SOME SOCIAL REQUISITES OF DEMOCRACY: ECONOMIC DEVELOPMENT AND POLITICAL LEGITIMACY

 SEYMOUR MARTIN LIPSET( University of California, Berkeley )

The conditions associated with the existence and stability of democratic society have been a leading concern of political philosophy. In this paper the problem is attacked from a sociological and behavioral standpoint, by presenting a number of hypotheses concerning some social requisites for democracy, and by discussing some of the data available to test these hypotheses. In its concern with conditions-values, social institutions, historical events-external to the political system itself which sustain different general types of political systems, the paper moves outside the generally recognized province of political sociology. This growing field has dealt largely with the internal analysis of organizations with political goals, or with the determinants of action within various political institutions, such as parties, government agencies, or the electoral process.2 It has in the main left to the political philosopher the larger concern with the relations of the total political system to society as a whole.

 

I. INTRODUCTION

A sociological analysis of any pattern of behavior, whether referring to a small or a large social system, must result in specific hypotheses, empirically testable statements. Thus, in dealing with democracy, one must be able to point to a set of conditions that have actually existed in a number of countries, and say: democracy has emerged out of these conditions, and has become stabilized because of certain supporting institutions and values, as well as because of its own internal self-maintaining processes. The conditions listed must be ones which differentiate most democratic states from most others.

A recent discussion by a group of political theorists on the "cultural prerequisites to a successfully functioning democracy" points up the difference between the approach of the political sociologist and the political philosopher to a comparable problem.' A considerable portion of this symposium is devoted to a debate concerning the contribution of religion, particularly Christian ethics, toward democratic attitudes. The principal author, Ernest Griffith, sees a necessary connection between the Judeo-Christian heritage and attitudes which sustain democratic institutions; the other participants stress the political and economic conditions which may provide the basis for a consensus on basic values which does not depend on religion; and they point to the depression, poverty, and social disorganization which resulted in fascism in Italy and Germany, in spite of strongly religious populations and traditions. What is most striking about this discussion is its lack of a perspective which assumes that theoretical propositions must be subject to test by a systematic comparison of all available cases, and which treats a deviant case properly as one case out of many. In this symposium, on the contrary, deviant cases which do not fit a given proposition are cited to demonstrate that there are no social conditions which are regularly associated with a given complex political system. So the conflicts among political philosophers about the necessary conditions underlying given political systems often lead to a triumphant demonstration that a given situation clearly violates the thesis of one's opponent, much as if the existence of some wealthy socialists, or poor conservatives, demonstrated that economic factors were not an important determinant of political preference.

The advantage of an attempt such as is presented here, which seeks to dissect the conditions of democracy into several interrelated variables, is that deviant cases fall into proper perspective. The statistical preponderance of evidence supporting the relationship of a variable such as education to democracy indicates that the existence of deviant cases (such as Germany, which succumbed to dictatorship in spite of an advanced educational system) cannot be the sole basis for rejecting the hypothesis. A deviant case, considered within a context which marshals the evidence on all relevant cases, often may actually strengthen the basic hypothesis if an intensive study of it reveals the special conditions which prevented the usual relationship from appearing.4 Thus, electoral research indicates that a large proportion of the more economically wellto-do leftists are underprivileged along other dimensions of social status, such as ethnic or religious position.

Controversy in this area stems not only from variations in methodology, but also from use of different definitions. Clearly in order to discuss democracy, or any other phenomenon, it is first necessary to define it. For the purposes of this paper, democracy (in a complex society) is defined as a political system which supplies regular constitutional opportunities for changing the governing officials. It is a social mechanism for the resolution of the problem of societal decision-making among conflicting interest groups which permits the largest possible part of the population to influence these decisions through their ability to choose among alternative contenders for political office. In large measure abstracted from the work of Joseph Schumpeter and Max Weber,5 this definition implies a number of specific conditions: (a) a "political formula," a system of beliefs, legitimizing the democratic system and specifying the institutionsparties, a free press, and so forth-which are legitimized, i.e., accepted as proper by all; (b) one set of political leaders in office; and (c) one or more sets of leaders, out of office, who act as a legitimate opposition attempting to gain office.

The need for these conditions is clear. First, if a political system is not characterized by a value system allowing the peaceful "play" of power-the adherence by the "outs" to decisions made by "ins" and the recognition by "ins" of the rights of the "outs"-there can be no stable democracy. This has been the problem faced by many Latin American states. Second, if the outcome of the political game is not the periodic awarding of effective authority to one group, a party or stable coalition, then unstable and irresponsible government rather than democracy will result. This state of affairs existed in pre-Fascist Italy, and for much, though not all of the history of the Third and Fourth French Republics, which were characterized by weak coalition governments, often formed among parties which had major interest and value conflicts with each other. Third, if the conditions facilitating the perpetuation of an effective opposition do not exist, then the authority of officials will be maximized, and popular influence on policy will be at a minimum. This is the situation in all one-party states; and by general agreement, at least in the West, these are dictatorships.

Two principal complex characteristics of social systems will be considered here as they bear on the problem of stable democracy: economic development and legitimacy. These will be presented as structural characteristics of a society which sustain a democratic political system. After a discussion of the economic development complex (comprising industrialization, wealth, urbanization, and education) and its consequences for democracy, we shall move to two aspects of the problem of legitimacy, or the degree to which institutions are valued for themselves, and considered right and proper. The relations between legitimacy and the effectiveness of the system (the latter primarily a function of economic development) will be followed by a discussion of the sources of cleavage in a society and the ways in which various resolutions of historically crucial issues result either in disruptive forms of cleavage or in cross-cutting affiliations which reduce conflict to a manageable level. Finally, the bearing of these various factors upon the future of democracy will be assessed.

No detailed examination of the political history of individual countries will be undertaken in accordance with the generic definition, since the relative degree or social content of democracy in different countries is not the real problem of this paper. Certain problems of method in the handling of relationships between complex characteristics of total societies do merit brief discussion, however.

An extremely high correlation between aspects of social structure, such as income, education, religion, on the one hand, and democracy, on the other, is not to be anticipated even on theoretical grounds, because to the extent that the political sub-system of the society operates autonomously, a particular political form may persist under conditions normally adverse to the emergence of that form. Or, a political form may develop because of a syndrome of fairly unique historical factors, even though major social characteristics favor another form. Germany is an example of a nation in which the structural changesgrowing industralization, urbanization, wealth, and education-all favored the establishment of a democratic system, but in which a series of adverse historical events prevented democracy from securing legitimacy in the eyes of many important segments of society, and thus weakened German democracy's ability to withstand crisis.

The high correlations which appear in the data to be presented between democracy and other institutional characteristics of societies must not be overly stressed, since unique events may account for either the persistence or the failure of democracy in any particular society. Max Weber argued strongly that differences in national patterns often reflect key historical events which set one process in motion in one country, and a second process in another. To illustrate his point, he used the analogy of a dice game in which each time the dice came up with a certain number they were increasingly loaded in the direction of coming up with that number again.6 To Weber, an event predisposing a country toward democracy sets a process in motion which increases the likelihood that at the next critical point in the country's history democracy will win out again. This process can only have meaning if we assume that once established, a democratic political system gathers some momentum, and creates some social supports (institutions) to ensure its continued existence. Thus a "premature" democracy which survives will do so by (among other things) facilitating the growth of other conditions conducive to democracy, such as universal literacy, or autonomous private associations. This paper is primarily concerned with explicating the social conditions which serve to support a democratic political system, such as education or legitimacy; it will not deal in detail with the kinds of internal mechanisms which serve to maintain democratic systems such as the specific rules of the political game.'

Comparative generalizations dealing with complex social systems must necessarily deal rather summarily with particular historical features of any one society within the scope of the investigation. In order to test these generalizations bearing on the differences between countries which rank high or low in possession of the attributes associated with democracy, it is necessary to establish some empirical measures of the type of political system. Individual deviations from a particular aspect of democracy are not too important, as long as the definitions unambiguously cover the great majority of nations which are located as democratic or undemocratic. The precise dividing line between "more democratic" and "less democratic" is also not a basic problem, since presumably democracy is not a quality of a social system which either does or does not exist, but is rather a complex of characteristics which may be ranked in many different ways. For this reason it was decided to divide the countries under consideration into two groups, rather than to attempt to rank them from highest to lowest. Ranking individual countries from the most to the least democratic is much more difficult than splitting the countries into two classes, "more" or "less" democratic, although even here borderline cases such as Mexico pose problems.

