cling to the belief that politics in evil and still unredeemed societies must embody an inherent motion toward socioeconomic polarization, apocalypse, and redemption can protect their vision from doubts, which evidence of the kind I have just discussed might otherwise create, by pointing to the persistence of rich and poor, the divergence in their "real interests," the obvious economic origins of many conflicts, and, above all, the expectation that the process of polarization will occur in some indefinite future. None of these arguments can be proved false; in particular, locating apocalypse and redemption in the future surely helps to preserve this faith, as it has others.
It would seem reasonable that as the United States lurched recklessly through the process of becoming an advanced industrial society and then went on to become a postindustrial society, the processes of polarization, apocalypse, and redemption, if they were ever to occur anywhere,
H
Robert A. Ihibl
Introduction
9
would suiily have occurred (here first of all. I5ur oven I lie Civil War docs nut (ii I lie schema. The inconvenient experience of rhe United States hits hern deull with hy ignoring it, by treating it as a unique case, or by postponing the American apocalypse (and redemption) to a yet more distant future.
An alternative test is provided by Japan, which like Russia has in less than a century traveled the path from an agrarian-based feudal society to n highly industrial economy and society — and soon, no doubt, to a post-indusirial economy and society. As in Western Europe in the second half of the nineteenth century, so too in Japan during the first third of the t wenticrh century, industrialization created some of the familiar conditions lor the exacerbation of twentieth-century class conflict. As Leiserson de-scrihes Taisho Democracy, it appears to have consisted of a competitive oligarchy undergoing a measure of liberalization and democratization; Inn before this evolutionary process was completed and consolidated, it was interrupted by Showa Fascism, which repressed the public expression of many latent conflicts. It is conceivable, certainly, that Taisho Democracy would ultimately have seen the emergence of political conflict highly polarized along clear-cut socioeconomic lines, but no such development seems to have occurred by the time Showa Fascism took over.
Japanese politics since World War II appears to be more polarized than American politics. At the level of electoral conflicts, the Liberal-Democrats and conservative independents gain the preponderant support of large business, farmers, and the national bureaucracy, while their opponents gain more supporters among labor, youth, intellectuals, and the lower middle classes. But as Leiserson's analysis makes clear, not only is this too simple a description, but the results have hardly been a polarization of political conflict along clear-cut socioeconomic lines. Voting patterns are a tangle of socioeconomic interests, party loyalties, personal loyalties, variations in organizational effectiveness, and deference to immediate or remote authority figures. The parties themselves are collections of factions. And some of the most important lines of cleavage, as Leiserson shows, are differences in ideological perspectives that are, it appears, only loosely related to socioeconomic position.
In Soviet Russia and in the socialist countries of Eastern Europe, except Yugoslavia, the barriers to the public expression of political preferences are comparatively high. Yet, like other countries, these socialist countries contain within their boundaries people who differ in language, religion, ethnic group, status, economic position, policy views, ideology, personal loyalty, organizational loyalties, and the like. The widespread defection among certain nationality groups in the Soviet Union during World War II may reveal something of the persistence of ethnic and national iden-
liiics. One of the most powerful arguments against introducing a plural puny system in Yugoslavia is the danger that parties would form around ci hnic groups and intensify ethnic conflicts.
Where the barriers to the public expression of preferences are high, as in the Soviet Union, one can only guess at the patterns that conflict would take were the differences allowed to express themselves. In Yugoslavia, where the barriers have been very much reduced, the variety of interests claiming attention is remarkable. And during the brief period when the barriers were dropped in Czechoslovakia, diversity flourished like long-dormant flowers in a desert after a rain. Thus it is a reasonable conjecture that whenever the barriers to the public expression of preferences are reduced in Communist countries, in the long run (should they get through the short-run period of turmoil that may ensue) they will face a problem not so much of polarization as of diversity and possible segmentation.
The experiences of the countries described in this volume seem to support what I referred to at the beginning of this chapter as a well-grounded observation: differences in political preferences among the people of a country tend toward diversity and multiplicity rather than bipolarity. To many readers, the proposition, which is hardly a novel one, will seem so nearly self-evident that any appeal to experience of the kind I have just made is an exercise in demonstrating the obvious. Yet some readers, I am sure, will remain unconvinced.
Admittedly I have done no more than assert the existence of a very general and ubiquitous tendency, and a statement of a general tendency is a long way from a precise proposition in a comprehensive theory. At a minimum, a satisfactory theory would say something about the variations in the strength and characteristics of this tendency, the conditions under wrIiich polarization, even if an infrequent process, would be likely to occur, the causes and consequences of crosscutting rather than reinforcing cleavages, and doubtless many other matters. But perhaps it is enough here simply to give some substance to the assertion that the general tendency exists over long periods and in a great many countries at different stages of socioeconomic development and with different political regimes and economic systems.
