It is a great honor for me to give this second Johan Skytte prize lecture. Since the citation refers to my work on the "fragility of democracy," it is appropriate to turn again to this theme. It is also timely to give a contemporary view of the concern of Johan Skytte when he wrote "unus rex, una lex, et grex unus," by discussing democracy in multi-national and multicultural societies, as well as democracy and federalism.

The scholar interested in democratic politics at this/;/? de siede and dawn of the 21st century faces new and challenging tasks. In 1995. 117 of 191 states claimed by a simplistic criterion of holding competitive elections, fair or not so fair, to be democracies, that is 61 percent of the countries of the world. Of those 117 states, only 76 satisfy certain requirements that would allow us, by any reasonable definition of democratic political institutions and processes, to define them as working political democracies, 40 percent of the countries of the world (Diamond 1996). Without falling into the fallacy of establishing idealist desiderata which lead to the problem of failure by definition, as Richard Rose so rightly noted, we have to be careful in accepting the claim of many states to be democracies. Even if we were to limit ourselves to the 76, more or less, they include such a wide range of societies and types of institutions, more or less adapted to them, that raise problems we were not dealing with when we focused on the stable, older, western European, American and Commonwealth democracies.

The horizon of Lijphart's book Democracies focused on 21 countries will have to be broadened (Lijphart 1984). In doing so, we will have to focus on many new and different problems to better understand under what conditions democracies will succeed, even in knowing many of their imperfections.

---

*Juan J. Linz is Sterling Professor of Political and Social Science at Yale University. Department of Sociology. P.O. Box 208265. New Haven. CT 06520-8265. USA.

This broader perspective will also call our attention to some of the preconditions that made possible the original development of western democracies. Such a comparative analysis will not be able to elude some problems that were not central to the thinking about democracy, like the relationship between state, nation and democracy, secularization, religion and democracy and, even more, state, constitutionalism, rule of law, bureaucracy and democracy. Much of our thinking has been shaped by the debates about capitalism, socialism and democracy or, more specifically, democracy and the class conflict, ignoring many of the other conflicts in society.

Much more consideration has to be given to the types of democratic institutions (Sartori 1994). In the past, we could limit ourselves to the European parliamentary democracies and the United States as one of the few presidential democracies. The fact that Latin America, with its traditional presidencies, has joined the world of democratic political systems, that two of the new Asian democracies, Taiwan and Korea, are presidential and that new political systems in Eastern Europe have opted for presidentialism, has already expanded our horizon (Linz & Valenzuela 1994; Shugart & Carey 1992). We will have to explore more closely unitary versus federal democracies and start thinking about federal nation states versus multinational federal states. Besides, the fact that so many of the new democracies are economically and socially underdeveloped or undeveloped, that in those countries presumably demo-power generates hopes for economic and social development, both of the country and among the citizens, raises new questions about the relationship between the performance of democratic politics and the legitimacy of democratic institutions. It also forces us to think more about what democracy can actually accomplish and what has to be accomplished by other social institutions.

Moreover, we have to rethink the working of western representative democracies once the argument "that democracy is the worst form of government, except all those other forms that have been tried from time to time" loses some of its power, as people forget the other forms which have contributed so much to our commitment to democracy in spite of all its failures (Rose & Mishler 1996a).¹ As Claus Offe has noted, democracy will be forced to justify itself more by its own success than by comparison with unsuccessful forms of government, even though we can expect the memory of other forms of politics to sustain a belief in democracy (Offe 1994). We need a new and better understanding of the contradictions and tensions in existing and stable democracies which may be inherent to the democratic process and which we have tended to ignore, particularly the role of political parties, representation and politicians in real democracies.

This is a big agenda and the first step has to be to question some
assumptions that easily creep into our thinking without analysis. At the same

time, we must proceed to a more systematic comparative study of existing democracies, both systematically and by in-depth analysis of particular democracies guided by broader questions derived from comparative thinking (Poggi 1978).²

I would like to give a few examples of what I mean. We have forgotten how important the existence of a reasonably modern state (as defined by Max Weber or, more recently, by Gianfranco Poggi) that enjoy some legitimacy and efficacy in the view of the citizens is as a precondition for the functioning of democratic institutions, whose purpose is to legitimate those governing the state (Gellner 1983). To put it bluntly, without a state there can be no democracy. In many parts of the world, the state is questioned or practically non-existent and the patterns of the Rechtsstaat are so inadequate that those democratically elected to govern it will not be able to use it for the purposes of good government.

Another assumption that with few exceptions has guided our thinking about democracy has been that democratic states were at the same time nation states, and the corollary assumption that every nation could or should attain the roof of statehood (to use Gellner's conceptualization), the principle of national self-determination (Linz 1993; 1995). For reasons I have developed elsewhere, we have to face the fact that most states in the world will not be like the traditional nation states (Diamond 1992; Hadenius 1992). They will be, in many cases, multi-national states and certainly multicultural, multi-lingual and multi-religious, in which the traditional conception of an exclusive national identity cannot be the basis of a stable and free polity. We need infinitely more attention to the conditions under which a multi-national, multi-cultural state can function democratically.