Efforts to classify all countries raise a number of problems. Most countries which lack an enduring tradition of political democracy lie in the traditionally underdeveloped sections of the world. It is possible that Max Weber was right when he suggested that modern democracy in its clearest forms can only occur under the unique conditions of capitalist industrialization.8 Some of the complications introduced by the sharp variations in political practices in different parts of the earth can be reduced by dealing with differences among countries within political culture areas. The two best areas for such internal comparison are Latin America as one, and Europe and the English-speaking countries as the other. More limited comparisons may be made among the Asian states, and among the Arab countries.

The main criteria used in this paper to locate European democracies are the uninterrupted continuation of political democracy since World War I, and the absence over the past 25 years of a major political movement opposed to the democratic "rules of the game."9 The somewhat less stringent criterion employed for Latin America is whether a given country has had a history of more or less free elections for most of the post-World War I period. Where in Europe we look for stable democracies, in South America we look for countries which have not had fairly constant dictatorial rule (See Table I). No detailed analysis of the political history of either Europe or Latin America has been made with an eye toward more specific criteria of differentiation; at this point in the examination of the requisites of democracy, election results are sufficient to locate the European countries, and the judgments of experts and impressionistic assessments based on fairly well-known facts of political history will suffice for Latin America.

II. ECONOMIC DEVELOPMENT AND DEMOCRACY

Perhaps the most widespread generalization linking political systems to other aspects of society has been that democracy is related to the state of economic development. Concretely, this means that the more well-to-do a nation, the greater the chances that it will sustain democracy. From Aristotle down to the present, men have argued that only in a wealthy society in which relatively few citizens lived in real poverty could a situation exist in which the mass of the population could intelligently participate in politics and could develop the self-restraint necessary to avoid succumbing to the appeals of irresponsible demagogues. A society divided between a large impoverished mass and a small favored elite would result either in oligarchy (dictatorial rule of the small upper stratum) or in tyranny (popularly based dictatorship). And these two political forms can be given modern labels: tyranny's modern face is Communism or Peronism; oligarchy appears today in the form of traditionalist dictatorships such as we find in parts of Latin America, Thailand, Spain or Portugal.

As a means of concretely testing this hypothesis, various indices of economic development-wealth, industrialization, urbanization and education-have been defined, and averages (means) have been computed for the countries which have been classified as more or less democratic in the Anglo-Saxon world and Europe and Latin America.

In each case, the average wealth, degree of industrialization and urbanization, and level of education is much higher for the more democratic countries, as the data presented in Table II indicate. If we had combined Latin America and Europe in one table, the differences would have been greater."

The main indices of wealth used here are per capita income, number of persons per motor vehicle and per physician, and the number of radios, telephones, and newspapers per thousand persons. The differences are striking on every score, as Table II indicates in detail. In the more democratic European countries, there are 17 persons per motor vehicle compared to 143 for the less democratic countries. In the less dictatorial Latin American countries there are 99 persons per motor vehicle, as against 274 for the more dictatorial ones.'2 Income differences for the groups are also sharp, dropping from an average per capita income of $695 for the more democratic countries of Europe to $308 for the less democratic ones; the corresponding difference for Latin America is from $171 to $119. The ranges are equally consistent, with the lowest per capita income in each group falling in the "less democratic" category, and the highest in the "more democratic" one.

1. A large part of this table has been compiled from data furnished by International Urban Research, University of California, Berkeley, California.

2. United Nations, Statistical Office, National and Per Capita Income in Seventy Countries, 1949, Statistical Papers, Series E, No. 1, New York, 1950, pp. 14-16.

3. United Nations, A Preliminary Report on the World Social Situation, 1952, Table 11, pp. 46-8.

4. United Nations, Statistical Yearbook, 1956, Table 139, pp. 333-338.

5. Ibid., Table 149, p. 387.

6. Ibid., Table 189, p. 641. The population bases for these figures are for different years than those used in reporting the numbers of telephones and radios, but for purposes of group comparisons, the differences are not important.

7. United Nations, A Preliminary Report .. ., op. cit., Appendix B, pp. 86-89.

8. United Nations, Demographic Yearbook, 1956, Table 12, pp. 350-370.

9. United Nations, Statistical Yearbook, 1956, op. cit., Table 127, pp. 308-310. Figures refer to commercially produced energy, in equivalent numbers of metric tons of coal.

10. United Nations, A Preliminary Report... , op. cit., Appendix A, pp. 79-86. A number of countries are listed as more than 95 per cent literate.

11. Ibid., pp. 86-100. Figures refer to persons enrolled at the earlier year of the primary range, per 1,000 total population, for years ranging from 1946 to 1950. The first primary year varies from five to eight in various countries. The less developed countries have more persons in that age range per 1,000 population than the more developed countries, but this biases the figures presented in the direction of increasing the percentage of the total population in school for the less developed countries, although fewer of the children in that age group attend school. The bias from this source thus reinforces the positive relationship between education and democracy.

12. Ibid., pp. 86-100.

13. UNESCO, World Survey of Education, Paris, 1955. Figures are the enrollment in higher education per 1,000 population. The years to which the figures apply vary between 1949 and 1952, and the definition of higher education varies for different countries.

14. Obtained from International Urban Research, University of California, Berkeley, California.

15. Ibid.

16. Ibid.

Industrialization-indices of wealth are clearly related to this, of courseis measured by the percentage of employed males in agriculture, and the per capita commercially produced "energy" being used in the country, measured in terms of tons of coal per person per year. Both of these indices show equally consistent results. The average percentage of employed males working in agriculture and related occupations was 21 in the "more democratic" European countries, and 41 in the "less democratic," 52 in the "less dictatorial" Latin American countries, and 67 in the "more dictatorial." The differences in per capita energy employed in the country are equally large.

The degree of urbanization is also related to the existence of democracy.  Three different indices of urbanization are available from data compiled by International Urban Research (Berkeley, California), the percentage of the population in places of 20,000 and over, the percentage in communities of 100,000 and over, and also the percentage residing in standard metropolitan areas. On all three of these indices of urbanization, the more democratic countries score higher than the less democratic, for both of the political culture areas under investigation.

Many have suggested that the better educated the population of a country, the better the chances for democracy, and the comparative data available support this proposition. The "more democratic" countries of Europe are almost entirely literate: the lowest has a rate of 96 per cent, while the "less democratic" nations have an average literacy rate of 85 per cent. In Latin America, the difference is between an average rate of 74 per cent for the "less dictatorial" countries and 46 percent for the "more dictatorial." The educational enrollment per thousand total population at three different levels, primary, postprimary, and higher educational, is equally consistently related to the degree of democracy. The tremendous disparity is shown by the extreme cases of Haiti and the United States. Haiti has fewer children (11 per thousand) attending school in the primary grades than the United States has attending colleges (almost 18 per thousand).

The relationship between education and democracy is worth more extensive treatment since an entire philosophy of democratic government has seen in increased education the spread of the basic requirement of democracy. As Bryce wrote with special reference to Latin America, "education, if it does not make men good citizens, makes it at least easier for them to become so."  Education presumably broadens men's outlooks, enables them to understand the need for norms of tolerance, restrains them from adhering to extremist and monistic doctrines, and increases their capacity to make rational electoral choices.

The evidence bearing on the contribution of education to democracy is even more direct and strong in connection with individual behavior within countries, than it is in cross-national correlations. Data gathered by public opinion research agencies which have questioned people in different countries with regard to their belief in various democratic norms of tolerance for opposition, to their attitudes toward ethnic or racial minorities, and with regard to their belief in multi-party as against one-party systems have found that the most important single factor differentiating those giving democratic responses from others has been education. The higher one's education, the more likely one is to believe in democratic values and support democratic practices.  All the relevant studies indicate that education is far more significant than income or occupation.

These findings should lead us to anticipate a far higher correlation between national levels of education and political practice than in fact we do find.