A hasty reader might be tempted at this point to jump to some unwarranted conclusions. To say that polarization, like the plague, is rare is obviously not to say that it presents no dangers. And because polarization, though rare, is dangerous, one should not leap to the conclusion that a multiplicity of political interests, though more common, inevitably produces cohesion and stability. For diversity, like polarization, can also
Robert A. I hi hi
(■rente problems, even for polyaxcliiei. In what follows if is the problems which sccni to stem from diversity r.ither than from polarization that I want to stress. But first let me emphasize one further part of the argument.
Regimes and Interests
Out part of the argument I offered at the beginning was that the lower I h<- barriers ro, or the greater the opportunities for, expressing, organizing, kOd representing political preferences, the greater the number and variety of preferences that would be represented in policy making. Hence the number and variety of interests represented in policy making would be greater under a polyarchy than under a mixed regime and greater under a mixed regime than under a hegemony. Therefore the transformation of i hegemony into a mixed regime or polyarchy, or of a mixed regime into I polyarchy, would increase the number and variety of preferences and Interests in policy making.
If one were not careful about definitions, the argument could of course become circular. Even if a meticulous concern for definitions will enable us to avoid the latent circularity, die hypothesis is not easy to test. Yet Changes in regime occur, and the results do seem to bear out the argument. VV hen hegemonic regimes are suddenly displaced by regimes that provide greater opportunities for opposition — as in Spain after the flight of the liing in 1931 and the establishment of the republic, or in Italy and Germany with the fall of their dictatorships, or more recently in Ghana with the fall of Nkrumah — political preferences and latent oppositions that have dammed up spout forth like water through a collapsing dam. Thus Leiserson mentions how, after the military dictatorship was defeated in Japan, "the profusion of political groups which burst into the open as soon as the wartime straitjacket had been removed showed that the eclipse of oppositions had been only temporary."
During its "interrupted revolution," as Skilling calls it, Czechoslovakia provided an even more dramatic example of die abrupt surfacing of hitherto submerged interests and preferences. Ideological cleavages became visible as opponents of all change clashed with those who were ready to accept limited reforms, and moderate reformers differed not only with conservatives but with advocates of more radical, even revolutionary, alterations. Organized groups and associations flourished, pressed for advantage, demanded changes. Long-dormant or merely formal associations took on life- pre-Communist organizations like the Boy Scouts were revived. The cleavage between Czechs and Slovaks appeared as a basic organizational principle. With the virtual suspension of censorship, existing journals became controversial, new papers sprang into existence, and conflicting views were heard on radio and television. Then, almost as
liiirodiiftiifii
11
abruptly as the revolution had begun, with the Russian occupation these processes were reversed, and hegemony was gradually restored.
Yet, like Khrushchev's famous speech, the brief Czech revolution opened a window deep into the interior of a hegemonic system and made clear how illusory is the tranquility of hegemony, for contained within it are pent-up forces ready to burst forth whenever the barriers to expression and organization are lowered.
Inadequate as the institutions of polyarchy may be, the variety of political preferences or interests that are taken into account in policy malting appears to be much greater than in hegemonic or mixed regimes. After reading Kothari's description of India, one could readily imagine how a hegemonic regime in India might give decisive weight to certain kinds of preferences by overriding many interests that now assert themselves in the complex and endless negotiations, bargaining, demonstrations, and formation and dissolution of coalitions that characterize Indian politics. But it is difficult to imagine how any Indian regime, unless it were another polyarchy, could give representation to a greater variety of interests.
Yet it is this very tendency for political interests to crowd against the barriers to expression and to participate in policy making in increasing number and variety as the barriers are lowered that sometimes generates threats to regimes.
Threats to Regimes
hegemonies: self-fulfilung prophecy Hegemonies by definition are regimes that impose the most severe limits on the opportunities available to opponents of the government. They forbid the organization of all political parties, like the dictatorship in Argentina, or they establish a uniquely privileged party organization, like the Movimiento Nacional in Spain or the Communist party in the USSR and Eastern Europe.
Fully hegemonic regimes not only suppress all rival parties or convert them into mere appendages of the dominant party; they suppress factions within the dominant party as well. Internal party democracy, then, provides no alternative channel for expressing dissent. Thus in Spain, as Linz has pointed out elsewhere, "the internal organization of the Movimiento has not been ideologically, legally, or in practice democratic, nor have recent changes in its statutes changed this even on paper, nor is there much prospect that changes in that direction would in practice lead to any broad and decisive changes in participation through the party."1 In the USSR within a few years after the October Revolution all rival
1. Juan J. Linz, "The Inauguration of Democratic Government and Its Prospects in Spain" (unpublished ms., n.d.), p. 3.
11
Rolxit A. DM
parties hid been suppressed; and as Hnrghoorn points our, in 1921 t lie IVnili Party (congress adopted a resolution prohibiting "further activity within 1 he party by 'fractions,' based 011 'special platforms,' with their owii 'group discipline.' ... its final provision . . . empowered the party's Central Committee to expel from the party anyone engaged in such activity." Skilling summarizes the situation in Eastern Europe:
In Kastcrn Europe [in the fifteen years after Stalin's death] even peaceful advocacy of basic opposition to the regime and its policies was not tolerated and could usually express itself only through subterranean channels. In none of the countries was genuine opposition inside the ruling party permitted. Even in Yugoslavia, where the position and role of the party were significantly modified, the idea of a multiparty system or of an opposition party was explicitly rejected. The Yugoslav party's internal structure was amended, and widespread debate often occurred, but dissidence on basic issues among party members was not tolerated.