Many democrats, ignoring the realities of democratic politics in most of the democracies, still cling to the idea that religion should not be an issue or a factor in democratic political debate, but relegated to the sphere of the private, and that democracy requires separation of church and state. Implicitly, they believe that only a more or less secularized society can be a democratic polity. Reality in the western world, but even more in many non-western, non-Christian countries (particularly Israel, the Islamic countries and India), forces comparativists to think more about the range of alternatives in the relationship between religion and politics compatible not only with democracy but with a liberal democracy. What limits have to be established in a constitution to the wishes of a religious majority? What restrictions of freedom for minorities and non-believers affect or do not affect the core of the liberal democratic regime? These are both theoretical and practical questions which cannot be resolved by wishful political theorizing.

Our thinking about the preconditions and facilitating conditions for
democratic development and stability has been focused on the level of

economic development and the related social and cultural development, with many studies about the relationship between economic development, economic and social inequality, indices of human development and social capital and democracy. Those studies are excellent and important and should be pursued further in the comparative study of democracies, but they do not exhaust the social conditions for democracy. Since so much of our research was focused on western European societies and their historical development, the 20th century crisis of democracy, some of the other developments in the western societies that made democracy possible were forgotten. When we now turn to study democracies in other parts of the world, we realize that some of the preconditions which we did leave largely unanalyzed because they seemed unproblematic are at the core of the difficulties in making real democracies workable in societies which have now democratic political institutions. I am thinking of some of the data I have been analyzing from the Latinobarometro, an excellent survey study of contemporary Latin-American democracies.³ The comparison between post-Communist Eastern Europe and the new states emerging from the Soviet Union also makes it imperative to think about the absolute centrality of a functioning modern state and the importance of the rule of law and constitutionalism as a prerequisite for democratic politics.

No State, No Rechtsstaat, No Democracy

We had forgotten that democracy evolved or was introduced in societies where a modern state had developed over more than a century. A distinction between the private interests of the ruler and officials had largely been institutionalized, a more or less autonomous judiciary had gained the confidence of the citizens, reasonably disciplined and honest police forces served the state, taxes were collected according to laws that treated categories of citizens more or less equally and for public purposes, etc. The modern state, as it evolved, did not always respond to the highest standards of legality and fairness but progressively had been transformed into the modern Rechtsstaat. In the 19th century, before democratization, constitutionalism became more and more powerful as demand and reality, even in countries resisting democratization like Imperial Germany. The state and the authorities were bound in many of their activities by the constitution and the laws, and expected to respect them and forced to do so. All this happened before governments became accountable to legislators and voters. The new democracies which emerged at the turn of the century and in the first wave of democratization took over this modern Rechtsstaat and developed it further. They could rely on bureaucracies and police forces accustomed to respect the law and obey their superiors, now their democratic superiors. The tax

authorities could collect the funds which allowed the expansion of the
modern service and welfare state.

We now discover that, for a variety of reasons, these preconditions are not institutionalized and even less satisfied in countries where democratically elected politicians, politicians committed to democratic political processes and to satisfying the expectations of the citizens, take over the governing of the state. A weak, flawed, underdeveloped, corrupt, incompetent state apparatus is a poor instrument for democratic government. Much of the dissatisfaction of citizens with the way the democracy works in their country is really not dissatisfaction with democracy, but with how the state and its agents work. Democratic governments are responsible for how the state operates. They have some capacity to reform it. but the task of state-building is more complex, slower and difficult than the installation of the core institutions of political democracy. Already Schumpeter, in his list of five conditions for democracy, mentioned the availability of a serviceable, capable and relatively autonomous bureaucracy (Schumpeter 1975, 293).

The problem can be compounded by the latent hostility of many democrats towards bureaucrats, the fear that they could be an obstacle for the implementation of democratically enacted policies and, let us not forget it, many politicians' desire to occupy the positions held by bureaucrats, and the importance of patronage and clientilism in mobilizing support for democratic politicians.

Because of all this, the problem of bureaucracy and all the institutions of
the modem state, from the judiciary to the police, and its relationship to
democracy should be at the top of the agenda.

Finally, we have to ask ourselves to what extent the acceptance and support to democratic institutions by significant sectors of the population, even in our western democracies, initially was not based on a belief in the desirability of those institutions. To what extent did the complex web of social institutions, beliefs and values make democracy possible? One is reminded of the architectural designs of Nervi, with their daring structures which we can understand only by looking at photographs, which also show the complex wooden scaffolding into which the concrete was poured. Democracies were born within the complex scaffolding of institutions which made their construction possible. Perhaps this accounts for the somewhat surprising fact that a disproportionate number of stable democracies in the interwar years had also been constitutional monarchies.⁴ We have to give much more thought to the non-democratic elements in a polity that enjoy the allegiance of the people and whose support or. sometimes initially, tolerance for democracy made possible the institutionalization of democratic processes. A re-reading of Bagehot and his emphasis on the deferential elements of the British constitution should have made us aware of this.