Germany and France have been among the best educated nations of Europe, but this by itself clearly did not stabilize their democracies. It may be, however, that education has served to inhibit other anti-democratic forces. PostNazi data from Germany indicate clearly that higher education is linked to rejection of strong-man and one-party government.

If we cannot say that a "high" level of education is a sufficient condition for democracy, the available evidence does suggest that it comes close to being a necessary condition in the modern world. Thus if we turn to Latin America, where widespread illiteracy still exists in many countries, we find that of all the nations in which more than half the population is illiterate, only one, Brazil, can be included in the "more democratic" group.

There is some evidence from other economically impoverished culture areas that literacy is related to democracy. The one member of the Arab League which has maintained democratic institutions since World War II, Lebanon, is by far the best educated (over 80 per cent literacy) of the Arab countries. In the rest of Asia east of the Arab world, only two states, the Philippines and Japan, have maintained democratic regimes without the presence of large antidemocratic parties since 1945. And these two countries, although lower than any European state in per capita income, are among the world's leaders in educational attainment. The Philippines actually ranks second to the United States in its proportion of people attending high school and university, while Japan has a higher level of educational attainment than any European state."9 Although the various indices have been presented separately, it seems clear that the factors of industrialization, urbanization, wealth, and education, are so closely interrelated as to form one common factor.20 And the factors subsumed under economic development carry with it the political correlate of democracy.2'

Before moving to a discussion of the inner connections between the development complex and democracy, mention may be made of a study of the Middle East, which, in its essential conclusions, substantiates these empirical relationships for another culture area. A survey of six Middle Eastern countries (Turkey, Lebanon, Egypt, Syria, Jordan, and Iran), conducted by the Columbia University Bureau of Applied Social Research in 1950-51, found high associations between urbanization, literacy, voting rates, media consumption and production, and education.22 Simple and multiple correlations between the four basic variables were computed for all countries for which United Nations statistics were available, in this case 54. The multiple correlations, regarding each as the dependent variable in turn, are as follows:

In the middle East, Turkey and Lebanon score higher on most of these indices than do the other four countries analyzed, and Lerner points out that the "great post-war events in Egypt, Syria, Jordan and Iran have been the violent struggles for the control of power-struggles notably absent in Turkey and Lebanon, where the control of power has been decided by elections."24 One of Lerner's contributions is to point to the consequences, for overall stability, of disproportionate development in one direction or another, and the need for coordinated changes in all of these variables. Thus, he compares urbanization and literacy in Egypt and Turkey, and concludes that although Egypt is far more urbanized than Turkey, it is not really "modernized," and does not even have an adequate base for modernization, because literacy has not kept abreast. In Turkey, all of the several indices of modernization have kept pace with each other, with rising voting participation (36 per cent in 1950), rising literacy, urbanization, etc. In Egypt, by contrast, the cities are full of "homeless illiterates," who provide a ready audience for political mobilization in support of extremist ideologies. On Lerner's scale, following the assumption of the functional interdependence of "modernization" factors, Egypt should be twice as literate as Turkey, since it is twice as urbanized. The fact that it is only half as literate explains, for Lerner, the "imbalances" which "tend to become circular and to accelerate social disorganization," political as well as economic.25

Lerner introduces one important theoretical addition, the suggestion that these key variables in the modernization process may be viewed as historical phases, with democracy a part of later developments, the "crowning institution of the participant society," one of his terms for a modern industrial society.

His view on the relations between these variables, seen as stages, is worth quoting at some length:

The secular evolution of a participant society appears to involve a regular sequence of three phases. Urbanization comes first, for cities alone have developed the complex of skills and resources which characterize the modern industrial economy. Within this urban matrix develop both of the attributes which distinguish the next two phases-literacy and media growth. There is a close reciprocal relationship between these, for the literate develop the media which in turn spread literacy. But, literacy performs the key function in the second phase. The capacity to read, at first acquired by relatively few people, equips them to perform the varied tasks required in the modernizing society. Not until the third phase, when the elaborate technology of industrial development is fairly well advanced, does a society begin to produce newspapers, radio networks, and motion pictures on a massive scale. This in turn, accelerates the spread of literacy. Out of this interaction develop those institutions of participation (e.g., voting) which we find in all advanced modern societies.26

Lerner's thesis concerning the functional interdependence of these elements of modernization is by no means established by his data, but the material presented in this paper offers an opportunity for research along these lines. Deviant cases, such as Egypt, where "lagging" literacy is associated with serious strains and potential upheaval, may also be found in Europe and in Latin America, and their analysis, a task not attempted here, will clarify further the basic dynamics of modernization, and the problem of social stability in the midst of institutional change.

A number of processes underlie these correlations, observed in many areas of the world, in addition to the effect, already discussed, of a high level of education and literacy in creating or sustaining belief in democratic norms.

Perhaps most important is the relationship between modernization and the form of the "class struggle." For the lower strata, economic development, which means increased income, greater economic security, and higher education, permit those in this status to develop longer time perspectives and more complex and gradualist views of politics. A belief in secular reformist gradualism can only be the ideology of a relatively well-to-do lower class.27 Increased wealth and education also serve democracy by increasing the extent to which the lower strata are exposed to cross pressures which will reduce the intensity of their commitment to given ideologies and make them less receptive to supporting extremist  ones. The operation of this process will be discussed in more detail in the second part of the paper, but essentially it functions through enlarging their involvement in an integrated national culture as distinct from an isolated lower class one, and hence increasing their exposure to middle-class values. Marx argued that the proletariat were a revolutionary force because they have nothing to lose but their chains and can win the whole world. But Tocqueville in analyzing the reasons why the lower strata in America supported the system paraphrased and transposed Marx before Marx ever made this analysis, by pointing out that "only those who have nothing to lose ever revolt.

Increased wealth is not only related causally to the development of democracy by changing the social conditions of the workers, but it also affects the political role of the middle class through changing the shape of the stratification structure so that it shifts from an elongated pyramid, with a large lower-class base, to a diamond with a growing middle-class. A large middle class plays a mitigating role in moderating conflict since it is able to reward moderate and democratic parties and penalize extremist groups.

National income is also related to the political values and style of the upper class. The poorer a country, and the lower the absolute standard of living of the lower classes, the greater the pressure on the upper strata to treat the lower classes as beyond the pale of human society, as vulgar, as innately inferior, as a lower caste. The sharp difference in the style of living between those at the top and those at the bottom makes this psychologically necessary. Consequently, the upper strata also tend to regard political rights for the lower strata, particularly the right to share in power, as essentially absurd and immoral. The upper strata not only resist democracy themselves, but their often arrogant political behavior serves to intensify extremist reactions on the part of the lower classes.

The general income level of a nation will also affect its receptivity to democratic political tolerance norms. The values which imply that it does not matter greatly which side rules, that error can be tolerated even in the governing party can best develop where (a) the government has little power to affect the crucial life chances of most powerful groups, or (b) there is enough wealth in the country so that it actually does not make too much difference if some redistribution does take place. If loss of office is seen as meaning serious loss for major power groups, then they will be readier to resort to more drastic measures in seeking to retain or secure office. The wealth level will also affect the extent to which given countries can develop "universalistic" norms among its civil servants and politicians (selection based on competence; performance without favoritism).

The poorer the country, the greater the emphasis which is placed on nepotism, i.e., support of kin and friends. The weakness of the universalistic norms reduces the opportunity to develop efficient bureaucracy, a condition for a modern democratic state.29

Less directly linked but seemingly still associated with greater wealth is the presence of intermediary organizations and institutions which can act as sources of countervailing power, and recruiters of participants in the political process in the manner discussed by Tocqueville and other exponents of what has come to be known as the theory of the "mass society."30 They have argued that a society without a multitude of organizations relatively independent of the central state power has a high dictatorial as well as a revolutionary potential.

Such organizations serve a number of functions necessary to democracy: they are a source of countervailing power, inhibiting the state or any single major source of private power from dominating all political resources; they are a source of new opinions; they can be the means of communicating ideas, particularly opposition ideas, to a large section of the citizenry; they serve to train men in the skills of politics; and they help increase the level of interest and participation in politics. Although there are no reliable data which bear on the relationship between national patterns of voluntary organizations and national political systems, evidence from studies of individual behavior within a number of different countries demonstrates that, independently of other factors, men who belong to associations are more likely to hold democratic opinions on questions concerning tolerance and party systems, and are more likely to participate in the political process-to be active or to vote. Since we also know that, within countries, the more well-to-do and the better educated one is, the more likely he is to belong to voluntary organizations, it seems likely that the propensity to form such groups is a function of level of income and opportunities for leisure within given nations."

