FEDERAL
theory
Max Frenkel
| Centre for Research on Federal Financial Relations | The Australian National University, Canberra |
© Max Frenkel (Zuchwil/Switzerland) 1986/2007
| CHAPTER | PAR. | |
| A. | Introduction | 001 |
| B. | Antagonism or Harmony | 036 |
| C. | Democracy | 044 |
| C.1 | Forms of Democracy | 057 |
| C.11 | Democracy as a regime | 060 |
| C.12 | Democracy and Democratization | 075 |
| C.121 | Opinion Polls | 090 |
| C.13 | Democracy as autocracy in Disguise | 100 |
| C.2 | Functional Aspects of Democracy | 106 |
| C.21 | Democracy and Learning in political systems | 111 |
| C.211 | Maintenance of Learning Capacity | 114 |
| C.212 | Problem Solving Capacity | 121 |
| C.22 | Democracy and Stability of Political Systems | 134 |
| C.221 | Conflict Resolution | 137 |
| C.222 | Legitimation | 151 |
| C.2221 | Legitimation and the Majority Principle | 156 |
| C.2222 | Legitimation and Political Culture | 167 |
| C.23 | Democracy's Assigned Functions | 183 |
| C.231 | Equality | 186 |
| D. | Federalism | 195 |
| D.1. | The Concept of Pluralism | 201 |
| D.2 | The Concept of Federalism | 210 |
| D.21 | Federalism and its Complements | 221 |
| D.22 | Federalism and Subsidiarity | 231 |
| D.3 | Federalism in Government | 241 |
| D.31 | Confederation | 247 |
| D.32 | Federation | 254 |
| D.321 | The Sovereignty Problem | 272 |
| D.3211 | Sovereignty | 274 |
| D.3212 | Sovereignty and Federation | 289 |
| D.33 | Federation and Confederation | 302 |
| D.34 | A Bevy of Federalisms | 312 |
| D.341 | The Alphabet of Federalism | 316 |
| D.342 | New, Co-operative, and Creative Federalisms | 339 |
| D.3421 | New Federalism | 339 |
| D.3422 | Co-operative Federalism | 346 |
| D.3423 | Creative Federalism | 354 |
| D.4 | A Brief Survey of Federal History | 364 |
| E. | Functions of Federalism | 404 |
| E.1 | Federalism and the Political System's Learning Capacity | 416 |
| E.11 | Maintenance of the Political System's Learning Capacity | 417 |
| E,111 | Relief of Decision Centres | 418 |
| E.112 | Competition and Division of Powers | 424 |
| E.113 | Innovation | 434 |
| E.12 | Solving Problems in Pluricentral Organizations | 443 |
| E.121 | Costs of Federalist Solutions | 447 |
| E.1211 | Costs of Information | 452 |
| E.1212 | Costs of Mobility | 455 |
| E.1213 | Costs of Scale | 457 |
| E.1214 | Spillovers | 472 |
| E.122 | Quality of Federalist Solutions | 476 |
| E.1221 | Quality as Technical Adequacy | 485 |
| E.1222 | Quality as Public Satisfaction | 501 |
| E.123 | Large and Small-scale Democracy | 519 |
| E.13 | Intellectual Æsthetics | 531 |
| E.2 | Federalism and the Political System's Stability | 540 |
| E.21 | Stabilization and System Reliability | 541 |
| E.22 | Conflict Resolution | 551 |
| E221 | Integration of Individuals | 558 |
| E.222 | Integration of Groups | 570 |
| E.2221 | Limits of Federalist Integration | 588 |
| E.23 | Legitimation | 592 |
| E.231 | Compact and Compromise | 594 |
| E.232 | Political Culture | 602 |
| E.3 | Assigned Functions of Federalism | 614 |
| E.31 | Citizen Participation | 617 |
| E.32 | Civic Liberty | 625 |
| E.4 | Function of Middle Levels | 644 |
| F. | Genesis and Durability of Federations | 663 |
| F.1 | Genesis of Federations | 664 |
| F.2 | Durability of Federations | 678 |
| Constitutions | ||
| Bibliography | ||
| Index | ||
| End |
A. Introduction
001 In their discourses on government, Plato and Aristotle discussed all those problems which were important to an Attic citizen if he were to understand and order his polity. This encyclopædic approach was also used in theories of government that were developed in the Middle Ages (Rehm L650/261). Only very slowly public law became a science proper (L650/261). But the theory of government remained an overarching theoretical concept in continental Europe right to the end of the nineteenth century (e.g. Bluntschli L75/1, Jellinek L382/4).
002 To-day, at least in Europe, such a unity is at best a historical recollection. The sciences concerned with the state have been differentiated and specialized. Public law, economics, political science, political sociology, geography, planning, and other academic disciplines have developed their own systems of reference for theory and analysis. To some degree the scientific dialects have become rites of passage, restricting access to the mysteries of special knowledge to the initiated few. Interdisciplinary discussion has become difficult and is neglected, assertions to the contrary notwithstanding.
003 Public law as a scientific discipline has been part and parcel of this development. On the one hand it has benefited from compartmentalization. Thus the impressive system of the Viennese doctrine of so-called pure law came into being. And thus it became possible continuously to hone and refine legal techniques.
004 On the other hand, compartmentalization has also brought disadvantages to public law doctrine. In continental Europe, lawyers traditionally do not deal with the state as an abstraction only. They also look at it as a reality. Yet most public law specialists are not interested in empirical studies. They prefer an intuitive approach. Whilst this may be possible where the scholar is thoroughly acquainted with the object of his studies and where he is aware of his ideological biases, both these prerequisites do not always obtain. But worse still is the habit of deducing normative findings from general principles such as democracy or federalism without having first analysed the extra-legal content of such concepts. And yet the instruments for such theoretical and empirical analysis are there. They can be adapted from the methodology of political science and, to some extent, of political economy. Even some of the necessary information can be found there.
005 All scientific disciplines that analyse government have one thing in common. They all claim that their findings are relevant to policy making. They all want to furnish basic information for political and administrative decision making. But as these findings are only relevant to parts of the problem and as they may very often be interpreted contradictorily, what really happens is that researchers do not furnish basic data for decision making but produce arguments for preconceived ideas. Obviously, this selective dealing with scientific data by political office-holders can never be and should never be quite eliminated. But the complementarity of the different sciences and their overlapping has become so evident that something along the lines of the old «integrated» theory of government is desirable.
006 The Swiss economist Frey (L268/117) for instance criticizes lawyers for not having devised ways of organizing special districts (i.e. voluntary or compulsory associations of local governments) without reducing democracy. Already the words used - special district and democracy - exemplify the different connotations they have for economists, lawyers and political scientists. Without mastering the different vocabularies nobody can solve the organizational problem.
007 The best starting point for such an attempt at integration is systems theory or systems thinking, because in a technical sense there are several competing systems theories.
008 Systems theory is a change of scientific paradigm in the sense proposed by Kuhn (L444). The former and still now predominant approach in the humanities emphasized relations of cause and effect. Systems theories -preceded by Gestalt psychology, Pepper's contextualism and Henderson's theory of balance- are different. They look for interrelations within a whole, the system, whose parts must not be treated in isolation. Systems theories do not build hierarchies. They analyse patterns. The first influential theoretician was von Bertalanffy in the thirties. At about the same time Parsons introduced systems thinking into social science. In connection with Wiener's cybernetics (systematic study of communications and control in organizations of all kinds: Deutsch L166/76) and computer technology systems theory experienced a boom after the Second World War. Thence its technical flavour which sometimes appears as exaggeratedly modernistic (for a critical presentation of systems theories' development see Lilienfeld L482/7ss,196ss).
009 Systems theory tends to be a theory of integration (Schmieg L300/444). It is able to show and analyse systemic relationships between findings of different scientific disciplines. This is achieved by tying them together in an abstract network. The latter's parameters are different as between different schools of thought. Systems theory dispenses to a large degree with causal relationships, whereas the usual attempts at integration try to relate in a consistent way different scientific findings to a few grand basic hypotheses.
010 The degree of abstraction of the frame of reference used is of necessity high. A system for instance is defined (Hall/Fagen in L300/444) as «a set of objects together with the relationships between these objects and their attributions.» But because in everyday language nearly everything is a system - e.g. the political system, the education system, the public transit system, the economic system etc. - two distinct dangers arise. One is that every time something can be understood, if only in a banal way, to be part of a «system», it is presented as «systems theory» and glorified or mystified with cybernetic terms (for examples see Lilienfeld L482/232ss). The other danger is the mixing of different scales of abstraction. Thus there is talk about the political system, understood in accordance with Easton (L188/21) as those structures and interactions of society by which authoritative decisions on values are arrived at. Against this some critics then hold (e.g. L482/216) that such a political system does not exist in reality or, in other words, that it does not conform with the parliamentary or some other political system of quite another scale of abstraction. This is a pitfall which is very difficult to avoid in discussion.
011 Systems thinking analyses relationships. It shows the functions that parts of systems have in connection with other parts and with the system as a whole. But such an integration is not intended to achieve consistency or to solve social conflicts. Systems thinking is for instance wholly alien to efforts to achieve harmony between «federalism rightly understood» and «democracy rightly understood» (e.g. Lang L451/322, Messmer L542/230s, Usteri L825/347). Systems thinking will regard both federalism and democracy as forms of governance, different but in several respects closely related. Starting from there it will then analyse the role of both organizational patterns in the political system as a whole, showing amongst other things the prerequisites for harmony and for conflict. The ways to do this will differ with various schools of thought. But the scientific as well as the political value of such an approach is integration, integration in this case standing for the processing of findings of different scientific disciplines so that their homogeneity with respect to specific systemic elements becomes apparent and allows comparisons.
012 Considered this way the advantage of systems thinking is its aptness as an interdisciplinary metascience. In data processing parlance, systems thinking is the program used to achieve compatibility of different computers. It is therefore obvious that systems thinking is no substitute for the sciences of law, economics, political science etc. But all social sciences can profit from the attempt to make the knowledge of other disciplines bear fruit for their own work through the techniques of systems thinking.
013 I was introduced to systems thinking by two key books: Karl W. Deutsch's «The Nerves of Government» (L166) and Norbert Wiener's «The human use of human beings» (L860). My professional insights into the realities of Swiss federalism, my study of economics, political science, legal and other theories of federalism, visits abroad, and contacts with foreign scholars made me look for a vantage point for a global view of the phenomena I was confronted with. I started to play with systemic approaches, designing for my own use the first elements of a systems theory. It was only later that I realized certain parallels with Parsons's theory, as for instance with his functions of pattern maintenance, integration and adaptation (L604/4s). But parallelism is all there is, not identity. Parsons's aims are directly connected with his interests in sociological research. Mine have little to do with the explanation of empirical data. I was looking for a reference model for interdisciplinary integration.
014 During the seventies many people worked with hierarchies of goals within global conceptions. I, too, attempted to take material goals of government as starting points for my attempts at integration. But because such goals are of necessity value oriented the approach bore, to my mind, little fruit.
015 It is a very old philosopher's thesis that human society has a value-oriented goal. This goal is justice. Thus Plato (L125/68s), thus some thinkers in the Middle Ages (Cassirer L125/51), thus in more modern times Elyot( L211/1) and Madison (L334/51), thus also in our own century Burckhardt (L109/228). Related to justice, or just another version of the same idea is the goal of general welfare or happiness (Kelsen: «Justice is social happiness» L413/6). For further examples, see Aristotle (L34/7/13), Bentham (L49/184), Madison (L334/43), Sismondi (L755/93).
016 Another goal of government which is often postulated is the full development of the personality either of individuals (e.g. Imboden L386/56, Marcic L509/314, Mill L547/206) or of society as a whole (e.g. Bluntschli L74/362s). Sometimes the safeguarding of private property is mentioned (Locke L489/124). For Hegel the State is an end in itself (L348/258), the reality of the moral idea (L348/257, compare Cassirer L125/274s).
017 It is comparatively easy to agree that justice is the ultimate end of society and of government. It is more difficult to define the contents of justice. The material definitions usually introduce elements which are themselves in need of explanation. Thus the famous first phrase of Justinian's Institutions, which may be traced back to Aristotle «Iustitia est constans et perpetua voluntas ius suum cuique tribuens», is unambiguous only for those for whom «ius» (law, right) is an evident reality in itself (e.g. Marcic L509/176). Wiener, mentioned above as the «father» of cybernetics, made this comment (L860/105): «Empirically, the concepts of justice which men have maintained throughout history are as varied as the religions of the world, or the cultures recognized by anthropologists. I doubt if it is possible to justify them by any higher sanction than our moral code itself, which is indeed only another name for our conception of justice.» But if justice is a formula the contents of which are defined by social value judgments, definitions might be as positivistic or legalistic as, for instance, the one by Hobbes (L363/15): «And the definition of Iniustice, is no other than the not Performance of Covenant. And whatsoever is not Unjust, is Just.» Thus the stress is no longer put on the aims of the procedure but on the procedure itself. Hobbes's law is just in itself; whatever happens legally realizes justice.
018 It may nevertheless be questioned whether such a counter-intuitive and ultimately paradoxical use of words has any cognitive value.
019 The former Prussian Prime Minister and political theoretician Brecht once undertook the task of listing all the tenets of belief taken as yardsticks for justice over the centuries. He found twelve of them: 1) egalitarianism, 2) liberalism, 3) belief in revelation, 4) conservatism, 5) authoritarianism, 6) majoritarianism, 7) hedonism, 8) group allegiance, 9) harmonism, 10) transpersonalism, 11) dedication to duty, and 12) maintenance of peace and order (L86/181ss). He then went on to isolate those elements which had always been part of the idea of justice, namely (L86/477s): 1) truth, 2) generality of the value system, 3) equal treatment of all who are equal according to accepted values, 4) no limitations on freedom other than those accepted by the value code, and 5) respect for the basic necessities of life in the strict sense of the term.
020 As to justice I accept positivism insofar as I have my own idea of justice as well as recognizing several others. But there is none which I would proclaim to be evident for everyone, even if there might be such a thing as true justice. My personal idea of justice has to do with the values of equal opportunity (materially and procedurally), human dignity and tolerance. This is my view of the aim of government. But it is certainly not the aim of government (compare Kelsen L414/24). For me there is no such thing as a unique goal or system of goals of government capable of being stated with more than a formal content. But this also means that it becomes impossible to start my attempt at integration from a value-oriented viewpoint if I want to claim evidence beyond the restricted circle of those who might share my values.
021 Bertrand de Jouvenel once stated his ideas about goals of government in a very lucid way (L375/14): «It is a very important matter to make up our minds whether the civil society we live in is to be thought of as dedicated to a purpose. If so, our institutions do not correspond. Personally, my feeling is that civil society should not be regarded as dedicated.... On the other hand we may reasonably think of the administration as a dedicated body from which we 'get' good public service just as we 'get' science from the republic of scientists. This perhaps is not a morally pleasing picture (though I would say in favor of such a state of affairs that it allows us to join dedicated societies of our own choice, which would not be the case if civil society were a dedicated body) but it seems to me to fit the facts. Certainly there is not the least implication in our institutions and mores that a man must make himself into a partner of a general purpose.»
022 All that may validly be said about goals of government is encapsulated in the concluding sentence of Woodrow Wilson's «The State» (L863/1536): «This, then, is the sum of the whole matter: the end of government is the facilitation of the objects of society.» These objects are manifold and conflict-ridden. It is not possible consistently to harmonize the demands of both crusading teetotallers and wine-makers. It is only possible to optimize such demands, that is, to balance them into a state of but partial satisfaction.
023 Under such a formal definition of the ends of government the state is a subsystem of society. Systems theory does not normally concern itself with the state but rather concentrates on the political system. This helps to avoid the dangerous confusion of abstract notion and concrete representation of a particular regime (compare no.010). The political system is the totality of those structures and procedures as well as their connexions by which value judgments are authoritatively arrived at, i.e. with an accepted and, if necessary, also enforceable claim to validity within the social system (see also Easton L188/21,153). In its legal essence the political system is the social subsystem of governance.
024 If we try to go a little further and to formulate an encompassing abstract end for the political system, we arrive at the axiomatic systemic statement of self-preservation. To avoid a possible misunderstanding: If we talk about self-preservation we do not mean the preservation of a particular social state of affairs obtaining somewhere in real life. This state of affairs, this regime may be changed. Perhaps it even has to be changed to maintain the function of governance necessary for the survival of the social system. Politics as such have to be preserved, not particular policies. For this axiomatic self-preservation of the political system the latter's structures and procedures all have their role to play. They thus have systemic functions within the political system.
025 I make a distinction between two different sets of such systemic functions. I differentiate between functions which have to do with the system's learning capacity and those which are concerned with its stability. I am fully aware that this difference is neither rigid nor compelling. Learning may be seen as a means to maintain stability and vice versa. But I think that this difference nevertheless corresponds to certain centres of gravity of individual functions. As a third set of functions I then introduce «willed» functions. These are functions which society assigns through the political process to certain structures and procedures. Such functions are systemic only in an indirect way. As examples, we may take the functions of guaranteeing individual liberties, participation in decision making, etc. The reason for treating these functions separately, and not «translating» them into systemic ones first, are purely practical, making the presentation easier.
026 According to Deutsch (L166/129) every system designed to steer and maintain itself autonomously is in need of a constant flux of information about its environment, its past, its own functioning and its parts. If this flux is stopped for any length of time - as the result of, say, information suppression, secrecy or bureaucracy - a loss of adaptability may ensue leading to gradual or to sudden collapse. A system's learning capacity is its capability to process information adequately and to make the necessary adjustments (see also Schmidt L300/213). Learning in this sense combines several closely related functions which will form the basis of this book's analysis: the maintenance of learning through structures and procedures, the perception of social as well as other demands, and, finally, the efficient processing of such demands. These considerations will be elaborated upon below.
027 As to the stability of a system, there are three groups of functions: systemic maintenance of stability (e.g. crisis management through organizational duplication involving so-called redundancies), resolution of material as well as demographic conflicts (the latter through integration) and, finally, legitimation of the system.
028 Maintaining stability through redundancies is the answer to the question whether it is conceivable to group elements which are unreliable individually into a reliable whole. This may be achieved if there is a sufficient and adequately organized overlap or duplication through redundancies. The system has to be designed in such a way as to prevent the failure of some of its parts to cause the breakdown of the whole. Wherever such failures could become dangerous for the system the parts will be duplicated and arranged along parallel lines so that several mutually independent information channels are provided for (compare Landau L449/187s).
029 In the context of conflict resolution, demographic integration will be important. This requires the focusing of individual or group loyalties to a given political system, thereby defusing or diminishing potential demographic conflicts (for political integration see also Deutsch L166/150,241). But we shall also look at the reduction of material conflicts as a prerequisite to stable government (see Eckstein L190/256).
030 In the long run, the state as a system can survive only if its citizens or subjects accept the given government. This is one of the older tenets of the science of government (see for instance Hooker L367/1/10/8, Jellinek L382/426). Governance will be accepted if this is an evident social duty. Government, then, is legitimized or, in cybernetic terms (Deutsch L166/152): «In this sense, a 'legitimacy myth'... is an effective set of interrelated memories that identify more or less clearly those classes of commands, and sources of commands, that are to be given preferential attention, compliance, and support, and that are to be so treated on grounds connecting them with some of the general value patterns prevailing in the culture of its members» (see also Duverger L186/32ss). «In the context of cultural legitimation, then, a society is self-sufficient to the extent that its institutions are legitimized by values that its members hold with relative consensus and that are in turn legitimized by their congruence with other components of the cultural system, especially its constitutive symbolism» (Parsons L604/10).
031 Finally, we shall look at the so called assigned functions of a political system: goals are ascribed to the system which are not strictly necessary for the latter's maintenance but which give expression to a society's values and ambitions. It is this kind of function which is usually at the centre of political debate. Critical discussion of federalism or of democracy, for example, is very often influenced by such values, as for instance with equality in the one case or minority rights in the other. (Readers with a penchant for economics might understand the word «assigned» better if they translate it as «meritorized», L566/65.)
032 A considerable part of the commonly proffered catalogues of the ends of government is of this assigned kind. For an economist, for instance (Frey L268/25s), the tasks of the state result from the imperfections of a pure market economy: «1. the provision of public services (mainly of infrastructure in its broadest sense); 2. economic stabilization; 3. distribution policy to adjust disparities in income and wealth resulting from the play of market forces; and, 4. order and security (creation and maintenance of the market-place's legal framework).» (For other catalogues see Weidenbaum L462/7s and Wilson L863/1475ss.)
033 It is obvious that a political system faced with such an array of tasks cannot possibly pursue one unique goal. It will constantly have to balance diverse demands it is confronted with. Any government, therefore, will be unable to achieve more than some medium degree of citizens' satisfaction. It will usually strive to contain dissatisfaction within tolerable limits. «The best definition of the best Government is, that it hath no inconveniences but such as are supportable - inconveniences there must be» (Halifax L330/203). The state is, as has already been pointed out, an optimizing system. As such it not only realizes values and goals. Through the continuing interplay of social groups it also defines the values and goals which are to be realized (for the relevant theories of Bentley, Lamprecht, Dewey and Lippmann see L579/20ss). Some of these goals will have long-range validity, others will only be pursued for a short time. But one thing is clear: given an optimizing decision-making system willed goals will never in their totality be consistent. There will always be contradictions.
034 This, then, is in a few short words the frame of reference used for the analysis here presented. But I would not dare to declare it exempt from ideology. By choosing systems thinking rather than a value-oriented theory for my approach I have taken a stand. And there is ideology very often in reflections on details. The author's ideological standpoint might be characterized as liberal in the traditional European sense, conservative and populist. But the greatest danger inherent in functionalism is probably another one. It is a certain inclination to abstain so completely from passing judgment and to explain everything systemically, so that the status quo might finally become an end in itself and will thereby be legitimized. Nevertheless, as long as one is aware of this problem, systems thinking still presents, to my mind, the best basis for identifying and integrating findings of different scientific disciplines.
035 The monograph you are reading is essentially a translation from the German original of the first volume «Föderalismus» of a treatise «Föderalismus und Bundesstaat» (L264a). The second volume «Bundesstaat» (L264b) contains mainly an exposition of comparative constitutional law and policy, based on the conceptual foundations laid with the present book. Part of the monograph was written in 1978 during my stay as a Visiting Fellow of the Centre for Research on Federal Financial Relations at the Australian National University in Canberra. I not only profited there from the intellectual stimulation of what was then probably the finest, most productive and most international centre for research on federalism. I also had the benefit of the far-reaching knowledge of the late Professor Russell Mathews, then the Centre's director. I should also like to mention those scholars of federalism which over the years have become friends and who have influenced me greatly, if not always in the same direction (I am an optimizer, too, insisting on the right to make my own mistakes). I refer specifically to the late Professor Daniel J. Elazar from Jerusalem and Philadelphia, John Hayes from Chichester and Professors Ellinor and Vincent Ostrom from Bloomington, Indiana. – Looking back in 2007 on what I published in 1986 I think it wasn't too bad. But by then I already had an advantage over many of the so-called experts on federalism: I had studied federalism on a worldwide scale and I knew what I was writing about.
B. Antagonism or Harmony
036 Federations are federalism's typical manifestation in the world of politics (no.255). For the moment we may leave the question as to what degree the federal principle has in fact been realized in these federations and how far it has remained an empty formula. The fact is that this model is either so attractive or that it corresponds so well to given facts that more than 52 per cent of the world's surface with 40 per cent of the world's population is organized under federal regimes (no.389, Duchacek L180/195ss). But applications of the federal principle may also be found in most unitary states, because it is a natural consequence of sociologically different political units (Elazar L205b/20ss, also Frey L268/40, Friedrich L275/1, Jerusalem L385/6, Imboden L387/21, Schollenberger L727/272, Schwartz L736/1s). Even communist China has made provision for autonomous regions (V4,95ss, Domes L177/122ss) with a say in central decision making (V59/1).
037 Today's federations are: Argentina, Australia, Austria, Brazil, Canada, the Comores, Czechoslovakia, the Federal Republic of Germany, India, Malaysia, Mexico, Nigeria, Pakistan, Switzerland, the Soviet Union, Tanzania, the United Arab Emirates, the United States of America, Venezuela and Yugoslavia. If «classic» federations are mentioned below this is meant to denote those that are committed to European/Anglo-American constitutionalism, e.g. Australia, Austria, Canada, Germany, Switzerland and the United States of America. (The Federated States of Micronesia with internal self-government as a U.S. administered trust territory are not treated in this text.)
038 But this does not mean that federalism has no critics. Far from it. Criticism of federalism, like federalism itself, enjoys a long tradition. Only the emphasis has been shifting somewhat since the advent of the welfare state. In earlier centuries it was primarily the weakness of federations that was negatively commented upon (see for instance James Harrington in the seventeenth century, L340/261, and Tocqueville in 1835, L806/553). To-day federations are blamed for being insufficiently efficient. And, above all, they are reproached with their obvious problems in fully accommodating another great principle of government, democracy.
039 There are obvious relationships between democracy and federalism. Those following Proudhon's ideology might even say that democracy is achievable only in federations. «Thus constituted only Federation is able to solve, in theory and in practice, the problem of harmony between Liberty and Authority by giving to each of these principles its just measure, its true attributions and all its initiative. Only Federation, therefore, guarantees, by absolutely respecting both the citizen and the state, order, justice, stability, peace» (L634/546). And from an empirical point of view Elazar is able to say: «Not only do all modern federal systems claim to be democratic and seek democratic legitimacy, but there is likely to be general agreement among true democrats that the great majority of those polities held up as models of democracy (France is the great and important exception), are either federal in form or extensively utilize federal principles» (L205b/47, see also Duchacek L181/6, Hahn L328/272).
040 But it is still more obvious that there is also considerable strain in the relationship between federalism and democracy. Riker states this in an extreme way (L660/142): «To one who believes in the majoritarian notion of freedom», as he obviously does himself, «it is impossible to interpret federalism as other than a device for minority tyranny.» It is not necessary to go as far as Riker does. It is sufficient to point out that federalism by its very nature will always isolate areas where minorities (or localized majorities) will outweigh national majorities. When is that minority tyranny, and when is it a just political order?
041 It is possible to answer the question with rhetoric. This up to now has more or less always been the case in European analyses of the tensions between democracy and federalism. The first step is then to define both concepts so that they are of necessity mutually exclusive: «If democracy is understood in terms of the absolute and unrestrained rule of the majority of a given political community, then a stark and unresolved conflict between federalism and democracy must be acknowledged» (L277/54). «Whether you want it or not, modern democracy, your democracy, is centralizing. Federalism and this democracy are fundamentally incompatible. It is the incompatibility of two opposing philosophies of man and of life» (de Reynold L654/89). The next step is consistent. The extreme is exposed. The solution is the «correct» understanding of both terms. «But if democracy is looked at correctly, that is to say in the light of liberty, and if one, therefore, subordinates the principles of popular sovereignty and of equality to the idea of liberty, then there cannot be but positive relationships between federal organization and democracy» (Usteri L825/347, also L826/11). And Lang states in a similar train of thought (L451/322): «To postulate an antinomy between Democracy and Federalism is accurate only with a view to their degenerate forms.»
042 This is but stating the obvious. Antinomy, the clearly incompatible, is not the problem. The problem is a tension or a grey area between two principles which intuitively on the one hand are designed to achieve similar ends but on the other hand very often clash in their actual application. This tension is not relievable. It would be a doomed endeavour to understand democracy on the base of a theory of federalism and vice versa (for such an attempt see Lang L451). Both federalism and democracy are means to certain ends. Both are principles for the organization of governance. They are organizational principles and not ends in themselves. Both principles may assume different forms, and these different forms are unequally apt to realize individual social aspirations. It is the difference in these aspirations that may make for tensions between specific forms of federalism and democracy. Using the systems framework presented in the Introduction I shall now try to elucidate the functions which democracy and federalism assume in the political process. This will indicate that, like their causes, there are parallels as well as strains and conflicts. Some of the prerequisites for the functionality of the one or the other of these principles will, I hope, become clearer. But as the political system is an optimizing system which cannot fall back on a hierarchy of ends of government (no.022), it will not be possible to make valid generalizations about the «right» relationship between federalism and democracy. Such assertions could only be made in a concrete context, and even then an ideological bias is often inevitable. The aim of the present monograph is to make this evident. At the same time it also attempts to make accessible empirical and theoretical information which is to a large extent widely dispersed.