The question is: Can democratic institutions gain full legitimacy in and by themselves isolated from the legitimacy of the broader social order and the legal system, from identification with the state or a nation or the national identities incorporated into the state, the religious elites or symbols of tradition, the economic system, etc.? The reciprocal interactions between the different realms of society, legitimizing a social order of which democratic institutions are only one part, should be at the center of our attention.

Constitutionalism and Democracy

In a democratic political system, the constitution occupies a unique place and the invocation of the constitution and the allegiance to it have been an essential component in the stability of democracies which otherwise were experiencing serious crisis. In Italy, the constant invocation of the participation of all the parties including the Communists in the drafting and approval of the constitution, the parties that were part of the "Arco Constituzionale," served to build a minimum of consensus and trust in an otherwise divided and polarized society. In Germany, in the absence of a single state encompassing the entire German nation in the presence of the ambivalence toward nationalism, the term Verfassungspatriotismus, the identification with the Grundgesetz, served as an element of political integration in the Federal Republic. In Spain, the approval by all major parties, including the Catalan Nationalists, and the overwhelming majority of the electorate of the 1978 constitution, is one of the bases of stability of Spanish democracy and the framework in which certain issues, which otherwise could be divisive, can be dealt with. The contrast between the 1978 and the 1932 constitution of the Republic could not be greater. Today it is an element of consensus, while during the Republic it was an element of contention.

A democratically approved constitution, built on a consensus of major political forces, provides the institutional framework within which political competition can take place. It also reduces that element of incertitude which is associated with democracy, the fear of the decisions of a temporary majority. Since a constitution defines certain issues beyond the control of a temporary majority (by requiring qualified majorities, complex processes requiring time and sometimes referenda to change basic constitutional laws), it gives protection to permanent minorities or those who might be in the minority, without having to wait to redress the decisions of the majority in the next election. The constitutional provisions thereby encourage those who might be fearful about their interests to participate in the democratic political game.

Democracies are and have to be constitutional democracies (Holmes 1995,
10). The lack of a constitutional spirit, of an understanding of the centrality

of constitutional stability is one of the weaknesses of the new democracies in the former Soviet Union and some countries of post-Communist Eastern Europe. A manipulative approach to the making and interpretation of constitutions is and can be a weakness of democracies where the tradition of the Rechtsstaat and constitutionalism is weak. Even the fully justified change of a constitutional provision that people might consider desirable, inevitably becomes an element of instability and distrust of politicians. A case in point is doing away with the "no re-election rule" in some Latin American presidential regimes.

Responsiveness, Responsibility, Accountability as Dimensions of Democracy

I would like to select one example of the inherent tensions in the conception of democracy we find both in democratic theory and in the political reality of democracies. Democracy, without any question, involves the notion of responsiveness, of taking into account the wishes, desires, expectations, values and interests of the electorate. However, democracy will not be possible without governments being responsible, that is, considering the consequences of their policies for the society as a whole, not only their supporters nor even the present generation of citizens. Responsibility involves respect for constraints under which decisions have to be made, the resources available, the stability of the economy, the linkages between a national and global economy and the myriad of other technical matters on which there is knowledge available although not necessarily to voters (Sartori 1995a).⁵

The priming of responsiveness over responsibility and accountability of elected politicians can contribute to the trend toward "poll democracy" and "mirror democracy," localist and interest politics, noted by students of American democracy. Accountability, an essential characteristic of democracy, should not be for representing specific interests but for a whole range of policies of statewide import.

Inevitably there can be conflicts between being responsive to the electorate and acting responsibly. It is up to the politicians elected by the people to deal with those conflicts and minimize them. The responsiveness to the party members, to the electorate supporting them, to the climate of opinion in the society which is of the essence of democratic politics and the basis for electoral success, cannot serve as justification for irresponsible policies. On the other hand, arguments based on conceptions of the welfare of the society in the long run, bolstered by technocratic knowledge, cannot always be successfully invoked. We know from surveys that people want the government to be responsive to their interests, that people might vote in terms of their own interests, but also know that people say that certain

matters should be left to the technocrats, unaware of the contradiction
between both demands.

Pluralistic theory, the emphasis on the inputs in politics, the demand side of the political process, tends to ignore the independent role of the politician and political leadership. Neither the pluralistic view of politics being fundamentally representation of the variety of interests and groups in the society, nor the new visions of a civil society of active citizens in search of reasonable solutions, nor the communitarian views of society, have an adequate place for the role of parties, politicians and political leadership. The misuse of public opinion research and focus groups to find out what the people think or wish and the justification of platforms, electoral promises and policies by such a simple-minded notion of responsiveness, ignores the fundamental role and duty of political leaders to influence, shape and sometimes oppose ill-conceived ideas that may be dominant in the public.

Sartori has stated the problem very well:

A government that simply yields to demands . . . turns out to be a highly irresponsible
government, a government that does not live up to its responsibilities. A representative is not
only responsible to, but also responsible for .... (Sartori 1995a).