It is obvious that democracy and the conditions related to stable democracy discussed here are essentially located in the countries of northwest Europe and their English-speaking offspring in America and Australasia. It has been argued by Max Weber among others that the factors making for democracy in this area are a historically unique concatenation of elements, part of the complex which also produced capitalism in this area. The basic argument runs that capitalist economic development (facilitated and most developed in Protestant areas) created the burgher class whose existence was both a catalyst and a necessary condition for democracy. The emphasis within Protestantism on individual responsibility furthered the emergence of democratic values. The greater initial strength of the middle classes in these countries resulted in an alignment between burghers and throne, an alignment which preserved the monarchy, and thus facilitated the legitimation of democracy among the conservative strata. Thus we have an interrelated cluster of economic development, Protestantism, monarchy, gradual political change, legitimacy and democracy.32 Men may argue as to whether any aspect of this cluster is primary, but the cluster of factors and forces hangs together.

III. LEGITIMACY AND DEMOCRACY

In this section I turn to an examination of some of the requisites of democracy which are derived from specifically historical elements in this complex, particularly those which relate to the need of a democratic political system for legitimacy, and for mechanisms which reduce the intensity of political cleavage. These requisites are correlated with economic development, but are also distinct from it since they are elements in the political system itself.

<Legitimacy and Effectiveness>

In the modern world, as the previous section has attempted to document, economic development involving industrialization, urbanization, high educational standards, and a steady increase in the overall wealth of the society, is a basic condition sustaining democracy; it is a mark of the efficiency of the total system.

But the stability of a given democratic system depends not only on the system's efficiency in modernization, but also upon the effectiveness and legitimacy of the political system. By effectiveness is meant the actual performance of a political system, the extent to which it satisfies the basic functions of government as defined by the expectations of most members of a society, and the expectations of powerful groups within it which might threaten the system, such as the armed forces. The effectiveness of a democratic political system, marked by an efficient bureaucracy and decision-making system, which is able to resolve political problems, can be distinguished from the efficiency of the total system, although breakdown in the functioning of the society as a whole will, of course, affect the political sub-system. Legitimacy involves the capacity of a political system to engender and maintain the belief that existing political institutions are the most appropriate or proper ones for the society. The extent to which contemporary democratic political systems are legitimate depends in large measure upon the ways in which the key issues which have historically divided the society have been resolved. It is the task of these sections of the paper to show first, how the degree of legitimacy of a democratic system may affect its capacity to survive the crises of effectiveness, such as depressions or lost wars and second, to indicate the ways in which the different resolutions of basic historical cleavages-which determine the legitimacy of various systems-also strengthen or weaken democracy through their effect on contemporary party struggles.

While effectiveness is primarily an instrumental dimension, legitimacy is more affective and evaluative. Groups will regard a political system as legitimate or illegitimate according to the way in which its values fit in with their primary values. Important segments of the German army, civil service, and aristocratic classes rejected the Weimar Republic not because it was ineffective, but because its symbolism and basic values negated their own. Legitimacy, in and of itself, may be associated with many forms of political organization, including oppressive ones. Feudal societies, before the advent of industrialism, undoubtedly enjoyed the basic loyalty of most of their members. Crises of legitimacy are primarily a recent historical phenomenon, following the rise of sharp cleavages among groups which have been able, because of mass communication resources, to organize around different values than those previously considered to be the only legitimate ones for the total society.

A crisis of legitimacy is a crisis of change, and therefore its roots, as a factor affecting the stability of democratic systems, must be sought in the character of change in modern society. It may be hypothesized that crises of legitimacy occur during a transition to a new social structure, if (a) all major groups do not secure access to the political system early in the transitional period, or at least as soon as they develop political demands; or, if (b) the status of major conservative institutions is threatened during the period of structural change. After a new social structure is established, if the new system is unable to sustain the expectations of major groups (on the grounds of "effectiveness") for a long enough period to develop legitimacy upon the new basis, a new crisis may develop.

Tocqueville gave a graphic description of the first general type of loss of legitimacy, referring mainly to countries which had moved from aristocratic monarchies to democratic republics:

" . . epochs sometimes occur in the life of a nation when the old customs of a people are changed, public morality is destroyed, religious belief shaken, and the spell of tradition broken .... . The citizens then have "neither the instinctive patriotism of a monarchy nor the reflecting patriotism of a republic; . . . they have stopped between the two in the midst of confusion and distress."

If, however, the status of major conservative groups and symbols is not threatened during this transitional period even though they lose most of their power, democracy seems to be much more secure. Striking evidence of the link between the preserved legitimacy of conservative institutions and democracy is the relationship between monarchy and democracy. Given the role of the American and French republican revolutions as the initiators of modern democratic political movements, the fact that ten out of 12 of the stable European and English-speaking democracies are monarchies seems -a rather ludicrous correlation. Great Britain, Sweden, Norway, Denmark, the Netherlands, Belgium, Luxemburg, Australia, Canada, and New Zealand are kingdoms; while the only republics which meet the twin conditions, of stable democratic procedures since democracy was instituted, and the absence of a major totalitarian movement in the past 25 years, are the United States, Switzerland and Uruguay. Nations which have moved from absolutism and oligarchy (linked to a state church) to a democratic welfare state, while retaining the forms of monarchy, more frequently seem able to make changes while sustaining a continuous thread of legitimacy for their political institutions.

The preservation of the monarchy has apparently retained for the system the loyalty of the aristocratic, traditionalist, and clerical sectors of the population which resented increased democratization and equalitarianism. And, by more graciously accepting the lower strata, by not resisting to the point that revolution might be necessary, the conservative orders won or retained the loyalty of the new "citizens".  Where monarchy was overthrown by revolution, and orderly succession was broken, those forces aligned with monarchy have sometimes continued to refuse legitimacy to republican successors down to the fifth generation or more.

The one constitutional monarchy which became a Fascist dictatorship, Italy, was, like the French Republic, relatively new and still illegitimate for major groups in the society. The House of Savoy alienated the Catholics by destroying the temporal power of the Popes, and was also not a legitimate successor in the old Kingdom of the Two Sicilies. Catholics, in fact, were forbidden by the church to participate in Italian politics until close to World War I, and the church rescinded its original ban only because of its fear of the Socialists. A similar attitude was taken by French Catholics to the Third Republic during the same period. Both Italian and French democracy have had to operate for much of their histories without loyal support from important groups in their society, both on the left and on the right. Thus, one main source of legitimacy lies in the continuity of primary conservative and integrative institutions during a transitional period in which new social institutions are emerging.

The second general type of loss of legitimacy is, as indicated above, related to the way in which societies handle the "entry into politics" problem. The determination of when new social groups shall obtain access to the political process affects the legitimacy of the political system, either for conservative or for emerging groups. In the 19th century these new groups were primarily industrial workers; the "entry into politics" crisis of the 20th century typically involves colonial elites, and peasant peoples. Whenever new groups become politically active (e.g., when the workers first seek access to economic and political power through economic organization and the suffrage, when the bourgeoisie demanded access to and participation in government, when colonial elites demand control over their own system), comparatively easy access to the legitimate political institutions tends to win the loyalty of the new groups to the system, and they in turn can permit the old dominating strata to maintain their own status integrity. In nations such as Germany, where access was denied for prolonged periods, first to the bourgeoisie and later to the workers, and where force was used to restrict access, the lower strata were alienated from the system, and were led to adopt extremist ideologies which, in turn, alienated the more established groups from an acceptance of the workers' political movement as a legitimate alternative.

Political systems which denied new strata access to power except through revolutionary means also inhibited the growth of legitimacy by introducing millenial hopes into the political arena. Groups which feel obliged to push their way into the body politic through forceful means tend to overexaggerate the possibilities which political participation afford. Their hopes are for far more than the inherent limitations of political stability permit. Consequently, democratic regimes born under such stress will not only face the difficulty of being regarded as illegitimate by those groups loyal to the ancien regime, but may be also rejected by those whose millenial hopes were not fulfilled by the change.