043 Federalism and democracy are not considered to be legal concepts. Both terms may have legal implications. They may also in a given situation be treated as constitutionally relevant value choices. But they are only legal terms in such cases where positive law confers this status upon them (no.215), thereby referring back to general theories for their eventual interpretation.
C. Democracy
♦ 044 «Democracy is a concept that virtually defies definition» (Lijphart L481/4).
045 If there existed such a thing as an accepted definition of democracy it would be much easier to trace its relation with federalism. But there is no such definition. «Democracy shows an increasing tendency to become synonymous with everything in society that is good, beautiful and true. Each of the other models of order or conflict has its followers in certain political parties and scientific doctrines as well as its adversaries in others. Only democracy is claimed by everybody.... All systems - social situation and level of development notwithstanding - are, according to a Unesco survey..., advocates of democracy» (Beyme L65/12, also L579/26, L702/8).
046 All that may be said about the concept in general terms is that it has an organizational and/or a value-oriented component. In organizational terms democracy is a system of governance that is essentially based on the autonomy (Kelsen L413/205,284) or on the participation of all members of the system. Certain forms of participation are foremost: universal suffrage and majority decision. From a value-oriented perspective, a system is democratic if it pursues specific aims of equality (chances, influence etc.) and of liberation from dependence.
047 Democracy is associated with several thousand years of occidental history. But democratic institutions are not only to be found in western history. There are also Japanese (L35/55, L819/51) and African examples. Yet there is no denying that it is the occidental Greco/Judeo/Christian tradition which moulded the traits of modern worldwide democracy.
048 For ancient forms of democracy immediacy was essential, that is the universal competence of the assembly of politically equal citizens (L382/719s). The fact that it was only the free who enjoyed political rights and not also the slaves (L494/479) did not worry the Greeks (with the exception of the human rights doctrine of the Stoa, L125/100). But what was important to them was rotation in office (L34/1)12). Whenever possible vacancies were filled through the casting of lots or through sequential service. Elections were, because of their «aristocratic» element, not favoured (L382/719).
049 Democracy was not favoured by Aristotle, who was probably the most influential of the political philo-sophers of antiquity. He considered democracy to be a degenerate form of government favouring only the have-nots. What we call democracy nowadays was to him «polity»: the minding of public business by the majority in the general interest (L34/3/7). But Aristotle, like other Greek thinkers (e.g. Heraklit L354/33,47) was no unconditional advocate of polity. He feared demagogy and therefore held polity to be best suited for an agrarian population (L34/6/4): «Because of all populations the agrarian one is the best.... As on the one hand people do not own much they have little spare time, and thus assemblies do not often congegrate. And as, on the other hand, they still own all they need... they do their work and do not desire other people's property. They prefer to pursue their toil to dealing with public business or holding office...»
050 Such direct democracy was and is only possible in small communities (Bryce L99/489, Madison L334/10, Montesquieu L551/8/20, Rousseau L677/577). As the city states grew and as they were unable to create representative democracy, classical antiquity brought forth autocratic territorial government. This was «legally» achieved by combining different offices in one man as in Rome (L544/340). It is true that the early Christian thinkers, influenced as they were by the teachings of Aristotle and the Stoa, reintroduced the idea of democracy. But as their democracy was oriented to the hereafter and not to this life (Cassirer L125/106s, Marcic L510/193) they could not prevent monarchy, aristocracy and autocratic regimes in general from becoming the norm.
051 It was probably the spreading of the reformation doctrine of religious resistance to the state (Loewenstein L494/278), and with it the idea of the social compact backtracing political governance to a contractual basis, which opened the way for the transfer of sovereignty from the autocrat to the people (no.281ss) and for the democratic theories of the Enlightenment, the English, American and French revolutions. Regimes are legitimate because they rest on the will of the citizens. «That all power is vested in and consequently derived from the people» (Virginia bill of rights sect.2, L250/6). «The oil of anointing seems unawares to have dripped from the head of the one to the heads of the many, and given sacredness to them also and to their decrees» (Spencer L762/151).
052 The classical philosophers' theory of the social compact solves the problem of how to justify the political duty to obey by creating a fiction (Cassirer L125/173, Sismondi L755/87) of an original contract amongst absolutely free individuals to subordinate themselves to a sovereign in order to secure life and possessions. For Hobbes this sovereign or depository of supreme power in the commonwealth was normally a monarch (L363/18, Oakeshot L584/40ss). For Rousseau the community as such was sovereign (L677/569, L678/1/6): «It is not possible to constitute government in any other legitimate way» (L678/3/17).
053 Rousseau gives the following text for his social compact (L678/1/6, similar L77/569): «Each of us gives his person and all his powers in common under the supreme direction of the general will; and we shall receive as one body every member as an indivisible part of the whole.» (For Hobbes see L363/17, L401/25.) Further information on the religious foundations of the social compact and the latter's significance for the legitimization of federal government is to be found in no.598ss.
054 Rousseau was without doubt the most influential theoretician of modern democracy. Apart from the social compact his distinction between the «general will» (volonté générale) and the «will of all» (volonté de tous) is important. It is still nowadays at the bottom of many controversies about democracy. The «will of all» is arrived at by adding the determinations of everybody. It is a procedure. This procedure should produce the «general will». This general will is a quality because it always corresponds to the general welfare: «will to attain common survival and general welfare ... public welfare is always evident and it needs but common sense to perceive it» (L678/4/1). Thus there is always a «general will», and Rousseau transcends Aristotle. But the problem is obviously the conformity of the will of the many with the general will. «There is often quite a difference between the will of the many and the general will. The latter's aim is always the common interest. The former is only related to particular interests. It is but the sum of such interests. But if from these you take away the extreme ones which mutually destroy each other the sum of the differences will be the general will» (L678/2/3). With this «statistical» method which, incidentally, is analogous to the one used by classical economists, Rousseau tries to demonstrate the agreement of both kinds of public will. But he does see limits: «If the people sufficiently informed do not communicate with each other (the conditions are evidently contradictory) there shall always result the general will from the great number of small differences... But if there are machinations, if there are partial associations to the prejudice of the community, if the will of each of these becomes general in relation to its members and particular in relation to the state: then one may say that there are no longer as many voters as there are people but only as many as there are associations. The differences become less numerous and give a less general result» (L678/2/3). We are therefore confronted with the situation that there is always a «general will». But the «will of the many» will not always perceive it (in legal theory the analogy would be the difference between true, or natural, and positive law: L414/298). Thus the democrat Rousseau paves the way for those who claim to be able to detect the «general will» which is always superior to the «will of the many», even when it is different from the latter. History's first illustration of this was the terror of the French Revolution. For Kägi it is «one of the most farreaching facts in the history of Democracy and of the Rule of Law that modern democracy's most influential theoretician is not part of the rule of law tradition but part of the absolutist school of occidental political thinking» (L405/110).
055 With the English Revolution of 1688 and the French Revolution of 1789 democracy triumphed over absolutism. The American Revolution of 1776 was a landmark for the autonomy of individuals and of peoples, another aspect of democracy. The nineteenth century's role was the enlargement, the development and the consolidation of democratic practices. Some of the important stages were Jackson's presidency in the United States (populism instead of elitism), the reform laws of 1832 in Great Britain and the French Revolution of 1848 (extension of the franchise) and America's civil war 1861-65 (priority of national over local democracy).
056 The twentieth century brought about the general emergence of the democratic principle already mentioned (no.045) in all the countries of the world, at least by way of lip service. It is doubtful whether Rousseau was aware of such a possibility (L678/3/4): «If there were a people of gods it would govern itself democratically. Government as perfect as this does not correspond to man.»
C.1 Forms of democracy
♦ 057 «The democratic idea is too grand to be trivialized by restricting itself to only one form of authority» (Dahl L156/67).
058 All the states of the world claim to have a democratic regime. It is not surprising, therefore, that democracy, depending on different theories, may have many faces, too. To facilitate the later exposition of functions it may be advisable to discuss briefly the main forms of democracy.
059 I make a distinction between democracy as a type of regime, democracy as a mechanism of democratization and democracy as autocracy in disguise. This is not meant to be a typology (such as is proposed by most theoreticians of democracy, e.g. Dahl L156/76s, Lehmbruch L465/97, Lijphart L481/105s). My distinction has but a practical purpose, namely the illustration of different centres of gravity: classical structures, modern postulates for democratization, hidden autocracy.
C.11 Democracy as a Regime
060 In the German-speaking world at least there exists a language convention distinguishing between the form of the state («Staatsform») and the form of government («Regierungsform»). The form of the state stands for the organization of the supreme political authority or formal organization of the body politic. The form of government characterizes actual governance arrangements. Democracy is used in both sets of terms which makes it somewhat confusing. Since the time of Aristotle (L34/3/7), monarchy, aristocracy and democracy have represented state forms. But democracy is also, and foremost, a form of government (oligarchy and autocracy are other such forms). It is in this sense that I shall henceforth use the term, referring to a republic when designating the analogous state form.
061 A typology for the democracy group of forms of government has often been proposed. But these proposals are usually not very convincing and none of them has been able to prevail in scientific parlance. «Democracy, like the other forms of government, is... a matter of degree, and the possible variations of degree are almost infinite» (Field L240/92s).
062 Loewenstein (L493/69s) gives a survey of the main types and subtypes of constitutional democracy corresponding more or less with legal terminology: «1. If the people organized as electorate is the primary depositary of power we talk about 'direct democracy'. 2. 'Assembly government' is the name of that type where parliament representing the population at large is the paramount power. 3. 'Parliamentary government' designates a state of balance between an autonomous parliament and an autonomous executive in such a way as to integrate the latter in the former: the members of the executive... are at the same time also members of the assembly.... 4. If parliament and the executive remain independently autonomous under a constitutional mandate to cooperate this may be achieved by coordination rather than by integration. Because of the usual predominance of the leadership function in the executive this type of government is called 'presidentialism'... 5. Finally, constitutionalists are wont to talk about the Swiss system of government as a type in its own right under the name of 'directorate government' because of the collegiate structure of the executive.»
063 Current usage can only be claimed for the terminological pair of direct and indirect or representative (Dahl L156a/27: polyarchy) democracy. Under direct democracy the electorate directly exercises so-called state sovereignty. Under indirect democracy this is done through representatives. The distinction is a clear one. But the clarity is only apparent.
064 As Rousseau pointed out (L678/3/4): «Taken in the accepted sense of the word true democracy has never existed and will never exist. It is against nature that a multitude governs and that the few are governed. It is inconceivable that the people stay continuously assembled to mind the public affairs. It is easy to see that for this purpose they have no choice but to institute committees without changing thereby the form of administration.» Direct democracy in its absolute form does not even exist in local communities or in the five Swiss cantons always mentioned because they still practise the «Landsgemeinde». But depending on the political system we do find more or less markedly direct democratic institutions. If we talk about direct democracy in the common usage of the term we say that the main political decisions are taken by the population at large, either in an assembly or through the ballot-box (L493/73). Which are the main political decisions? These are largely a matter of convention. But they must involve considerably more than elections and the adoption of a constitution. Usually laws will also have to be submitted to this procedure.
065 Duverger's «new» distinction between direct and mediated democracy, where direct democracy stands for the popular election of the head of government, may be understandable in the context of French constitutional policy. But it is of little relevance.
066 Who is part of the «demos» that makes decisions in direct democracies? That is the age-old question. Never has it been the whole population. Franchise qualifications such as sex (mainly in former times), age, citizenship, domicile, wealth, and literacy have meant that even in the most «direct» of democracies only a part of the population constitutes the electorate. Schindler therefore asks (L713/28): «is it, then, really possible to talk about identity when, in truth, only the so-called active citizens commit the whole population?» He gives his own answer: «As a matter of fact there is no identity of persons. But this is without significance. There is identity of interests, aims and views: these are the same that hold in the electorate as well as in the population at large. The disenfranchised population as a whole has no political will different from that of the electorate.» Considering that this was written in Switzerland and at a time when women's vote was still unknown there, Schindler's assertion appears to be a rather bold one. At the most it might be said that the factors influencing decision making, e.g. the mass media, are the same for the whole population. Still, this would hardly hold in the case of youth and of immigrant foreigners. Who should be part of the electorate is, in the final analysis, connected with the question about what will be accepted as legitimate governance in a given political system. For reasons which democratic procedure itself determines, the electorate will never comprise the whole population even in the most direct of democracies.
067 Direct democracy can be practised in a meaningful way only in restricted space. «As in a free republic every free man should be governed by himself it would be necessary that the people as a whole should exercise the powers of legislation. But this is impossible in big states, and it is subject to many inconveniences in small ones. It is necessary that the people have representatives do all they cannot do themselves» (Montesquieu L551/1/11). - «Democracy in the classical sense at the base of European parliamentary systems since the French Revolution is a system of delegating the powers of the people in such a way that the delegates are continuously accountable before those their delegation stems from» (Lenk L475/32). - «Democracy, as we understand it, means that the people must be free, the free choose the rulers, and the rulers govern according to the wishes of the people» (Jennings L383/204).
068 Direct democracy is replaced by indirect or representative democracy. But is this still democracy? It is evident that, looked at from the outside, elections are all there are (as for instance in Schumpeter's definition L475/35). This from the viewpoint of Imboden creates tension which in the final analysis is fatal: «All power emanates from the people.... This dogma sounds so very evident and it is usually accepted without examination. But it is strangely contradictory. It says that the people rule and yet don't rule;... The people don't govern directly. They only govern indirectly and in theory: in between the ruled and the presumed rulers there are the authorities which as representatives are the real actors» (L387/16ss). «Is it conceivable that the state always remains subject to this tension? May it always be something else than it purports to be? Isn't there a natural necessity that the face of an institution corresponds to the political ideas behind it?» (L387/25). This is a real question. But the doubt expressed will probably only be justified in situations where there is too much of a gulf between democratic rhetoric and oligarchic reality. In all other cases there are at least regular elections to remind the representatives -as well as those they represent- of their ties with the will of the people (Bryce L99/603). The legitimizing function (no.151ss) will remain unimpaired as long as the great majority of the electorate see themselves adequately represented by their representatives, or if this system at least corresponds to their idea about the correct political process.
069 It is possible to augment the control of the representatives and thereby their ties with the demands of those they represent through plebiscitary elements (Neidhart L782/609) as is the case in Switzerland for instance. But the cost to be paid is a sometimes significant retardation of decision making.
070 Every kind of organizational differentiation of government which may be traced back, sometimes over several stages, to popular election is indirect democracy. But it only makes sense to talk about democracy where this connection between office holders and popular election is still evident, albeit indirectly. A distinctive mark of all representative democracies is a popularly elected parliament which amongst other functions of policy making has at least control over the purse.
071 A special form of representative democracy is the delegation system. It is an attempt to keep the distance between the representatives and those they represent as close as possible. For details see no.078ss.
072 The role of elites is one of the basic problems for direct as well as indirect democracy. «Every kind of politics, and not in the last instance the most liberal and the most democratic politics, are shaped and supported by elites» (Kopp L442/257). «Thus Free Government cannot be, and has in reality always been an Oligarchy within a Democracy» (Bryce L99/603, see also Thoma L801/67, to the contrary Arendt L32/276). According to the American political economist Ostrom the existence of an elite is unavoidable for any political system, be it democratic or not (L592/108s): «First, individual human beings are assumed to be the basic material or units which form any political system.... Each person, in the final analysis, is assumed to be the judge of his own self-interest and can be expected to act in a way that will enhance his net welfare potential. Second, decision rules are the basis for ordering relationships in any association. Decision rules are propositions assigning decision-making capabilities to those who participate in social relationships... Third, decision rules are neither self-generating nor self-modifying, but depend upon individual persons both for their formulation and alternation. ... Fourth, decision rules are not self-enforcing but depend upon the assignment of extraordinary powers to some persons to enforce decision rules in relation to other members of the community.... Fifth, given the conditions specified in the above assumptions, then we must necessarily conclude that any form of organization capable of establishing and enforcing ordered social relationships among a large community of persons will necessarily depend upon a radical inequality in the assignment of decision-making capabilities to those who exercise the prerogatives for allocating and controlling the decision-making capabilities exercised by others. Conditions of political inequality must necessarily exist in any political association.» We might add that the postulate of popular rule in the sense of an equal influence of everybody is also unrealistic inasmuch as this presupposes such an interest of everybody. This is manifestly not the case just as not everybody has an interest in music, in sports or in stamp-collecting. To discuss public affairs, to keep informed, to use the appropriate instruments or to run for office, all these are quite exacting in time and intellect. In a functionally differentiated society the pursuit of political interests cannot escape being a specific function of a specific group.
073 Scharpf (L702/58) perceives «two factors which might explain a low level of political participation: on the one hand discrepancy between the number and complexity of political decisions as against the capacity of each citizen to absorb and process information, and on the other hand competition of politics with additional relevant interests of the individual. Even if they might be considered trivial, especially in the face of the pretensions of participatory theories both points have to emphasized.» «Looked at from the point of view of individual self-determination political participation does not appear as a particularly promising chance for the active shaping of concrete conditions of life. The least that can be said is that participation is usually competing with a multitude of 'private' endeavours which might be much more effective. In an empirical perspective the political commitment of the few is much more in need of an explanation than the relative disinterest of the many» (L702/61ss, also L31/222).
074 Democracy's problem with elites is not the existence of an elite as such. The problem is rather guaranteeing open access to elites (Ostrom L593/231) so that the latter have a democratic legitimacy, too, not being based just on the accidents of birth, school or wealth. This, amongst other things, is the aim of many of democracy's institutions: fixed terms in office, the division of powers, party primaries, federal non-centralization. The other problem is the safeguarding of these democratically legitimized elites' access to all important decision-making arenas so that decisions are not arrived at outside democratic structures. Here again, specific institutions support this goal in democracies: the press, omnicompetence of the constitution, parliamentary questions, publicity and transparency of procedures etc. That both postulates remain to a considerable degree postulates is not an indictment of democracy. It is rather the result of the fact that in any pluralist society the system of government itself is a result of the optimizing (no.022) of conflicting interests.
C.12 Democracy and Democratization
075 The elitist character of democratic institutions has not escaped criticism. The radically democratic left, for instance, (L702/25) reintroduces a classical idea (no.048) by taking citizen participation as a value in itself and giving it priority over other ends such as efficiency (see also Barry L49/173s.).
076 Eschenburg (L217/150) describes this approach: «Is it too bold to conceive the introduction of the parliamentary system on the basis of the general and equal franchise as the completion of the first phase of democratization? Its start was very soon overlapped by a second phase of democratization, i.e. the introduction of social safeguards for working people, the numerically biggest but economically weakest class.... Since the mid-sixties there appears to be a third phase of democratization: democratization of democracy.... Democratization means to bring more democracy about. Its thrust is against oligarchic and monocratic elements. Its demands are: more power to parliaments, public access to committee meetings, the introduction of plebiscites about personnel and other matters, participation in administrative affairs, extension of the transparency of political procedures. Democratization means power-sharing and, at the same time, reduction of domination wherever there is domination. In addition to that the democratization of instruments to attain political decisions is postulated and practised.»
077 Leftist criticism of parliamentary democracy comes from two sides (Beyme L65/15): «1) A moderate position asks for the revitalization of forms of direct democracy and for more autonomy for basic groups (with original vote, referenda, rotation in office, publicity and control). 2) The radical position demands the introduction of the delegations (soviet) system on the basis of socialist property and with a binding mandate, recall, rotation, publicity, ban on factions, criticism and self-criticism as well as control.»
078 The delegations or soviet system is the most thorough-going of these models of democratization. Kevenhörster describes it in the following way (L417/96): «The voters organize themselves in basic groups in enterprises, living communities or administrative units. Most proponents favour enterprises. The basic groups articulate political positions. They also take the political decisions. All public positions are elective. A global competence for all questions is ascribed to members. It is desirable that society is totally politicized. All officeholders are accountable to their electors even for matters of detail. Their mandate is tied to specific instructions. They are submitted to continuous scrutiny and they may be recalled at any time. The delegates work as such without pay. Their average income must not be higher than that of the electors to preclude the emergence of an economically autonomous ruling group. All workers are in turn participating in the administration of the state. The possibilities for re-election are drastically reduced. Rotation in office is designed to make corruption and the accumulation of influence impossible. The traditional civil service no longer exists. The powers are not separated but fused because in a society with a politically and socially organized electorate institutional division of powers has no function.» «The aims of the delegations model are exacting: - permanent and unconditional accountability of the elected, - permanent participation of electors in personnel and other matters, - debureaucratization of decision making procedures, - disappearance of parties and factions, - abolition of the electorate's alienation from the state» (L417/95).
079 It would be wrong to dismiss the delegations system as something artificial. It is impressive to see the regularity with which, after revolutions, soviets have been constituted, very often against the wishes of the revolutionary leaders (see Arendt L32/256ss, L33/216s). But this model's typical problem is the near-impossibility of organizing it in larger polities which need several tiers of decision making. There are attempts to accommodate these needs with differentiated delegations and indirect elections, delegations electing their delegations on the next higher level (L32/278, L702/64). But this necessarily restrains basic group influence on central decision making (L417/97, L702/64), thereby weakening one of the main arguments for the system's introduction. «Compared with direct elections of representatives to parliaments the delegations model would in fact bring about a considerable weakening of the ruled's democratic influence and control vis-à-vis the rulers» (Scharpf L702/64).
080 The only actual example of the delegations model is Yugoslavia (C132ss, Stefanovic L772/15ss). But the institutions have been modified better to achieve a certain stability and -this is most important- they have been combined with the Marxist model of democratic centralism (no.100ss).
081 The ideology of democratization is based - even more so than classical direct democracy ideology - on the assumption that citizens may be held in a state of continuous political activity. That is against all experience (no.072s). «It is therefore necessary for participation theory to interpret the empirical diagnosis of political apathy obtaining with the large majority of the population as... surmountable. It is difficult to see how this premise might be justified» (Scharpf L702/57, also Ellwein L208/124). «Unreflected participatory romanticism courts the danger to end as mass pseudoparticipation, manipulated by totalitarian groups, if its excessive aspirations fail to be realized. All that has a chance to survive in such circumstances is the reduced form of democratic centralism» (Beyme L65/29). «Democratic theory which does not totally lose contact with reality because of normative participatory postulates - as such justified - will have to accept that even enlightened citizens remain consumers of services and passive onlookers in areas where they could and should participate» (Beyme L65/29, also L65/15).
082 A further difficulty with the delegations model is its relation to conflict resolution. Grosser (L229/69): «It is inevitable that in a society marked by heterogeneity of interests more participation and basic group autonomy will intensify conflict. Yet the weakening of the representation principle through elements of direct democracy -especially the binding mandate and office rotation- will lessen the system's ability to solve cflicts via the balancing of interests.»
083 Finally, it should not be forgotten that the modern state is so complex that it cannot do without a relatively complicated and large administrative machinery. To apply the delegations model to the administration might have anti-democratic results. Löwenthal points out (L217/156): «that wherever we have to do with functional organs of society whose goals, whose aims, whose ends are prescribed by society as a whole we cannot in a meaningful way democratize them in themselves. A ministry the civil servants of which democratically decide is not a ministry of a democratic government.»
084 For all these reasons Beyme formulates four limits to democratization (L65/25ss): «1) Participatory models have to take the specific requirements of individual subsystems into account. It may well be that the basic group or delegations model could be quite adequate... for not too complex multifunctional organizations without being equally well suited to the organization of a large combine.... 2) The potential for democratization of organizations may vary considerably with their dependence on the accountability mechanisms of other organizations... 3) Limits to democratization may result from the plurality of status hierarchies in society. To demand total participation of all citizens on every level is to apply a one-dimensional model of stratification which makes a distinction only between owners of means of production and workers. Modern theories of stratification are multi-dimensional.... 4) Finally, limits to democratization may also result from the plurality of organizational allegiances.»
085 It is interesting to note that American ideologies of democratization have a different point of departure from European ones. In the United States we have at the base of the demands the realization that up to now it has not been possible to sensibly improve the situation of underprivileged groups. There is a theory that social welfare programmes have been to a large degree unsuccessful because they are run by the privileged. Thence, it would appear necessary to give the underprivileged the decisive influence for deciding of the use of the money earmarked for them. «Only by generating political power, therefore, could the poor alter the structure of community influence and thus open up opportunities currently denied them - for example by breaking the dependence of the social service agencies on the local middle class» (Salamon L693/263). Since the sixties there have been provisions in some of the federal social laws that the programmes have to be administered with «maximum feasible participation» of the local poor. Democratization took the form of «community control». This means that central government programmes were administered through a centrally designed local political ad hoc forum which ignored the traditional local structure in favour of the underprivileged (blacks, Puerto Ricans etc., see Eisinger L197/144, Sundquist L792/85). The reviews of this approach are controversial. But there is agreement that «maximum feasible participation» has influenced the traditional power structures of the communities concerned in a much bigger way than had been anticipated by the politicians who decided on this provision (as to this «time-bomb effect» compare Salamon L693/264).
086 Rein describes this kind of democratization with its basic dilemma (L652/222s): «American democracy has to reduce this basic imbalance of elites and non-elites by changing the power disparity between them. It tries to achieve that by helping underprivileged groups to better articulate their demands and preferences; by helping them to organize protests where they may express their moral right to justice and equality; by giving the poor in addition to collective action the support of the legal machinery to put them in a position to sue institutions which denied them their rights... The logical conclusion to this approach is an anomaly: there will necessarily be two kinds of democracy - one for the poor and one for the rest.» In the final analysis this, too, is an elite theory, only this time it is an elite of the underprivileged. Whether such interventions have in the long run a stabilizing or a destabilizing influence on the democratic system is an open question.
087 But this attempt, too, cannot bypass the problem that the groups it is designed to activate do not develop the expected permanent citizen's interest. This seems to be even more the case with the underprivileged than it is elsewhere. The policy courts very quickly the danger of being exploited for other interests. Rein (L652/22s): «It is quite astonishing how bureaucrats try to organize lower income groups to put through their own goals. Redevelopment authorities create block groups, welfare offices in the ghettos organize community work, schools further PTAs (parent/teacher associations). In a similar way town planners try to mobilize streets to take the wind out of the opposition's sails etc.» (for further examples see L792/65).
088 This American experience with democratization is mainly influencing local political structure. It is an important factor changing the quality of American federalism. For a further discussion of this kind of structural intervention see the German companion volume to this book (L264b/866ss).
089 Scharpf considers this to be the real possibility of the democratization movements (L702/66s): «If it is true that political participation of the great majority of the population is hardly ever greater than the participation in general elections then, from a participatory point of view, it is important to add to the importance of elections for the political process. If, beyond participation in elections, the political systems provides possibilities for active participation in other decision-making procedures it, secondly, is important that these possibilities are equally opened up to all willing and able to actively commit themselves. And if there are social areas where the postulate of universal participation may be approximated such prospects, thirdly, should be made use of to the highest possible degree.»
C.121 Opinion Polls
♦ 090 «To legally organize public opinion or the 'will of the people' - that is the goal modern democracy's constitutions seek to attain with their great array of instruments» (Preuss L632/244).
091 Democracy's methods of measuring public opinion (elections, ballots, popular initiative etc.) record real public opinion but imperfectly. That is evident. The mechanisms are clumsy and not frequently applied. They do not take into account those who do not have the vote or who do not vote (the latter very often constituting a majority). The questions put to the voters are very simple and do not allow differentiated answers. To-day's developments in mass communications and data processing technology render polls possible which are much more sophisticated than the ones devised in the nineteenth century, when political rights were introduced. It cannot come as a surprise, then, that there are many who propose to put these new techniques to the use of democracy, to further democratize democracy.