Politicians cannot excuse themselves from responding to and before the electorate by invoking technical arguments provided by the experts or putting the blame on them (nowadays very often), on actors outside of the political system, the European Union, or the world bankers. Their actions would be constrained, but it is their task to make the citizens understand those arguments and gain their approval for the resulting policies.

The study of democracy has to start with an analysis of the governmental process and the role of the politicians and political leaders in their relation with the public and with the experts. There has to be constant flow, not only from the bottom up but from the top down, to create an electorate responsive to the political leadership and ready to support responsible policies.

Fortunately, the problem is not as difficult as we might assume. From surveys we know that people perceive their own personal situation and that of their family differently from the situation of the country. They can do personally well and think that the country is going badly, that the economy is not working, and they can themselves feel disadvantaged and still believe that the country is moving and that the economy is doing relatively well. More importantly, it seems, from a preliminary analysis of such data, that the evaluation of the performance of democracy in their country, of how it is working, depends more on the evaluation of the economy as a whole than on that of their personal situation. This element has been called the sociotropic factor in politics (Kinder 1981; Kinder & Kiewiet 1981). It means that there is a breathing space between the difficulties encountered by individuals and their families and the macro-policies and their effectiveness. This is

particularly important when we think of new democracies, both in Latin America and in post-Communist Eastern Europe where, as Richard Rose has emphasized, patience is needed for the success of any reforms and successful development (Rose & Mishler 1996b).

Contrary to the image of demo-power resulting from self-asserting people, from the revolution of rising expectations, there is considerable evidence of the willingness under certain conditions of people to postpone their demands and to project their hopes into the future. That is the basis of the fact that the serious downturn of the economies in post-Communist Europe, even by comparison with the previous Communist regime clearly perceived by the people, has not resulted in a more negative evaluation of the new democratic political system. Interestingly enough, this is not the case in Russia, Belorus and the Ukraine (Linz & Stepan 1996, 434-^57). We need to know much more about how expectations and reality, about how the past, present and future, are perceived and understood by democratic electorates. There is strong evidence that legitimate democratic institutions, clean democratic processes, a state that serves collective rather than personal interests of the rulers, a state in which the courts and the police are trusted, can survive and handle situations in which democracy cannot satisfy the demands of the people. While legitimacy is not unrelated to efficacy, there is no immediate, tightly-coupled relationship between efficacy and legitimacy of democratic institutions.

Democracy is not based simply on acquiescence nor performance but ultimately on the belief in the desirability of democratic institutions even when there is distrust of particular incumbents, as long as there is no generalized sense of distrust of the system. In the past, the availability of ideological alternatives to democracy which could offer themselves as a solution undermined that overall trust, but today, unless the political leadership, the politicians, justifiably lose the trust of the people, the democratic institutions might continue to be legitimate, even in crisis periods.

Much of our research on democracies has focused on the electorate. The easy availability of election statistics stimulated the original work of the great Swedish political scientist, Tingsten, and of public opinion research, to which another Scandinavian, Stein Rokkan. provided much intellectual support, accounts in part for that focus on the electorate. With it. we have tended to focus on responsiveness. On the other side, in the advanced industrial democracies, a large body of research has centered on public policy, and some cooperation between political scientists, sociologists and economists has allowed us to better understand the conditions leading to different policy outcomes, particularly in connection with the study of the welfare state. However, this dual perspective has neglected one of the cores of democratic politics: the politicians. Except for interesting but not very

illuminating studies of the social background of political elites, we know too
little about democratic politicians.

In my view, we need to know much more about what motivates politicians and their crucial role in responding to the demands of the voters, a response that very often is a result of what Carl Friedrich calls "the rule of anticipated reactions," and their ability to articulate those demands and even to shape them, to reformulate them to create public opinion rather than respond to it. We also need to focus more on how they choose between alternatives presented to them by economic theorists, advisers, experts and bureaucrats. We need to understand more about how they select among all the demands, pressures and advice coming to them, and how they mediate between those different pressures, deflect them and transform them.

A good example of the problem is the age-long fear already seen by Aristotle that the mass of the population and therefore of the electorate would be pressuring for distributive policies. Only recently, O'Connor, in Fiscal Crisis of the State, (1973) reformulated that age-old fear of the conservatives, as Sartori did in his interesting essay on "How Far Can Free Government Travel?", expressing his concerns about the exporting of democracy to poorer societies (Sartori 1995b). The theme was invoked some decades ago by the fears of overload of democratic governments in the book by Crozier, Huntington & Watanuki (1975): the fear that redistribution would threaten the necessary process of accumulation of capital in market economies. Now there is evidence that people can be patient, can accept a certain level of inequality if there is hope for improvement, Albert Hirschman's famous tunnel effect (Hirschman 1981, 39-58). The recent work by Rose and Mishler on the patience factor in the post-Communist societies of Eastern Europe shows that people can understand the need to postpone gratification and for sacrifice in certain circumstances. In fact, my own research on economic mentalities of Spaniards shows people are ready to accept that extraordinary profits of an enterprise might be devoted to expansion or modernization of the enterprise rather than redistributed to shareholders or to those who work in it, or paid in taxes.