Nor are interest organizations permitted to function autonomously. The Official organizations tend to be transmission belts for the regime— so much so that, as in Spain, the official trade unions have come to be displaced by illegal workers' councils.
Yet even -in highly hegemonic regimes, opposition continues. Perhaps in all regimes, no matter how representative they may be, conflicts are bound to occur among the most powerful. These conflicts may be no more than surreptitious, lethal struggles by men who seek to win the dictator's favor and avoid his wrath. But beyond these struggles for power and place among the courtiers, there may also be, to quote Barg-hoorn, "efforts by members of the highest party and governmental decision-making bodies to. change the personnel or policies of the party-state." As he and Skilling show, this kind of factional opposition has persisted in Communist countries, and it is clear from Linz's essay that factional opposition has also been a salient characteristic of Franco's dictatorship. Factionalism may spread beyond the inner circle as allies are sought in lower echelons of the party, the bureaucracy, the controlled press, or among intellectuals. In addition to the struggles among factions, there is the seemingly unavoidable maneuvering among leaders who wish to protect or improve the position of specific segments and institutions of the society: the military, heavy industry, consumer goods, education, sciences, the arts. Factional and interest-based opposition by leaders essentially loyal to the regime verges on a more fundamental opposition to basic policies or institutions — for example, by nationality groups who feel themselves unfairly treated, by intellectuals to whom a thaw following a harsh winter of repression is but a prelude to a libertarian regime, to economists
Introduction
•3
who believe that a decentralized price-system economy must replace rigidly centralized planning. Beyond these oppositions, there are opponents who wish to subvert the regime and who are prepared to participate, if need be and if the opportunity arises, in conspiracies, violence, and revolution.
From the perspective of those who uphold a hegemonic regime, factional opposition may seem less dangerous than the others, while the kind of opposition that Barghoorn calls "subversive" and Skilling "integral" will no doubt seem most dangerous of all. Yet each shades off imperceptibly into the other. How can a fine be drawn between "safe" and dangerous oppositions? The hegemonic regime, in fact, creates a self-fulfilling prophecy:
Since all opposition is potentially dangerous, no distinction can be made between acceptable and unacceptable opposition, between loyal and disloyal opposition, between opposition that is protected and opposition that must be repressed. Yet if all oppositions are treated as dangerous and subject to repression, opposition that would be loyal if it were tolerated becomes disloyal because it is not tolerated. Since all opposition is likely to be disloyal, all opposition must be repressed.
I may have overdrawn the point. Yet highly hegemonic regimes seem unable wholly to escape the force of this self-fulfilling prophecy, particularly if they are endowed with an official ideology that claims a kind of divine right to rule based upon the exclusive possession of political truth and virtue. They move within a small orbit of toleration and repression: as toleration begins to set free the latent forces of opposition, the regime's leaders become fearful and clamp down. Their fear is not necessarily irrational, for the hitherto repressed opposition might surge out of control. Thus, because toleration may sow the seeds of the regime's destruction, it contains the seeds of its own destruction.
The self-fulfilling prophecy, then, holds so long as the basic premise is accepted that all opposition imperils the regime. Suppose, however, that leaders in a hegemony were to abandon this premise and try to establish a protected area within which opposition would be tolerated. If they could tolerate some opposition, could they indefinitely enforce any limits to toleration — short of the wide limits set in polyarchies?
This, in fact, is the dilemma of the mixed regime.
the dilemma of mixed regimes: repression or explosion? Although
there are a substantial number of polyarchies and full hegemonies among the national regimes that govern various countries, mixed regimes are more numerous. The frequency of mixed regimes partly reflects the fact that they make up a rather undiscriminating residual category in my
'•f Uobai A. Ihihl
classification. They vary from highly conipL-iinvc oligarchies where public I "incM.iiinii, including party organizations, is well protected but restritn.il ro a small elite (as in Britain by the end of the eighteenth century) to systems like that of Tanzania, where in 1965, under universal suffrage and the dominance of a single party, two candidates ran against one another in every parliamentary district, apparently in honestly conducted elections. During the late Meiji period and even under the Taisho Democracy, Japan was evidently ruled by a mixed regime. Yugoslavia should probabiy be classed as a mixed regime. During its brief revolution Czechoslovakia was rapidly being transformed from a full hegemony into a mixed regime.
In mixed regimes, some oppositions can engage in at least some forms of public contestation, "loyal" oppositions are tolerated, and (as in late eighteenth-century Britain or a century later in Japan) their leaders may even succeed from time to time in peacefully displacing some of the incumbents in the highest political offices.