092 Opinion surveys, where statistically representative samples of the population are polled, are well known in the political realm. They are privately commissioned in most democracies before important elections and ballots (in some countries, for example in France, the final dates for the publication of their results are regulated by the state). They are also used by politicians as a barometer for the public mood (for a critical voice about this new industry of 'Big Opinion' see Boorstin L79/18). «In the opinion poll we have an instrument that extends the legitimizing function of vote counting to a larger segment of the political process» (Edelman L191/119, compare also L743/55).
093 But the procedure in use to-day is questionable for political purposes: There is no control over the framing of the question. Yet it is a fact that slight modifications of questions, the formulation of set answers as well as the actual polling procedures may be of great importance to the outcome. Apart from statistical imprecision it is probably these elements which account for the deviant results found very often with competing polls. Furthermore, the population's low level of information may yield results which are anything but clear.
094 This has been demonstrated by, among others, Pimlott with his example of the U.S. «Metallic Metals Act» (L618/256). This nonsensical and non-existent bill elicited opinions from 70 per cent of those who were asked. Of the respondent,. 21.4 per cent were favourably disposed; 58.6 per cent would have preferred the states to be responsible; 15.7 per cent considered such a measure to be necessary abroad but not for the United States; 4.3 per cent were against it outright. This poll is perhaps significant for the state of public opinion on federalism and the constitutional division of functions. But such an attempt at interpretation just further illustrates the method's limitations.
095 Advocates of opinion polls (demoscopy) for political decision making also misjudge the difference between opinion and demand as analysed by Easton (L188/42s): «it is not at all unusual for members of a political system to feel, for example, that they should have more public housing and yet, in the political context of the moment, they may also hold the opinion that they ought not to press for government action. They may recognize the urgent need for housing, express themselves in an opinion survey as being in favor of having something done about it. But because of the state of the economy or the priorities of other urgent matters, they may not be prepared to take a position one way or another as to whether political action should be taken, that is, to state their position as demand.»
096 The main argument against constitutive use of opinion polls for political decision making is their misconception of the political process. They are apolitical. Imboden (387/87): «If the idea is to measure with the help of modern opinion surveys which opinion prevails in a given collective this procedure may correspond formally... to the majority formula. And yet it is the complete opposite of democratic decision making. In actual democracy the demos... is 'advised'. It considers the advice of public authorities or... political parties. It listens to those persons whose opinion seems to be important. The majority finally counted is the result of an exchange of opinions or of a debate. Democratic decision is not just the arithmetical sum of non-reflected statements. It is an integrationary discussion.... Individual opinion is confronted with the opinions of those representative for society.... The technique of demoscopy, the statistical collection of set and unreflected 'opinions' threatens to choke true democracy.»
097 For demoscopy to be a legitimizing instrument for democratic decision making it would probably have to meet four conditions: 1) The procedure would have to be government controlled. 2) The questions would have to be set in a democratically acceptable way. 3) The organization of the survey would have to guarantee a prior public debate. 4) To arrive at democratically legitimized results the polling of all citizens, not just of a sample, might well be a prerequisite.
098 There are three considerations which illustrate the problematic nature of such a procedure. 1) Multiplication of surveys and the concomitant need to digest the relevant information might make citizens politically still more apathetic than they are now. 2) It is not realistic to expect political parties and interest groups to finance any number of information campaigns. Government financed campaigns, on the other hand, would introduce the problems of censure as well as the blowing up of fringe positions. 3) It is difficult to see why such expense should be entered into only to facilitate the participation of those citizens who are so little interested in public affairs that they stay away from the ballot-box, thereby doing the very thing which corresponds to their level of information.
099 The sum of the matter is that the instruments of direct democracy are only superficially the means of surveying public opinion. Their foremost function is the legitimation of decisions through a democratic procedure, i.e. one that all citizens have access to (compare no.155ss). This is not to deny the possibility that the traditional ways for the expression of political opinion will be changed with the introduction of modern data transmission techniques. Under such circumstances one of the results might be the facilitation and increase of ballots. But private ballots also would certainly greatly increase and compete with the public ones. In all probability this would destabilize the legitimizing function of the latter and to some extent even destabilize democracy.
C.13 Democracy as Autocracy in Disguise
100 We have seen that every democratic regime contains elitist elements. But these elements are not part of the democratic ideology (with the possible exception of representation). Yet there are ideologies which try to present an organic linkage between democracy and autocracy. The best known example of these «guided democracies» is democratic centralism.
101 To better understand democratic centralism it is necessary to remember that, according to Marxist doctrine, democracy «will have to disappear with the withering away of the state» (Lenin L473/5). But until the full introduction of communism, dictatorship of the proletariat is a necessity. Democratic centralism tries to combine dictatorial rule over the class-enemy with democracy amongst class-comrades. «Only a government of democratic centralism can fully express the will of all the revolutionary people and most effectively fight the enemies of the revolution.... The state system - joint dictatorship of all revolutionary classes. The political structure - democratic centralism» (Mao L506/25, see also article 3/1 of the 1982 Constitution of the People's Republic of China, article 3 of the Soviet and article 18/1 of the Czechoslovakian Constitutions).
102 Grosser describes democratic centralism as follows (L229/71): «The electorate is limited to the confirmation of candidates selected by the party leadership and to non-binding discussion of alternatives. The decisions are the prerogative of central office. Minorities have no right to organize themselves in order to achieve a reversal of a decision (prohibition of factions). It is possible to reconcile democratic centralism with the principle of legitimation by popular rule only in so far as one accepts with Rousseau a true 'general will' which the leaders will objectively and scientifically discern and to which every well meaning and sufficiently informed person would agree» (for Rousseau's volonteé générale see no.054). In theory it might be feasible to organize the preparation of decisions democratically and to have the decisions themselves autocratically arrived at. If there always is objective and perceivable truth then the only function of the democratic process would be the furnishing of data for the decision and not the latter itself. Yet the history of communist governments with their leaders denouncing their predecessors' mistakes is too obvious an illustration of the weakness of such a theory. Democratic centralism is centralism first and foremost. Its learning capacity is therefore quite restricted in a cybernetic sense (no.114ss).
103 Authoritarian democracy is not confined to Marxist states. It is or was also to be found in South America (L496/51) in Franco Spain, in Salazar's Portugal etc. But the western ideal of strong leadership also indicates authoritarian elements in conceptions of democracy. As to illustrations we may think of Gaullist France and of the United States under Kennedy, Johnson and Nixon.
104 A similar trend is «depoliticization» of political decisions by handing them over to «experts instead of politicians». But because there is no such thing as an impartial decision as such (no.122) and even less so when we have to deal with value judgements, the conclusions of experts are also political conclusions. Technocracy is an authoritarian style of government.
105 Is democracy suited to developing countries? There, we are faced with a dilemma. On the one hand the prerequisites for democracy of a more than narrowly local nature are severely lacking. On the other hand democratic culture cannot come about without democratic institutions (Field L240/127). Under such circumstances some kind of guided democracy seems to be required for a transition system. This would mean an authoritarian regime practising democracy in limited areas with the idea of expanding these areas according to the developments in political expertise of the population. After all this was the way western democracy gradually grew (the stages there were demographical rather than functional). Such a regime is sometimes called tutelary democracy. It is a fine theory. It also has the advantage of offering one of the few possibilities of limiting inflation of demand in underdeveloped countries (see Easton L188/115). But, alas, there is often a complete lack of the ruling class's will to democratize. «Instead, political reality shows that the concepts of 'tutelary democracy', 'guided democracy' (e.g. Sukarno in Indonesia), 'basic democracies' (in Pakistan) etc. have primarily served political leaders to hide their authoritarian practices behind a 'democratic' cloak» (Berg-Schlosser L61/154). Only the future will show whether this has been a transitional, perhaps inevitable, stage.
C.2 Functional Aspects of Democracy
♦ 106 «Democracy is a form of government that is never completely achieved» (MacIver L499/175).
107 If we take democracy to be a postulate or a principle of political organization we are faced with a question. What are its ends? The formal approach yields little in the way of an answer: «In conclusion, it can be said that majority rule is indeed the quintessence of democracy, if it is not taken to mean 'majority of one'» (Friedrich L277/49). Yet to point to majority decision is but a very terse description. It is no explanation. Still, it has always been a goal of democratic theory to justify democracy. Many diverse deductions have been tried: truth, justice, rule of law, popular sovereignty, individual personality development. A consensus about the «correct» set of values has, understandably, not been arrived at. There is, to follow Ellwein (L208/136)), only some sort of minimal consensus in particular areas: «The basic characteristic of modern constitutional democracy is the phenomenon of a largely shared anthropological point of departure without any concomitant agreement on final goals. The common basis is founded in the democrats' conviction that man is destined to freely rule himself. There is no complete agreement as to the direction of man's self-rule and as to whence this freedom is bestowed on him.»
108 Confronted with this problem modern political science has a tendency to cease making deductions from general principles and to attempt a description taking mechanisms as a reference. «But modern democracy research pays a heavy price for such progress: Democracy is but a specific political method; its institutions are formally nothing more than a system of balance; and in the end it is sufficient to adequately perceive the conditions for balance to correctly steer the apparatus. This conception of social engineering takes democracy to be a model that, given the necessary adaptations, may be completely isolated from its social background and transferred to any given situation» (Habermas L323/317s).
109 A review of theories of democracy shows that most of them are one-dimensional or mini-dimensional, whatever their scientific nature. They set out to explain the nature of democracy with the help of one or just a few parameters. By doing this they expose themselves to the reproach of being insufficiently general or complex. Such blame is significant for the state of modern theoretical discussion (Beyme L65/23). Sometimes, therefore, the state of the art is said to be primitive (Barry L49/193).
110 Fortunately, it is not necessary to adopt one of these theories or to propose a new one to be able to present both harmony and tension between democracy and federalism in federations. According to the systems thinking approach presented in the introduction, we may look at democracy - and likewise at federalism (no.198) - as a principle to organize governance within a political system. Various justifications for democracy may then be analysed for their systemic functions. To do that we shall use three main functions or bundles of functions: democracy's contribution to the system's learning capacity (no.111ss); its contribution to the system's stabilization (no.134ss); and the connection between democracy and assigned functions (no.183ss). This same approach will be applied later to the functions of federalism (no.404ss). As the main emphasis of this monograph is laid on federalism, the latter part will be more extensive. It will also primarily be there where we consider the interrelationships of democratic and federalist ends.
C.21 Democracy and Learning in Political Systems
111 It has been shown in the Introduction that the study of systems is of special interest to cybernetics. Cybernetics is the systematic exploration of communication and control in all kinds of organization (Deutsch L166/76, also Wiener L860/15s). The emphasis is on the marshalling and processing of information by a system. Looked at from this point of view the concept of learning capacity is of paramount importance.
112 «Learning is taken to be a process of communication. It is dependent on the structure of the flow of information, the information channels' capacity and the effectiveness of the mechanisms for steering and control. It is a decision center's learning capacity that determines its viability and its pathology. Social systems can only survive if they are capable of social learning. With other words: if it is able to effect necessary changes in its structure of preferences in order to guarantee successful operation in the future. ... Learning becomes pathological for instance through the coincidence of will and power; will being the desire not to learn and power the faculty of not having to learn. Thus, important decision makers will quite often learn but in a pathological way, i.e. through painful experience. This we should want to avoid» (Schmidt L300/213s, see also Deutsch L166/92).
113 For the purposes of the following description I shall differentiate as components of learning its maintenance and its problem solving capacity. This approach will be further refined later for the analysing of the relevant functions of federalism (no.402ss.).
C.211 Maintenance of Learning Capacity
114 It is already evident that, from a statistical viewpoint, the participation of the many in decision making introduces great variety into social learning. Institutions providing for citizen access to decision-making force the ruling elite to be confronted with undesirable information even in democracies with authoritarian traits. This will not be the case in perfect autocracies or oligarchies.
115 Democratic procedures create a quasi-continuous flow of new findings, arguments or other data for decision making. Such procedures are ballots, elections, political freedoms (as for instance the right to the free expression of opinions, freedom of the press, petitions, the right to associate freely with one another, freedom of instruction), public hearings, etc.
116 On the other hand, information overload may well clog a system's capacities for processing information. This may happen through excessively long discussion in preparation for decisions (for this reason Steiner, L774/105s, to my mind wrongly, criticizes Switzerland for her direct democratic ballot process). But overload may also result from an inflationary accumulation of demand by the mechanisms of competitive democracy (Scharpf L705/14s). Because of the last phenomenon and also because in small countries the number of personal interests to be taken into account is more restricted (Lijphart L481/68), small states are more suited than larger ones to the extension of institutions of direct democracy. (The problem of information overload will be treated once again in connection with functions of federalism, no.418ss.)
117 The means with which the foregoing democratic procedures further learning may all be discussed under the heading of competition. Competition results in variety of arguments. That is why it was and is so important to enlightened liberalism «to structure the composition, the organization and the procedure of representative authorities in such a way that they provided for a maximum of competition of ideas while at the same time bringing about a close nexus between parliamentary deliberation and open discussion by an enlightened public» (Scharpf L702/22s).
118 Thus, competition is essential for democracy. The specific difference between other systems of government and democracy «is the latter's potential to question governance» (Lenk L475/43). «The concept, therefore, stands in the first place for the critical reflection of all rule and of all power to decide other people's possibilities, actions and thoughts» (Ellwein L458/56). In most countries the mechanism for such critical reflection is party competition in elections (in small democracies, as for instance in Switzerland, direct competition of arguments in ballots is also possible). Elections which offer the chance of a change in the composition of the ruling majority are one of the answers to the age-old problem (e.g. Constant L147/133 and Mill L547/68) that a ruling majority may be just as tyrannical and unwilling to learn as an autocrat.
119 Elections, therefore, have an important role to play in many definitions of democracy, as for instance in the one of Lipset (in L586/263): «Democracy in a complex society may be defined as a political system providing for regular constitutional opportunities for a change of the rulers, and as a social mechanism enabling a maximum part of the population to influence important political decisions via their choice amongst several candidates for political office.»
120 To justify democracy by reference to the competition of opinions reminds us of the division of powers doctrine. Certainly the latter is primarily designed to guarantee individual liberty (see Frenkel L253/103, Kägi L401/248, Kopp L441/475). But in a cybernetic sense division of powers can also be an instrument for keeping information channels open and thereby for maintaining the learning capacity of a political system (see also Dahrendorf's definition of democracy, L157/116). This is a function which, if attributed to the separation of parliament and executive, has to-day become somewhat problematic in countries where there is for all practical purposes political identity of the parliamentary majority and the executive (see also no.633).
C.212 Problem Solving Capacity
♦ 121 «It is not without reason that the voice of the people is likened to the voice of God» (Machiavelli according to L737/703). «Whereas the people taken apart, are but so many private interests, but if you take them together, they are the publick interest; the publick interest of a Common-wealth... is nearest that of mankind, and that of mankind is right reason» (Harrington L340/141s).
122 The goal of political learning is to make a «correct» decision. But a correct decision per se does not exist or, if it exists, it is not perceivable as such for everybody. There is only that which is reputed to be correct. The basis for this reputation may be varied. One such basis is democracy. Democracy legitimizes «truth» through the democratic procedures used to determine it.
123 In principle, democratic opinion formation offers equal chances for all arguments to be heard. This consideration, enhanced by the expectation of some sort of adding up of truth, leads to Aristotle's justification of democracy (L34/3/11): «That it is rather the many who should have supreme power than the best which are only few, this is an assertion which may evidently be defended as possessing a certain probability and perhaps even a certain veracity.... Because with the many each one of them owns a modicum of virtue and of insight. And if they assemble the many will be as one man with a multitude of feet, hands and sense-organs. The same will hold true for attitudes and insights, too.» Certainly, experience offers abundant illustrations for the problematic nature of such a theory.
124 When for instance in 1958 Egypt and Syria voted on their merger in a new United Arab Republic there were only 286 out of nearly seven and a half million votes cast against such union (L493/273). But already at the end of 1961 this never fully implemented state disintegrated again into its two component units. It appears that democratic procedure (if it really can be called that) produced but a relative or very provisional truth. And it is obvious that not only truths might be added but errors, too (see Montesquieu L552/86, Tocqueville L806/352).
125 But, as has been pointed out before, it is not really truth we are concerned with. We are examining a decision-making process in an optimizing social system (no.033) which accords the status of importance to those demands which are directly or indirectly carried by a majority. In this perspective it may be said with Downs (see L49/172) that the system of competing political parties offers a practicable method for a society to aggregate their members' preferences for different political measures expected from their government. The rulers are able to maintain their position as long as they also take into account the demands of others. «For elites that are actively engaged in the political process support from non-elites is an important, often decisive factor. Therefore, established elites in such an output-oriented model always fear the entrance of competing new elites which represent heretofore neglected group interests. Given this situation, established elites will try to anticipate the interests of potentially mobilizable elites and to base their own recruitment on a broad spectrum of social strata and groups» (Scharpf L702/39, see Ostrom L593/231).
126 According to Dahl/Tufte the ideal political system meets two criteria: «the criterion of citizen effectiveness (citizens acting responsibly and competently fully control the decisions of the polity); and the criterion of system capacity (the polity has the capacity to respond fully to the collective preferences of its citizens).»
127 As to the first criterion of citizen effectiveness, we may assert that, from the viewpoint of what has been said before, democratic systems are able to process the most varied demands. Democracy's problems in this field have to do with the way opinions and demands are generated. Convictions do not come from nowhere. They are arrived at by personal experience, interests and environmental influences. The main problem, therefore, is manipulation of opinions. «Considering that the majority of politically active electors have already opted for a political party leaving only the undecided as a propaganda target propaganda will have to concentrate its efforts on this comparatively apolitic and wavering group. This group, then, will not be subjected to arguments about the relationships between everyday problems and political decisions. It is cheaper and quicker to try to ensnare it with psychologically loaded advertising» (Fetscher L239/167). «If we consider democracy not just as a political system, but as a set of institutions which do aim to make everything available to everybody, it would not be an overstatement to describe advertising as the characteristic rhetoric of democracy. One of the tendencies of democracy, which Plato and other anti-democrats warned against a long time ago, was the danger that rhetoric would displace or at least overshadow epistemology; that is, the temptation to allow the problem of persuasion to overshadow the problem of knowledge. Democratic societies tend to become more concerned with what people believe than with what is true, to become more concerned with credibility than with truth» (Boorstin L79/28s, also L98/109ss).
128 The problem, perhaps, is not so much that opinions are influenced by rhetoric but that they are focused on something which may have little or nothing to do with the matter at hand. This may still be acceptable for elections because they are primarily votes of confidence, despite all propaganda efforts. But with the kinds of ballots which are common in Switzerland and the United States, reduction to votes of confidence or competition among purely symbolic formulas is more questionable. Two sets of preconditions may be responsible for this situation: First, the image of the competent voter informing himself about the details of a bill is pure fiction. It presupposes a degree of civic training and of political interest not obtaining with more than ten per cent of the population and unlikely to be bettered even through the most well-intentioned programmes, given the competition from other interesting pursuits. And then we have to realize that, since the victory of democracy in the second half of the nineteenth century, politics, as exemplified by legislation, has had a tendency to favour empty formulas. These permit the greatest possible number of participants in the process to see their ideas translated into reality, even if the ideas are mutually contradictory (see also Glaser L295/9). This explains the spreading of notions like «optimal», «considering also», «ordinarily» etc. Constitutions as well as other laws increasingly adopt journalistic traits. This in turn makes advertising rhetoric appear almost relevant.
129 One of the most interesting monographs on symbolic politics has been written by Edelman. To his eyes essential elements of politics are of a purely symbolic nature (L191/172): «The basic thesis is that mass publics respond to currently conspicuous political symbols: not to 'facts', and not to moral codes embedded in the character or soul, but to the gestures and speeches that make up the drama of the state. The mass public does not study and analyze detailed data about secondary boycotts, provisions for stock ownership and control in a proposed space communications corporation, or missile installations in Cuba. It ignores these things until political actions and speeches make them symbolically threatening or reassuring, and it then responds to the cues furnished by the actions and the speeches, not to direct knowledge of the facts. It is therefore political actions that chiefly shape men's political wants and 'knowledge', not the other way around.»
130 This proneness of democratic public opinion to manipulation (it is the same with other regimes but of less consequence there) leads to something which might be called excessive learning. This appears as a nervous vacillation of political demand, the desire to see every whim effected at once. The age of instant coffee is the age of instant solutions everywhere (based, of course, on long-range global master plans).
131 Already Tocqueville foresaw this inflation of demand in democracies (L806/332s): «In democratic societies there is also some sort of agitation without precise aim. They are struck by a kind of permanent fever bringing forth all sorts of innovations, innovations which are nearly always expensive.... In democracies where the sovereign is always needy his favour can only be acquired by improving his welfare. This can hardly ever be achieved without money. What's more: when the people themselves start thinking about their position a multitude of needs arise which were not perceived before and which can only be satisfied with the recourse to the public coffers.»
132 Scharpf's explanation makes sense (L705/16): «Public services are mostly financed not with servicerelated fees but with general taxes. For the individual taxpayer the 'price' of a service used is no longer identifiable. Thus on the one hand cost-awareness as a factor limiting demand no longer works: a public service, for the consumer, is a quasi-free good. There is in the public sector as opposed to the private sector a general tendency to excessive demand bringing about political postulates for an expansion of public services offered. On the other hand there is by no means a corresponding willingness to contribute to the costs. One reason for this is the possibility of redistribution where it is not the beneficiaries that have to bear the costs of the additional supply but other groups. Still, if we consider that the net redistributory effects of public budgets are rather small and tendentially even regressive, then the main cause for this state of affairs seems to be the maximum separation of the determination of the sum total of the budget (which fixes the tax ceiling) from the earmarking of the monies for the various categories of services. Each political group will find in the budget so many unnecessary or even undesirable expenses which might quite easily be struck out to make place for its own pet projects that the correlation between the latter and the tax ceiling doesn't at all appear evident» (see also Pawlowsky L607/35ss).
133 The result is a conflict between both the criteria mentioned above (no.126), citizen effectiveness and system capacity. It may be said that the standard for meeting citizens' demands will itself be subject to the democratic process and that it will thereby contribute to the problem solving capacity of the system even here. Still, the constraints placed by the scarcity of financial and other means on the solution of problems democratically perceived to be in need of a solution may eventually endanger democracy itself. According to Lipset (L49/72), every regime is supported by the two convictions of its legitimacy and of its efficiency. The history of democracies shows that the absence of the second conviction will either lead to the introduction of oligarchical elements or to the system's abandonment in favour of an authoritarian one. For Scharpf, therefore, one of the foremost tasks of modern democratic theory is (L702/24s) «the theoretical reflexion of the welfare and planning state considering both technological and political/democratic rationality.»
C.22 Democracy and Stability of Political Systems
♦ 134 «In whatever way one organizes and balances power in a democratic society it will always be very difficult to believe in that which the masses reject and to profess what they condemn. This favours marvelously the stability of beliefs» (Tocqueville L807/356).
135 This ironical dictum of Tocqueville's refers to a kind of system stabilization which is only marginally related to the systemic functions I shall now be treating (there is perhaps a certain connection with political apathy's stabilizing influence, see no.174s). We will now be concerned with democracy's role in the stabilization of the political system, i.e. to guarantee it a high probability of survival (as to the concepts of political stability or of stable democracy see L49/60ss, L480/71, L481/4). But beware: we are not talking about a specific regime but about the political system as such (compare no.023s).
136 When we discussed the social frame of reference we included under the heading of systems stabilization (no.027) three bundles of functions: systemic maintenance of stability, conflict resolution and system legitimation. In order to avoid unnecessary duplication I shall now only talk about one aspect of conflict resolution and about legitimation. That is because only here do we have to do with basic problems which are not treated or not adequately treated in connection with federalism. On the other hand, everything which has to be said about other stabilization functions may be contrasted with the relevant functions of federalism (no.541ss and 551ss).
C.221 Conflict Resolution
137 A stable political system is one that successfully reduces and processes tensions which are always in evidence (Eckstein L190/256). Conflicts or tensions may have primarily material causes or primarily demographical ones. Insofar as democratic procedures reduce material conflicts, this is mainly achieved by their legitimizing attributes (see especially the comments on the majority principle no.156ss). Therefore, the resolution of material conflicts will not be discussed here.
138 Demographic tensions, on the other hand, are based on the pluralist structure of society (as to pluralism see no.201ss). «In a segmented society clearly the most pressing problem is to find a form of association which will maintain the unity and peace of the state without threatening the existence of the social and cultural segments» (Nicholls L579/62, compare L491/33). Questions which individual groups consider as essential for their very existence cannot be decisively solved by majority decision in the overarching system, because groups which are continuously in a minority position do not accept the system as a legitimate decision mechanism in such matters. The integration of different population groups in a democracy has to be achieved by forms which, whilst being of a democratic nature, still modify democracy to a considerable degree. For this we know several, partially overlapping models, the most important of which I shall now briefly discuss.
139 The assimilation model, which would have the effect of making group differences disappear in a generalized society, has little chance of success in democracies (Lijphart L481/44s). History shows (Vile L835/11) that assimilation has only been successful in a few authoritarian states, as for instance Tudor England, absolutist France and Stalinist Russia. In democracies assimilation may only be brought about in a very slow and gradual way through the success of other models. In the short run it is only possible with very specific groups, as for example in Switzerland with foreigners who want to become citizens and from whom a severe naturalization procedure demands full personal assimilation.
140 The model of cross-cutting cleavages: This idea hails back to Bentley, Truman and Lipset (L189/70s, L481/10). It explains the low level of conflict obtaining in some plural systems as being due to the fact that group allegiances criss-cross. This makes for a continuous shifting of the fronts of conflict, which prevents individual tensions from becoming strong enough to endanger the stability of the whole system. Further consideration will be given to this model when we deal with federalist integration of groups (no.582ss).
141 Based on his analysis of Norwegian democracy, Eckstein developed a community model (L189/194): This involves «systems in which cohesion exists mainly despite division, due to overarching sentiments of solidarity, whatever their source may be. Community, it is essential to note, does not make for cohesion in the same way as consensus. The latter obviates the need for searching out agreements, while the former makes for cohesion rather through norms that facilitate the quest for agreements - for example by putting a high value on agreeing as an end in itself, even where men are manifestly divided (as happens in most societies when events, such as war, induce temporarily strong sentiments of solidarity), or by cultivating a certain political considerateness of others: deference to their expertise or experience, sympathy with their special interests, or reluctance to raise issues that might exacerbate feelings of hostility of any kind». «In short, Norway's great political divisions and its great political cohesion are perfectly consistent if both have a single source rather than being results of contradictory pulls in Norwegian life, if both result from unusually compelling demands for solidarity that operate on all levels of social life» (L189/176).
142 Segregation models are based on the observation that, contrary to widespread opinion, increased contact amongst culturally and linguistically different groups does not reduce tensions; it may even heighten them. This will especially be the case if there are minorities with a feeling of deprivation (see Duchacek L181/27, Lijphart L481/88s). The segregation models try to have as few points of contact or of friction as possible amongst groups by granting to each of them maximum autonomy and by reducing the need for consensus to a few central affairs.
143 Of this model there are two basic varieties. The better known one is federalism. The greater part of this book is devoted to it. The other variant, which will be explained below, is called consociationalism». There are other segregationary approaches which may be mentioned without further explanation. The first of these is the Ottoman Millet system (no.205), which underlies the Cypriot and Lebanese Constitutions. It differs from consociationalism by its mechanistic integration at the centre of elites that have little will to co-operate. Corporativism also should be mentioned (see no.206s), even if segregation in this case is restricted to relatively few areas.