Indeed, in the modern democracies of the advanced industrial societies, we see that with the expansion of a leveled middle-class society with few highly privileged and a relatively small number of underprivileged, relatively speaking, although better off than past generations, there is a spreading sentiment that redistribution might involve higher taxes which will inevitably affect them. Without giving up the ideological support for the welfare state, the obligation of the state to provide many public goods to everybody, they also support more and more those who advocate lower taxes and give up Keynesianism for neo-liberalism. There is a new demagoguery of fiscal restraint, lower taxes and opposition to redistribution which may gain the support of majorities in the electorate. We do not know to what

extent those positions are the product of ideological formulations, reasonable expert opinion or of politicians thinking they are responsive to the mood of a broad middle-class society, a climate of opinion that is far from the fears of redistributionist demands.

The Tension Between Pluralism and Integration

At the core of democratic politics is the recognition of social pluralism, not only in democratic theory but in institutional declarations like Article I of the Spanish 1978 Constitution which makes explicit reference to political and cultural pluralism. However, democratic societies and political systems are based on a shared identity of the vast majority of the population with the state, and a sense of solidarity of the people expressed in many declarations like: "The Spanish nation proclaims its will to, the Cortes adopt and the Spanish people ratify, the new democratic constitution." There is an inherent tension between the recognition of pluralism in its many expressions, political, cultural, sometimes national, religious and the search for some sense of unity of common identity, shared symbols and sometimes national pride. In all democracies, the question arises how much pluralism is compatible with some form of integration, of loyalty to the institutions of the state, and some sense of either nation or what I have called "state/nation." In addition, there is a tension between a liberal conception based on the rights of individuals to their own distinctiveness and the conditions of freedom in a liberal society, and a conception of pluralism that privileges a pluralism of groups, communities and institutions to which people might or might not belong. This, in some cases, limits the freedoms of individuals to choose not to identify with those groups or institutions privileged by the pluralistic society and polity (Offe 1996, unpublished paper).

The combination between alternative emphasis on pluralism and societal consensus and between individualistic pluralism and group pluralism can generate four different models of democratic polity. One extreme is democracy striving for homogeneity limiting social, cultural and religious pluralism in search for a monocolor rather than a rainbow society. Today we are going to consider such a search for homogeneity and one dominant shared identity as threatening to democracy. This was the Rousseaunian idea that inspired the French Revolution in its hostility to any corps intermédiaires between the individual and the polity, the Jacobin conception of national sovereignty and nation building, the idea of a secularized society in which religion would be tolerated as a private matter, but exclude it from the public sphere. This Rousseaunian idea also underlies the conception of a statist welfare state in which the community would assume responsibility for the well-being of all its citizens. Some of the problems of contemporary

democracy, as we have already noted, are derived from the persistence of this model of society and polity that its advocates consider totally compatible with a government emerging from the free choice of the citizens and accountable to the body politic through the democratic processes, a body politic constituted by individual free citizens. In many conversations with Turkish politicians committed to a progressive democracy based on the thought of Attaturk and ultimately to French social thought, this conception of democracy and freedom is articulated constantly and used to deny the existence of a Kurdish problem.

How different democracies fit into the other three types resulting from these two major dimensions we have defined or, more correctly, how they have placed themselves in the attribute space derived from this multidimensional view, approaching one or another extreme, is one of the great problems that students of contemporary democracies have to deal with. It is the problem at the root of many of the conflicts and tensions in not only new democracies in some parts of Asia, but in the Balkans, in Spain and even in old established democracies like the United States and those of western Europe confronted with growing immigrant, ethnic and cultural minorities.

Federalism and Democracy

The relationship between federalism and democracy should be at the top of our agenda. We need to analyze more carefully the theoretical issues derived from the existence of federal distribution of power and the pursuit of democracy, both political democracy and, even more, some of the goals we associate with democratic government and the values of equality, freedom and solidarity a democratic polity should strive for. We need to analyze from this perspective not only the theoretical arguments for or against federalism, but how federalism in the different democratic federal states has actually worked in relationship to those values. It may be true that there is no alternative to federalism in many states, particularly multi-national, multiethnic, multi-cultural states, but this should not exclude an evaluation of the implications from the point of view of the realization of democracy. The same could be said for federalism in state/nations or nation/states, particularly some large states like the United States.

In the analysis of federalism, it should be clear that federalism should not be confused with decentralization of some decision making powers or the administration within a political unit in which the state retains the power, and therefore the electorate of the state, to modify relatively easily the distribution of power and resources within the political unit. As Robert Dahl has highlighted in a brief but brilliant essay, federalism implies that the demos of the larger unit devolves its power to decide on the demos of a

smaller unit, renouncing to the power to make decisions in the reserved domain of that unit (Dahl 1986, 114-126). It could be argued that ultimately the demos of the larger unit has freely and democratically decided to do so and that therefore federalism does not mean a loss of power by the demos of the larger unit. However, this is not true, since the federal compact institutionalized in the constitution, once enacted, is not subject to the simple majority decision neither legally nor politically.