If the barriers are, by definition, lower in mixed regimes tlian in hegemonies, and if as a result a greater variety of interests engage in public contestation, it is also true that by definition in mixed regimes the barriers are higher than in polyarchies. For example, even if opposition is well protected, the whole political game may be legally restricted to a tiny segment of the people, as in eighteenth-century Britain or in Japan until after World War II; and even if everyone is allowed to participate in the game (at least nominally), the rules of the game impose certain marked limits on the right to form political organizations, to contest elections, and so on. In Czechoslovakia under Dubcek, the Action Program of the Central Committee called for democratization of the party but rejected the more radical argument that socialist democracy would be impossible without a plural party system. Of Tanzania, probably the most liberal of the one-party African states, Foltz writes:
Because of the safeguards against coordination between different candidates, and because the party moved after the [1965] election to tighten restrictions against party members speaking publicly out of turn, the elections do not seem to have materially changed the conditions under which Tanzanian factions operate nor opened up acceptable means for party members to discuss major differences publicly.
Under mixed regimes, political preferences are repressed that would make themselves felt if the barriers to public contestation or participation were lower. In competitive oligarchies, the preferences of great mass segments—peasants, rural laborers, artisans, middle strata, and some commercial or industrial segments — are largely unorganized, unexpressed, and unrepresented. Only the extension of effective political rights to these excluded segments enables them to enter into the existing system of public
introduction 15
contestation. This was the familiar historical process by which more and more strata of society were brought into the oligarchical but already somewhat liberalized systems in a number of European countries in the nineteenth century.
Today, however, highly exclusive regimes — competitive oligarchies in the historical sense — are comparatively rare. More typical is the mixed regime with a broad citizenship but limits on public contestation, particularly on the right to form opposition parties. Where one party is privileged and others are prohibited, intraparty democracy is sometimes offered as an alternative to plural parties. Foltz quotes Nyerere's statement in 1963: "A National Movement which is open to all — which is identified with the whole nation — has nothing to fear from the discontent of any excluded section of society, for there is no such section." Yet when this alternative is examined theoretically or tested against practice, it proves to be, I believe, a rather romantic exaggeration of the possibilities.
If one makes reasonable assumptions about political motives and behavior and tries to specify carefully the kinds of guarantees and institutions that would be required for intraparty democracy in a one-party state to serve as effectively for the expression of political preferences as a plural party system in a polyarchy, the institutions and guarantees one must specify turn out to be remarkably like those of a plural party system in a polyarchy. (The theoretical exercise is highly illuminating; if the reader is unconvinced, he may wish to try it for himself.)
The catch in the theory of intraparty democracy in a one-party state is that the kinds of guarantees that would fully protect the expression, organization, and representation of interests within a party would also allow the formation of de facto parties, even if these masqueraded as factions in the single party. Conversely, if opponents of government are not protected in their right to form opposition parties in order to challenge the conduct of the government in elections, it is difficult to see how opposition within the dominant party can be fully protected.
In practice, it is probably not accidental, then, that organized party factions do not seem to exist in anything like the same degree in one-party states as they do in some countries where oppositions can express themselves not only in party factions but also in opposition parties. The highly organized factions in the Liberal-Democratic party of Japan, or the Christian Democratic party in Italy, seem to have no counterpart in one-party countries. As Foltz points out, since the time Nyerere made the statement quoted above,
Tanzania's case itself has shown . . . that, as the regime comes to be beset with more and more problems, the old emphasis on upward communication of grievances that was useful to the party in the na-
Robert A. Ihibl
tlonalisi period is shaiply reversed. Downward communication from i lie top party leadership becomes the rule and is reinforced by the leadership's right to approve nominations to lower party offices and particularly by the president's ability to give his personal investiture.
Thus in mixed regimes as in hegemonies, when the barriers are lowered, oppositions, interests, and political preferences previously repressed or inhibited spring forth to engage in public contestation. Leaders in these countries of course know this perfectly well — whatever they may say in public. And the presence of these repressed forces creates a genuine danger to the regime for which it can find no easy solution. . Since at least limited opportunities are available to oppositions, some oppositional elements will almost certainly use their opportunities to press for a further lowering of the barriers. In the competitive oligarchies of the nineteenth century it was inevitable that the restricted suffrage should come under attack; in one country after another, leaders were faced with the alternative of either yielding peacefully to demands for broadening suffrage or else repressing the organizations that arose first among the middle classes and later among the workers. Today the question of suffrage is less controversial; attempts to deny suffrage to a special group, such as the discrimination against Negro voting in the southern United States (now ending), are comparatively rare. Unlike their predecessors in liberal regimes, therefore, oppositions in mixed regimes today do not usually need to force their way into already existing systems of public contestation. An opposition in a mixed regime is less likely to be seeking inclusion in the system or effective citizenship and more likely to be seeking greater liberalization, more opportunities for public contestation. And just as it was possible for social strata excluded from citizenship or suffrage to find ways of exploiting the opportunities available for contestation in order to press their demands for suffrage, so in mixed regimes today oppositions can doubtless find ways of using the limited system of public contestation in order to press their demands for liberalization.