144 The word «consociational» may be traced back to Althusius's «consociatio». In the 20th century it was first used by Apter (L155/107, L481/1,161s). As an explicit model consociational democracy (or accommodation democracy as it is sometimes also called) is based on Lijphart's analysis of Dutch politics (L480). But it may also be used outside the Netherlands to explain political stability in countries where population groups are to a considerable degree segregated, as for instance in Austria, Belgium, Canada, Colombia, Israel, Switzerland and Uruguay. The main characteristic of consociational democracy is a clear mutual distrust between population groups coupled with an elite policy of co-operation and peaceful settlement of conflicts. «The essential elements of the politics of accommodation are the division of society into a number of distinct and self-contained segments or blocs, overarching co-operation at the elite level, strong deference to the bloc leaders, and, as a result, a high degree of political stability» (Lijphart L480/197). For us it is less important to know whether the roots of consociational democracy are to be found in elite policy (Lijphart L480/196) or in the historical experience from earlier centuries and thus in a certain form of political culture (Lehmbruch L465/93, see also Daalder L155/121); or whether the model still adequately describes to-day's political situation of the Netherlands (see L480/197ss, L492/205s). The latter it quite obviously does no longer. Our interest is for the model as such.
145 The condition for the consociational model is a deeply divided population without cross-cutting allegiances. This, at least until the sixties, was the case in the Netherlands with her divided existence of Catholics, Calvinists, socialists and liberals (L480/23): «Deep religious and class divisions separate distinct, isolated, and self-contained population groups. Social communication across class and religious boundary lines is minimal. Each group has its own ideology and its own political organizations: political parties, labor unions, employers' associations, farmers' groups, newspapers, radio and television organizations, and schools - from kindergarten to university» (Lijphart L480/1, also L481/56, L670/49). Yet television later on made strict segregation over the long run impossible. This medium, therefore, contributed to the eventual pulling-down of barriers between the various population groups or «pillars» (Lorwin L491/58s).
146 «A key element of this conception is the lack of a comprehensive political consensus, but not the complete absence of consensus. There must be a minimum of agreement on fundamentals. Dutch national consensus is weak and narrow, but it does contain the crucial component of a widely shared attitude that the existing system ought to be maintained and not be allowed to disintegrate. The second key requirement is that the leaders of the selfcontained blocs must be particularly convinced of the desirability of preserving the system. And they must be willing and capable of bridging the gaps between the mutually isolated blocs and of resolving serious disputes in a largely nonconsensual context. The politics of accommodation resembles politics at the international level» (Lijphart L480/103s). «The first and foremost rule of the Dutch political game is that politics should not be regarded as a game at all. It is... a business.... Probably the second most important rule that governs the Dutch political business is the pragmatic acceptance of the ideological differences as basic realities which cannot and should not be changed» (L480/123s).
147 The elites are supported in their function of finding arrangements and compromises by various political «rules». Thus coalition-type governments and authorities are quite typical for consociational democracies (L481/25). Where the theory of games usually lets coalitions be just as big as necessary to achieve a simple majority (L84/217s), consociational coalitions are to the largest possible extent grand coalitions thereby giving minorities a veto-position (L481/26s,37). Proportionality is the basic pattern for representation and for the allocation of offices and of financial means (L480/127s, L482/25). To better solve political problems they are «depoliticized» through economic or legal formulas (L188/262ss, L480/129) or discussed in a climate of utter confidentiality as to the compromise mechanism (l480/131ss).
148 Consociationalism has its political price. Stability is achieved at the expense of a sometimes technically non-rational use of scarce resources (L491/51). There is also a certain immobility (L491/51s). Still, the innovations finally arrived at are then rather stable (L481/51s). Politics are «dull» because they largely lack the drama of conflict between the government and the opposition, keeping public interest at a low ebb (L260/328, L480/137s). Consociational democracy accepts elite government as a fact. But this can only be considered a shortcoming if it is compared with some ideal but non-existent form of democracy (L481/49s). And, finally, there is a danger in consociational regimes that citizen dissatisfaction is not absorbed by changes in government but produces a generalized discontent with the regime as such. «Although this is indeed a serious weakness, it does not have to be fatal. If voter disaffection is mobilized by new political parties, these may be antisystem or antiregime parties but they are not necessarily antidemocratic. And because the typical electoral system of consociational democracy is proportional representation, it is easy for new parties to gain a voice in the political process. This is what happened when consociational government began to break down in the Netherlands in the late 1960s. The relative ease with which consociationalism can be discarded makes the persistence of a democratic regime more likely. This argument can also serve as a final reply to the various charges of the insufficiently democratic character of consociational democracy: when these weaknesses are felt to be increasingly onerous, and particularly when they are regarded as less and less necessary because a society has become less plural, it is not difficult to move from a consociational to a more competitive democratic regime» (Lijphart L481/52).
149 Lijphart's consociational democracy is primarily an analytical concept. It is a description of a certain type of regime to be found in various countries. But Lijphart goes further. He ascribes also normative value to his model: «This book's message to the political leaders of plural societies is to encourage them to engage in a form of political engineering: if they wish to establish or strengthen democratic institutions in their countries, they must become consociational engineers» (L481/223). «Consociational democracy is both an empirical and a normative model.... The argument that consociational democracy can serve as a normative model challenges the pervasively pessimistic mood of our times and is deliberately unconventional» (L481/1,3). Such an application Lijphart proposes for instance for South Africa (L481/236). This proposition has greatly influenced South African intellectuals, as I have been able to see myself in the years 1979 and 1981. The new South African Constitution is an attempt to achieve consociational democracy for the white, the Indian and the brown population groups -but not for the blacks- in a mechanist way. But consociationalism arranged for by one of the partners does not conform to the prerequisite of elite compromise. And the question is still open whether the different elites can count on the kind of deference from their followers which the model stipulates as necessary to achieve accommodation (see also Nicholls L579/57s).
150 I consider both consociational democracy and federalism to be models of segregation (no.143). There are obvious similarities amongst them, as Lijphart pointed out himself (L481/42s): «As a theory, federalism has a few significant parallels with consociational theory: not only the granting of autonomy to constituent parts of the state, which is its most important feature, but also the overrepresentation of the smaller subdivisions in the 'federal' chamber. Federal theory can therefore be regarded as a limited and special type of consociational theory. Similarly, federalism can be used as a consociational method when the plural society is a 'federal society': a society in which the segmental cleavages coincide with regional cleavages. Because government at the subnational level is in practice always organized along territorial lines, federalism offers an especially attractive way of implementing the idea of segmental autonomy.»
C.222 Legitimation
151 There is no logical necessity for man to obey the state. On the contrary, there is a fundamental «contradiction between state domination and human autonomy» (Scharpf L701/9). True, fear of punishment is a powerful inducement to obedience. But those who trace the latter back uniquely to the former (as, for example, Raz L644/186) fail to see that fear cannot be sufficient reason for a whole population's obedience. At least in a long-range perspective «force is always on the side of the governed» (Hume in L170/77). Government must be perceived as «correct», «right», or «legitimate» to have a good chance of permanent success.
152 Easton (L188/278): «A member may be willing to obey the authorities and conform to the requirements of the regime for many different reasons. But the most stable support will derive from the conviction on the part of the member that it is right and proper for him to accept and obey the authorities and to abide by the requirements of the regime. It reflects the fact that in some vague or explicit way he sees these objects as conforming to his own moral principles, his own sense of what is right and proper in the political sphere.» Jouvenel (L398/323): «The idea of democracy means that the rulers are legitimized through their harmony with public opinion. This harmony does not have to be perfect; but neither must it sink below a certain level.»
153 The ultimate foundation of state and law is irrational (Burckhardt L109/186, also Eckstein L190/266s, Weber L850/70ss). Legitimacy of regimes is no abstract quality. It depends on subjective attitudes. «Legitimacy is the quality of a regime to conform to the theory of government accepted as true.... Thus defined, legitimacy is of a subjective nature» (Duverger L186/32). Or, in a cybernetic sense, legitimacy is (Deutsch L166/152) «an effective set of interrelated memories that identify more or less clearly those classes of commands, and sources of commands, that are to be given preferential attention, compliance, and support, and that are to be so treated on grounds connecting them with some of the general value patterns prevailing in the culture of the society, and with important aspects of the personality structures of its members.» (As to the distinction between legitimacy and legality, not treated here, see e.g. Blank L300/215ss, Friedrich L277/113 and Marcic L510/197,322.)
154 Friedrich (L277/114) makes a distinction between four kinds of legitimation: religious, legal (philosophical), traditional and procedural. The meaning of the first three concepts is self-evident. Democracy is the main type of procedural legitimation. To-day this form of legitimacy is for all practical purposes the only one universally accepted. «In the modern world there is in the long run no other justification left for governance than the idea of democratic legitimation» (Abendroth L1/156). For the social system the democratic myth has thus also the function of justifying governance. From a systemic point of view, democracy thereby guarantees the continuity of the system. It has a stabilizing influence, making it possible to defuse conflict without undue force (see Barry L49/173).
155 If we look at the elements of democracy's legitimizing strength we perceive that citizen participation (compare no.184) plays a vital role. «Any kind of government can claim to rest on 'the will of the people,' whether it be oligarchy or dictatorship or monarchy. One kind of government alone rests on the constitutional exercise of the will of the people» (MacIver L499/198). It may be that the actual exercise of participation is limited, both by citizen interest and by factual possibilities. «We no longer share the belief, originally looked upon as self-evident, that each citizen can participate in all important decisions. The multitude as well as the complexity of such decisions make this impossible. Bureaucracy and expertocracy are also democratic phenomena» (Ellwein L208/120, also Scharpf L703/168). However, in the final analysis democratic regimes may trace back their legitimation to ultimate citizen consent. This is the old idea of the social compact (no.051ss,598ss). But consent is not directed towards some legal text. It is a shared conviction that the political rules of the game are «right», that they are rules which organize the «volonté de tous» (no.054).
C.2221 Legitimation and the Majority Principle
156 The gathering of information in the political system is learning (no.114ss). Learning prepares decisions. The latter are no longer learning. On the contrary: the specific quality of decision, will in a cybernetic sense, is not being influenced by additional information. «Will resembles the 'deadline' in a newspaper: it could be called the internally-labeled preference for predecision messages over postdecision ones. The 'moment of decision' might then be seen as that threshold where the cumulative outcome of a combination of past information begins to inhibit effectively the transmission of contradictory data» (Deutsch L166/105s). But what should be the rules for arriving at political decisions? It is obvious that unanimity of all participants will only rarely be required (as in some African and Asian village democracies, L482/166s). Yet for the non-persuaded to follow the opinions of somebody else there has to be a compelling conviction that no alternative action is possible. Such a conviction may be based on the fear of force. As such, it does not interest us here. But as a rule the conviction will be based on the fact that the decision emanates from some authority which in the context of a given political culture is considered to be legitimate for this sort of decision.
157 In democracies, this authority is the totality of the citizens. It is not, strictly speaking, their majority. The majority is but the procedure to determine the will of the totality (compare Arendt L32/164). This is easy to demonstrate. If for instance there would always be the same majority imposing its will on the same minority, we could expect sooner or later a legitimacy crisis. But as there is in reality always a succession of «winners» and «losers», this fact somehow recedes into the background and it is the majority decision as such that appears to be the essential quality of democracy.
158 Sometimes the concept of democracy is reduced in a one-dimensional way to this procedure as for instance by Schumpeter (according to L702/36) who holds democracy to be «this state of institutions to achieve political decisions where some acquire jurisdiction through a competition for the votes of the people.»
159 There is no objective answer to the questions as to how large a majority has to be to assume identity with the totality and as to whose voice is to be counted. This again is subject to the beliefs held in a given political culture. To-day general (no.046) and equal suffrage is the rule.
160 Ruys (L685/136): «Democracy is defined as being a system of assigning a bundle of public goods such that each participant has an equal weight in the determination of the social benefit of that bundle.»
161 From an academic point of view it is evident that we may question the theory of the equal weighting of votes. In early forms of democracy, theory and practice favoured weighted over counted votes. This is no surprise if we consider the interests of the bourgeoisie that struggled for democratic rights. Thus for instance Mill (L547/281): «It is also important, that the assembly which votes the taxes, either general or local, should be elected exclusively by those who pay something towards the taxes imposed. Those who pay no taxes, disposing by their votes of other people's money, have every motive to be lavish and none to economise. As far as money matters are concerned, any power of voting possessed by them is a violation of the fundamental principle of free government; a severance of the power of control from the interest in its beneficial exercise. It amounts to allowing them to put their hands into other people's pockets for any purpose which they think fit to call a public one.» Mill also proposed the weighting of votes (L547/284s): «It can always be taken into the calculation, and counted at a certain figure, a higher figure being assigned to the suffrages of those whose opinion is entitled to greater weight. ... The only thing which can justify reckoning one person's opinion as equivalent to more than one is individual mental superiority; and what is wanted is some approximate means of ascertaining that. If there existed such a thing as a really national education or a trustworthy system of general examination, education might be tested directly. In the absence of these, the nature of a person's occupation is some test.» Notwithstanding such ideas universal and equal suffrage is nowadays the rule everywhere, even in large illiterate states where it becomes something of a farce. Nevertheless, maximum
equality is logical for democratic systems.
162 The assumption that majorities are right and may impose their will on minorities is not logically compelling (for the medieval discussion of this point see L98/164). It is the majority of one that especially provokes contradiction: «The nation is in doubt, the assembly is uncertain. And now one of the people's representatives crosses from the right to the left without evident motive thereby tipping the scales; he is making the law. And this law, this expression of some fantastic will, this is to be considered the will of the people!... And my holy duty to oppose tyranny yields to the almighty ball of a fool» (Proudhon in L328/289)! We might even say that truth is more likely to be with the minority, as again did the utilitarian Mill playing on a very much older theme (L547/107): «On any of the great open questions just enumerated, if either of the two opinions has a better claim to be encouraged and countenanced, it is the one which happens at the particular time and place to be in a minority. That is the opinion which, for the time being, represents the neglected interests, the side of human well-being which is in danger of obtaining less than its share.» But here,too, the radical solution, simple majority, has carried the day. Still, this has not been so to the same degree as for general and equal franchise. Constitutional law for instance presents many cases of some qualified majority requirements.
163 Again, it stands to reason that a logically compelling justification for qualified majority requirements is not to be found. But it is possible to give rational arguments for them. Public choice theory nevertheless makes a very interesting attempt: «According to Buchanan and Tullock, a representative individual wanting to form an organization to provide a public good would need to take two types of costs into account: 1)external costs which are defined as those costs an individual would expect to bear as a result of decisions which deviate from his preferences and impose deprivations upon him; and 2)decision-making costs which are defined as the expenditure of resources, time, effort and opportunities forgone in decision making.... If a constitutional decision maker were a cost minimizer, and the two types of costs described above were an accurate representation of the costs he perceives, we would expect him to prefer the constitutional choice of a decision rule where the two cost curves intersect. When the two cost curves are roughly symmetrical, some form of simple majority vote would be a rational choice of a voting rule. If expected external costs were far greater than expected decision-making costs, an extraordinary majority would be a rational choice of a voting rule» (Ostrom L592/67s). In other words: If in a political system we have to anticipate difficulties resulting from a simple majority decision to be greater than those entailed by the enlarging of the basis for consensus, the introduction of some qualified majority would seem to be appropriate.
164 The opposition against qualified majorities usually refers to public welfare (this is Rousseau's general will, no.054) or to the «undemocratic» preference of a minority introduced by qualified majorities. Thus Lord Brougham (L92/305): «This clumsy device is liable to the manifest objection among others that it must frequently prevent measures being adopted which the public good requires. It also tends to... give a minority undue weight and influence.»
165 But the most difficult problem that any attempt to validate democratic decision rules has to face is the forming of the opinions which, in their aggregate, constitute a majority and are therefore called the will of the people. Dahl stipulates the conditions that have to be met if we are to be able to talk about majority will in a strict sense (according to L277/10): «1) every member of the organization expresses his preference for one of the alternatives, for example by voting; 2) these expressions are treated as identical in weight; 3) the alternative with the greatest number of preferences is the winner... 4) members may add an alternative not previously identified; 5) all members possess identical information about the alternatives; 6) the preferred alternatives become operational and are executed by those in authority.» Dahl adds that «no human organization has ever met or is ever likely to meet these conditions». If we were to complete condition 5 by Edelman's thesis about the symbolic contents of political information (no.129), the democratic process would appear as pure fiction. But this would be a misconception of its function. Democratic procedures are a sham only if they are measured against their rhetoric. In reality they are but a means to submit fundamental political decisions to the largest possible forum, in order to bestow legitimacy upon them through the acquiescence of the ruled (and also to open up additional channels of information, no.028).
166 If we conceive of the majority principle as a legitimation mechanism, we have nevertheless to realize that it is not unquestioningly accepted as such in all democratic regimes. In plural democracies (see especially the discussion of consociationalism, no.143ss), the case for simple majoritarianism is opposed by the fact that it might place specific groups in a permanent minority. It is more likely that majoritarianism may have to be defused at the outset through special institutions to protect these groups (veto, coalition-type government, etc., see also L481/145).
C.2222 Legitimation and Political Culture
♦ 167 «Every society is held together by amyth-system, a complex of dominating thought-forms that determines and sustains all its activities. All social relations, the very texture of human society, are myth-born and myth-sustained» (MacIver L499/4).
168 We have arrived at the conclusion that democracy legitimizes governance. But if we say that we presume a specific kind of political culture, what is political culture? Gabriel Almond who, together with Sidney Verba, was one of the pioneers of modern empirical research in this field, once defined «political culture as consisting of cognitive, affective, and evaluative orientations to political phenomena, distributed in national populations or in subgroups» (in L19/26, compare L61/26ss, L206/5ss, L516/9s, L606/52s and especially L671/4ss).
169 There is a very close relationship between forms of government and political culture. According to Eckstein's congruence theory (L189/186ss, L190/234ss) the stability of a given regime is greater the more its characteristics and especially its authority patterns coincide with those of society at large. To take as an example British political institutions, we immediately see that they are a creation of British political culture and, in turn, have their influence on the latter. They cannot be transplanted to places where the foundations are not the same without having different results. The unsuccessful attempts to introduce Westminster institutions in Africa and in Asia are obvious demonstrations of this truth.
170 There exist different types of political culture. Almond for instance makes a distinction between a parochial, a subject and a participant type. The different forms of government, according to him, result from the mix of these types (L61/49ss). Eckstein proposes a democratic, an authoritarian and a constitutionalist pattern (L190/235).
171 We are mainly interested in the democratic type of political culture. It demands a certain level of intellectual abilities. «Democratic government, although based on such a simple and natural idea, always presupposes a very civilized and very well-informed society» (Tocqueville L806/328). «Another condition which seems at first sight necessary or highly desirable for the working of democracy is a minimum standard of education, amounting at least to literacy» (Field L240/131s). But intellectual abilities are not enough. If we look to Verba's definition we will also ask for certain kinds of values and behavioural attitudes. It is not easy to say what is part of these attitudes and what is not.
172 Almond and Verba have described this type of culture as civic culture. Barry (L49/56): «To put it in a very simple way civic culture is characterized by a balance of deference to public authority and steadfast independence. In this sense the ideal democratic citizen would believe in the legitimacy, general competence and good will of political authorities whilst being convinced at the same time that he has the right (if not the duty) to influence their decisions and to remedy abuses.»
173 Empirical studies have shown that this civic culture is regarded very differently by the population at large and by their elites. It has even been found «that the great majority of the population has a surprisingly low commitment to democratic values» (Bachrach L328/309). Scharpf points to the same problem when he states (L702/34) that «even under the favourable conditions of Anglosaxon or Scandinavian democracies participation research has always confirmed the existence of a large majority of politically apathetic citizens; and opinion surveys have not only demonstrated the low level of information and of political interest obtaining with the majority of the population, they have also shown how little perceptions of a broad consensus on basic values and rules in a pluralist democracy correspond to reality.» It appears that the population at large, for whom democracy offers more access to political influence than any other form of government, is less interested in it than is the elite (Lipset talks in this context of «working class authoritarianism», L702/41).
174 Given these facts official policy tries, at least superficially, to activate citizens through information campaigns and other measures. Contrasting with this a behaviourist school of democratic theory talks about «stabilizing apathy». «This hypothesis of stabilizing apathy has three very different sources: 1) a psychoanalytic trend suspecting behind every active politician a neurotic being using politics for pathological relief - a thesis as developed e.g. by Erich Fromm... probably too much impressed by dictatorial regimes; 2) a theory of modern mass society as developed in David Riesman's 'The lonely crowd'... where he described the dangers of manipulated pseudo-activities by externally guided mass man; 3) a theory of stable government as developed by Berelson, Lipset and other researchers with failed democracies in mind (as for instance the German Weimar Republic). For them these collapses were the result of, amongst other influences, over-participation of doctrinaire groups» (Beyme L65/16).
175 Such deferential apathy is also described by Lijphart (L480/144) in his study of Dutch democracy: «Neither the ideologically stylized pattern of elite-mass communications nor the high degree of elite dominance of the bloc organizations can fully explain the persistent allegiance of the rank and file of the blocs to their leaders. The people must have an inherently strong tendency to be obedient and allegiant - regardless of particular circumstances. This tendency will be referred to as deference. This term is here used in its broadest meaning: an individual's acceptance of his position both in the social hierarchy and on the scale of political authority, accompanied by a low level of participation and interest in politics. For the masses this entails respect for and submission to their superiors. These deferential attitudes are strong among the Dutch. Hans Daalder... speaks of the 'mixture of both deference and indifference which has tended to characterize the attitude of most Dutchmen toward authority'.» Eckstein is another of those who have studied this phenomenon's function for the system of democracy (L189/169): «Functional deference clearly narrows the range of groups and men among whom agreement must be sought... and helps identify men who can speak and act authoritatively in special areas of policy. This enables the system to act with some ease even in many nonconsensual areas (that is, those not governed clearly by communal norms)... At the same time it is a kind of deference highly consonant with democratic norms, since it does not militate against diffuse egalitarian sentiments, restricts authority to highly limited fields, disperses it among many men and structures, and compartmentalizes it in leagues» (this study's focus is Norway).
176 Democratic culture, like any other political culture, is thus characterized by the existence of stratified groups who participate differently in political life. Dahrendorf (L157/108): «There is continuous exchange between the different groups. But at any given time we may still discern the following ones: a) the latent public of the non-participating, be it that they lack motivation and interest, be it that they are hindered in their decisions by cross-pressures; b) the passive public comprising those who sporadically make an appearance in the political process as members of the public or as voters but whose initiative does not transcend the occasional question in a political meeting or nominal membership of an organization; c) the active public of those who participate regularly with their own ideas in the political process, who belong to organizations, enter upon public office, and who in their public utterances deplore the non-participation of the others. The dimensions of these different publics vary according to conditions that have been largely elucidated by science. But normally the active public is very much smaller than the other two. At no time and under no definition is it probable that it comprises more than one percent of the electors.» There is only one thing that makes this stratification democratically acceptable: transition is basically open to everybody (compare Berg-Schlosser L61/52).
177 Such findings - restriction of active participation to a small group, authoritarian attitudes in democratic systems - together with the already mentioned convergence of social and political authority patterns necessary for stable government, may explain why neither ideal democracies nor ideal democratic cultures exist in the real world. The democratic form of government always exhibits oligarchic and even autocratic elements. «This lends support to the still older theory of political science that mixed government is the most stable form of government - and puts it also in a new light» (Eckstein L190/267).
178 In this analysis of political culture we have up to now neglected one essential element. Populations must not only be seen as assemblies of individuals. They are also assemblies of groups (see also no.579). Practically all states have a plural structure of population (for pluralism see no.201ss). They are based on political part-cultures showing more or less great differences. An important part of the general political culture, therefore, is formed by the values and rules governing the living together of these groups. In democracies this coincides practically always with the renunciation in some form (federalism, consociationalism etc.) of full reliance on the majority principle. Plural democracy is a concept which shows an inherent tension. It is this tension that we shall examine below in connection with a specific kind of plural democracy: federalism.
179 Lijphart's models of democracy (L481/105s) are attempts to do justice to pluralism. They are based on two dimensions: population structure and elite behaviour. Lijphart distinguishes between four types of democratic systems, depending on whether the political culture is homogeneous or pluralist, and on whether elite behaviour is oriented towards co-operation or conflict. According to Lijphart, homogeneous culture and co-operation-oriented elite behaviour make for a depoliticized model of democracy. The same kind of elite behaviour in a pluralist setting results in consociational or accommodational democracy (no.144ss), where group leaders able to commit their followers will arrange for the peaceful resolution of conflicts. If elite behaviour is conflict-oriented, we have the models of centripetal democracy with homogeneous political culture and of centrifugal democracy with pluralist political culture. The latter is either a transitory state of affairs on the way to disintegration or it is associated with continuous crises of government. In centripetal democracy we may expect one elite or coalition of elites to impose its will on the others. That is why some authors (see L61/139s) even postulate tensions to be a necessary prerequisite for stability and the long-term survival of democratic structures.
180 Finally, of great importance for the continuance of democratic culture are the information channels transmitting and consolidating this culture. Amongst them are the mass media. «It is obvious that the mobility of an ideology as well as the receptivity of the ruled is in the first place dependent on the technological intensity of communication techniques. It comes as no wonder, then, that to-day's democratic state has to consider the control or at least the neutralization of the media a guarantee to its survival» (Loewenstein L494/253). «Pluralism of information media is an element of the pluralism of government, next to the pluralism of political parties. Besides, the political parties' pluralism would remain illusory and formal were it not accompanied by information media pluralism» (Duverger L187/220).
181 The mass media's contribution to democracy is somewhat different from official ideology's ideal democracy. A noted opinion surveyor (Dichter) assessed the situation in the United States in devastating terms (L171/207s): «If during the day the radio offers serialized soap operas, television fairy tales and popular tabloids stories with very strong mass appeal this only promotes latent tendencies to mental apathy, stereotyped reactions and standardized response.... There is a near kinship between moral routine reactions as called forth by these media and the blind following of fascist and communist ideologies' dogmas.» One could perhaps accept this with a shrug and point to the theory on stabilizing apathy (no.174) if the values transmitted corresponded to the accepted image of democracy. But this is very often not the case. «The essential problem of mass communication for to-day's society is the negative influence on the beliefs which are at the basis of this society. This results in the disintegration of values.... The so-called crisis of democracy is largely the outcome of developments in this field» (Friedrich L276/67).
182 The twentieth century brought the universal break-through of the principle of democracy (no.056). It is still very much an open question what the form of the twenty-first century's political culture will be.
C.23 Democracy's assigned Functions
183 Assigned functions are understood to be those that have no systemically necessary connection with the survival of political governance or organization per se (see no.031s). Such functions are assigned by a given social value system to a specific regime, in our case to the democratic regime.
184 The distinction is not always very easy. Citizen participation for instance is a core element of democracy. In one form or another it is essential for the functions of both learning and stabilization. But citizen participation is also deemed to have humanitarian ends. Thus Borter (L81/55): «This principle of citizen participation may be justified from an anthropological as well as from a political viewpoint. Participation facilitates in the political realm the preeminence of human development and self-determination, both committed to the ideas of freedom and equality.» For Tocqueville citizen participation had a pedagogic function (L806/369): «Democratic government makes the idea of political rights descend right to the lowliest citizen... This, to my mind, is one of its greatest merits.» These functions are what I call assigned functions. I will not treat these specific functons in detail here. Relevant considerations are examined in the corresponding discussion of assigned functions of federalism (no.614ss).
185 Next to the pedagogical idea of citizen participation, the most typically democratic of the functions assigned to democracy is equality.
C.231 Equality
♦ 186 «I think that democratic populations have a natural liking for liberty. Left to themselves they will search for it, they will love it, and they will feel grieved if it is removed from them. Yet for equality they have a passion ardent, unquenchable, eternal, and invincible. They want equality in liberty, and if they cannot have that they still want it in servitude» (Tocqueville L807/133s).
187 The idea of equality is part of the core of democratic ideology. But which equality? Aristotle (L34/5/1) made a distinction between «two kinds of equality. One is arithmetic and the other one according to one's due. By arithmetic equality I want to say that according to quantity or to size something is the same and equal. By proper equality I hold something to be equal according to its proportions.»