We tend to forget that there are democratic decisions which are within the agenda of the demos and can be reversed by the electorate whenever, after a limited time period, it or its elected representatives so decide. There are, however, other democratic decisions, like those concerning the existence of a state, the secession of part of a state and generally a federal constitutional system, which cannot be reversed by an electorate. That is why decisions about secession by 50.3 percent of those voting are so questionable. On the other hand, it is questionable to establish thresholds to avoid such largely, if not completely, irreversible decisions by small majorities, majorities of the voters and not of the electorate, since the threshold established might prejudge already the outcome, particularly in multi-national societies.

The greatest difficulty for democratic federalism is, as Dahl wrote:

The fact that one cannot decide from within democratic theory what constitutes a proper unit for the democratic process, like the majority principle, the democratic process presupposes a unit. The criteria of the democratic process presupposes the rightfulness of the unit itself. If the unit itself is not a proper or rightful unit, then it cannot be made rightful simply by democratic procedures (ibid., 122).

In view of this, federal states have often been created by non-democratic or only indirectly democratic decision making, and often without a chance for the electorate of the larger unit to approve the change. In fact, sometimes only the people of the smaller unit are granted the power to make the decisions if they want to support a particular federal arrangement or even opt for secession, although those affected are not only citizens of the smaller unit.

The problems derived from the legitimacy of the unit, both the larger and the smaller unit, are closely linked with the extent to which there is a dual identification with both units, the sentiment among a large number of people to be part of the smaller unit but also of the larger unit, and the willingness of the people of the larger unit to respect the rights granted to the demos of the smaller unit.

These complex problems of legitimation of the units within which democratic processes should take place are often forgotten when politicians and scholars advocate federalism as a solution to the problems of democracy today. Without much more careful and thoughtful analysis of the relationship between democracy and federalism within the existing federal states, we

should not rush to think that trans-state, which many would call transnational, power can only be legitimated by federal democratic institutions. Without a previous sentiment of identity strong enough to provide such a trans-state authority with power, democratic elections would not legitimate that authority. To repeat Dahl,

The criteria of the democratic process presupposes the rightfulness of the unit itself. If the
unit itself is not a proper or rightful unit, then it cannot be made rightful simply by
democratic procedures (ibid.).

A democratic election of a parliament in Strasbourg, the representation of the citizens of Europe in that parliament, the accountability of the Commission (the government of Europe) to that parliament, unless the citizens of Europe would feel their European identity as equal or stronger than their present state or nation-state identity, would not legitimate that authority. Some of those who speak loosely of the democratic deficit of the European Union should think more about the assumptions underlying democratic decision making processes. The fact that they have worked well within nation/states and work, although with some difficulties, in multinational federal states, should not lead to the conclusion that they would work equally well in the European Union (Lepsius 1990a, 247-255).

Trans-state, trans-national communities, although exercising considerable constraining powers, are not directly but indirectly legitimated democratically. A more direct democratic legitimation would assume a broader sense of solidarity equally if not stronger than that uniting the citizens of a state (Lepsius 1990b, 256-269). One can wonder if that sense of solidarity can easily be created beyond the existing states when already it is questioned in the multi-national states and has not been easily achieved even in the historically created states. Let us not forget the solidarity of the family - the "amoral familism" of Italian southerners - the Campanilismo that has plagued Italian politics, the emerging selfish nationalism of the Lega in Italy, the notion that all politics is local politics, should make us wonder about the conditions under which people are ready to accept the decisions of the demos in any political unit.

We should also keep in mind that all the theories of federalism emphasize the importance of a state-wide party system, how a federation like Nigeria tries to impose the requirement that candidates would have a certain support in a number of states, and how complicated democratic politics becomes when the formation of coalitions requires the support of a nationalist regional party.⁶

Democratic theoreticians of federalism emphasize the central importance of an independent judiciary authority, a constitutional court respected by all parties, to make decisions in case of conflict between the component units. It has been argued that with the repatriation of the constitution to Canada and

the disappearance of the jurisdiction of the Privy Council in London, some of the problems of Canadian federalism became more acute (Bednar, Eskridge, Jr. & Ferejohn 1996). This role of the courts in federal states should not escape democratic theorists because it implies that a body not directly and often not even indirectly elected has decisive final authority.

While federalism in multinational democracies serves to give recognition and representation to nationalist demands and directly or indirectly provides a legitimation to the federal state, the federalist recognition of multinationalism isolates the extremist separatist nationalists but also strengthens nationalism, the moderate nationalist parties and the "nation-building" policies (on language, etc.) in the sub-unit.

This should make those who favor a truly federal Europe - with real power in the Strasbourg parliament - think if such a parliament, rather than reflecting large European ideological groupings, like social democrats and popular parties, would not be composed largely by nationalist parties representing the nationalists of the present states and the state-less nations in those states. Instead of expressing dissatisfaction with their governments, with European unification, with the policies of the Union, by voting for protest candidates like the Greens or the Left, nationalism could become the main political force. How to make policies with a parliament composed by nationalist-particularist-representatives is a question that cannot be ignored.