Viewed from the perspective of those who want to maintain a mixed regime and avoid both hegemony and polyarchy, the danger is that by yielding to the demands of oppositions for liberalization they will trigger a runaway explosion out of which polyarchy will emerge or, if repression is needed to bring liberalization to a halt, hegemony.
Just as it took only moderate foresight in the nineteenth century to see that it would be difficult to limit inclusion short of universality, so it is reasonable now to conclude that in a mixed regime some oppositions will not rest content with liberalization until the threshold to polyarchy has been crossed. In mixed regimes with a privileged party, a crucial step toward or even across that threshold will take place when an opposition
Introduction
'7
is permitted to organize a rival party to challenge the hitherto dominant party in fairly conducted elections; for once this step is taken, the opposition might then be able to win an election, displace incumbents, and alter their most treasured policies and institutions. Leaders of a regime who do not regard this as a fearful danger to be avoided at all costs scarcely require the limitations of a mixed regime.
Can oppositions in a liberalized regime be expected to refrain from pressing demands for an opposition party and from mobilizing behind it those who feel inadequately represented? As we have seen, the doctrine of intraparty democracy is unlikely to provide enough representation in practice to satisfy all the major interests of a country. Some interests will therefore hope to organize a party more immediately responsive to their particular views. Democratic ideology also continues to exert a powerful force for plural parties. After all, the institutions of polyarchy were developed partly as a response to democratic ideas. Since the institutions of polyarchy now exist and do not need to be invented anew, the idea of plural parties as an integral part of democracy is hardly unfamiliar, nor can it easily be dismissed as ridiculous or unwarranted.
Can the process of liberalization be launched and then checked just short of this critical threshold? Neither the Yugoslavian example nor the case of Czechoslovakia's interrupted revolution provides a clear-cut answer. The process of liberalization has gone much further in Yugoslavia than in any other Communist country (if we except the brief Czech interlude). The regime has definitely established the principle, however, that it will not tolerate public demands for an opposition party, much less the actual existence of a rival party. To pass over this last threshold to polyarchy is seen as perilous, not merely because "antisocialist" forces might gain strength, but perhaps even more because of the conviction, which seems to be widely held, that each of the different nationalities would soon have its own party and that plural parties would thus sharpen ethnic conflict and seriously endanger national cohesion. In short, some Yugoslavs argue, polyarchy means national fragmentation — and their fears can hardly be written off as simply a rationalization for the status quo. Yet as liberalization proceeds further in Yugoslavia, can the regime remain poised indefinitely on the last threshold to polyarchy? There is the risk that the demand for plural parties will grow so strong that the regime must yield, with all the dangers this transformation is thought to carry with it, or else step back firmly from the threshold and in no uncertain terms suppress demands for further liberalization.
Because the revolution in Czechoslovakia was terminated by the Russian occupation, no one can know how far it would have gone. The Action Program indicates clearly that Dubcek and the Central Committee intended to stop short of polyarchy and opposition parties. Yet the process
BRNO
iK Robert A. I hi hi
of liberalization w:is gathering SOCh great speed and momentum before (lie Soviets moved in that one wonders whether it could have been brought to a skidding halt at the crucial threshold.
In view of the danger inherent in arriving at this threshold and then living to restrain the forces that would pull the country across it, it will often seem wiser — from the perspective of those who support a mixed regime — to confine public contestation to fairly narrow limits. Yet if repression is needed to maintain these boundaries, as it probably will be, the mixed regime confronts a painful and inescapable dilemma: by increasing discontent and disloyalty, repression also increases the chance for a runaway explosion — and so the need for even more repression. ,   Thus mixed regimes are prone to oscillate between liberalization and repression. By suitable dosages of each, skilled leadership may successfully avoid both full hegemony and polyarchy. Yet the moment of truth will nrrive for any mixed regime if, in the process of liberalization, it should ever reach the threshold to polyarchy and a substantial group of spokesmen should begin to make a public demand for an opposition party.
two sources of discontent in polyarchies By definition polyarchies are systems that offer the lowest barriers to the expression, organization, and representation of political preferences and hence provide the widest array of opportunities for oppositions to contest the conduct of the government. But polyarchal regimes are not immune to discontent.2 For even in polyarchies, the political preferences of some people—perhaps of many people — remain unsatisfied. Let me focus here on two reasons why this happens.
i. Inequalities. In the first place, even in polyarchies some people have I grounds for believing that their interests are inadequately expressed, organized, and represented. For one thing, there are no generally accepted, objective criteria for deciding when an interest is fairly or adequately taken into account in political processes. Everyone is prone to emphasize the importance of his own preferences; but even if one were content with mere equality, it is by no means clear what equality among preferences would require. Since political philosophers have not been able to agree on whether individuals or interests should be equal — or on whether majority rule means an absolute majority of people, or of interests, or both, or a qualified majority of one or the other — one should scarcely expect ordinary citizens to agree on criteria for fair representation of interests. In any case, one scarcely needs a sophisticated view of politics to be aware
2. William Gamson's Power and Discontent (Homewood, 111.: Dorsey Press, 1968) has been helpful for the following discussion, though I have not followed his theoretical scheme.