188 The kind of equality nineteenth century's liberalism stood for was proportional equality (see Burke L111/56s, Cooper L149/152, Leibholz L470/174, Preuss L632/39s, Schindler L713/56s). Equality according to one's due is to be understood as an attempt at a synthesis of individual autonomy and public governance. It is «buffered» equality. «Only then may we talk about 'true' democracy... when the definition of a democracy also contains material elements which will be of necessity normative, resting on the individual's ethical conviction and humanitarianism, as for instance the guarantee of liberal and social rights, minorities' rights, formal as well as substantive rule of law etc.» (Berg-Schlosser L61/155).
189 For emerging liberalism proportional equality was compatible with quite a few inequalities, especially in the field of the franchise (see no.161). This can be explained in the historic context of the interests of the bourgeois class. But democratic ideology also had to come to terms with conservative objections pointing to natural inequalities. Thus for instance Carlyle (L122/315s): «I have to complain that, in these days, the relation of master to servant, and of superior to inferior, in all stages of it, is fallen sadly out of joint.... Certainly, by any ballot-box, Jesus Christ goes just as far as Judas Iscariot; and with reason, according to the New Gospels, Talmuds and Dismal Sciences of these days.»
190 But it is in the nature of the democratic system that the yardstick for measuring equality cannot come from outside the system. The definitions of equality, of due, and of justice are also subject to the democratic process. The characteristic element of modern democracy is the dismantling of inequalities of every kind. «I prefer to describe a democratic society as one which is governed by a spirit of equality and dominated by the desire to equalize, to give everything to everybody» (Boorstin L79/102). The result of this development which, incidentally, the French Revolution confirmed soon afterwards, had been foreseen by Tocqueville (L807/185): «But man will never achieve equality he will be satisfied with.»
191 «The Jacobine Republic introduced this unitary ideal into the laws and habits of France which later a minister of the second Empire so well expressed when he, looking at his watch, flattered himself on the fact that he was able to say just which exercise all the pupils in all the schools of France were doing according to their curriculum» (Viatte L834/180s).
192 Hippel (L361/233) attempted to contrast modern democracy, which he described as quantitative democracy, with the qualitative, moral order of classical democracy. But mass society's democracy is also value oriented with its ideal measure (Imboden L387/23s), its idea of justice. In the final analysis both classical and modern Western democracy aim to enhance the individual person. The difference lies with the methods used and especially with the role of the state in this process. As modern democracy seeks to achieve this end through the highest possible degree of social and political equality, where the state has to play a major role it is impatient with formalities and with individual interests which impede the achievement of the objectives of the majority (see Kägi L403/181, L405/137). And as political freedom is nothing but a special kind of protection for certain areas of individual autonomy from majority dictate, it is particularly endangered in egalitarian democracies.
193 «Democracy cannot be conceived without equality, but it can be conceived... without liberty» (Schindler L713/60). Or Lenin (L474/38): «In reality, democracy excludes liberty.» - Tocqueville (L807/432ss): «Let's imagine which could be the traits of a new kind of despotism in the world: I see an uncounted multitude of men, all similar and equal, in constant agitation to procure themselves little vulgar pleasures with which to fill their souls.... Above them there is an immense tutelary power that unassisted assumes the task to provide for their pleasures and to look to their future. This power is absolute, detailed, orderly, farsighted, and benign.... I have always thought that this kind of servitude... which I have just described could be combined much better than many believe with some of the outward appearances of liberty and that it would not be impossible for it to be established even in the shadow of popular sovereignty.»
194 It is part of the logic of equality that it applies to individuals rather than to groups. «The image that corresponds most closely to the democracy's vision of society is a social atomism: the atoms themselves being simple and perfectly interchangeable are suitably governed by a few majestic generalities» (Koritansky L443/69, see also Tocqueville L807/137). Social atomism is the antithesis of pluralism, where groups are the building blocks. Pluralist democracy stands for the equal treatment of varied groups. This will of necessity result in a certain measure of inequality towards individuals. The concept of equality in the modern, individualist, and all-embracing sense is clearly strained if it has to include the acceptance of the autonomy of plural groups. This tension becomes obvious precisely in the relations of democracy, modern democracy especially, with federalism.
D. Federalism
195 As was the case with democracy (no.045), the term federalism has no universally accepted meaning. «Federalism has now become one of those good echo words that evoke a positive response but that may mean all things to all men, like democracy... or peace. We see the term applied to almost any form of pluralism and cooperation within and among nations» (Duchacek L180/190). Federalism is a generic concept for very diverse forms of the organizational combination of more or less autonomous units (Walper L843/11) or for the process leading to either such combinations or to their opposite, disaggregation (Duchacek L180/189).
196 It is understandable that proposals are made from time to time for a standard vocabulary of federalism for research purposes (e.g. Baker L45/1, Landau L449/173). I do not believe in such an enterprise. It would be unlikely to bear more fruit than the scholastic quarrels in former centuries. All that may in fairness be insisted on is that it becomes clear in the context of an author's thesis what he means by using the word federalism and that he applies a certain economy to his use of concepts by not calling the same thing under different names (as would be the case, for example, if pluralism were to be equated with federalism). But the last expectation is possibly already too ambitious.
197 The following considerations are not designed to bring forth the definitive definition of federalism. My only aim is to bring some clarity to conceptual relationships and to show what we are talking about. It goes without saying that I shall try to use consistent terminology. But I make no purist claims. Judged by its results my approach is to steer a middle course between the tendency to expand the concept to all forms of pluralist decision making (L180/51, L550/1111, L654/109s, L843/11) and the opposing tendency to use federalism in such a narrow way that only a very small part of the existing federations may still be called federal (L528/10).
198 For the purpose of explaining federalism, the point of departure is the observation that it is a principle for the organization of decision making in an association of groups. The particularity of this association is that the groups as such have a special function in central decision making and they possess also a certain autonomy. Such a system is pluralist (no.201ss). It is only when the basis of group diversity is territorial that we call it federalist (no.210ss). If the territorially organized groups are at the same time states under the law of nations we may then talk about a confederacy or a confederation of states (no.247ss). And if both the association and its members are states -the latter, it is true, in a limited sense- then we are confronted with a federation (no.254ss). This classification at least has the merit that each definition is deduced from a more general one by the introduction of a strict minimum of additional limitations. However, a proviso must be made for the concept of federation. This is not only a scientific term. It may also be a de iure term defined by a given constitution. This means that it is logically possible that there is federation without federalism (no.256) because the concepts are defined on different levels.
199 This hierarchy of concepts is simply a useful aid for classification. It conveys nothing, or very little, about the essence, the meaning, and the ends of federalism. It is static. The dynamic aspects of federalism and its relation to democracy will only be seen when we analyse the functional aspects of federalism (no.404ss).
200 There are advantages to be gained from this distinction between concepts and functions. For one thing it makes it possible to clarify the interdependence of terms which up to now has been the case mostly for word-pairs but not for the whole array of terms (see Frenkel L258/1s). Thus it is feasible to answer the question «what is federalism?» (as to some doubts see Fleiner L782/1028) without falling prey to some empty empiricism (see Hunger L373/10), and without becoming entangled in the confusion caused by the difference between reality and normativity (Dennewitz L161/84). Furthermore, the distinction provides a useful framework for the understanding of the characteristics of other definitions. Such characteristics usually emphasize specific goals implicit in the definitions. Alternatively, federalism could be defined by isolating specific institutions that establish the federal character of a political system. This approach is not very successful. It either leads to a vague concept (e.g. Livingston L485/272) or to a somewhat arbitrary one, as for instance Hughes (L370/13,16) who does not admit of the very concept of federation. (To Hughes federations are but a subspecies of the constitutional type of government, differentiated by some institutional elements of confederacy which alone has true conceptual reality. This approach reminds us of the notion of federation as a decentralized unitary state, no.228s.)
D.1 The Concept of Pluralism
201 A system for decision making is pluralist if it is an entity composed of groups each of which enjoys a certain independence and which, together, participate in an ordered and permanent way in the formation of the central entity's will.
202 This definition shows a certain relationship to some other definitions, for example those of Kremendahl and by Presthus. «Pluralism is the existence of a multitude of varied social interests and their organized representation in political decision making. The struggle for influence takes place within a framework of mutually agreed rules» (Kremendahl L229/286). «We shall here understand it as a social system of politics where the state, a multitude of private groups, interest organizations, and individuals represented by such associations share power.... To state it briefly: Pluralism stands for a system which distributes political power amongst public law corporations, in addition to which it is split up between the public realm and a plurality of private groups and individuals» (Presthus L631/81).
203 The decision-making system defined here does not necessarily have to be in the public realm. It may be any association at all. However, at the centre of theories of pluralism we nearly always find the state and thus political pluralism. The groups may represent any deliberately formed human associations. These may be based on ethnic, professional, religious, geographic, voluntary and other affiliations. That they enjoy a certain independence, that is autonomous regulation, is both a result of their organization as a group and a practical prerequisite for their participation in central decision making. This participation has to be secured and permanent; it must not be left to chance. The guarantee is to be found in the existence of regular mechanisms for participation with their own sanctions. (As to the connection between autonomy and participation see Kelsen L413/284s.)
204 Elements of pluralism may be found in practically all states. Some of the best known empirical studies are Dahl's in the United States city of New Haven, Lijphart's in the Netherlands (L480) and Lorwin's in Belgium and in the Netherlands (L491s). But Austria (Diamant L532/150, Lehmbruch L465/93) and Switzerland (Steiner L773s) also offer interesting opportunities for studies in this context.
205 The actual Lebanese Constitution, which is largely in abeyance because of Lebanon's entanglement in the Middle East conflicts, is a marked example of a pluralist but non-federalist model. Following the Ottoman millet tradition (L205a/5), the Constitution neatly distributes public power among most of the different denominational population groups (L180/106, L183/96s, L465/93, L481/148ss, L582/23). This pluralist approach was also a characteristic of the former Estonian Constitution (Bülck L104/12, L275/3) as well as of the Cypriot Constitution from 1960 until the coup d'état of 1974. The Cyprus Constitution attempted to achieve a settlement between the Greek and the Turkish population groups (compare articles 1 to 5; L180/105, L183/97s, L481/158ss). A similar approach was to be found in Appartheid South Africa. The Constitution in existence before 1982 in no way corresponded with the plural structure of society. Appartheid reform plans were mostly oriented towards a pluralist concept (see L656/409s,455, L870/53ss, Government Gazette 3.4.1979 pp.12 and 30).
206 Another variety of pluralist political organization is the corporativist or corporatist state (both expressions are used). Under corporativism or corporatism representation in at least one house of parliament is organized on the basis of non-ethnic groups, especially professional groups (see Manoilesco in L719/44 and already Hegel L348/255). D'Annunzio's Constitution for Fiume after the first World War was of this kind. But the favour corporativism enjoyed with the fascists in Germany, Italy and on the Iberian peninsula made the concept fall into disrepute, along with fascism itself (L579/54s, L719/8). Nevertheless, we find a modern application in the Yugoslav Constitution, where the Federation Council is composed of representatives from enterprises, local communities and social organizations.
207 Recently corporatism or neo-corporatism has had a renaissance as an analytical tool for political science. It has been used, for example, by Schmitter. For him it is not a subset of pluralism but stands on the same level with the latter (L719/15, like monism and syndicalism L719/16s). To this end Schmitter has constructed a definition of pluralism which is markedly different from the one used here (L719/15). He conceives corporatism as (L719/13, critical: Lehmbruch L467/149s) «a system of interest representation in which the constituent units are organized into a limited number of singular, compulsory, noncompetitive, hierarchically ordered and functionally differentiated categories, recognized and licensed (if not created) by the state and granted a deliberate representational monopoly within their respective categories in exchange for observing certain controls on their selection of leaders and articulation of demands and supports.» Schmitter readily admits that this definition is a construct (L719/45), the elements of which will not be fully implemented in any real state (L719/14). His examples - Brazil, Chile, Denmark, Portugal, Switzerland, and Yugoslavia (L719/18) - lead to the conclusion that for him corporatist structures may be found in all pluralistically organized political systems. This somewhat diminishes the analytical value of his concept. However, corporativism or corporatism seems to be very much the darling of contemporary political science (for a detailed exposition see Lehmbruch L468/1ss).
208 The theory of pluralism is multi-faceted. The models of pluralism and of democracy are very often closely intertwined. For an overview of different, mainly British and American, schools of pluralism see Nicholls (L579). The functions ascribed to pluralism are very often the same as those assigned to federalism. Therefore, a considerable part of the presentation below (no.404ss) is also applicable to pluralism.
209 In the foreground of theories of pluralism we usually find the problem of how to manage intergroup conflict, i.e. the problem of the integration of groups into the overarching system. In this context there are two mechanisms which have been closely analysed: separation or segmentation with elite-accommodation (see no.142ss) and cross-cutting cleavages (see no.582ss).
D.2 The Concept of Federalism
210 A system for decision making is federalist if it is an entity composed of territorially defined groups each of which enjoys relatively high autonomy and which, together, participate in an ordered and permanent way in the formation of the central entity's will.
211 It is the territorial element that distinguishes federalism from pluralism and defines the former as a subvariety of the latter. The main characteristic of the groups, then, is their relationship to some specific part of the globe's surface. «To paraphrase Gabriel Almond, a territorial interest group is a group of people 'linked by a particular bond of territorial concern and advantage and aware of these bonds» (Duchacek L181/3).
212 When other authors make a distinction between pluralism and federalism they practically always use territory as a basis for their differentiation (e.g. Aubert L37/402, Duchacek L180/51,192, Ermacora L214/22, Friedrich L277/51, King L419a/19, Livingston L485/2s, L486/23, and Lorwin L491/44). It appears that this territorial element is one of the most vital components of political structures generally. This may be one of the reasons for the comparatively large number of federal organizations relative to other plural ones. «This is so because territory provides one of the primary foci of identity, not in the sense of any romanticized 'territorial imperative,' but simply because where one lives is a basic way of establishing a social as well as a physical location» (Schwartz L736/107). «Politically, territorial organization has an attractive function, as it aims at and often succeeds in aggregating various special and functional interests into a comprehensive and cooperative whole.... In general, people seem to support and identify with a spatially defined community more readily than with a nonterritorial interest or organization» (Duchacek L181/24s).
213 The definition proposed here does not specify autonomy but requires «relatively high autonomy». This recognizes ordinary usage whereby an entity with a small degree of autonomy would be called decentralized rather than federal (as to this distinction see no.222ss). «Relatively» at the same time introduces a factor of judgment, which explains why the federal character of a given political system may be assessed differently by different observers. And lastly we must be careful not to interpret autonomy in an absolute sense. It is always relative autonomy, autonomy within the framework of a system (see Altenstetter L21/5).
214 As a word federalism may be traced back to foedus, the Latin term for union. The English term was coined during the Civil War in 1645. Its present meaning originated in 1777 during the American Revolution (Elazar L198/xiiis, also L424/580). The French «fédéralisme» was first used by Montesquieu (Voyenne L837/63, but «fédération» dates back to the 14th century). In French parlance the concept is mainly connected with the Revolution and the struggle between Jacobins and Girondists (L675/97s, see no.381s). The modern European use of the word usually stresses the autonomy of the units (see Deuerlein L165/153s). The English term «federal», on the other hand, rather points to central concerns. This is certainly the case for the United States and Canada. It is increasingly so for Australia (Alexander L17/vi). When the word «federalism» is used in the present book it is in a more scientific sense, in relation to the system as a whole. (See the comments on federation and confederation, no.302.)
215 Federalism is not a legal term (Dennewitz L161/21, Harbich L339/46, Loebenstein L487/834, de Reynold L654/110; Messmer is inaccurate L542/110). It is an organizational principle in the political or social realm. Confederation and federation are both legal terms (no.241ss). Federalism might be a legal concept if a given constitution would make it a binding mandate for legislation, administration and the judiciary (for an illustration see L525/20/I/20). In such a case the general meaning of the word would not suffice for its interpretation. The legal construction would have to be arrived at in the context of the specific constitutional arrangements.
216 There exists a certain relation between the definition here proposed and those given by, amongst others, Friedrich (L277/54) and Livingston (L485/9). Friedrich: «In short, we can properly speak of federalism only if a set of political groupings coexist and interact as autonomous entities, united in a common order with an autonomy of its own.» Livingston: «Federal government is a form of political and constitutional organization that unites into a single polity a number of diversified groups or component polities so that the personality and individuality of the component parts is largely preserved while creating in the new totality a separate and distinct political and constitutional unit.»
217 Some authors stress additional requirements. Schechter for instance (L710/3s) emphasizes contractual relationships as the historic basis of federalism. Ostrom (L593/205) asks for a minimum of democratic control; and many scholars introduce goals or rule of law prerequisites into their definitions. I prefer not to include such elements but to look at them in the context of the functions of federalism (no.404ss). A more recent tendency considers federalism to be mainly a process (Altenstetter L20/5, Friedrich L275/3) or a model (Vile L835/4,6s). This may be understandable as a reaction to the emptiness (Vile L835/2) of many modern definitions. However, both approaches are not of much help for classification.
218 In the United States the disparity of the various concepts of federalism (compare the alphabet in no.316ss) has resulted in a certain academic preference for the more clearly definable term «intergovernmental relations» as a substitute for federalism (Beer L57/50, Wright L873/1s). It is obvious that this term covers but a segment of the subject.
219 If we look at federal systems we will find federations like Australia, Germany, and Switzerland. But there are also nongovernmental entities we may call federal, such as the League of Red Cross Societies, trade unions, and political parties (especially in federations). And finally associations of states are federal systems, as for instance the United Nations and the Council of Europe.
220 The economic theory of federalism (see no.223, 472ss, 516ss) shows a tendency to equate federalism with pluralism. Kirsch (L421/13s) is a case in point: «There are, then, two ways to transcend the actual state of the theory of federalism. One is to include decision processes that take place within collectives. The other approach extends the field of inquiry to non-state, non-regional associations.» For the sake of conceptual clarity it might be better to call this an economic theory of pluralism.
D.21 Federalism and its Complements
221 Federalism, particularism, centralism, unitarianism, or decentralization are terms very often used together. Let us see how we may differentiate them. (American texts very often make little distinction between intergovernmental relations and federalism. Yet the former are but one aspect of federalism, Frenkel L264b/1477, Gaus L783a/44ss.)
222 The terms just mentioned are not opposites. We may imagine a continuum of links ranging from no connection at all to complete unity. It is on such a continuum that we must localize our terms. Viewed from the one end of the row - unrelated diversity - to the other -unity- the process will appear as one of centralization or consolidation. Centralization is usually more concerned with the concentration of procedures, consolidation with the concentration of structures. But we may also look at the continuum from the other side, from unity to diversity. Then we are faced with decentralization, particularization, disintegration, or separatism. Again these are processes. The concept of federalism occupies a fairly large range somewhere in the centre of the continuum, where particularist and unitary elements overlap. If we transfer these considerations to government we may say that confederation as well as federation are to be found in some place between the complete (theoretical) independence of states and their (equally theoretical) consolidation in a totally undifferentiated unitary state. Confederation is nearer to the one, federation is nearer to the other side (see also L161/92, L174/4, L221/8ss, L251/1, L339/20, L370/5s, L413/310,316, L486/24, L654/21s, L685/111, L700/16, L708/138, L744/177, L862/136).
223 Political economists have shown a special interest in the concept of decentralization. They see it as the common denominator for federalism, democracy, and the market economy. «All three types of decentralization are about the distribution of decision-making powers amongst a great number of persons or organizations. In the market economy we will have the consumers and employees on the one and the producers and employers on the other side which are the autonomous actors; in democracy these are the citizens, in federations the regional member-states» (Frey L268/17, also L58/22, L267/362, for a more subtle analysis compare Winkler L868a/277ss).
224 Sometimes a distinction is made between decentralization, deconcentration and devolution. The essence of deconcentration is the transfer of powers from central authorities to subordinate, specialized, and usually local ones which are still part of the same hierarchy (L231/5s, L503/23, L538/348, L742/52, L855a/151). Decentralization, then, would stand for the transfer of powers to local communities with a certain autonomy. The same differentiation is meant when authors talk about administrative and geographic decentralization (L377/86). Transfer to autonomous local units is also known as devolution, especially in British parlance. It could be said that decentralization is the general concept which comprises both administrative deconcentration and political devolution (L503/23). But usage is very loose.
225 «There is a theoretic discussion about the difference between 'federalism' proper and 'decentralization'. As a matter of fact there exists no difference of quality but only one of degree: decentralization is reduced federalism, federalism is decentralization very emphasized» (Duverger L186/71). If we are only concerned with the logical graduation of legal normativity we may accept this idea of Duverger and others. But it does not do justice to the political phenomenon of federalism. I agree with Diamond (L169/136): «Now the very word decentralization implies the existence of a real center from which things are to be decentralized, and not only the existence of that center, but also its priority or supremacy.... As we have seen, the essence of federalism lies in the fact that it rests upon arguments as to why a group of polities ought, despite certain common interests, to remain decisively themselves and ought not to form a nation. Decentralization, on the contrary, presupposes a nation and rests upon arguments merely as to how the nation ought to be organized so as to achieve liberty or other desired qualities.»
226 If we look to social reality rather than to abstract terms we observe that federal structures are usually brought forth by some process of centralization and not by decentralization. It is not unity that is the basis but particularism. For federations, to take the typically federal form of government, it is not decentralization that is essential but non-centralization. «True federal systems must be non-centralized systems. Even when, in practical situations, there seems to be only the thin line of the spirit between non-centralization and decentralization, it is that thin line which determines the extent and character of the diffusion of power in a particular regime» (Elazar L198/xii, compare Elazar L205b/14, Richardson L637/6-4/4).
227 There are some additional terms which complement decentralization, for instance self-organization and self-management. The differences are similar to those between non-centralization and decentralization (Pernthaler L615/19ss). Federalism is based on the assumption that its constituent parts are to a considerable degree self-organized and self-managed.
228 The constitutional law analogy to the general discussion about federalism and decentralization is the distinction between federations and decentralized unitary states (L244/45, no.292). Preuss calls this the «eternal question with no possible answer» (L632/145).
229 To conceive of federations as decentralized unitary states is in keeping with a purely norm-logical approach (such as Kelsen's so-called pure theory of law). From this point of departure it is quite probable that a clear-cut distinction will never be possible (see Usteri's attempt in L825/236). The whole discussion is very much steeped in the whimsicalities of European constitutional law traditions. To call a federation a decentralized unitary state introduces a polemical element into the discussion. From a pure theory of law viewpoint, the unitary state is a tautology. As legal systems states are, according to this theory, always unitary. If logical consistency were the aim -and Kelsen's school aims for it- federations might be called decentralized states. To talk about decentralized unitary states makes sense only if we wish to introduce a pointedly anti-federal note.
230 I quite like the graphic description of the difference between administrative decentralization and political federalism which was given by the French politician Barrot who, when commenting on the decentralization decrees of 1852 and 1861, sarcastically said (L837/294): «It's still the same hammer that pounds, they have but shortened the handle.»
D.22 Federalism and Subsidiarity
231 The concepts of federalism and subsidiarity are related to each other. But they do not coincide. Federalism is concerned with the structure of the decision-making process, with the co-existence and co-operation of different autonomies (no.210). Subsidiarity has to do with the allocation of tasks within a given political structure in such a way that the process ascends from «bottom» to «top». The principle of subsidiarity may be optimally implemented within a federal structure. But the latter is not indispensable (see Dennewitz L161/87, Ehringhaus L194/33, Hesse L356/2s). Subsidiarity is suited to every kind of pluralism and also to the relationship between the private and the public sector (compare Jellinek L382/259, Utz L827/84, Wilson L863/1531s). It also has its place in unitary states, e.g. for problems of decentralization, because the element of participation in central decision making, essential for federalism, plays no role for the subsidiarity principle.
232 Like federalism, subsidiarity is not a legal concept. When the Basic Law (Grundgesetz) of the Federal Republic of Germany was being drafted, an attempt was made to make this notion part of the law of the land (L339/117s). But this was not successful. Within the European Union's legal framework subsidiarity is a concept of law. But it is questionable to what degree its application corresponds to the bottom-up approach usually associated with the term.
233 What is the meaning of the principle of subsidiarity? The term has two different if related connotations. Both derive from the Latin word «subsidium». One meaning has to do with «reserve», i.e. with something that applies only if some primary condition is lacking. The other meaning is connected with the concept of help: subsidy.
234 The first meaning is the one mainly used in European theories of government. It signifies that one should always start with the smallest group or even with individuals when allocating functions. This level is always competent insofar as it can assume a function effectively and without damage to others. Only those problems that exceed the group's possibilities are to be assumed by the next more encompassing group. «As a supreme principle of social philosophy the subsidiarity principle is nothing else than a kind of jurisdiction rule, for the jurisdiction between communities and finally also between the individual and society» (Marcic L509/431).
235 Thus defined, subsidiarity corresponds with Héraud's concept of «exact adequation» (adéquation exacte, L668/211). Héraud himself does not admit this because he burdens his definition of subsidiarity with an unnecessary element of delegation (L668/208ss).
236 In its other sense the term is that used by the Freiburg (Switzerland) social philosopher Utz: «The principle of subsidiarity is a law concerning assistance: society's to the individual, or the more encompassing group's to the smaller one, as e.g. the state helping the families or free associations (subsidium = assistance)... According to the principle of subsidiarity society should assist the individual where the latter cannot fully achieve his goals with his own means» (L827/17s). «The general welfare is based on solidary subsidiarity. This is to say that in the interest of the public welfare society as a whole has the duty to come to the aid of the individual according to his needs» (L827/61).
237 The philosophic foundation of subsidiarity may in both cases be the same. The view-points are different. In the first case it is «bottom to top», in the second one «top to bottom». This is the same with political parlance. The subsidiarity principle in its theory of government meaning is invoked when the non-centralization of a function is to be defended. The other meaning is emphasized when there is a need to justify financial subsidies.
238 Quite often the principle of subsidiarity is interpreted as an achievement of Catholic social teaching (e.g. Marcic L509/429, Stadler L764/12ss) by pointing to Pope Pius XIth's encyclical «quadragesimo anno». But this is too narrow an understanding. Based on the doctrines of the division of powers and of the social compact there is a Western tradition of subsidiarity which has probably been more influential than the religious one (L764/13). Classical antiquity was familiar with the idea, as is evidenced for instance by Aristotle's consideration of the advantage of independence over unification (L34/2/2).
239 In the context of this text, subsidiarity is not taken as an end of or a justification for federalism (see the doubts of Hangartner L336/31, Lerche L477/74s). Subsidiarity is first and foremost a principle for the allocation of functions in federations. Despite other authors' doubts, it is helpful to use subsidiarity as a principle for assigning the burden of proof in discussions about the division of functions in federations. As a matter of fact, it is more functional than most other approaches (such as national interest or spillover theories). From a federalist perspective, subsidiarity introduces a bias towards non-centralization. Given the overwhelmingly centralizing pull of modern institutions this is not an undesirable counterweight. For a more detailed discussion of the application of the principle of subsidiarity see the German companion volume to this monograph (L264b/957ss).
240 The legal embodiment of subsidiarity in federal constitutions is, to a certain degree, the states' rights clause, for instance the 10th Amendment to the United States Constitution: «The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.» In all federations with the exception of India and of Canada the constitution presumes that tasks not assigned to the central government are of the member states' jurisdiction. Because of very special circumstances the situation is practically the same in Canada, although the Constitution says something different (see L264b/996 and L684/1ss). The Yugoslav Constitution (C116/3) even contains a presumption in favour of local government.
D.3 Federalism in Government
241 I have repeatedly emphasized that federalism is an organizational principle that is not restricted to the political sphere. Nevertheless, it is here that scientific interest is most acute. Two specific forms are in the foreground of discussions: confederation (no.247ss) and federation (no.254ss).