The way in which the Helvetic Confederation has handled the problem of the creation of a new canton could be one example of how to achieve this goal. A careful analysis of that experience would also show how difficult it would be to reach such a consensus within and between the members and the citizens of a future federal European community.

I have extended myself too far and, at the same time, too little on the
problem of federalism to highlight how it should be at the top of our agenda.

Conclusion

The response to many of the problems that this multi-dimensional view of contemporary societies and polities generates, ultimately depends also on our conception of what a democratic state can and should do. There is a tension between a statist conception of democracy and one based on the idea that civil society should assume a wide range of functions. An idea that underlies certain ideological conceptions, like those of the Catholic social doctrine of subsidiarity, the variants of neo-corporatism and ideological federalism. It is no accident that the Spanish politicians and thinkers in the 19th century and the first decades of this century inspired by Proudhon established a link between anarchism and extreme federalism. From such a perspective, the state should have a minimal role, a conception that brings

neo-liberalism in its extreme formulations close to those ideological currents linking sometimes with a critique of contemporary forms of democracy, relying on statewide elections and parties legitimating democratic state government. Paradoxically, some of those critiques of the democratic state, representative democracy and political parties as an instrument to make it possible, link with a participative democracy conception which advocates a democratization of the whole society and all its institutions and forms of direct democracy.

I want to stress that students of democracy should not analyze this whole range of problems from a purely theoretical perspective. Political philosophy is not always a good guide of thinking on these complex matters. It is necessary to look at the implications in the different types of societies and even in each one specifically with such a framework in mind for a better understanding of the real implications: what difference do those different actions of a democratic polity make in real life with the kind of citizens we have in our societies rather than some ideal conception of citizenship.

More concretely, how do these different conceptions approach asyntotically the guiding values of democracy: freedom, equality, and perhaps we should not forget, fraternity? As I mention these three words which we find on every public building in the French Republic, we should bear in mind that this was the Jacobin ideal of democracy. The fact that today our thinking about a pluralistic democracy is not particularly enthusiastic of the Jacobin model is evidence of the centrality of this set of problems for the study of democracies. Indeed, most democracies in the world do not respond to that original Jacobin model. This does not mean that that model did not have its virtues, but certainly it has proven unrealizable in most societies.

Offe has recently stated very well our task of linking theory and research:

It is necessary, to proceed in the context of political theory not only legal and philosophical value and legitimation discourses, but in addition to analyze and evaluate in empirical-social science perspective the contribution of specific institutional components of the constitutional reality to the satisfaction of the challenges (listed in his paper). [It also has] to explore the institutional transformations that possibly can contribute to meet those challenges; a comparative functional analysis of democratic institutions and the existing and imaginable variants (Offe 1996).

Within the house built on the blueprint of the core institutions for political democracy, as we have defined it, we have a house of many mansions. But we should always be clear minded about how far some of the conceptions of the various inhabitants of the house are threatening to the overall building of a democracy. We have to explore how the different demands emerging from different conceptions of democracy relate to the democratic core that defines the boundary between democracy and all other forms of polity. Without clarity on this question, we risk much confusion.

Let me close with a text from Hölderlin about the state which I think could
also apply to our work on understanding and improving real democracies. In
Hyperion, the German poet wrote:

You accord the state far too much power. It must not demand what it cannot extort. But what love gives, and spirit, cannot be extorted. Let the state leave that alone, or we will take its laws and whip them into the pillory! By Heaven! he knows not what his sin is who would make the state a school of morals. The state has always been made a hell by man's wanting to make it his heaven.

The state is the coarse husk around the seed of life, and nothing more. It is the wall around
the garden of human fruits and flowers.

[Du räumst dem Staate denn doch zu viel Gewalt ein. Er darf nicht fordern, was er nicht erzwingen kann. Was aber die Liebe gibt und der Geist, das Läast sich nich erzwingen. Das lass er unangetastet, oder man nehme sein Gesetz und schlag es an den Pranger! Beim Himmel! der weiss nicht, was er sündigt, der den Staat zur Sittenschule machen will. Immerhin hat das den Staat zur Hölle gemacht, dass ihn der Mensch zu seinem Himmel machen wollte.

Die rauhe Hülse um den Kern des Lebens und nichts weiter ist der Staat. Er ist die Mauer
um den Garten menschlicher Früchte und Blumen.] (Hölderlin 1970, 607; Santner 1990).

---

NOTES

1. Rose quotes the original Statement by Churchill in the House of Commons in 1947, Hansard, 11 November, 1947, col. 206, in which he said: "Many forms of government have been tried and will be tried in this world of sin and woe. No one pretends that democracy is perfect or all wise. Indeed, it has been said that democracy is the worst form of government, except all those other forms that have been tried from time to time."

2. Claus Offe in an unpublished paper, "Homogeneity and Constitutional Democracy. Group Rights as an Answer to Identity Conflicts?" starts with an excellent analysis of why the state is a precondition of a political regime - concretely democracy - and democracy as such can not deal with some of the problems of stateness. His analysis converges with the one developed here and in the book I co-authored with Alfred Stepan (Linz & Stepan 1996, ch. 2, 16-37).