Iiilit/tliii lion
that even when a polyarchy meeis sonic such rough criterion :is "one man, one vote," there remain great inequalities in the effective representation of different individuals and groups. The fact is too familiar to require documentation here.
One obvious source of political inequalities is the massive differences in the political resources of different individuals and groups. In extreme cases, within the confines of the same state, one stratum of the population may govern itself through a polyarchy and at the same time impose a hegemony on other strata. The political regime of Athens may have been a direct "democracy" among adult male citizens (I use quotes because I do not know how much inequality of power existed even among citizens), but it was a hegemony vis-a-vis the slaves. The aristocratic republic of Venice was perhaps a kind of constitutional "democracy" among the fifteen hundred to two thousand families who were privileged to participate in making the laws and electing the doge, but it was a hegemony (even though a constitutional one) vis-a-vis the people of Venice. Before the abolition of slavery — and for a century afterward in the South — the United States was a polyarchy for whites but a hegemony for most blacks.
These are anomalies, however, and not the main source of inequalities in polyarchies. Indeed, it is reasonable to argue that these anomalous cases do not indicate defects in polyarchy per se but simply demonstrate the existence of dual political systems in which hegemony is partly substituted for polyarchy. But even disregarding these cases, no polyarchy has ever eradicated large differences in the political resources of its citizens. In every polyarchy, therefore, these differences help to generate inequalities in the effective representation of individuals and groups. Again, this fact is so widely known that it needs no supporting evidence here.
In addition to these ancient and well-known causes of political inequalities, representation itself presents technical problems in ensuring equality that have never been satisfactorily solved. The size of the nation-state, even if it happens to be as small as most new countries in Africa or the smaller polyarchies of Europe, puts a great distance between ordinary citizens and national policy makers, complicates communication, and makes for distortions in the representation of interests. Moreover, in many polyarchies the systems of representation — for example, the single-member district with a winner-takc-all principle — have built-in biases, such as marked exaggeration in the proportion of representatives going to the largest party or to regional majorities. Even the ideal schemes proposed by legal pluralists and advocates of functional representation (such as the Guild Socialists) and the various methods of proportional representation do not, so far as I am aware, guarantee perfect equality in representation. It should be remembered too that in the absence of perfect con-
Roben A. I hi hl
•.I'iimin, no system of representation anil decision making can produce Hohjtious perfectly satisfactory to everyone. Even in a society of political equals there might he a permanent minority whose preferences always i;in counter to those of the majority.
Finally, the more a polyarchy represents the full variety of preferences m n country, the more laborious becomes the task of aggregating and then resolving or reconciling all these preferences in order to arrive at decisions. Since many citizens prefer that their government not be im-iiiobili/.cd by conflict, a regime may, in order to satisfy this particular preference, have to find ways to ignore many interests that would otherwise lead to deadlock.
To be sure, over a fairly extended period so many people might be at !e;ist partly satisfied in so many of their most salient political preferences (even though they might be dissatisfied in particular cases) that discontent would not be high. This result sometimes seems to be considered inherent in democratic procedures, but obviously it is not. I have already suggested the hypothetical case of the permanent minority for which no democratic rules provide a satisfactory solution. This of course represents an extreme form of persistent polarization, which, as I have argued earlier is comparatively rare. In practice, discontent is less likely to be associated with polarization between a permanent majority and a permanent minority than with fragmentation or segmentation in political conflicts.
2. Polarization and Segmentation. This last consideration points directly to a second reason why the greater opportunities that polyarchies provide for the expression, organization, and representation of interests do not protect them from serious discontent — and why they would not necessarily be immune from severe dissatisfaction even if they were considerably more democratic than in fact they are. For even if decisions were arrived at among political equals according to exact rules of democratic procedure, the outcome of decisions, in a country with a great diversity of political preferences, could rarely satisfy everyone.
To facilitate discussion, it will be helpful at this point to call attention to several aspects of political conflict: One is the level of anatagonism. Although antagonism is admittedly difficult to measure, particularly across countries or cultures, as a theoretical (and potentially operational) concept it seems indispensable. The number of sets of antagonists involved in a single conflict is a second aspect. To simplify the discussion, I intend to consider only the distinction between conflicts involving two sets of antagonists (bipolarity) and those involving more than two (multipolar-ity). A third aspect of conflict is the extent to which the composition of the sets of antagonists in one conflict is identical with the composition of the sets involved in other conflicts. If they are substantially the same, the conflicts are reinforcing or cumulative; if substantially different, the
Introduction
conflicts arc crosscut ting.11 A fourth aspect is the duration of any particular pattern, that is, the extent to which all of these aspects persist unchanged over time. In what follows, for the sake of simplicity I usually ignore the question of duration by assuming a more or less persistent pattern over a fairly substantial period of time. Although the possible patterns are innumerable, table i.i shows some important possibilities.