242 Elements of federalism are also to be found, however, in many non-federations. Very often, for instance, territorial units like provinces or communes enjoy a relatively high degree of autonomy and even participate in central decision making. In federations there is also a relatively high probability that the member states will show traits of federalism in their organization. Even if they are not themselves federations (as some were in the USSR and in Yugoslavia), local governments are very often considered to be the political pillar of federalism; as for instance in Switzerland (see Bäumlin L54/143). In unitary states with demographically different units we may also find elements of federalism. Even if, to give an example, we know very little about the actual structure of the People's Republic of China, geography, the existence of 54 national minorities (L176/14) and the Constitution (e.g. art.4,95ss), it is evident that we may expect some federal traits, democratic centralism notwithstanding (see Bryce L99/559, also Donnithorne L177a/3ss). Another illustration of federal structures in a unitary state are the Italian regions, especially those with a special statute (L182/101ss, L852/introduction; in this context see also my reflections on partially deconcentrated states, no.403). The Spanish Constitution reads very much like that of a federation. Federalism in this «unitary» state appears to be at least as real as in Argentina, to take a «real» federation. Or consider the special status of Northern Ireland and of Scotland in Great Britain (L180/121, L182/98, L485/269s, L505/331s). Federalism is undoubtedly more universal than federations are!
243 The quasi-state entity of the Roman Catholic Church also shows elements of federalism. Thus there exists a kind of confederation of the fifteen Benedictine congregations which are themselves federally structured.
244 Political federalism, therefore, is not tied to the existence of member states or of other territorial subdivisions of a specific kind. The «groups» I mentioned in my definition of federalism (no.210) may be varied and of different influence. In this I concur with Ostrom (L593/205) who prefers «to characterize a 'highly federalized' political system as one which has a rich structure of overlapping jurisdictions with substantial autonomy among jurisdictions. I see no point in confining discussions... to state governments and the Federal government. The federal principle can be (and is in the American context) used to organize numerous concurrent regimes.»
245 Because federalism is a political rather than a legal concept (no.215), the usual imagery representing it as a pyramid and thereby as a hierarchy is erroneous. The hierarchy of legal norms which has been taken over from legal theory is not essential even there (see L264b/877). To illustrate federalism as a structure of more or less overlapping centres of power is much more accurate than to point to the pre-eminence of a higher level in a pyramid (compare Elazar L201/283, Jerusalem L385/5). The adjoining illustration is taken from Elazar (L205b/15).
246 In this context we should introduce Grodzin's famous analogy between federalism and the marble cake (L310/3s, compare no.328): «The federal system is not accurately symbolized by a neat layer cake of three distinct and separate planes. A far more realistic symbol is that of the marble cake. Wherever you slice through it you reveal an inseparable mixture of differently colored ingredients. There is no horizontal stratification. Vertical and diagonal lines almost obliterate the horizontal ones, and in some places there are unexpected whirls and an imperceptible merging of colors, so that it is difficult to tell where one ends and the other begins. So it is with federal, state, and local responsibilities in the chaotic marble cake of American government.» This is quite different from the Frenchman Scelle's assertion (L700/14): «Federalism is, let's repeat it, the contrary of anarchy because it is hierarchy: hierarchy of norms, hierarchy of institutions, of governments, and of administrations.»
D.31 Confederation
247 A system for decision making is a confederation if it is an entity composed of states, i.e. of territorial governments independent under the law of nations which, together, participate in an ordered and permanent way in the formation of the central entity's will.
248 States being comprehensive territorial units, they have by their nature great autonomy. It is therefore not necessary to emphasize this element (no.210) in the above definition.
249 Confederations (we will also encounter the terms «leagues» or «unions») are not complete unions. They are restricted to a few specific common tasks. It is thus possible - and normal - for a state to be a partner in several confederations. An earlier German doctrine considered certain ends to be conceptually essential: «The confederation ('Staatenbund') is a permanent treaty based union of independent states with the twin goals of common defense of the union's territory and the keeping of peace amongst the united states. Additional aims may be agreed upon (Jellinek L382/762).» It is true that internal and external security were at the basis of most old confederations. But to-day we are also talking about confederations, even if they have nothing to do with security, being for instance concerned with economic problems only, or with very restricted technical ones (e.g. navigation on common water-ways). Still, not every international treaty institutes a confederation or union. This occurs only when some sort of common decision making is institutionalized. This institutionalization, incidentally, will sometimes go very far and the common functions may be so important that the new entity is treated not unlike a state in international relations (as with the United Nations and the European Communities).
250 From an historical perspective confederations are forerunners of federations. The inventors of modern federations, the founding fathers of the American Constitution (no.379), set out to combine the kind of loose federalism they considered the ancient Greek unions to have had with the unitary or national principle. The result was a new kind of regime, federation. This explains why the partisans of the 1787 Constitution were very often attacked as anti-federalists: «They are national men, and their opponents, or at least a great majority of them, are federal, in the only true and strict sense of the word» (L80/6, compare Diamond L169/135s). A modern parallel to this discussion is Hughes's theory granting conceptual reality only to confederation (no.200, for his differentiation of loose, narrow and very narrow confederation see L370/11ss).
251 The main terminological problem is the distinction between confederations and federations. It is comparatively recent (no.250). I will come back to this after analysing the concept of federation (no.302ss). Let me just point out that English and French parlance are less rigid than is the German use of the words «Staatenbund» (confederation) and «Bundesstaat» (federation).
252 I do not make a distinction between alliances and confed erations. If such a one had to be proffered, it would probably be based on the level of institutionalization and on the fact that alliances are usually temporary, whereas confederations are permanent (compare Flora L247/11, Nelson L576/8). But both unions are soluble and transitions are fluid. Nor do I deal with the other kinds of associations of states such as personal unions, real unions and the formerly important relations of suzerainty (see Jellinek L382/750ss, Rivier L664/94ss).
253 To-day all states of the world which are independent under the law of nations belong to at least one confederation. The few countries which are not yet members of the United Nations, Taiwan for instance, are bound to other nations through a multitude of different treaties.
D.32 Federation
254 A system for decision making is a federation if it is itself a state, i.e. a territorial government independent under the law of nations, composed of territorial units with a relatively high degree of autonomy which therefore, or for historical reasons, are also called states and which participate in an ordered and permanent way in the formation of the central government's will.
255 The federation is the typical form of federalist government (compare Dennewitz L161/105, Huber L782/1158, Jellinek L382/785; for a different theory, see Hughes no.200). Federation is a legal term (Frenkel L264b/699, Harbich L339/46, Macmahon L501/v,3). Like most legal terms the concept of federation has two aspects. First it is a subject for constitutional law or theory in general (for a contrary view: Hippel L361/250). However, this is not enough to understand fully the practical applications of the concept. We have to consider also its use in a given constitution. The concept of federation as we find it in the basic law thus has two sources. By using this term (the actual wording is different) the constitution refers to general theory. But this theory is then complemented, and sometimes considerably amended, by the actual form of federation (see Hempel L352/209).
256 It may very well be the case that a given federation is not a federation in the sense of the general theory. Thus the federal character of communist and Latin American federations is denied by many scholars. And it was indeed difficult to consider the two member republics of former Czechoslovakia as enjoying «relatively high autonomy». Nevertheless, the fact that separate authorities do exist may create the fertile soil necessary for a certain federalist evolution. As to the Yugoslav Constitution, it was the most dogmatically federalist basic law ever devised (see for instance the member republics' absolute veto power in article 295). We may perhaps question the effectiveness of such federalist elements. In any case it is advisable to check closely the constitutional realities before passing judgement on the «federality of federations» (compare L26/163s, L339/46, L838/51).
257 My definition of federation may sound unusual. Yet it is not so different from Wheare's, the one usually found in textbooks (L858/10, also L859/19): «By the federal principle I mean the method of dividing powers so that the general and regional governments are each, within a sphere co-ordinate and independent.» Unlike Wheare I emphasize the applicability of the federal principle to the non-governmental realm, and in the governmental one I do not restrict it to federations as Wheare's definition implicitly does.
258 Wheare's definition may be traced back to Garran (L858/14): «A form of government in which sovereignty or political power is divided between the central and the local governments, so that each of them within its own sphere is independent of the other.» Birch has slightly restated Wheare's formula (L173/7): «a federal system of government is one in which there is a division of powers between one general and several regional authorities, each of which, in its own sphere, is coordinate with the others, and each of which acts directly on the people through its own administrative agencies.» In a similar vein Duchacek (L180/194): «By a federal system we mean a constitutional division of power between one general government (that is to have authority over the entire national territory) and a series of subnational governments (that individually have their own independent authority over their own territories, whose sum total represent almost the whole national territory.» Further refinements have been added by Vile (L63/4): «Federalism is a system of government in which central and regional authorities are linked in a mutually interdependent political relationship; in this system a balance is maintained such that neither level of government becomes dominant to the extent that it can dictate the decisions of the other, but each can influence, bargain with, and persuade the other. Usually, but not necessarily, this system will be related to a constitutional structure establishing an independent legal existence for both central and regional governments, and providing that neither shall be legally subordinate to the other. The functions of government will be distributed between these levels (exclusively, competitively, or co-operatively), initially perhaps by a political process, involving where appropriate the judiciary...» These definitions, incidentally, show that Anglo-Saxon writers do not usually make a clear distinction between federalism and federation.
259 In German definitions, as in many French ones, the problem of sovereignty is very often an essential element. «The federation is a sovereign state composed of a plurality of states whose power is based on its politically united member states. It is a public law association of states instituting governance over the united states in such a way that the members are always these states which in their totality rule or participate in ruling and which singly are subjects in certain areas» (Jellinek L382/769). For Merk (L833/21/115) federations are «legal associations of several states creating an aggregate state with an aggregate state population, an aggregate state territory, and an independent and supreme dominion, in such a way that according to its constitution these member states have independent state powers not derived from the central government, that they participate in the decision making of central government in a supreme authority, but that they are subjected to the domination of the central government in the areas of the latter's jurisdiction.» The problem of the locus of sovereignty in federations bedevils a great part of German and French legal discussion about the concept of federation. This, together with the related debate about the status of member states, will be treated in more detail below (no.272ss).
260 Let us now consider the individual elements of my definition: First, the association has to be a state, i.e. a state as conceived by the law of nations. This establishes a differentiation from a confederation (for details see no.302ss). It also points to the existence of an encompassing legal structure (compare the detailed description in Frenkel L264b/875ss).
261 The parts are member states, i.e. territorial governments which are designated as states. This parlance is somewhat questionable because it is obvious that such entities differ from the nation states which usually participate in international relations. Many European authors, therefore, refuse to talk of states in this context (see Giacometti L244/43s, Öhlinger L587/15, Usteri L825/209). This has a close connection to the discussion about sovereignty in federations (no.272ss). On the other hand we must not fail to notice that the constitutions very often talk about states, republics or Länder. Science thus has no choice but to accept that there exists some state quality to be defined in member states, even if we lack legal yardsticks for the clear-cut differentiation from provincial self-government (in political reality this may be simpler).
262 The characteristic elements of a member state are a) a full and autonomously appointed structure of authorities not only administrative but also political, b) direct constitutional settlement of organization and essential tasks (see article 29 of the West German Basic Law providing for legislative rearrangement of state boundaries), and c) a specific role in central decision making. The level of central control (see the difference between «decentralized» and «non-centralized» no.226) is also important. But an evident distinction between member states and provinces is not possible. In addition to this, there is the problem that in some countries (such as Canada by article 5 of the Constitution) member states are called provinces. In the final analysis the difference between member states and autonomous provinces is largely one of terminology and of little importance so long as it is not used as the basis for normative deductions and it is clear what one talks about. The particularity of a member state as compared with an autonomous province cannot be deduced from the term but only from the fact that it is part of a federation.
263 The term «state» is used in Australia, Brazil, India, Malaysia, Mexico, Nigeria, the United States of America, and Venezuela. In the composite «council of states» it also appears in the French and in the Italian texts of the Swiss Constitution. Here the common term is «cantons» or «Stände». The term «republic» was used in Czechoslovakia, the USSR, and Yugoslavia. Austria and the Federal Republic of Germany speak of «Länder», Argentina, Canada and Pakistan of «provinces». The Comores are made up of «isles». The United Arab Emirates contain «emirates». Finally, Tanzania just talks about «Zanzibar» and sometimes, by way of contrast, of «Mainland Tanzania».
264 According to the definition the autonomy of member states should be relatively high. If something is «relative» it is based on a value judgment. This explains why we often find differences of opinion as to some federation's «genuineness» or its «quasi» status (e.g. Wagner L838/29, Wheare L858/20, L859/20s). It is obvious that it is quite difficult to place an exact borderline between «sufficient» and «insufficient» autonomy. The observer's ideological viewpoint will colour his assessment. (The question is one I deal with further in the context of the criteria for the allocation of functions in federations, L264b/886ss). The field is open-ended at both sides. «A given federation is federalist if its member states, in their relations with the federal government, enjoy some full autonomy which may be heightened to such a degree that there is but a theoretical difference to confederation. Where, on the other side, a federation shows unitary traits, the member states' independence is more or less restricted in favour of the federal government. This may reach a point where the difference to a unitary state is, again, but theoretical» (Jerusalem L385/5).
265 The definition's last component is participation. This concerns a large variety of forms which are treated in detail in my German companion volume (Frenkel L264b/1302ss). Let us just say at this point that there are different variants of participation. A member state may either participate as a structured political unity through its authorities, or it may appear as an essentially statistical base as is the case with electoral boundaries. It is also necessary to look not only to legal texts but to the political realities of participation, which are sometimes much more important than the texts (e.g. career patterns in political parties etc.). Finally, we must not overlook that there is a relationship between the relatively high degree of autonomy postulated and participation. Both are dependent on and mutually guarantee each other (see Wagner L838/29). To emphasize but the element of autonomy (see Bennett L59/10, Schollenberger L725/170ss) is too narrow. On the other hand, it is difficult to conceive that participation would be of much importance in the long run if it were not based on autonomous political activities in the member states. These, again, presuppose a considerable degree of independence. There is a tendency in modern German theory to give priority to participation to the exclusion of autonomy. This results from an interesting analysis of modern German federalism (Hesse L356/25). However, this analysis is perhaps too narrow (see Hempel L352/200, Tiemann L804/323ss), and the conclusion mentioned is certainly not warranted in a normative sense.
266 Participation has to take place in an «ordered and permanent way». This is to say that it must not be at the mercy of the central government. As a rule it will be organized by the constitution (Burckhardt L109/163). It will also happen regularly, i.e. in the ordinary process of decision making, and not as an exception.
267 As tasks are interrelated the central government also participates in the decision making of member states. But this is a by-product. Participation of the centre in member state decision making is not an element of federalism. It may even open the way to the destruction of autonomy. This form of participation is a form of decentralization rather than non-centralization (no.226). Ermacora (L435/87) has postulated the conceptual necessity that member states participate in the governance of the central government and vice versa, but the latter element is very questionable.
268 From the viewpoint of the definition, it is irrelevant whether a federation has been brought into being by an aggregation of formerly separate states or by partial disintegration of a unitary state (compare no.674ss). Fischer (L241/44) prefers to call only the first entity federation. For the second case, he proposes to talk about a federal unitary state («bundesstaatlicher Einheitsstaat»). This is neither necessary nor practical. If such a distinction were really to be made it would be more sensible to refer to the level of autonomy and of participation than to the origin of the federation.
269 Usually the member states of a federation have equal rights without regard to their economic or demographic strength (Macmahon L501/9). But there are deviations. We find them in multi-level federations (e.g. USSR and Yugoslavia) and especially in the organization of participation in central decision making (compare L264b/1301ss). The principle of equality is of some political importance but it is not conceptually essential.
270 Whether the member states themselves are federally organized is, again, not conceptually essential. It may even be said that one of the real-type characteristics of member states, which sets them apart from decentralized regions in unitary states, is the ability to change the boundaries and the political structures of local government. According to the classic doctrine as developed in the United States (Dillon's rule L315/123, L371/140; compare L264b/759) local governments are «mere tenants at the will of the legislature». Member states may be organized as unitary states. This usually leads to two conclusions. In the first place the central government in theory does not have the power to deal directly with local government (see L632/137s, L782/426,1162s, and especially Frenkel L264b/766ss, 866ss). In the second place, a member state is free in granting autonomy to local government. The first conclusion may be politically sound as a means of bolstering state autonomy and discouraging central circumvention of state governments. But it is wrong to assume that member state autonomy is not compatible with an original, constitutionally acknowledged, right to exist for local government. As to the second conclusion it fails to grasp that this is compatible with federation but hardly compatible with the fact that federations should ideally be an embodiment of the federal principle. This federal principle should be pervasive where its prerequisites are to be found at the local level. Or as the Swiss de Reynold put it (L654/105s): «It has been the mistake of our federalists to have always remained at the surface only of federalism, to have confounded federalism with cantonal sovereignty, to have never been able to apply federalism within the cantons themselves. Where has centralism really started from? Where has étatism begun? In the cantons and even in the communes; contamination and bad examples come from there.» Jefferson, too, perhaps the most eminent of early American constitutional thinkers, made local «elementary republics», «ward democracy» the cornerstone of his structure of government (see Arendt L32/249ss).
271 It is not always easy to establish, on the basis of abstract definitions, whether a given state is a federation. Duchacek has therefore proposed ten criteria (he does not weight them) for this purpose. All of them are illustrations of one or other elements of the definition of federation (L180/207): «The ten yardsticks of federalism (the first four aim at measuring the reality and 'indestructibility' of the union) that we will now use are:
1: Has the central authority exclusive control over diplomacy and defense as befits a nation-state in its relations with other nations? (Compare the relevant analysis in Frenkel L264b/1553ss.)
2: Is the federal union constitutionally immune against dissolution by secession?
3: Is the exercise of the central authority as it reaches all citizens directly independent of the individual approval and resources of the component units?
4: Who has the ultimate control over amendments to the federal constitution?
5: Are the component units immune to elimination of their identity (antedating or postdating the union) and authority?
6: Is the collective sharing in federal rule making adequately secured by equal representation of unequal units in a bicameral system? What are the constitutional provisions for collective sharing in the executive and judiciary rule implementation?
7: Are there two independent sets of courts, one interpreting and adjudicating the federal laws and the other the state laws? (Again, compare L264b/1066ss.)
8: Is there a judicial authority in the central authority but standing above the central authority and the component units to determine their respective rights?
9: Have the component units retained all the powers that the constitution has not given the central authority? And are these retained powers significant or marginal?
10: Is the territorial division of authority clear and unambiguous?
D.321 The Sovereignty Problem
♦ 272 «A federation is a unity composed of several states which continue to exist as sovereign states» (Schollenberger L726/4). - «It is obvious and uncontradicted by science that now even less than before (under the imperial constitution of 1871) there can be talk about some länder sovereignty» (Anschütz L29/41). - «The citizen faces sovereignty in the central government and in the member states together. They are sovereign both together» (Bäumlin L54/145). - «All federalisms, to the extent that they oppose absolute concentration of power at the expense of territorial autonomy (and international organization), are doctrines of anti-sovereignty; but at the same time that they oppose sovereignty, they ascribe it to themselves, and is precisely the source of the oddity of federal theory» (Riley L637/6-3/18).
273 The discussion «carried on with much relish and little success» (Triepel L812/51) about the locus of sovereignty in federations is one of those debates that are kept alive by the fact that there are divergent views about their starting point and in this case about sovereignty itself. It strictly depends on its definition whether sovereignty is divisible or not and whether it is comprehensive or limited. It is therefore appropriate to arrive at some conceptual clarity about sovereignty and the scientific function of this term (no.274ss) before engaging a discussion about sovereignty and federation. It is my thesis that for the theory of government the concept of sovereignty is no longer functionally necessary and that it is therefore superfluous (no.287). Nevertheless, for the sake of completeness I shall also deal with the federal consequences of various doctrines of sovereignty (no.289ss).
D.3211 Sovereignty
♦ 274 «The historical origin of sovereignty is a political idea which was later consolidated as a legal one» (Jellinek L382/435).
275 The term sovereignty may be traced back to the medieval Latin «superanus» or «superanitas» and the old French «souverain»: one who takes precedence over others (L626/6). In Pufendorf's works we occasionally find the neologism «suverenitas» (L638/13*). The history of the concept goes back to antiquity (see L382/436ss). But its modern use is connected with the first real theory of sovereignty put forward in the 16th century by the Frenchman Bodin (L626/8). Hobbes in his Leviathan emphasized its absolutist core even more (L363/43).
276 Throughout the history of sovereignty two elements intertwine: the sovereign as holder of supreme power, and sovereignty as supreme power itself (compare L240/61,76, L626/11s, L644/41s, L825/86s).
277 Sovereignty as supreme power per se is a logical abstraction. It is a concept of legal methodology. Sovereignty in its legal meaning is comprehensive legal power, not delegated and thus supreme. Legal power which is not further deducible corresponds to the idea of the state (compare L109/417, L414/377, L500/272) or of the community of nations (see L413/287, L414/389). Whether it is the state that is sovereign, or the community of nations, or both, is a problem of legal theory (L413/284ss). In German public law doctrine sovereignty is usually tied in its abstract sense to the state. To be consistent this state must also be taken as an abstraction. (To describe a given state, e.g. Australia, as sovereign would point to its being the holder of supreme power, see below.)
278 This abstract concept of sovereignty was present in the teachings of the early theoreticians. But it took second place to the concept of the holder of supreme power. It had to take second place because the idea of sovereignty evolved as a political weapon, first in the conflict between church and state, and later in the struggle between absolutist monarchy and the estates of the realm. The focus, obviously, was on who was sovereign, who was to hold supreme power.
279 An interesting early abstract definition of sovereignty was given by the seventeenth century English politician Halifax (L329/135): «For instance, there can be no government without a Supreme Power; that power is not always in the same hands, it is in different shapes and dresses, but still, wherever it is lodged, it must be unlimited, it hath a jurisdiction over everything else, but it cannot have it above itself. Supreme Power can no more be limited than infinity can be measured, because it ceaseth to be the thing; its very being is dissolved when any bounds can be put to it.»
280 For the political conflict, abstract supreme power was of less consequence than its holder. This holder is the sovereign. Thus Hobbes (L363/17): «And in him consisteth the Essence of the Common-wealth; which (to define it,) is One Person, of whose Acts a great Multitude, by mutuall Covenants one with another, have made themselves every one the Author, to the end he may use the strength and means of them all, as he shall think expedient, for their Peace and Common Defence. And he that carryeth this Person, is called Soveraigne, and said to have Soveraigne Power; and every one besides, his Subject.» Or in the words of Pufendorf (L639/5/4): «For one to be sovereign ruler he not only must not have a superior, but all the others must obey him without demur or recourse.» This anticipates elements of the concept of sovereignty as put forward by Austin (based on Bentham L644/9), who was probably the legal thinker who had most influence on the development of Anglo-American law. For Austin the sovereign was this person or group of persons who as a rule are obeyed by the majority of the subjects and who as a rule do not have to obey a superior (L579/11, for a further discussion of this concept see Field L240/62ss, Raz L644/6ss). «He is sovereign who in a given political community is legitimated to exercise power or who in the final analysis does exercise it» (Loewenstein L493/4).
281 I have already intimated (no.278) that the historical bases for the doctrine of sovereignty were the struggles between church and state, absolutism and estates. These conflicts were described in detail by Jellinek (L382/441s). Two aspects proved to be of particular importance: the legitimation of the holder to exercise sovereignty as well as the unity and the unlimitable nature of sovereignty.
282 As to legitimation, medieval theory had the choice of two authorities. It was either possible to attribute imperial or princely dominion to divine will (for the Etrusco-Roman sources see L18/188). Or sovereignty could be based under natural right on the consent of the ruled («lex regia», Digests 1/4/1, L494/17s), that is on the social compact (no.051ss). It was this second approach which the French and English monarchomachs used in their conflict with absolutist monarchy after the latter's victory over Papal aspirations. Althusius was its prime apologist (compare Denzer L639/195ss, Duverger L186/37, Riley L637/6-3/32).
283 To be useful as a political weapon, sovereignty also had to be conceived as uniform and unlimitable in its exercise. It would not have been expedient for the instrument in this struggle for supremacy to contain restricting elements which might have limited one's own claims. «For the same reason that sovereignty is inalienable it is also indivisible» (Rousseau L678/2/2). This explains Harrington's (L340/17), Hobbes' (L363/29), Pufendorf's (L638/30*), Rousseau's and other thinkers' aversion to limitation of sovereignty and, especially, to the doctrine of the separation of powers. When the sovereignty of the people triumphed over monarchical sovereignty in the French Revolution, one absolutism was simply replaced by another one (compare Arendt L32/156, Paine L600/69).
284 For further constitutional developments the concept of sovereignty had become dysfunctional. Power was now exercised «in the name of the people» and no longer «in the name of the king». The problem of how to safeguard individual interests against state encroachment remained. This became the new focus for constitutional policy. The absolutist content the concept of sovereignty had assumed during the earlier struggles hindered the breakthrough of liberalism. «The dogma of sovereignty's indivisibility was one of the big obstacles to the evolution of the rule of law and constitutional doctrine» (Kägi L405/117). «The abstract recognition of the sovereignty of the people does in no way augment the sum of individual liberties; and if this sovereignty is given an extent it should not have liberty may be lost despite this principle or even because of this principle» (Constant L147/2). The liberal thinkers used two approaches to deal with sovereignty (both having their own traditions). They considered state sovereignty to be limited in its extent. It had to respect an area of individual liberty where even the sovereign majority had no right to intrude. The second course was the theory of the separation of powers. The exercise of sovereignty was to be functionally and personally distributed for the sake of guaranteeing individual liberty.
285 What function has the concept of sovereignty to-day? It is minimal for both constitutional and international law. In both cases it is pointed out that sovereignty in a sociological sense does not exist and has never existed. The omnicompetence of the state or its authorities is considered to be a legal fiction (compare L156a/128s, L373/18, L863/1448). I concur. Nevertheless, this does not dispose of the question whether the concept is in any way necessary for legal purposes.
286 If we look to public international law, we discern the ubiquity of sovereignty. But the term is never used to say something which in other words might not be said with more precision. Wherever it should decide a conflict this concept does not work, because the conflict is usually about matters that are at the same time essential conceptual elements. The famous controversy about the precedence of state or international law, for instance (see Kelsen L413/384ss, L414/203s,283), cannot be solved by pointing to the sovereignty of the one or other legal order. This would be circular reasoning because the result differs according to the definitions. Nor is it of any scientific value to call states, which are independent under the law of nations, sovereign. If one would deduce from this that non-sovereign states, e.g. member states in federations, have no status in public international law, this would again be circular reasoning using some custom-made definition of sovereignty. To say what it is supposed to say, the term «directly under the law of nations» is more precise and quite sufficient. And to paraphrase final jurisdiction in a given field, modern international law rightly prefers the concept of competence over sovereignty (L688/6). Sovereignty, because of its history, provokes erroneous associations in a world where final jurisdiction very often lies with international organizations. To sum up, for public international law the concept of sovereignty has become unnecessary if not misleading.
287 For constitutional law our findings are similar. It is true that the concept of sovereignty is less used here. But where it is used -as for instance in the discussion about federations, no.289ss- its usefulness is again limited by the fact that it leads to circular reasoning. Let us take, for instance, the problem of legislative delegation. This is treated differently in the United States on the one side and in the United Kingdom and Australia on the other. In the USA, the legislative power is conceived to be delegated to Congress and thus, under Anglo-American legal theory, there is no way to delegate it further. In the United Kingdom and Australia, it is an original function of parliament which may be delegated. To base these findings on the sovereignty or non-sovereignty of parliament is to ask whether the powers of parliament are to be considered original or derivative. The notion of sovereignty certainly does not add further clarification. When sovereignty was introduced into the system of law it led to conflicts about the very elements which should be universally accepted to make the concept useful. Because the necessary conceptual agreement is not likely to be found, and because such agreement is not necessary since there exist adequate terms to describe whatever sovereignty should stand for, it is advisable to dispense with the latter concept.
288 I fully agree with Portus who says (L626/18): «Can we not drop the term altogether from political discussions? The lawyers will not agree to drop it. It is too hallowed by ancient usage. It is embedded in too many documents. Very well. Let us leave it to the lawyers along with 'torts' and 'decrees nisi' and 'mandamus' and 'feme sole'. And let plain people like us stick to the less ambiguous word 'power'. For sovereignty is a concept which has been brought out of its original historical setting, and endued with metaphysical attributes of supremacy to law, irresponsible prerogative, and with almost divine right, to tease and bewilder us moderns with notions of national righteousness, ineffable privilege, and moral omnipotence.» Unfortunately, it is unlikely that sovereignty will disappear. Its very ambiguity makes it a near priceless asset for political debate.