3. Latinobarometro 1996, Total Data Crosstabulations by Country, June 1996, Directed by Marta Lagos.

4. Of the 13 monarchies in interwar Europe, seven were stable democracies, and of the 16 republics, five were stable. See Linz 1991, 231-277.

5. On the relationship between responsiveness, responsibility, accountability and ideological consistency, see Linz 1966, VII-CXIX, LXXIX-SCIX.

6. On the creation of a 19-state federation - to weaken the ethnic and regional solidarities that cursed the First Republic - and the condition of winning at least one fourth of the votes in two thirds of those 19 states, see Diamond 1988, 33-92; see also Horowitz 1985, 601-628.

REFERENCES

Bednar, J., Eskridge, W. M. & Ferejohn, J. 1996. A Political Theory of Federalism.
Unpublished paper. April.

Crazier, M. J., Huntington, S. P. & Watanuki, J. 1975. The Crisis of Democracy. New York:
New York University Press.

Dahl, R. 1986. "Federalism and Democratic Process," in Democracy, Liberty, and Equality.
Oslo: Norwegian University Press.

Diamond, L. 1988. "Nigeria: Pluralism, Statism and Struggle for Democracy," in Diamond,
Linz & Lipset, eds., Democracy in Developing Countries, Africa (vol. 2). Boulder, CO:
Lynne Rienner.

Diamond, L. 1992. "Economic Development and Democracy Reconsidered," in Marks, G. &
Diamond, L., eds., Reexamining Democracy, Essays in Honor of Seymour Martin Lipset.
Newbury Park: Sage.

Diamond, L. 1996. "Is the Third Wave Over?" Journal of Democracy July, 20-37.

Hirschman, A. O. 1981. "The Changing Tolerance for Income Inequality in the Course of
Economic Development," in Essays on Trespassing, Economics to Politics and Beyond.
Cambridge: Cambridge University Press.

Holmes, Stephen. 1995. Passions and Constraint, On Theory of Liberal Democracy. Chicago:
Chicago University Press.

Horowitz, D. L. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press.

Hölderlin, F. 1970. "Hyperion," in Sämtliche Werke und Briefe. München: Carl Hansen. Vol.
1.

Kinder, D. R. 1981. "Presidents, Prosperity and Public Opinion," Public Opinion Quarterly
45:1, 1-21.

Kinder, D. R. & Kiewiet, D. R. 1981. "Sociotropic Politics: The American Case," British
Journal of Political Science 11, 129-161.

Lepsius, M. R. 1990a. "Ethnos and Demos," in Interessen. Ideen und Institutionen. Opladen:
Westdeutscher Verlag.

Lepsius, M. R. 1990b. "Der europaische Nationalstaat: Erbe und Zukunft." in Interessen,
Ideen und Institutionen. Opladen: Westdeutscher Verlag.

Lijphart, A. 1984. Democracies. Patterns of Majoritarian and Consensus Government in
Twenty-One Countries. New Haven: Yale University Press.

Linz, J. i. 1966. "Michels e il suo contributo alia sociologia politica." introduction to Michels,
R. La sociologia del partito politico. Bologna: II Mulino.

Linz, J. J. 1991. "La Crisis de las Democracias." in Cabrera, M.. et al., eds.. Europa en Crisis
1919-1939. Madrid: Pablo Iglesias.

Linz, J. J. 1993. "State Building and Nation Building," European Review 1. 355-389.

Linz, J. J. & Stepan, A. 1996. Problems of Democratic Transition and Consolidation: Southern
Europe, South American and Post-Communist Europe. Baltimore: The Johns Hopkins
University Press.

Linz, J. J. & Valenzuela, A., eds. 1994. The Failure of Presidential Democracy. Baltimore:
The Johns Hopkins University Press.

O'Connor, J. 1973. Fiscal Crisis of the State. New York: St. Martin's Press.

Offe, C. 1994. Der Tunnel am Ende des Lichts: Erkundungen der Politischen Transformation
im Neuen Osten. Frankfurt: Campus.

Offe, C. 1996. "Homogeneity and Constitutional Democracy. Group Rights as an Answer to
Identity Conflicts." Unpublished paper.

Poggi, G. 1978. The Development of the Modem State. Stanford: Stanford University Press.

Rose, R. & Mishier, W. 1996b. "Political Patience in Regime Transformation: A Comparative
Analysis of Post-Communist Citizens." Paper. APSA Meeting. San Francisco.

Santner, E. L., ed. 1990. Hyperion and Selected Poems. New York: Continuum.

Sartori, G. 1995a. "Failure of Representation or Failure of Understanding." Paper. Forum
International des Sciences Humaines Symposium. "The Crisis of Representation." Paris.

Sartori, G. 1995b. "How Far Can Free Government Travel?", Journal of Democracy 6:3,
101-111.

Schumpeter, J. A. 1975. Capitalism, Socialism and Democracy. New York: Harper and Row.