France seems to offer an example of such a phenomenon. Right-wing clericalists have viewed the Republic as illegitimate, while sections of the lower strata still impatiently await millenial fulfillment. Many of the newly independent nations of Asia and Africa face the problem of winning the loyalties of the masses to democratic states which can do little to fulfill the utopian objectives set by nationalist movements during the period of colonialism, and the transitional struggle to independence.

We have discussed several conditions bearing upon the maintenance, or the initial securing of legitimacy by a political system. Assuming reasonable effectiveness, if the status of major conservative groups is threatened, or if access to the political system is denied at crucial periods, the legitimacy of the system will remain in question. Even in legitimate systems, a breakdown of effectiveness, repeatedly or for a long period, will endanger its stability.

A major test of legitimacy is the extent to which given nations have developed a common "secular political culture," national rituals and holidays which serve to maintain the legitimacy of various democratic practices.3" The United States has developed a common homogeneous secular political culture as reflected in the veneration and consensus surrounding the Founding Fathers, Jefferson, Lincoln, Theodore Roosevelt and their principles. These common elements to which all American politicians appeal are not present in all democratic societies. In some European countries, the Left and the Right have a different set of symbols, and different historical political heroes. France offers the clearest example of a nation which has not developed such a common heritage. Thus many of the battles involving use of different symbols between the left and the right from 1789 down through much of the 19th century are "still in progress, and the issue is still open; everyone of these dates [of major political controversy] still divides left and right, clerical and anti-clerical, progressive and reactionary, in all their historically determined constellations."

As we have seen, nations may vary in the extent to which their political institutions are viewed as legitimate by different strata. And knowledge concerning the relative degree of legitimacy of a nation's political institutions is of key importance in any effort to analyze the stability of these institutions when faced with a crisis of effectiveness. The relationship between different degrees of legitimacy and effectiveness in specific political systems may be more graphically presented in the form of a four-fold table, with examples of countries characterized by the various possible combinations.

Societies which fall in box A, those which are high on the scales of both legitimacy and effectiveness, will clearly have stable political systems. Nations like the United States, Sweden, and Britain satisfy the basic political needs of their citizens, have efficient bureaucracies and political decision-making systems, possess traditional legitimacy through long-term continuity of the key symbols of sovereignty, the monarchy or constitution, and do not contain any important minorities whose basic values run counter to those of the system.37 Ineffective and illegitimate regimes, those which would be found in box D, must, of course, by definition be unstable and break down, unless they are dictatorships maintaining themselves by force such as the governments of Hungary and eastern Germany today. The political experiences of different countries in the early 1930's illustrate the effect of varying combinations of legitimacy and effectiveness. In the late 1920's, neither the German nor the Austrian republics were held legitimate by large and powerful segments of their populations, but nevertheless remained reasonably effective.38 In the four-fold table, they fell in box C.

When the effectiveness of the governments of the various countries broke down in the 1930's, those societies which were high on the scale of legitimacy remained democratic, while countries which were low such as Germany, Austria, and Spain, lost their freedom, and France narrowly escaped a similar fate. Or to put the changes in terms of location in the four-fold table, countries which shifted from A to B remained democratic, while the political systems of those which shifted from C to D broke down. It remained for the military defeat in 1940 to prove conclusively the low position of French democracy on the scale of legitimacy. It was the sole defeated democracy which furnished largescale support for a Quisling regime.39 Situations such as those discussed above in which either legitimacy or effectiveness is high while the other is low demonstrate the utility of this type of analysis. From a short-range point of view, a highly effective but illegitimate system, such as a well governed colony, is more unstable than regimes which are relatively low in effectiveness and high in legitimacy. The social stability of a nation such as Thailand-even with its occasional coups d'etats-stands out in sharp contrast to the situation in the neighboring former colonial nations of Southeast Asia. The link between the analysis of legitimacy and the earlier discussion of the contribution of economic development to democracy is evident in the processes through which regimes low in legitimacy may gain it, and conversely in those which are related to the collapse of a legitimate system. Prolonged effectiveness which lasts over a number of generations may give legitimacy to a political system; in the modern world, such effectiveness mainly means constant economic development. Thus those nations which adapted most successfully to the requirements of an industrial system had the fewest internal political strains, and either preserved their traditional legitimacy, the monarchy, or developed new strong symbols of legitimacy.

The social and economic structure which Latin America inherited from the Iberian peninsula prevented it from following the lead of the former English colonies, and its republics never developed the symbols and aura of legitimacy.

In large measure, the survival of the new political democracies of Asia and Africa is related to their ability to sustain a prolonged period of effectiveness, of being able to meet the defined instrumental needs of their populations.

<Legitimacy and Cleavage>

Prolonged effectiveness of the system as a whole may, as in the cases of the United States and Switzerland, eventually legitimate the democratic political system. Inherent, however, in all democratic systems is the constant threat that the conflicts among different groups which are the lifeblood of the system may crystallize to the point where societal disintegration is threatened. Hence, conditions which serve to moderate the intensity of partisan battle, in addition to effectiveness, are among the key requisites for a democratic political system.

Since the existence of a moderate state of conflict is an inherent aspect of a legitimate democratic system, and is in fact another way of defining it, we should not be surprised that the principal factors determining such an optimum state are closely linked to those which produce legitimacy viewed in terms of continuities of symbols and status. Essentially the character and content of the major cleavages affecting the political stability of a society are largely determined by historical factors which have affected the way in which maj or issues dividing society have been solved or left unresolved over time.

In modern times, three major issues have emerged in western states. The first was the religious issue: the place of the church and/or various religions within the nation. The second has been the problem of the admission of the lower strata, particularly the workers, to "citizenship," the establishment of access to power through universal suffrage, and the legitimate right to bargain collectively in the economic sphere. The third has been the continual struggle over the distribution of the national income.

The significant general question here is this: were these major issues dealt with one by one, and each one more or less solved before the next arose, or did the problems accumulate, so that historical issues and sources of cleavage mixed with newer ones? Resolving tensions one at a time contributes toward a stable political system; carrying over issues from one historical period to another makes for a political atmosphere characterized by bitterness and frustration rather than by tolerance and compromise. Men and parties come to differ with each other, not simply on ways of settling current problems, but rather by fundamental and opposed weltanschauungen. They come to see the political victory of their opponents as a major moral threat; and the total system, as a result, lacks effective value-integration.

The religious issue, the place of the church in the society, was fought through and solved in most of the Protestant nations in the 18th and 19th centuries, and ceased to be a matter for serious political controversy. In some states, such as the United States, the church was disestablished and it accepted this result.

In others, such as Britain, Scandinavia, and Switzerland, religion remains statesupported, but the state churches, like constitutional monarchs, have only nominal sway and have ceased to be major sources of controversy. It remains for the Catholic countries of Europe to provide us with examples of situations in which the historic controversy between clerical and anti-clerical forces, sparked by the French Revolution, has continued to divide men politically down to the present day. Thus in countries such as France, Italy, Spain, and Austria, being Catholic has meant being allied with rightist or conservative groups in politics; while being anti-clerical (or a member of a minority religion) has most often meant alliance with the left. In a number of these countries, newer issues, when they emerged, became superimposed on the religious question; and for conservative Catholics, the fight against Socialists was not simply an economic struggle, or a controversy over social institutions, but a deeprooted conflict between God and Satan, between good and evil.40 For many secular intellectuals in contemporary Italy, opposition to the church legitimates alliance with the Communists. As long as religious ties reinforce secular political alignments, the chances for democratic give-and-take, and compromise, are weak.

The "citizenship" or "political equality" issue has also been resolved in various ways. Thus the United States and Britain gave citizenship to the workers in the early or mid-nineteenth century. Sweden and a number of European nations resisted through the beginning of the 20th century, and the struggle for citizenship became combined in these countries with socialism as a political movement, thereby producing a revolutionary socialism. Or to put this in other terms, where the workers were denied economic and political citizenship rights, their struggle for redistribution of income and status was superimposed on a revolutionary ideology. Where the economic and status struggle developed outside this context, the ideology with which it was linked tended to be that of gradualist reformism. In Hohenzollern Germany, for example, the workers were denied a free and equal suffrage in Prussia until the revolution of 1918.