Table 1.1 Some Patterns of Conflict
Conflicts	Sets of Antagonists	Levels of Antagonism Low High	
	" Constant		
Reinforcing -	Bipolar	Moderate bipolarity	Polarization
	Multipolar	Moderate multipolarity	Severe segmentation
Crosscutting ^Changing		Moderate crosscutting conflicts	Moderate segmentation
Obviously, conflicts are relatively easy to handle as long as the level of antagonism remains low. They become dangerous to a regime when the level of antagonism is too high. Persistently high levels of antagonism seem to be less associated with purely socioeconomic differences than with differences in language, religion, ethnic or racial identity, and ideology. Japan offers an interesting confirmation of the hypothesis that while the harsh injustices of the early stages of industrialization and urbanization do generate bitter conflicts (it was these that Marx witnessed in nineteenth-century Europe), the severity of strictly economic conflict tapers off during the stage of late industrialization, while purely ideological differences, such as those that tend to surround foreign affairs or ideas of the good society, may become sharper. Leiserson observes that:
A new initiative by the government along the "reverse course" [i.e., to reverse the reforms introduced by the Occupation] or in foreign policy will produce a major political crisis — refusals to participate in investigatory commissions; demonstrations and strikes; physical violence in the Diet — but bread-and-butter policies are disposed of fairly calmly.
3. I refrain from introducing other aspects that would unnecessarily complicate the discussion here. Over a series of conflicts, for example, the number of sets of antagonists is a function of the third variable and two others that I do not present here: (a) the number of characteristics, each of which forms a single, separate dimension of conflict, such as class, religion, region, and language; and (b) the number of antagonistic groups formed along a particular dimension of cleavage.
ii Robert //. Ihihl
As ruble r.i indicates, depending on whether cleavages reinforce or cut across one another, high antagonism may result either in polarization or in segmentation.
That polarization is a profound threat to any regime seems so little open to doubt as to need no demonstration here. Polarization seems particularly dangerous for polyarchies, since they lack, the coercive forces and the willingness to repress severe antagonisms. If polarization occurs ■i along regional lines, the best that can be hoped for as a solution consistent I with polyarchy is peaceful separation, as between Norway and Sweden 'in 1905. But if separation is a threat to the prevailing concept of the nation, as the threat of secession was to the idea of the Union in the United States in the late 1850s, or as it would be in India today, the result is likely to be an attempt to coerce the separatists. Indeed, the most 1 likely short-run consequence of polarization in a polyarchy seems to be Jcivil war, as in the United States in 1861, in Austria in 1934, and in Spain in 1936. If polarization persists, a hegemony is likely to emerge as a kind of permanent means of coercion and, perhaps, of long-run pacification.
. Yet if my earlier argument is roughly correct, segmentation is a good deal more likely than polarization. Like polarization, segmentation also carries with it a threat to polyarchy. For, like polarization, segmentation 1 can also endanger nationhood whenever regional conflicts are reinforced I by antagonisms arising from such differences as those of language, religion, or socioeconomic condition. Language and region do often coincide, as in much of India; if these differences are reinforced by still others, such as socioeconomic condition, regions may seek not only autonomy but separation. If Spain and Yugoslavia, for example, were to become polyarchies, they might be threatened by segmentation along regional lines, I for in both countries differences of region, language, tradition, and economic status tend to reinforce one another.
Even where this is not the case and segmentation carries no threat to nationhood, it is likely to foster a form and style of politics that will produce discontent and cynicism about polyarchy. Such an enormous effort must be invested in building easily fragmented coalitions that the game of politics is reduced to — or at least is widely thought to be — little more than a narrow struggle for partisan advantage. Moreover, without
Isome system of mutual guarantees, every major segment must live in perpetual fear lest a coalition from which it is excluded should override its most important interests. If, on the contrary, every major segment manages to win a veto over policy, no coalition can ever deal firmly with any pressing problem. Immobilisni is Hie classic lot of the segmented polity. And immobilism combined with a total transformation of politics into the art of maneuvering for partisan advantage is likely to breed dis-
illlrodiirlioii JJ
content, political cynicism, and, in time, demands for a new political order more suited to decisive action.
The level of antagonism in political conflicts might decline if politics and government were to become less salient, less important as a source of advantages and disadvantages. Thus one reason for what the late Otto Kirchheimer called the waning of oppositions in Western Europe after World War II was thought to be the rapid and widespread rise in personal incomes, which made conflicts over governmental policies seem less important. Or, as in the case of fiscal and monetary policies, the actions of government may become more technical and remote — arcane matters for specialists who can be left in charge as long as the machine functions satisfactorily and incomes keep increasing. But the decline in political antagonisms in some polyarchies after World War II appears to have been temporary. When economic controversies became less salient, ideological differences involving foreign affairs and the distribution of power and opportunity began to manifest themselves. After twenty years of rising incomes in France, fierce antagonisms among Frenchmen were revealed during the uprising of May 1968. And in many countries below the high levels of affluence reached by the late- or postindustrial countries, conflicts over economic policy still carry a considerable freight of antagonism. Moreover, some countries, like Argentina, may find themselves locked into a system in which segmentation makes it nearly im-4 possible for a polyarchal government to solve major economic difficulties;' the government's failure then perpetuates the antagonisms that help produce segmentation.