D.3212 Sovereignty and Federation
289 To-day the use of sovereignty as a catchword is restricted mainly to two fields. In international relations it is invoked to ward off foreign interference. This aspect does not concern us here. But sovereignty is also used in dealing with federation and its institutions.
290 With the exception of Canada, the problem of the locus of sover eignty was a topic for political argument in all pre-1900 federations. The states were not readily willing to give up their sovereignty in favour of federation. That is why in the United States the concept of dual sovereignty was invented (L59/195, L96/422). The Swiss federal Constitution explicitly mentioned cantonal sovereignty (article 1). In the North German Union and the German Empire, the literature was unanimous in holding that sovereignty was vested in the totality of the united sovereigns and governments (L106/19ss, for the Weimar discussion see L632/53). In Germany the prestige aspects of the debate were especially evident, because there were monarchs or sovereigns at the head of both the Empire and the Länder (Schindler L715/9s, L343/61). Another solution than the one arrived at would have been a slight to the prestige of these rulers. (A similar discussion took place in 1849 in Austria, L214/30.) Dual sovereignty was a doctrine near at hand. In 1792 the U.S. Supreme Court declared in Chisholm v. Georgia (L626/12): «the United States are sovereign as to all the powers of government actually surrendered. Each State in the Union is sovereign as to all the powers not conferred.» And in 1819 the Court restated this theory in McCulloch v. Maryland (L626/12): «the powers of sovereignty are divided between the government of the Union and those of the States. They are each sovereign with respect to the objects committed to it, and neither is sovereign with respect to the objects committed to the other.» (As to dual sovereignty compare Bluntschli L75/577s, Dubs L179/28, Hamilton L334/32, Tocqueville L806/239s.)
291 Such a doctrine may be justified if sovereignty is taken to mean original powers of governance in a sociological sense . Legal originality is a more delicate question. In particular, the fact that powers may be taken away from a member state even against the latter's will makes legal originality difficult to construe. But the essence of sovereignty is not original power but supreme power. Supreme power is a logical abstraction (no.277). As such it is not divisible. That is why many scholars strongly repudiate the doctrine of dual sovereignty. Thus Kelsen (L415/117): «To prove... member states to be sovereign was not easily possible even for a theory fully committed to 'life's realities'. It therefore fell back upon... the grotesque idea to 'divide' sovereignty among central state and member states. This doctrine of dual sovereignty, too, hardly merits repudiation.» Or Giacometti (L244/40): «If the Confederation is sovereign, i.e. if it is the supreme government on Swiss territory, it is not logical to ascribe sovereignty to the Cantons, too, because sovereignty as pre-eminence is a concept of relations.»
292 If, to pursue this approach, one agrees with the dominant view that sovereignty is an essential element in the definition of the state (although Anschütz L29/39, Jellinek L382/769s, Veiter L829/149 and others differ on this) then Giacometti's opinion necessarily leads to the rejection of the state quality of member states (L244/43). «If the Cantons therefore are not states even if they show the essential elements of a territorial government, they can only be intrastate governments, that is to say a certain level of decentralization by self-managing units. In other words they are self-managing units of the Confederation just as the local governments are the self-managing units of the Cantons. The Cantons are the Confederation's local governments.... It is also possible ('because of their qualified functions and autonomies, M.F.') to call them magnified self-managing units» (L244/44s). Federation (the Swiss Constitution uses the term confederation), thus, appears as a highly decentralized unitary state (L244/45, also Merkl L435/92). «It even seems to appear that a federation in the sense of a state composed of states is conceptually impossible» (Giacometti L244/45; as to the rejection of an independent concept of federation see L825/354).
293 This disapprobation of member states' sovereignty and state character corresponds to the so-called monistic theory of federation. A federation is nothing else than a special form of a decentralized unitary state (L435/92, compare L541/299). An illustration is Fleiner's definition of federalism (L782/1028): «Federalism is the renunciation of the supreme government to fully enforce its sovereignty as to decision, control, and implementation in favour of the autonomy of territorially circumscribed units.» (To-day Fleiner seems to lean towards divided sovereignty, L365/88s.)
294 The dualist doctrine of federation acknowledges the co-existence of the central state and the member states. The theoretical foundations are varied. They may be based on dual sovereignty (no.290), on the recognition of some special legal position as for instance non-sovereign statehood of member states (no.292), or on the attribution of full sovereignty together with functionally divided governance to both central and member states (compare L63/22, L339/23, L435/75, L751/8, L811/72s).
295 The trinomial theory of federation has been expounded most consistently by Kelsen. He starts from indivisible, abstract sovereignty. In federations sovereignty lies with the legal system as a whole («Gesamtrechtsordnung»). This consists of two co-ordinate partial legal systems («Teilrechtsordnungen»): central state and member states. «If the state is no longer personified but seen to be a legal order, or its personification, then both central state and member states are decentralized legal orders. Thus, based... on a global constitution attributing and delegating functional jurisdictions there are two kinds of partial orders: one partial order with spatial jurisdiction for the whole territory ('partial order' because only competent for specified fields), and several partial orders with spatial jurisdiction for partial territories. The first partial order is the 'central state', or (in confederations) the union. The others are the member or individual states. If we analyse the relations between these orders we will find that there is no subordination (because there is no delegation) between the central state and the members states. There is rather coordination» (L412/135, compare L833/6/57).
296 The trinomial theory of federation had its precursors in Haenel, Max Huber (L557/44) and Austin. Haenel (L435/113): «Not the member states, not the central states are states pure and simple, they are but political communities organized and acting in the way of states. The true state is only the federation as the totality of both.» Austin (L557/144): «the sovereignty of each of the united societies, and also of the larger society arising from the union of all, resides in the united government as forming one aggregate body».
297 The American Calhoun started with a concept of personalized sovereignty. He construed a middle form between the dualist and the trinomial concept of federation by making a distinction between the constituent and the legislative power. Only the former was sovereign to Calhoun, and it was an attribution of the people in the states. The state people, then, delegated it to the states as well as to the federal legislators (L541/279ss).
298 The trinomial theory of federation has been adopted by various scholars (e.g. Harbich L339/85, Loebenstein L487/828, Maunz L525/20/I/5, Monz L554/38), by Mexican constitutional theory (212/94), and for a time also by the German Constitutional Court (L339/38s). Irrespective of the sovereignty question -here one can dispense with this concept- the trinomial theory is best able to construe a consistent legal order in such a way as to recognize satisfactorily the phenomenon of federation with its actual political structure and its states rights' clause (no.240). Critics maintain that the joint state does not exist (Burckhardt L109/162, Grawert L304/68, Huber L782/1159, Koja L435/126, Usteri L825/219). To this we may respond that the central government and the state governments are merely the organs of the joint state. It is true that there is not usually a written joint constitution (the Swiss Constitution of 1815 was such a text, L406/113) but there is a joint legal order which is presupposed by both federal and member state constitutions. (Organized joint state institutions may be found, too, as for instance the Loan Council in Australia and the provisions for changing Länder frontiers in Austria and for amending the Federal Constitution in Switzerland. But this is not theoretically essential.)
299 Foreign policy also provides a good illustration of trinomial federation. It is neither the central state nor the member states which are direct subjects of international law but the whole or joint state. This joint state is bound by the actions of its organs. These are usually the central state's authorities, but may sometimes be member states' authorities (see L264b/1608ss). The trinomial theory is the approach which can most easily be reconciled with the international law peculiarities of federations. It comes therefore as no surprise that the German Constitutional Court's application of the theory was in the context of the so-called «concordate lawsuit» when intrastate validity of an international treaty was under litigation.
300 Strictly, trinomial theory is also a monistic doctrine. The legal order is hierarchical (not necessarily with relation to its partial orders, see L264b/883s). Sovereignty as abstract supreme power lies with the joint legal order. (Whether it is only the joint legal order that enjoys state quality may be left unanswered.)
301 The trinomial concept of federation relegates the member states as well as the central state out of supremacy in favour of an encompassing joint state. The dualist (based on non-sovereign units) and the monistic theories, on the other hand, attribute pre-eminence to the central state, and this not only in the sense of legal precedence. It is obvious that an author's preference for one or the other doctrine often runs parallel to his more or less centralizing bias. Öhlinger (L587/14s) for instance observes with relation to most of to-day's representatives of the so-called pure theory of law that «the vehemence with which this thesis (of the decentralized unitary state) is proffered will of necessity antagonize those who... make a qualitative difference between unitary states and federations. On the same level is the constant disqualification of the länder (member states) as non-states, even if this can only be a nominal definition in the context of pure theory of law. Here indeed a certain... anti-federalist emotion seems to come out» (against a similar reproach see Koja L435/66). Again, it is obvious that those who nurture an aversion to centralization will be inclined to construe their concept of federation in such a way as to ascribe to member states sovereignty, autonomy, statehood, or at least no qualitative discrimination (compare Riley L637/6-3/41). Although I am convinced of my own objectivity I am, perhaps, no exception.
D.33 Federation and Confederation
302 The distinction between federation and confederation has been of scholarly interest mainly in Germany in the nineteenth and the beginning of the twentieth centuries (Jellinek L382/762). Here and in continental European parlance in general the distinction is still quite well observed, even if a clear-cut scientific differentiation is no longer of much concern for to-day's authors. In Anglo-American theory, the distinction is not considered to be of much importance. The generic term federalism is used where a German or a Frenchman would specifically talk about federation as a form of government. Thus, Duchacek's yardsticks for federalism (no.271) apply to federations only. They would be virtually meaningless in the context of either federalism at large or confederation. However, the difference between confederation or confederacy on the one hand and federation or federal state on the other one is also known (see L863/1375). The corresponding terms in French are «confédération» and «fédération». Until the nineteenth century, both words had been used rather indiscriminately (it is interesting to note that the medieval prefix «con» once emphasized unity, thereby reversing modern usage, Voyenne L837/61s). The Swiss federation even to-day is styled Confédération in French and not Fédération (article 1 of the actual Federal Constitution of 1874, «Confederazione» in Italian; for Canada see L859/23).
303 The essential elements of the differentiation are the legal basis of the union, the relations between central authorities and the citizens, and the status of the member states in international law.
304 A confederation is based on a treaty between independent states. This treaty remains the foundation of the union's activities. Amendments to at least its more important provisions have to be accepted by all the members. The latter have in principle the right to leave the association (the right to secession). A federation's base, on the other hand, is a constitution. This may originally have been arrived at by treaty. But it is something new, a new public will, independent from the treaty makers' will. The constitution may be amended according to its own provisions for revision. The members have in principle no right to secede. The main practical sources of the differentiation are the basic text's amendment procedure and the members' right to secession.
305 However, it is questionable to say that the members of a fed eration have no right to secede and to take this as a major characteristic differentiating a federation from a confederation. Some federations, for example former USSR (article 72 of the Constitution) and Burma under its first Constitution (L180/297) allow the member states - at least in theory - to leave the union. Kelsen (L415/224) considers such a faculty to be fundamentally admissible. Some authors deduce the right to secession from the original treaty base even where it is not mentioned in the Constitution (see, for Austria, Pernthaler L616/134). If, with Bryce (L96/321), we explain the silence of the federal constitution on this matter as being due to the results of a compromise which was not to be burdened with certain delicate questions, then the absence of the right to secession is not a particularly distinctive mark of federation. (As to secession compare Frenkel L264b/732ss.)
306 The amendment procedure is probably a more decisive difference (thus Durand L185/167). But, again, we must not forget that a federation is able to make amendments to its constitution dependent on the agreement of all member states. This was for instance the case in Yugoslavia (article 402 of the Constitution). It makes little sense to describe as a confederation a state which in all other respects is fully integrated, simply because of this one element.
307 Federation differs «from confederation in its considerably greater number of central competencies» (Kelsen L415/207, also Dubs L179/5). But as it is impossible to state the exact degree of functions necessary to transform confederation into federation, this observation does not yield a criterion for differentiation.
308 The second characteristic mentioned above (no.303) is the central authorities' relations with the citizens. The subjects of a confederation's directions are usually the member states and not their citizens. To enforce its decisions it has to rely on the member states and not on its own authorities. In federations, central laws and decisions are directly binding on individual citizens. They may be enforced by central authorities. In Tocqueville's words (L806/225): «In America, the union has as subjects not states but ordinary citizens.» It is true, that this question played a major role in the discussions of the founding fathers of the American Constitution (e.g. Hamilton L334/15s, Morris L232/34). But, again, it is also conceivable that the treaty which establishes a confederation may give its authorities power to deal directly with the citizens (compare Hughes L370/6s). This is increasingly the case in modern, economically oriented confederations.
309 The criterion most used to distinguish between federation and confederation is the international status of the member states. If after union they remain independent states under the law of nations, they form a confederation. A federation exists if there is no such independence or if it is restricted to some specific matters (for details see Frenkel L264b/1574ss). Scholars usually give expression to this distinction by pointing to the sovereignty or non-sovereignty of member states (see L343/44s, L382/767, L542/72). But as has been shown above (no.290ss), it is neither necessary nor expedient to refer to sovereignty in this context. Nevertheless, the international law status of member states is probably the most suitable test for differentiating between the two types of association of states (it is also the basis of the definitions given here, nos.247 and 254), notwithstanding the fact that formally two of the USSR's republics (the Ukraine and Byelorussia, L771a/7s) are treated by the United Nations and some other transnational organizations as full and direct participants in international relations. It has to be acknowledged that this criterion also is not unambiguous. It comes close to circular reasoning because the distinction between a federation and a confederation is of some practical importance in determining international law status.
310 To use the internal or international law character of legal relations between the member states as the basis for the distinction (thus Wildhaber L862/256) is, in the final analysis, also a reference to the law of nations.
311 Whether we are dealing with a confederation or with a federation is usually not very difficult to decide in real life. Despite a common view to the contrary (e.g. L277/52), the concepts are difficult to distinguish only in theory. In reality an association is clearly one regime or the other. The distinction is more clear-cut than the one between a federation and a unitary state (compare Burckhardt L109/154, Hughes L370/6). A difficulty arises only in connection with the consolidation of certain transnational organizations (no.286). And it is precisely in this case that we must question the value of a sharp theoretical distinction based on the accidents of a specific historical evolution.
D.34 A Bevy of Federalisms
312 Federalism is a widely used term. As long as we are concerned only with structural and general questions, its meaning is more or less clear. But, as is the case with all concepts which contain an element of judgment, it is sometimes difficult to ascribe a meaning to borderline cases. In such instances it is customary to use epithets. With federalism, likewise, this is usual. Unfortunately, however, these hyphenated or qualified federalisms have not led to greater clarity (thus Landau L449/176). On the contrary, they have multiplied borderline cases. Sometimes linguistic usage is further complicated by including in the definition some goal or criticism ascribed to federalism. Such differentiations usually take the form of adjectives or of some other addition. They do not really inform us about the matter at hand - creativity, co-operation, rigidity, competition, etc. may be found everywhere - but indicate an author's basic opinion of his subject. «Co-operative» federalists for instance have little regard for federalism that is «old-style» and «dual». There is a danger that emotionally attractive adjectives may be used to confer pseudo-objectivity on a political programme. «Co-operation under some conditions is equivalent to collusion and may be detrimental to the aggregate welfare of the persons involved. The choice of value-laden words should not be used to prejudge findings and conclusions» (Ostrom L594/17).
313 Practically every text on federalism contains hyphenated or qualified federalisms. I have amused myself collecting them, and so far have found some 460 (compare Davis L158/204ss, Frenkel L255/725ss). The following list is by no means complete, notwithstanding its impressive length. It would be easy to extend it at will by self-made terms and catchwords taken from the daily press (rate federalism, censorship federalism, witch federalism, etc.).
314 A comprehensive list has been compiled by Stewart (L637/12-2/5ss). He divides his 326 federalisms into twelve categories. Although the source and the explanation are only given for a few of them (this was not the author's aim) I have included in my own list those of Stewart's terms I was not familiar with.
315 After presenting this catalogue of federalisms, I shall analyse in some detail the co-operative and creative varieties. These concepts nowadays dominate the «market» of hyphenated federalisms. Unlike most of the others, they are more than mere curiosities.
D.341 The Alphabet of Federalisms
316 Federalism à-la-carte: confederational character, restricted to certain areas (Béaudoin L56/13)
Abstract f.: purely analytic justification (Duclos L183/173s)
Achaean f.: (L637/12-2/23)
Active f.: ready for co-operation (Tschäni L578/1965/254)
Adaptive f.: title of a discussion of criteria for the allocation of functions in federations (L13)
Adjectival f.: see hyphenated f. (Stewart L637/12-2/9)
Administrative f.: see executive f.; term for f. of execution (L289/171, L423/56); contractually based (Wiltshire L864); role of administrations in German f. (Ellwein L209/232); unilateral change of state boundaries by the central government (Kis L422/127); a special form of federal management of assistance relationships (Newton L873a/218ss)
Africanized f.: (L637/12-2/23)
F. by aggregation: (L637/12-2/23)
Althusian f.: (L637/12-2/23)
American f.: republican federation (Diamond L169/134)
Amphictyonic f.: close union of social units (Sidjanski L748/preface)
Amplified f.: inclusion of cities into the system (Martin L513/171,173)
Analogical f.: see hyphenated f. (Stewart L637/12-2/10)
Anarchical, anarchist f.: Proudhon's f. based on the autonomy of the smallest possible units (L48a/14s, L172/2450, L637/12-2/23)
Antagonistic f.: (L637/12-2/21)
Anti-vacuum f.: national priorities are enforced by national intervention because the states deliberately remain inactive (Shuman L746/12)
Antiquated f.: (L637/12-2/24)
Areal f.: related to subject matters (Reeves/Glendening L647/135)
Aristocratic f.: pre-revolutionary France, especially before the «Fronde» movement (Voyenne L837/165s)
Articulated f.: see hyphenated f. (Stewart L637/12-2/10)
Asymmetrical f.: a federation composed of clearly distinguishable member units (Tarlton L173/16, L797/83); a form of association of states other than confederation and federation, as for instance Switzerland/Liechtenstein (Stevens L637/7-4/177ss)
Authentic f.: confederation (L169/135s); USSR f. (L422/77)
317 Balanced f.: financially balanced system (L789/966); pluralist distribution of power (L141/9); the antithesis of hegemonial f. (Harbich L339/152)
Bamboo fence f.: (L637/12-2/21)
Bargaining f.: political bargaining between levels of government about public services (Rosenthal L672/10, Fras L272/61s)
Biblical f.: (L637/12-2/23)
Blue-Light f.: (L637/12-2/24)
Boston f.: (L637/12-2/23)
Bourgeois f.: non-marxist f. in Soviet perspective (L422/77)
Buber's f.: (L637/12-2/23)
Bureaucratic f.: administrative co-operation without political control (L46/28)
318 Calhounian f.: see no.297 (L637/12-2/23)
Capitalistic f.: (L637/12-2/23)
Categorical f.: (L637/12-2/22)
Centralized f.: opposite of peripheral f. (Dikshit L173/231, Riker L660/5, L449/178)
Centralist f.: its end is maximum integration (King L419a/24ss)
Centralized f.: (L637/12-2/23)
Centralizing f.: (L637/12-2/23)
Centrifugal f.: aims for greater independence (Ermacora L214/22s)
Centripetal f.: aims for greater unity, e.g. in Europe (Ermacora L214/22); corresponds to creative f. (Desmond L746/12)
Channelled f.: see focused f. (Wright L872/5)
City f.: (L637/12-2/22)
Civil f.: relations between citizens (Saint-Just L690/144)
Classical f.: (L158/204)
Cliental f.: (L637/12-2/22)
Club f.: (L637/12-2/22)
Coerced f.: (L637/12-2/21)
Coercive f.: central direction (Cranston L153/ 128, Else-Mitchell L210/109, L519/xx,xxiv)
Coinage f.: Austrian pre-modern standards of coinage (Informationsblatt des Institutes für Föderalismusforschung 1/84)
Collective f.: see communitarian f. (Djordjevi? L175/4)
Colonial f.: as set up by European colonial powers, e.g. Afrique Occidentale Française (AOF), Afrique Equatoriale Française (AEF, Neuberger L205b/171)
Combined f.: see expanded f. (Djordjevi? L176/11)
Commercial f.: (L637/12-2/22)
Commonwealth f.: (L637/12-2/22)
Communal f.: Yugoslav f. (Dunn L637/12-2/19)
Communitarian f.: based on national and cultural pluralism (Djordjevi? L176/4); co-operative f. (Esterbauer L221/28)
Community defining f.: communal boundary redefinition through state or federal help/intervention (Hart L342/155s)
Compact f.: self-co-ordination (Schneider L721/30)
Comparative f.: scientific discipline
Compartmental f.: separation of levels of government (Sharp L535/287)
Competitive f.: emphasis on its competitive function (L594/17); mutual animosity and distrust amongst levels of government (Gordon L637/4-1/45); economic acceptance of f. as a system of competing public corporations (Frey L271/2s)
Complete f.: (L637/12-2/23)
Compromise f.: see unachieved f. (Djordjevi? L176/10)
F. by compulsion: (L637/12-2/21)
Concentrated f.: (L637/12-2/23)
Confederalism: (L637/12-2/23)
F. via confederation: (L637/12-2/23)
Confederational f.: see external f.
Confrontational f.: Canada 1968-1972, lack of intergovernmental co-operation (McWhinney L534/1)
Congregational f.: (L637/12-2/22)
Congressional f.: synonymous word for dysfunctional f. (Walker L842/57)
Consensual f.: (L637/12/2)
Conservative f.: see defensive f. (Bieri L68/89)
Constitutional f.: opposite of bureaucratic f. (L46/28); based on constitutional allocation of responsibilities (Duchacek L182/132)
Constructive f.: method of motivating Cantons to conclude compacts (Stranner Basler Zeitung 12.5.1982)
Consultative f.: (L637/12-2/21)
Consumer f.: (L637/12-2/22)
Contemporary f.: (L637/12-2/24)
Continental f.: (L637/12-2/22)
Contract f.: (Duclos L183/188, L449/173, Price L645/23)
Contractual creative f.: (L637/12-2/22)
Conventional f.: see old-style f. (Reagan L646/3)
Co-operative f.: see no.346ss
Co-optative f.: (L637/12-2/21)
Co-ordinate f.: opposite of co-operative f. (Mathews L521/2s)
Co-ordinated f.: see co-operative f. (Inns L388/39); also opposite of co-operative f. (Sawer L698/30); reorganization of federal relations and responsibilities (Schwartz L735/55)
Co-ordinative f.: co-operation and shared responsibility (Mathews L519/xx, L153/121, L864/x)
Corporate f.: (L158/204, L637/12-2/22)
Corporatist f.: based on organized (e.g. professional) interest groups (Brugmans L94/62, L545/123); voluntary union of nationalities as in Estonia after World War I (Friedrich L275/3)
Cosmopolitan f.: (L637/12-2/22)
Counterfeit f.: (L637/12-2/24)
Creative f.: see no.354ss
Creeping f.: grants to promote federal programmes in member states (L162/24, 84)
Criminal-justice f.: (L637/12-2/22)
Cultural f.: pluricultural co-operation (Lang L451/341); «another name for the fellowship of man in America» (Boorstin L79/60); see nationality f. (L422/37s)
319 Dead f.: (L637/12-2/24)
Decadent f.: (L637/12-2/24)
Decentralist f.: the emphasis is on the regime's national character with new forms of decentralization (Diamond L169/136); emphasis on the autonomy of smaller units (King L419a/39ss)
F. by decentralization: (L637/12-2/23)
Decentralized f.: see competitive f. (L271/2s.); also L449/178
Decentralized administrative f.: Swiss form of co-operative f. (Veiter L829/195)
Defensive f.: no consensus on national goals (Bieri L68/89); defence of state autonomy (Siegrist L751/138)
Degenerate f.: see particularist f. (Lang L451/34)
Delegational f.: the member states' autonomy is based on central delegation (L271/3)
Democratic f.: (L637/12-2/23)
Destructive f.: (L637/12-2/21)
Developing f.: (L637/12-2/23)
F. by devolution: (L637/12-2/23)
Dictatorial f.: (l637/12-2/23)
Diplomatic f.: confederation (Duclos L183/186, L184/43)
Direct f.: central programmes do not provide an option for the member states to join or not to join (Macmahon L501/109); direct central government - local government relations (Reagan L646/171)
F. of divided responsibility: see dual f. (Halstenberg L106/139)
Division of power f.: programmatical opposite to co-operative f. (L221/28)
Documentary f.: (L637/12-2/23)
Domestic f.: (L637/12-2/22)
Dominant f.: (L637/12-2/21)
Double f.: see dual f. (Scheuner L194/42)
Drawing-board f.: technocratic aspects of German and other territorial reforms (Frenkel Vaterland 28.7.1979)
Dual f.: strict separation of tasks and competition between central and state governments (Ehringhaus L194/42, L449/178); emphasis on separation of tasks and mutual independence (Elazar L198/20, L162/30)
Dualist f.: see no.294
Dynamic f.: (L449/173, L637/12-2/21), no.345
Dynastic f.: historical phase of German f. (Dennewitz L161/24)
Dysfunctional f.: resulting from centralist and conservative trends in the United States (Walker L842/53)
320 Ecclesiastical f.: (L637/12-2/22)
Economically optimal f.: (Theiler L799/101s)
Ecumenical f.: (L637/12-2/22)
Effective f.: efficient (L637/12-2/16)
Eighteenth-century f.: (L637/12-2/24)
Emergent f.: (L637/12-2/23)
Entranced f.: (Krüger L833/21/105)
Entrenched f.: f. with old social roots (Haqqi L338/11)
Epistomological f.: (L637/12-2/24)
Equal f.: (L637/12-2/23)
Equilibrated f.: see balanced f. (Lerche L339/152)
Essex f.: (L637/12-2/24)
Ethical f.: emphasis on standards of proper conduct (L637/12-2/14)
Ethnic f.: ethnic factors as the basic determinant (Esterbauer L205b/146s, Héraud L220/75s)
Ethno-territorial f. (integral or partial): correspondence of language to territory (Duchacek L637/7-4/15)
Euro-African F.: quasi-federal links, i.e. French Union of the Fourth Republic, French Communauté of Gaullist France (Neuberger L205b/171)
Evolving f.: see amplified f. (Martin L513/40)
Execution f.: increasing burden put on the Swiss Cantons by their execution of federal legislation (L288), no.647
Executive f.: emphasis on negotiations between the different levels' executives and administrations (L753/5, Smiley L757/330, L758/17, L136/101s)
Executive-legislative f.: central legislation, member state execution and judiciary (Merkl L180/270)
Expanded f.: association of a state with a federation (Djordjevi? L176/11); see amplified f.