This denial of "citizenship" facilitated the retention of revolutionary Marxism in those parts of Germany where equal suffrage did not exist. In Southern Germany, where full citizenship rights were granted in the late 19th century, reformist, democratic, and non-revolutionary socialism was dominant. The perpetuation of revolutionary dogmas in much of the Social Democratic party served to give ultra-leftists a voice in party leadership, enabled the Communists to win strength after the military defeat, and perhaps even more important historically, served to frighten large sections of the German middle classes. The latter feared that a socialist victory would really mean an end to all their privileges and status.

In France, the workers won the suffrage but were refused basic economic rights until after World War II. Major groups of French employers denied legitimacy to the French trade-unions, and sought to weaken or destroy them following every trade-union victory. The instability of the French unions, their constant need to preserve worker militancy to survive, gave access to the workers to the more revolutionary and extremist political groups. Communist domination of the French labor movement can in large part be traced to the tactics of the French business classes.

The examples presented above do not explain why different countries varied in the way they handled basic national cleavages. They should suffice, however, to illustrate the worth of a hypothesis relating the conditions for stable democratic government to the bases of diversity. Where a number of historic cleavages intermix and create the basis for weltanschauung politics, the democracy will be unstable and weak, for by definition such political views do not include the concept of tolerance.

Weltanschauung politics have also weakened the possibilities for a stable democracy, since parties characterized by such total ideologies have often attempted to create what Sigmund Neumann has called an "integrated" environment, one in which as much as possible of the lives of their members is encapsulated within ideologically linked activities. These actions are based on the assumption that it is important to isolate their followers from contact with "falsehood" expressed by non-believers. Neumann has suggested the need for a basic analytic distinction between parties of representation, which strengthen democracy, and parties of integration which weaken it.4' The former are typified by most parties in the English-speaking democracies and in Scandinavia, and by most centrist and conservative parties other than the religious ones. They view the party function as primarily one of securing votes around election time.

The parties of integration, on the other hand, are concerned with making the world conform to their basic philosophy or weltanschauung. They do not see themselves as contestants in a give-and-take game of pressure politics, in which all parties accept the rules of the game. Rather they view the political or religious struggle as a contest between divine or historic truth on one side and fundamental error on the other. Given this conception of the world, it becomes necessary to prevent their followers from being exposed to the cross-pressures flowing from contact with falsehood, which will reduce their faith.

The two major non-totalitarian groupings which have followed such procedures have been the Catholics and the Socialists. In general, in much of Europe before 1939, the Catholics and Socialists attempted to increase intrareligious or intra-class communications by creating a network of church- and party-linked social and economic organizations within which their followers could live their entire lives. Austria offers perhaps the best example of a situation in which two groups, the Social Catholics and the Social Democrats, divided over all three historic issues and separated the country into two hostile camps, which carried out much of their social activities in party or churchlinked organizations.42

The totalitarian organizations, Fascist and Communist, expanded the integrationist character of political life to the furthest limit possible. They outdo all other groups in defining the world in struggle terms, and in seeing the corrupting influences either of Judaism or capitalism as requiring the insulation of the true believers.

Efforts by democratic parties of integration to isolate their social base from cross-pressures are clearly disruptive of the requirements for a stable democracy in which there is shifting from one election to another, and in which issues between parties are allowed to be resolved over time. Isolation may intensify loyalty to the party or church, but it may also serve to prevent a party from reaching new strata. The Austrian situation also illustrates the frustration of the electoral process which results when most of the electorate is encapsulated within parties of integration. The necessary rules of democratic politics assume that conversion both ways, into and out of a party, is possible and accepted as proper. Parties which hope to gain a majority by democratic methods must ultimately give up their integrationist tendencies. The only justification for isolation from the rest of the culture is a strong commitment to the idea that the party possesses the only truth, that there are certain basic issues which must be resolved by the triumph of historic truth. As the working-class has gained complete citizenship in the political and economic spheres in different countries, the Socialist parties of Europe have dropped their integrationist emphasis. The only non-totalitarian parties which can and do maintain such policies are religious parties such as the Catholic parties, or the Calvinist Anti-Revolutionary party of Holland. Clearly, the Catholic and Dutch Calvinist churches are not "democratic" in the sphere of religion. They insist there is but one truth, as the Communists and Fascists do in politics. Catholics may accept the assumptions of political democracy, but never those of religious tolerance. And where the conflict between religion and irreligion is viewed as salient by Catholics or other believers in a one true church, then a real dilemma exists for the democratic process. Many political issues which in other countries may be easily compromised become aggravated by the religious issue, and cannot be settled.

The intense forms of cleavage developed by that cumulation of unresolved issues which creates weltanschauung politics is sustained by the systematic segregation of different strata of the population in organized political or religious enclaves. Conversely, however, it should be noted that wherever the social structure operates so as naturally to "isolate" individuals or groups with the same political disposition characteristics from contact with differing views, those so isolated tend to back political extremists.

It has been repeatedly remarked, for example, that workers in so-called "isolated" industries, miners, sailors, fishermen, lumbermen, sheep-tenders, and longshoremen, tend to give overwhelming support to the more left-wing tendencies. Such districts tend to vote Communist or Socialist by large majorities, sometimes to the point of having what is essentially a "one-party" system in the areas concerned. Isolation is created by the fact that the requirements of the job make workers in these industries live in communities which are predominately inhabited by others in the same occupation. And this very isolation seems to reduce the pressures on such workers to be tolerant of other points of view, to contain among themselves diverse strains of thought; and makes them receptive to extremist versions of the doctrine generally held by other less isolated members of their class. One should expect that the least "cosmopolitan"

(the most isolated) of every political predisposition, or stratum, will be the ones most likely to accept extremism. The political intolerance of farm-based groups in times of crisis may be another illustration of this pattern, since farmers, like workers in isolated industries, tend to have a more homogeneous political environment than do those employed in most urban occupations.4'

These conclusions are further confirmed by studies of individual voting behavior which indicate that individuals under cross pressures-those who belong to groups predisposing in different directions, who have friends supporting different parties, who are regularly exposed to the propaganda of different tendencies-are less likely to be strongly committed politically.44

Multiple and politically inconsistent affiliations and loyalties are stimuli that serve to reduce the emotion and aggressiveness involved in political choice. For example, in contemporary Germany, a working-class Catholic, pulled in two directions, will most probably vote Christian-Democratic, but is much more tolerant of the Social Democrats than the average middle-class Catholic.45 Where a man belongs to a variety of groups such that all predispose toward the same political choice, he is in the situation of the isolated worker, and is much less likely to exhibit tolerance of opposition opinions, or view the possibility of their coming to power with equanimity.

The evidence available suggests that the chances for stable democracy are enhanced to the extent that social strata, groups and individuals have a number of cross-cutting politically relevant affiliations. To the degree that a significant proportion of the population is pulled among conflicting forces, such groups and individuals have an interest in reducing the intensity of political conflict.46 As Robert Dahl and Talcott Parsons have pointed out, such groups and individuals also have an interest in protecting the rights of political minorities.47

A stable democracy requires relatively moderate tension among the contending political forces. And political moderation is facilitated by the capacity of a system to resolve key dividing issues before new ones arise. To the extent that the cleavages of religion, citizenship, and "collective bargaining" have been allowed to cumulate and reinforce each other as stimulants of partisan hostility, the system is weakened. The more reinforced and correlated the sources of cleavage, the less the likelihood for political tolerance. Similarly, on the level of group and individual behavior, the greater the isolation from heterogeneous political stimuli, the more that background factors "pile up" in one direction, the greater the chances that the group or individual will have an extremist perspective. These two relationships, one on the level of partisan issues, the other on the nature of party support, are linked together by the fact that parties reflecting accumulated unresolved issues will seek to isolate their followers from conflicting stimuli, to prevent exposure to "error," while isolated individuals and groups will strengthen the  intolerant tendencies in the political party system.  The conditions maximizing political cosmopolitanism among the electorate are the growth of urbanization, education, communications media, and increased wealth. Most of the obvious isolated occupations, mining, lumbering, agriculture, belong to the category of "primary" occupations, occupations whose relative share of the labor force declines sharply with economic development.48

Thus, we see again how the factors involved in modernization or economic development are linked closely to those involved in the historic institutionalization of the values of legitimacy and tolerance. But it should always be noted that correlations are only statements concerning relative degrees of congruence, and that another condition for political action is that the correlation never be so clear-cut that men cannot feel that they can change the direction of affairs by their actions. And this fact of low correlation means also that it is important for analytic purposes to keep variables distinct even if they intercorrelate. For example, the analysis of cleavage presented here suggests specific propositions concerning the ways in which different electoral and constitutional arrangements may affect the chances for democracy. These generalizations are presented in the following section.