Obviously, polyarchy is able to cope better with the dangers of segmentation if it can find more effective ways of reconciling the antagonistic groups. Reconciliation requires the investment of more energy and talent in the search for mutually satisfactory solutions to the issues in conflict. One system for mobilizing energy and talent in such a search is the large, catchall party that stands a good chance of putting together enough of the segments to win control of the government. In this situation, crass political incentives are mobilized in a search for solutions which, by being acceptable to a wide range of voters and interest groups, insure the dominance of the big, catchall party. The Congress party in India may be the world's leading example, as Kothari's description makes clear, and the Liberal-Democrats in Japan may well run a close second.
Unfortunately for the social engineer who opts for this solution, the large, catchall, integrating government party is more likely to be a product of historical accident than design. Nonetheless the extent to which a country's system of political parties is itself segmented seems to be influenced in some measure by the method of election. If on independence India had adopted a system of proportional representation in national elections, the
Robert A. Dahl
introduction
Congress party could never have won a majority in the I .ok Sabha; the fragmentation that befell the oppositions would have wrecked the Congress parry as well; and that extraordinary institution for expressing, representing, and reconciling an enormous variety of interests could not have carried out its task of national integration—a task for which no other institution in India seems to have been so well suited.
Another kind of incentive for mobilizing energy and talent in the search for mutually satisfactory solutions is provided by a system in which all the major segments have, de jure or de facto, a veto on the government's policies, as in Lebanon or the Netherlands; for the threat of a veto compels the leaders of all segments to throw themselves into the search for conciliatory solutions. Conceivably a country endangered by nationality conflicts, such as Yugoslavia, might be able to manage the transition to polyarchy if it were to develop some system of mutual veto among the nationalities.
Yet a system of mutual veto reduces one danger only to create another: immobilism. For often the only solution acceptable to everyone is the status quo; and even if this results in satisfactory short-run outcomes, the long-run outcome may be massive dissatisfaction with a system unable to confront and solve pressing problems of poverty, economic growth, welfare, redistribution, housing, and the like.
A special case of this system of mutual vetoes for reconciling antagonistic groups in a segmented polity requires a willingness to decentralize — investing a great deal of authority in the various cohesive segments — and at the same time to permit the settlement of the remaining major issues by the leaders of the various segments. In a broad sense, to be sure, this formula might be read as no more than a standard description of polyarchy; but what I have in mind is the system in the Netherlands as it has been described by Daaldcr and Lijphart.4 A system like this does not, of course, spring up overnight, and probably not every segmented society could successfully engineer it. It requires a commitment all round, but particularly among the leaders, to maintain the nation intact, to the seriousness of politics, to the importance of discovering mutually acceptable solutions, and to a process in which leaders are often allowed to negotiate solutions in secret, outside the usual contentions of politics, that are then willingly accepted by their followers.
The Dutch system suggests a more general kind of solution or partial solution: the development of a specialized set of institutions or arrangements for searching for mutually satisl.u imy solutions and ensuring their
4. See particularly Hans Daaldcr, "The Netherlands: Opposition in a Segmented Society," in Political Oppositions in Western Democracies, cd. Robert A. Dahl (New Haven: Yale University Press, 1967), and Amid Lijphart, The Politics of Accommodation (Berkeley: University of California Press, i</>8).
adoption. The classic site of this process was the national parliament and its network of personal, bloc, and party activities. But parliaments have not been able to handle the job by themselves, and their inability to overcome fragmentation and immobilism doubtless has often contributed to dissatisfaction with parliamentary democracy itself. Hence in practically every polyarchy, other institutions have emerged to take over a share — usually a very large share — of the task. These institutions are typically more hierarchical than legislative bodies, their top leaders are endowed with more political resources, they have greater legitimacy, they usually have more technical skills, and they are better able to operate in secret.
In some countries these institutions are the executive and bureaucracy, operating with very considerable independence of the legislature. In the Scandinavian countries, for example, the dangers of socioeconomic segmentation implicit in a very comprehensive organization of all the major economic interests in nationwide associations of workers, farmers, employers, and consumers is overcome by the periodic negotiating sessions at which contracts and policies are arrived at for the ensuing term under the aegis of the government. Despite the fact that the future of the economy depends on the results of these negotiations, the parliament can do little more than stand to one side until it can ratify the agreements that have been reached by the interest organizations themselves.
There may well be still other and better solutions to the dangers that segmented pluralism creates for polyarchies. Surely the invention of new institutions for political reconciliation — institutions consistent with democratic goals — should have a high place on the agenda of social scientists. For the prospect that existing polyarchies can do a better job of satisfying the claims of citizens, and that countries with hegemonic or mixed polities can in time be democratized, depends in no small measure on the creation of institutions for reconciling diversities.