IV. SYSTEMS OF GOVERNMENT AND DEMOCRACY

From the hypothesis that cross-cutting bases of cleavage are better for the vitality of democracy, it follows that two-party systems are better than multiparty systems, that electoral systems involving the election of officials on a territorial basis are preferable to systems of proportional representation, and that federalism is superior to a unitary state. In evaluating these propositions, it is important to note again that they are made with the assumption of all other factors being held constant. Clearly, stable democracies are compatible with multi-party systems, with proportional representation, and with a unitary state. And in fact, I would argue that such variations in systems of government, while significant, are much less important than those derived from basic differences in social structure of the sort discussed in the previous sections.

The argument for the two-party system rests on the assumptions that in a complex society, such parties must necessarily be broad coalitions; that they cannot seek to serve only the interests of one major group; that they cannot be parties of integration; and that in building electoral coalitions, they necessarily antagonize support among those most committed to them, and conversely must seek to win support among groups which are preponderantly allied to the opposition party. Thus, the British Conservative or American Republican parties must not so act as to antagonize basically the manual workers, since a large part of the vote must come from them. The Democratic and Labor parties are faced with a similar problem vis-d-vis the middle strata. Parties which are never oriented toward gaining a majority seek to maximize their electoral support from a limited base. Thus a peasant-oriented party will accentuate peasant group interest consciousness, and a party appealing primarily to small businessmen will do the same for its group. Elections, instead of being occasions on which parties seek to find the broadest possible base of support, and so to bring divergent groups to see their common interests, become events in which parties stress the cleavages separating their principal supporters from other groupings.

The proposition that proportional representation weakens rather than strengthens democracy rests on the analysis of the differences between multiparty and majority party situations. If it is true, as is suggested above, that "multi-partyness" serves to sharpen differences and reduce consensus, then any electoral system which increases the chance for more rather than fewer parties serves democracy badly.

Further, as Georg Simmel pointed out, the system of electing members of parliament to represent territorial constituencies, as contrasted with systems which encourage direct group representation (such as proportional representation), is preferable, since territorial representation helps to stabilize the political systems by forcing interest groups to secure their ends only within an electoral framework that involves some concern with many interests and the need for compromise.

Federalism serves to strengthen democracy by increasing the opportunity for multiple sources of cleavage. It adds regional interests and values to the others such as class, religion and ethnicity which cross-cut the social structure.

A major exception to this generalization occurs when federalism divides the country according to lines of basic cleavage, e.g., between different ethnic, religious, or linguistic areas. In such cases, as in India or in Canada, federalism may then serve to accentuate and reinforce cleavages. Cleavage is desirable within linguistic or religious groups, not between them. But where such divisions do not exist, then federalism seems to serve democracy well. Besides creating a further source of cross-cutting cleavage, it also serves various functions which Tocqueville noted it shared with strong voluntary associations. Among these, it is the source of resistance to centralization of power and a source of training of new political leaders; and it gives the "out" party a stake in the system as a whole, since national "out" parties usually continue to control some units of the system.

Let me repeat that I do not suggest that these aspects of the political structure as such are key conditions for democratic systems. If the underlying social conditions are such as to facilitate democracy, as seems true for Sweden, then the combination of multi-partyness, proportional representation, and a unitary state, do not seriously weaken it. At most they serve to permit irresponsible minorities to gain a foothold in parliament. On the other hand, where a low level of effectiveness and of legitimacy has operated to weaken the foundations of democracy as occurred in Weimar Germany, or in France, then constitutional factors encouraging multi-partyness serve to reduce the chances that the system will survive.

V. PROBLEMS OF CONTEMPORARY DEMOCRACY

The characteristic pattern of the stable western democracies in the mid-20th century is that of a "post-politics" phase-there is relatively little difference between the democratic left and right, the socialists are moderates, and the conservatives accept the welfare state. In large measure this reflects the fact that in these countries the workers have won their fight for citizenship and for political access, i.e., the right to take part in all decisions of the body politic on an equal level with others.

The struggle for citizenship had two aspects, political (access to power through the suffrage) and economic (institutionalization of trade union rights to share in the decisions affecting work rewards and conditions). The representatives of the lower strata are now part of the governing classes, members of the club. Political controversy has declined in the wealthier stable democracies because the basic political issue of the industrial revolution, the incorporation of the workers into the legitimate body politic, has been settled. The only key domestic issue today is collective bargaining over differences in the division of the total product within the framework of a Keynesian welfare state; and such issues do not require or precipitate extremism on either side.

In most of Latin and Eastern Europe, the struggle for working-class integration into the body politic was not settled before the Communists appeared on the scene to take over leadership of the workers. This fact drastically changed the political game, since inherently the Communists could not be absorbed within the system in the way that the Socialists have been. Communist workers, their parties and trade unions, cannot possibly be accorded the right of access by a democratic society. The Communists' self-image and more particularly their ties to the Soviet Union lead them to accept a self-confirming hypothesis. Their self-definition prevents them from being allowed access and this in turn reinforces the sense of alienation from the system (of not being accepted by the other strata) which workers in nations with large Communist parties have. And the more conservative strata are reinforced in their belief that giving increased rights to the workers or their representatives threatens all that is good in life. Thus, the presence of Communists precludes an easy prediction that economic development will stabilize democracy in these European countries.

In the newly independent nations of Asia, the situation is somewhat different. In Europe at the beginning of modern politics, the workers were faced with the problem of winning citizenship, the right to take part in the political game, from the dominant aristocratic and business strata who controlled politics. In Asia the long-term presence of colonial rulers has identified conservatism as an ideology and the more well-to-do classes with subservience to colonialism; while leftist ideologies, usually of a Marxist variety, have been dominant, being identified with nationalism. The trade unions and the workers' parties of Asia have been part of the political process from the beginning of the democratic system. Conceivably such a situation could mean a stable democracy, except for the fact that these lower-strata rights pre-date the development of a stable economy with a large middle class and an industrial society.

The whole system stands on its head. The left in the European stable democracies grew gradually in a fight for more democracy, and gave expression to the discontents involved in early industrialization, while the right retained the support of traditionalist elements in the society, until eventually the system came into an easy balance between a modified left and right. In Asia, the left is in power during the period of population explosion and early industrialization, and must accept responsibility for all the consequent miseries. As in the poorer areas of Europe, the Communists exist to capitalize on all these discontents in completely irresponsible fashion, and currently are a major party, usually the second largest in most Asian states.

Given the existence of poverty-stricken masses, low levels of education, an elongated pyramid class structure, and the "premature" triumph of the democratic left, the prognosis for the perpetuation of political democracy in Asia and Africa is bleak. The nations which have the best prospects, Israel, Japan, Lebanon, the Philippines and Turkey, tend to resemble Europe in one or more major factors, high educational level (all except Turkey), substantial and growing middle class, and the retention of political legitimacy by non-leftist groups. The other emerging national states in Asia and Africa are committed more deeply to a certain tempo and pattern of economic development and to national independence, under whatever political form, than they are to the pattern of party politics and free elections which exemplify our model of democracy. It seems likely that in countries which avoid Communist or military dictatorship political developments will follow the pattern developing in countries such as Ghana, Tunisia or Mexico, where an educated minority uses a mass movement expressing leftist slogans to exercise effective control, and holds elections as a gesture toward ultimate democratic objectives, and as a means of estimating public opinion, not as effective instruments for legitimate turnover in office of governing parties.5' Given the pressure for rapid industrialization and for the immediate solution of chronic problems of poverty and famine through political agencies, it is unlikely that many of the new governments of Asia and Africa will be characterized by an open party system representing basically different class positions and values.

Latin America, underdeveloped economically like Asia, is, however, politically more like Europe in the early 19th century than like Asia today. Most Latin American countries became independent states before the rise of industrialism and Marxist ideologies, and contain strongholds of traditional conservatism. The countryside is often apolitical or traditional, and the leftist movements secure support primarily from the industrial proletariat. Latin American communists, for example, have chosen the European Marxist path of organizing urban workers, rather than the "Yenan way" of Mao, seeking a peasant base.53 If Latin America is allowed to develop on its own, and is able to increase its productivity and middle classes, there is a good chance that many Latin American countries will follow in the European direction. Recent developments, including the overthrowal of a number of dictatorships, in large measure reflect the effects of an increased middle class, growing wealth, and increased education. There is, however, also the possibility that these countries may yet follow in the French and Italian direction rather than that of northern Europe, that the communists will seize the leadership of the workers, and that the middle class will be alienated from democracy.

The analysis of the social requisites for democracy contained in this paper has sought to identify some, though obviously far from all, of the structural conditions which are linked to this political system. It has been possible in a very limited fashion to attempt some tests of the hypotheses suggested. These preliminary efforts to apply the method of science to comparative political systems can still be considered only as illustrative since we can say so little about actual variations in national social structures. Considerably more research must be done specifying the boundaries of various societies along many dimensions before reliable comparative analysis of the sort attempted here can be carried out. Although the task obviously presents tremendous difficulties, it is only through such methods that we can move beyond the conventional semi-literary methods of giving illustrative examples to support plausible interpretations.

The data available are, however, of a sufficiently consistent character to support strongly the conclusion that a more systematic and up-to-date version of Aristotle's hypothesis concerning the relationship of political forms to social structure is valid. Unfortunately, as has been indicated above, this conclusion does not justify the optimistic liberal's hope that an increase in wealth, in the size of the middle class, in education, and other related factors will necessarily mean the spread of democracy or the stabilizing of democracy. As Max Weber, in discussing the chances for democracy in Russia in the early 20th century pointed out: "The spread of Western cultural and capitalist economy did not, ipso facto, guarantee that Russia would also acquire the liberties which had accompanied their emergence in European history.... European liberty had been born in unique, perhaps unrepeatable, circumstances at a time when the intellectual and material conditions for it were exceptionally propitious."

These suggestions that the peculiar concatenation of factors which gave rise to western democracy in the nineteenth century may be unique are not meant to be unduly pessimistic. Political democracy exists and has existed in a variety of circumstances, even if it is most commonly sustained by a limited cluster of conditions. To understand more fully the various conditions under which it has existed may make possible the development of democracy elsewhere. Democracy is not achieved by acts of will alone; but men's wills, through action, can shape institutions and events in directions that reduce or increase the chance for the development and survival of democracy. To aid men's actions in furthering democracy was in some measure Tocqueville's purpose in studying the operation of American democracy, and it remains perhaps the most important substantive intellectual task which students of politics can still set before themselves.

METHODOLOGICAL APPENDIX

The approach of this paper (as has already been indicated) is implicitly different from others which have attempted to handle social phenomena on a total societal level, and it may be useful to make explicit some of the methodological postulates underlying this presentation.

Complex characteristics of a social system, such as democracy, the degree of bureaucratization, the type of stratification system, have usually been handled either by a reductionist approach or by an "ideal-type" approach. The former approach dismisses the possibility of considering those characteristics as system-attributes as such, and maintains that qualities of individual actions are the sum and substance of sociological categories. For this school of thought, the extent of democratic attitudes, or of bureaucratic behavior, or the numbers and types of prestige or power rankings, constitute the essence of the meaning of the attributes of democracy, bureaucracy, or class.

The "ideal-type" approach starts from a similar assumption, but reaches an opposite conclusion. The similar assumption is that societies are a complex order of phenomena, exhibiting such a degree of internal contradiction, that generalizations about them as a whole must necessarily constitute a constructed representation of selected elements, stemming from the particular concerns and perspectives of the scientist. The opposite conclusion is that abstractions of the order of "democracy" or "bureaucracy" have no necessary connection with states or qualities of complex social systems which actually exist, but comprise collections of attributes which are logically interrelated, but characteristic in their entirety of no existing society. An example of this type of abstraction is Weber's concept of "bureaucracy," comprising a set of offices, which are not "owned" by the office-holder, continuously maintained files of records, functionally specified duties, etc. Another is the common definition of democracy in political science, which postulates individual political decisions based on rational knowledge of one's own ends and of the factual political situation.

Criticism of categories, or ideal-types, such as this, solely on the basis that they do not correspond to reality is irrelevant, because they are not intended to describe reality, but to provide a basis for comparing different aspects of reality with their deviations from the consistently logical case. Often this approach is quite fruitful, and there is no intention here of substituting another methodological approach in its place, but merely of presenting another possible way of conceptualizing complex characteristics of social systems, stemming from the multi-variate analysis pioneered by Paul Lazarsfeld and his colleagues on a quite different level of analysis.

The point at which this approach differs is on the issue of whether generalized theoretical categories can be considered to have a valid relationship to characteristics of total social systems. The implication of the statistical data presented in this paper concerning democracy, and the relations between democracy, economic development, and political legitimacy, is that there are aspects of total social systems which exist, can be stated in theoretical terms, can be compared with similar aspects of other systems, and, at the same time, are derivable from empirical data which can be checked (or questioned) by other researchers. This does not mean at all that situations contradicting the general relationship may not exist, or that at lower levels of social organization, quite different characteristics may not be evident. For example, a country like the United States may be characterized as "democratic" on the national level, even though most secondary organizations within the country may not be democratic. On another level, a church may be characterized as a "non-bureaucratic" organization, when compared with a corporation, even though important segments of the church organization may be as bureaucratized as the most bureaucratic parts of the corporation. On yet another level, it may be quite legitimate, for purposes of psychological evaluation of the total personality, to consider a certain individual as "schizophrenic," even though under certain conditions, he may not act schizophrenically. The point is that when comparisons are being made on a certain level of generalization, referring to the functioning of a total system (whether on a personality, group, organization, or society level), generalizations applicable to a total society have the same kind and degree of validity that those applicable to other systems have, and are subject to the same empirical tests. The lack of many systematic and comparative studies of several societies has obscured this point.

This approach also stresses the view that complex characteristics of a total system have multivariate causation, and also multivariate consequences, insofar as the characteristic has some degree of autonomy within the system. Bureaucracy and urbanization, as well as democracy, have many causes and consequences, in this sense.

On this view, it would be difficult to identify any one factor crucially associated with, or "causing" any complex social characteristic. Rather, all such characteristics (and this is a methodological assumption to guide research, and not a substantive point) are considered to have multivariate causation, and multivariate consequences. The point may be clarified by a diagram of some of the possible connections between democracy, the initial conditions associated with its emergence, and the consequences of an existent democratic system.

The appearance of a factor on both sides of "democracy" implies that it is both an initial condition of democracy, and that democracy, once established, sustains that characteristic of the society, an open class system, for example. On the other hand, some of the initial consequences of democracy, such as bureaucracy, may have the effect of undermining democracy, in turn, as the reversing arrows indicate. Appearance of a factor to the right of democracy does not mean that democracy "causes" its appearance, but merely that democracy is an initial condition which favors its development. Similarly, the hypothesis that bureaucracy is one of the consequences of democracy does not imply that democracy is the sole cause, but rather that a democratic system has the effect of encouraging the development of a certain type of bureaucracy, under other additional conditions, which have to be stated if bureaucracy is the focus of the research problem. This diagram is not intended as a complete model of the general social conditions associated with the emergence of democracy, but as a way of clarifying the methodological point concerning the multivariate character of relationships in a total social system.

Thus, in a multivariate system, the focus may be upon any element, and its conditions and consequences may be stated without the implication that we have arrived at a complete theory of the necessary and sufficient conditions of its emergence. This paper does not attempt a new theory of democracy, but only the formalizing, and the empirical testing, of certain sets of relationships implied by traditional theories, on the level of total social systems.