Expansive f.: (L637/12-2/23)
Experimental f.: experimental evolution of international relations (Portus L626/19)
External f.: confederational f. (Duclos L183/186, Merle L540/5,11)
Extralegal f.: (L637/12-2/23)
Extreme f.: extensive non-centralization (Kock L430/37)
321 Façade f.: (L637/12-2/24)
Failed f.: (L637/12-2/24)
Federating f.: (L627/10)
Federational f.: by contrast with regionalism, member states participate in sovereignty (Esterbauer L221/22)
Feudal f.: stresses degree of autonomy of programme specialists and programmes (Wright L637/12-2/16)
Feudal-functional f.: see picket-fence f. (L871/24)
F. as fiction: (L637/12-2/24)
Figurative f.: see hyphenated f. (Stewart L637/12-2/7)
Financial f.: (L637/12-2/22)
First New f.: Jane Perry Clark's introduction of the term New f. (Wright/White L873a/1, compare no.341)
Fiscal f.: multilevel decentralization in economic theory of f. (Beer L58/22, Bieri L68/1, Musgrave L566/614, Oates L585/4)
Flowering f.: (L637/12-2/21)
Focused f.: US grants system 1945-1960 (Wright L872/5)
Formal f.: f. of territorial units as distinct from economic interests (Miller L548/8); f. as a form of government (Reagan/Sanzone L646/170)
Fourth New f.: President Reagan's attempt at restoration of theconstitutional role of state and local government (Wright/White L873a/22ss)
Fragmented f.: see picket-fence f. (Wright L872/5)
Free f.: (L637/12-2/22)
Full f.: (L637/12-2/23)
Functional f.: pluralism (von der Gablentz L843/11, Miller L548/1, L116/14, L180/51); vertical functional interlacing (Beer L58/31); the central government decides on the volume of expenditures, the member states decide on their expenditure priorities (Hanusch L36/108s)
Functioning f.: more than just formal (Dikshit L173/24)
Fused f.: opposite of dual f. (Morstein-Marx L194/42, L158/204)
Fused-foliated f.: U.S. f. 1958-1968 based on federal planning and grants (Wright L872/5)
322 Generic f.: large in contrast to old-fashioned f. (Dexter L167/6s)
Genuine f.: (L637/12-2/23)
Geographical f.: territorially circumscribed f. (L116/14)
Good f.: (L637/12-2/22)
Governmental f.: (L637/12-2/23)
Group f.: (L637/12-2/22)
Guild f.: (L637/12-2/22)
323 Hamiltonian f.: (L637/12-2/23)
Hand-in-hand f.: interdependent responsibilities, e.g. shared tasks (Bundesblatt der Schweizerischen Eidgenossenschaft 1971 volume II/1018)
Hegemonial f.: one member state, e.g. Prussia, is strong enough to determine union policy (Dennewitz L161/24, L339/152)
Hierarchical f.: contrast to co-operative f. (Ermacora L214/193)
High f.: (L637/12-2/24)
Higher f.: (L637/12-2/23)
Historical f.: (L637/12-2/23)
Horizontal f.: federal Umbrella for several African colonies and protectorates (e.g. Afrique Occidentale Française, Afrique Equatoriale Française, Neuberger L205b/171), intergovernmental co-operation (Starr L368/26)
Hybrid f.: composite structure (L637/12-2/14)
Hyphenated f.: f. with epithet (Landau L449/176)
324 Ideal f.: (L637/12-2/23)
Ideological f.: based on opportunism not on history (Lenin L422/76)
Imperfect f.: (L158/204)
Imperial f.: (L637/12-2/23)
Implicit f.: highly decentralized unitary constitution (Young L481/193)
Imposed f.: stage of German history (Dennewitz L161/24,52)
Improper f.: (L637/12-2/24)
Impure f.: (L637/12-2/24)
F. in a broad sense: classification
F. in a narrow sense: classification
Incipient f.: (L158/204)
Income f.: (L637/12-2/22)
Industrial f.: (L637/12-2/22)
Instituted f.: federation, confederation (Duclos L183/75)
Institutional f.: see organic f. (Scelle L700/13)
Instrumental f.: (L637/12-2/22)
Integral f.: general principle of organization based on personalism (Kinsky L220/43ss, L420/26s, L695a/8)
Integrated f.: see organic f. (Dikshit L173/9, L449/178)
Integrative f.: (L637/12-2/23)
Intercantonal f.: regional co-operation in Switzerland (Schumann L730/271)
Intercommunal f.: commune based European union (Duclos L183/147)
Interdependent f.: mutual interdependence of levels of government (L158/204, L255/727)
F. of interests: excessive representation of special interests (Stadler L782/1033)
Interlocked f.: see co-operative f. (L180/318)
Internal f.: contrast to external f. (Merle L540/12); federation (Duclos L183/77)
International f.: proposed European peace treaties, e.g. those of Kant and of St.Pierre (Riley L637/6-3/18); international collaboration in functionally defined areas (Friedrich L275/7); confederation (Duclos L183/77)
International law f.: (Dennewitz L161/24)
Interregional f.: co-operation of European regions (Duclos L183/147)
Interstate f.: confederation (Burdeau L810/2, Duclos L183/77)
Intertwined f.: see co-operative f. (L180/318)
Intolerant f.: (L637/12-2/22)
Intrastate f.: federation (Burdeau L810/2, Duclos L183/77)
325 Jeffersonian f.: (L637/12-2/23)
Joint f.: (L637/12-2/23)
Judicial f.: the American system of parallel state and federal judiciaries (Winkle L868/68)
Juridical f.: (L637/12-2/22)
326 Kant's f.: (L637/12-2/23)
Kingly f.: (L637/12-2/23)
327 Late f.: largely achieved integration (Zeh L708/138)
Latin American f.: (L637/12-2/23)
Layer-cake f.: emphasis on the separation of levels of government (McLean L637/12-2/9s, L872/5, compare no.246)
Legal f.: see normative f. (Scelle L700/13); legally constituted f. (Saint-Just L690/144)
Liberal f.: (L172/2450)
Limited f.: (L637/12-2/24)
Limited dual f.: U.S.: the Taney Court's construction of the commerce clause (L8/47s)
Lip-service f.: (L637/12-2/24)
Little f.: (L637/12-2/24)
Local f.: (L637/12-2/22)
Locational f.: concept based on locale (L637/12-2/10)
328 Macro-f.: centre/state relations (Elazar L202/19)
Marblecake f.: see no.246, emphasis on merged functions (Grodzins L310/3s, McLean L637/12-2/9s, L873a/2)
Mk.1 (etc.) f.: ironical technical analogy to describe stages of American evolution (Davis L158/74ss)
Marketplace f.: (no.516ss, L637/12-2/21)
Marshallian f.: (L637/12-2/23)
Maryland f.: (L637/12-2/23)
Matrix f.: (L637/12-2/22)
Mature f.: (L449/173)
Maximal f.: (L637/12-2/24)
Medieval corporative f.: (L637/12-2/24)
Medievalist, medieval f.: corporative and regional basis (Riley L637/6-3/18, L637/12-2/24)
Medievalist-constitutionalist f.: (L637/12-2/24)
Mercantile f.: (L637/12-2/22)
Metaphorical f.: see hyphenated f. (Stewart L637/12-2/7)
Methodological f.: (L637/12-2/24)
Metropolitan f.: (L637/12-2/22)
Micro-f.: city/region relations (Elazar L202/19)
Military f.: (L158/204, L637/12-2/22)
Minimal f.: (L637/12-2/24)
Modern f.: see co-operative f. (Hicks L358/18, Riley L637/6-3/18) and U.S. f.
Monarchical f.: (L158/204)
Monistic f.: increasing centralization (McWhinney L533/16s, L457/63s); see no.293
Monoethnic f.: (L637/12-2/22)
Montesquieuan f.: (L637/12-2/23)
Multi-dimensional f.: not restricted to internal state and national relations (Djordjevi? L175/7)
Multilevel f.: more than the usual three levels of government, e.g. USSR (Duchacek L181/35)
Multinational f.: (L637/12-2/22)
Mutual f.: based on reciprocity (L637/12-2/16)
Mutual interest f.: (L637/12-2/22)
329 Naive f.: clear constitutional allocation of responsibilities (Kock L430/52)
Naked f.: un-hyphenated f. (no.349)
Narrow-minded f.: egotistic f. (Frenkel L260/337)
Nation-centred f.: tracing back the U.S. Constitution to popular and not to state will (Haskell L12/2s)
National f.: functional expansion of grants (Clark L375/9, L449/178); opposite to dual f. (Nwabueze L583a/42)
Nationality f.: Austro-Marxist non-territorial f. (L422/37s)
Nationality-territorial f.: the Soviet Union (Tchistyakov L422/77s)
Necessary f.: human rights are enforced independently of the member states' will (Shuman L746/12)
Negative f.: against limiting member state autonomy (L165/293)
Neighborhood f.: (L637/12-2/22)
Neo-f.: claims more influence, more self-co-ordination, and more federal grants at the same time (Siegrist L782/572)
New f.: see no.339ss
New England f.: (L637/12-2/23)
New-modelled f.: (L637/12-2/24)
«New» new f.: (L449/173)
New-style f.: (L637/12-2/24)
New urban f.: (L637/12-2/22)
Newest of new federalisms: focuses on large cities (Sharkansky L744/1)
Nigerian f.: (L637/12-2/23)
Nineteenth century f.: opposite to co-operative f. (Sawer L698/30)
Nixonian new f.: see no.342 (L637/12-2/23)
Nominal f.: (L637/12-2/24)
Noncentralized f.: see no.226 (L637/12-2/23)
Non-ethnic f.: units are not ethnic groups (Héraud L220/75)
Non-federalism: (L637/12-2/24)
Non-territorial f.: segmented political autonomy (L481/43)
Normative f.: f. as differentiation of the legal order (Scelle L700/13)
Not-full f.: (L637/12-2/24)
Novanglian f.: (L637/12-2/23)
330 Old f.: (L637/12-2/24)
Old-fashioned f.: emphasis on constitutional equality of the member states (Dexter L167/6s)
Old-line f.: (L637/12-2/24)
Old-style f.: clear division of responsibilities and mutual independence of central and state governments (Mathews L519/xx, Reagan L646/170)
Old urban f.: (L637/12-2/22)
Oligarchic f.: (L158/204)
One-dimensional f.: restricted to territorialism (Djordjevi? L175/3)
Ontological f.: (L637/12-2/24)
Operational f.: emphasis on service character (Ostrom L591/1)
Operative f.: functioning f. (Wildavsky L861/95)
Optimal f.: balancing constitutional acceptability and stabilization efficiency (Kock L430/98); optimal territorial organization of supply and demand (Pawlowsky L607/133)
Organic f.: central direction and guaranteed autonomy (Sawer L173/9); see communitarian f. (Djordjevi? L176/4); f. as governmental differentiation (Scelle L700/13); naturally evolved sharing of responsibilities (Wiltshire L866/74)
Organizational f.: harbouring elements of disintegration and secession (Stalin L766/71,80)
Orthodox f.: (L158/204, L637/12-2/23)
331 Pacific f.: (L637/12-2/22)
Pan-f.: pervasive (L637/12-2/14)
Pan-African f.: (L637/12-2/23)
Partial f.: (L158/204)
Participatory f.: see polyvalent f. (L175/1)
Particularist f.: emphasis on divisive aspects (Dennewitz L161/101, Lang L451/34, L91/393)
Partnership f.: see marblecake f. (L637/6-4/164)
Paternalist f.: critical of Canadian central government role (Newfoundland speech from the throne 12.7.1979)
Perfect f.: (L158/204)
Peripheral, peripheralized f.: confederational f. (Dikshit L173/231, Neal L573/3, Riker L660/5)
Permissive f.: member state autonomy depends on the central government's concession (Reagan L646/157ss, L449/173)
Personalist f.: see integral f. (L637/12-2/22)
Petrified f.: creation of states in Germany (Frantz L632/157)
Philosophical f.: based on an objective a priori (Laforet L448/39s)
Picket-fence f.: U.S. governors have no control over federal grants and thus are mere bystanders (Wright L871/24); interests aim to overcome federal/state compartmentalization (L46/28); disinclination to co-operate (Tschäni L578/1965/254)
Piecemeal f.: (L637/12-2/21)
Pinwheel f.: (L637/12-2/21)
Pluralist f.: centrifugal tendencies (McWhinney L533/17, L534/1)
Pluridimensional f.: see polyvalent f. (Djordjevi? L176/24)
Polar f.: the Federalist's f. based on complementary institutions (Eulau L226/102)
Polis f.: see small republic f. (L169/136)
Political f.: federation (Duclos L183/186, L184/43, Lang L451/341, Mogi L550/1112); f. as a form of government (Saint-Just L690/144)
Political-constitutional f.: (Laufer L456/24)
Poly-ethnic f.: several ethnic groups as basis for federation (Duchacek L181/35)
Poly-national/poly-ethnic f.: USSR (Kis L422/64)
Polyvalent f.: social, functional and participatory f. (Djordjevi? L175/1, L176/24)
Positive f.: renaissance of local self-administration (L165/293)
Postf.: (L637/12-2/24)
Post-Civil War f.: (L637/12-2/24)
Post-colonial f.: African federations after the colonial period (Neuberger L205b/172ss)
Postmodern f.: (L637/12-2/24)
Posttechnocratic f.: turning away from large entities (Elazar L201/239)
Practical f.: (L637/12-2/22)
Pragmatic f.: ad hoc redistribution of responsibilities (L8/2, Reeves-Glendening L647/136)
Preceptive f.: textbook f. (Fras L272/62)
Premodern f.: (L637/12-2/24)
Primitive f.: (L637/12-2/24)
Private f.: see contract f. (Reagan L646/171)
Processual f.: (L637/12-2/22)
Procrustean f.: U.S. structural intervention (Nathan L569/125, compare L264b/866ss)
Professional f.: see co-operative f. (Aubert L37/402)
Professionalized f.: see technocratic f. (Beer L57/80)
Professorial f.: (L637/12-2/22)
Proletarian f.: (L637/12-2/23)
Proliferated f.: see fused-foliated f. (L872/5)
Proper f.: (L637/12-2/23)
Proprietary f.: (L637/12-2/22)
Protean f.: (L637/12-2/21)
Protof.: (L637/12-2/24)
Proudhonian f.: see integral f. (Kinsky L420/7)
Provincial f.: (L637/12-2/22)
Pseudo-f.: evaluation of co-operative f. (Aubert L37/402)
Public f.: (L637/12-2/22)
Pure f.: (L637/12-2/23)
Purported f.: (L637/12-2/24)
332 Quasi-f.: lack of a conceptually essential element (Wagner L838/29, Wheare L858/20); not political-constitutional (Laufer L456/24)
333 Racial f.: (L637/12-2/22)
Rainbow cake f.: see marblecake f. (L637/12-2/10)
Real f.: real differences of member units (Saint-Just L690/144)
Realtype-actual f.: (Theiler L799/117)
Reformed f.: (L637/12-2/24)
Regional f.: territorial element (Walper L843/11); interprovincial co-operation (Fras L272/26)
Regiopolitical f.: (L637/12-2/22)
Religious f.: (L637/12-2/22)
Renewed f.: the Canadian Trudeau government's 1978 constitutional reform proposals (L797/84s)
Reoriented f.: see community defining f. (Hart L342/153)
Representational f.: (Eulau L225/153, L226/91)
Republican f.: (L637/12-2/23)
Respectable f.: (L637/12-2/22)
Responsive f.: (L637/12-2/22)
Restrictive f.: (L637/12-2/22)
Restructured f.: the Canadian Pepin/Robarts constitutional reform proposals (L136/85ss, L796/81ss)
Resurgent f.: (L637/12-2/24)
Revised f.: (L637/12-2/24)
Revolutionary f.: Marc's integral f. (L507/7)
Rigid f.: (L637/12-2/22)
334 Second New f.: the 1960s f. (Wright/White L873a/6ss, no.347)
Scientific f.: self-assessment of Soviet f. (L422/77)
Self-managing f.: Yugoslav f. (Djordjevi? L176/47)
Semi-f.: unitary state with federal institutions, e.g. Belgium (Lijphart L481/97)
Separative, separatist f.: mutual compartmentalization, no co-operation (Kubel L165/276, Patzig L389/146, L637/12-2/22)
Service f.: (L637/12-2/22)
Sham f.: (L637/12-2/24)
Shape f.: (L637/12-2/22)
Shared, sharing f.: see co-operative f. (L158/201, L180/318)
Small-city f.: (L637/12-2/22)
Small republicf.: ancient city state confederations, authentic f. (Diamond L169/136)
Social f.: (L158/204, L175/1)
Socialist f.: USSR, Yugoslavia, Czechoslovakia (Djordjevi? L176/18); USSR Constitution article 70/1
Socio-cultural f.: as a contrast to political f. (Khan L418/1, Livingston L637/12-2/19)
Sociological f.: see non-territorial f. (L481/98s)
Southern f.: (L637/12-2/23)
Soviet f.: (L637/12-2/23)
Spaghetti f.: (L637/12-2/21)
Stable f.: (l545/122)
Stalwart f.: (L637/12-2/22)
State f.: (L637/12-2/23)
State-centered f.: (L637/12-2/23)
State-emancipated f.: political f. not based on a nation state (Duclos L183/183)
State-oriented f.: the U.S. Constitution is considered to be based on a treaty of the member states (Haskell L12/2)
States' rights f.: (L637/12-2/23)
Strict f.: narrow construction of federal responsibilities (Brady L535/409s)
Student f.: (L637/12-2/22)
F. sui generis: analytical classification of the USSR (Kis L422/1)
Supple f.: emphasis on flexibility (Clark Devoir 30.12.1977)
Supranational f.: (L637/12-2/23, Samsinger L695a/9)
Symbolic f.: emotional, symbolic rather than rational-constitutional foundation, e.g. Austria (Gerlich L775/214)
Symmetrical f.: federation comprising completely homogeneous units (Tarlton L173/15s)
335 Technocratic f.:centralizing influence of the increasing professionalization of the public sector, intergovernmental collaboration (Beer L57/79s)
Technological f.: (L637/12-2/22)
Territorial f.: as a contrast to social f. (Haqqi L338/11)
Theological f.: (L637/12-2/24)
Third New f.: President Nixon's new f. (Wright/White L873a/10s, no.343)
Three-way f.: see two-way f. (L513/109)
Tie-wig f.: (L637/12-2/24)
Total f.: (L158/204)
Totalitarian f.: (L637/12-2/23)
Town f.: (L637/12-2/23)
Trade union f.: (L637/12-2/22)
Traditional f.: vertical separation of powers (Bastien L52/47)
Transnational f.: (L637/12-2/23)
Tribal f.: German historical stage (Dennewitz L161/24)
Trinomial f.: see no.295ss
Tripartite f.: (L637/12-2/23)
True f.: emphasis on non-centralization as against decentralization (Elazar L198/xii); secured existence of member states (Haqqi L338/61)
Twentieth century f.: (Sharp L535/287)
Twin-stream f.: (L637/12-2/21)
Two-level f.: (L637/12-2/22)
Two-sided f.: see dual f. (Brecht L194/42)
Two-tiered f.: consisting of «primary» federations (AOF and AEF, see colonial f.) within the «secondary» federal French Union (Neuberger L205b/171s)
Two-way f.: without inclusion of the local government level (Martin L513/109)
336 Ultra-f.: particularist f. (Lang L451/34)
Unachieved f.: Commonwealth, Australia, Canada (Djordjevi? L176/10)
Unco-operative f.: the central government acts only if required to do so by the constitution (Shuman L746/12, differing: Belknap L637/12-2/25ss)
Uncreative f.: (L637/12-2/22)
Undeclared f.: Belgian evolution (Lorwin L492/146)
Unifying f.: (L637/12-2/22)
Unitary f.: see delegational f. (Frey L271/3)
Unstable f.: status and existence of the member states are at the mercy of central government (Barschel L49a/37, Thoma L304/158); outmanoeuvring the separation principle through co-operation (Grawert L304/158)
Unworkable f.: ironical assessment of criticisms of the actual system (Ostrom L594/7)
Urban f.: urban studies in a federal context (Mathews L522)
Utopian f.: (L637/12-2/23)
337 Vertical f.: see Euro-African f.
Viable f.: (L637/12-2/22)
Village f.: (L637/12-2/23)
338 Water taps f.: see focused f. (Wright L872/5)
Welfare co-operative f.: (L637/12-2/22)
F. without Washington (Canberra etc.): emphasis on member state autonomy (Ehringhaus L194/96, Prybutok L637/12-2/79)
Working f.: justification of intergovernmental structures (Gregg L305/77)
World f.: (L637/12-2/23)
D.342 New, Co-operative, and Creative Federalisms
D.3421 New Federalism
339 Every federal system is dynamic. The balance between centre and units, once achieved, does not remain static. It has to adapt itself continuously to new circumstances. The conceptual elements essential for federalism -autonomy and participation- remain, but their contents change. A «new» federalism emerges which in turn will again be supplanted by a newer one.
340 Thus directly after the foundation of the federation in 1848 the Swiss Baumgartner (L53/vi) wrote about this «new federalism» to distinguish it from the old confederation.
341 As a concept with programme or technical contents, new fed eralism plays a major role especially in the United States, where hyphenated federalisms abound anyway. At the beginning of the twentieth century new federalism stood for a judicial doctrine that arose out of the fact that there were certain functional areas where neither federal nor state governments were able to be active because some part of the necessary competences always belonged to the partner's jurisdiction. The courts handled this situation using two complementary approaches. They assumed tacit state acceptance of the relevant federal legislation. And they upheld federal laws removing federal legal obstacles to make way for state action. Federal legislation, for instance, made it possible for the states to prohibit the sale of lottery tickets although such sales enjoyed the protection of the Federal Constitution's interstate commerce clause (188 USS 221, 1903, cited in Strong L787/507s,513, compare Clark L133/539ss, Wright/White L873a/1s).
342 This use of the adjective «new» was never very widespread. It is only in connection with co-operative federalism that we hear the term often in the United States, mostly as a synonym (see Martin L513/37, L789/970). With Nixon's presidency (L516/6s) the term came to be used as a special form of co-operative federalism, emphasizing the states' and local governments' role. In an Address to the Nation Nixon called it (8.8.1969, L302/41): «A co-operative venture among governments at all levels... in which power, funds, and authority are channeled increasingly to those governments which are closest to the people.» Nathan, one of the architects of Nixon's new federalism programmes, considered its main characteristics to be decentralization, the use of existing state and local authorities instead of creating new ones, priority of unconditional revenue sharing or block grants in the place of categorical federal grants, and a retreat from federal intervention to change state and local political structures (L569/111s). Seen in this way, new federalism was an attempt to correct an overly centralizing and structurally interventionist (compare Frenkel L264b/866ss) character of many co-operative federalism programmes. The goal was a more balanced assessment of federalism (see also Fürst L281/22s).
343 Wright/White call Nixon's new federalism the Third New Federalism (after President Johnson's Second and before President Reagan's Fourth New Federalisms, L873a/10s,22s). There are still other new federalism concepts, but their use is more isolated. Thus the combination of individual freedom and rational state activity (Mogi L550/1112). Or new forms of interracial co-operation in the United States, between Blacktown and Whitetown (Ferry L461/29). In discussions I have encountered the word as a generic term for federal experiments in the third world. But mostly «new» is just meant to emphasize the originality of an author's thoughts on federalism (e.g. L116/1).
344 Next to the United States, the term new federalism seems to be most popular in Australia, which imported the word from the U.S.A. (L768/7). Here we find two different new federalisms. The first kind was used by the Whitlam Labor Government (1972 to 1975) to designate national priorities nationally and regionally implemented. It was a policy to circumvent state obstacles to the federal government's broad social reforms. The second type of new federalism, that of the Fraser Liberal-National Country Party Government (1975 to 1983) was not unlike the Nixon brand, stressing efficiency and clear divisions of responsibilities. It also favoured revenue sharing and general purpose grants relative to specific purpose grants (Sawer L698/23, Starr L768/24).
345 With the advent of the Hawke Labor Government in 1983 a new kind of federalism might need to be invented. Perhaps dynamic federalism should be considered! It is a nice progressive word which up to now has been surprisingly little used.
D.3422 Co-operative Federalism
346 The term «co-operative» federalism was first used in the United States in 1938 by Clark in connection with New Deal grant methods (Beer L57/74). After 1936, the American courts would no longer construe the Federal Constitution's due process clause (amendment XVI/1) as a narrow limit to member state economic regulations. At the same time they refused to entertain taxpayers' complaints based on the allegedly lacking constitutional basis for public grants (L138/28). The point of departure was, as with new federalism, the assumption of mutual federal/state agreement (U.S. v. Bekins 58 Sup.Ct. 811ss, 1938; see Strong L787/504ss). The concept assumed greater political importance in the early sixties, when the federal government started to pursue far-reaching social and structural reforms. At the same time it acquired a new meaning described below.
347 In the sixties, the term also began to be adopted in other federations (see Ehringhaus L194/28). Thus in 1963 in Canada, it was used to describe the federalism programme of Prime Minister Pearson (L463/409, L534/1, L753/172s). In the Federal Republic of Germany co-operative federalism was first mentioned in 1966 at a meeting of the Freiherr-vom-Stein Association (Veiter L829/195). Its real introduction into political discussion can be traced back to the so-called Troeger Report of 1966 (L438/20s, L165/296s). At about the same time the term also appeared in Switzerland (Reich L782/1079), whence it was taken over by Austria (Ermacora L214/193). The concept was also introduced into Australian (Inns L388/22, Sawer L698/30), Indian (Haqqi L338/49) and Yugoslav (Djordjevi? L176/22) discussions.
348 The new epithet did not remain unchallenged. One type of objection pointed to its catchword character. Another took exception to the uncritical transfer of a concept from a constitutional system marked by the bifurcation of federal and state administrations to federations where, as is the case in Western European ones, administrative interlocking of levels of government was always a fact (Ehringhaus L194/17). And even for the United States Elazar (L198/20) showed that, contrary to traditional belief, a patterned sharing of responsibilities had been characteristic of the early years of union. Some other critics found fault with the hazy, legally elusive content of this neologism. Reid (L388/4) called it a hurrah-term. To Deuerlein (L165/284) it was an «embellishment entailed by the times». And some other authors referred to the fact that the idea of co-operation is already part of the term «fœdus», federation (Frenkel L255/728). Saladin (L782/1079) calls co-operative federalism an «ugly pleonasm». Laufer (L458/164) and Lang (L451/220) criticized the term as a linguistically confusing tautology.
349 Nevertheless, the task remains to analyse the essential elements ascribed to the new term. As I do not consider it to be systemic like «naked» federalism but rather programmatical (see Frenkel L255/729), I shall start the exposition with the term's actual acceptances. We will see that co-operative federalism influences both elements of federalism, autonomy as well as participation.
350 In the foreground we have co-operation. Practically all the attempts at defining co-operative federalism stress this aspect. Thus administrative collaboration (L148/23, L162/31), vertical co-operation (Deuerlein L165/298), functional institutions of co-operation (Ehringhaus L194/165s), forms or institutions of collaboration in federations (Frenkel L255/279), forms of voluntary or federal law-based co-operation amongst member states or between them and the central government (Häfelin L325/572), partnership collaboration (Kunze L446/130), balanced co-operation (Loebenstein L487/870), voluntary collaboration (Mathews L519/xx), new forms of co-operation (Laufer L458/164), consultation (Richardson L659/65), techniques for persuasion, diplomacy, and consultation (Simeon L753/172), multilevel relationships and polycentric collaboration (Tiemann L804/325,333), etc. The co-operation referred to is that affecting federal/state relations (this is at the centre of Australian discussion, Sawer L698/30), or interstate co-operation. Thus those Swiss who advocate resistance against central encroachment talk about the necessity for increased horizontal collaboration through co-operative federalism. But the proponents of additional central intervention also rely on this concept. This ambiguity and its adoption by quite diverse ideologies is a result of the emotional appeal of the term (who is not for co-operation!) and its being already an element in the concept of federalism. It is indeed the case that political texts confirm Laufer's impression (L457/178) that co-operative federalism is represented as a magical formula for the justification of each and any form of collaboration, whatever its real contents.
351 The partners co-operating are usually the executives and the administrations. Only rarely, by giving or withholding consent, do parliaments and populations become involved. That is because only executives and administrations are specialized and small enough to be able to negotiate. This has the result that it is mainly technical concerns which are the subject matter of co-operative federalism. But there is also a correlative tendency to transform political problems which traditionally are judged from ideologically different viewpoints, into technical ones which thereby are rendered administratively negotiable (compare my more detailed exposition of federal forms of collaboration in L264b/14288ss and of the interlacing of polities in L264b/1058ss).
352 But co-operative federalism also influences the allocation of functions and thus member state autonomy. Not for nothing the roots of this kind of federalism have to do with federal grants (no.346, Sawer L698/30) and thus with the central government's ability to affect the member states' priorities by imposing more or less stringent conditions. It is, in the words of Sharman (L745/13s), an inevitable result of the central government's grant jurisdiction. There are few authors who emphasize the need for a more clearly defined allocation of responsibilities (e.g. Loebenstein L487/870, Pearson L753/172s). The tendency is rather for the extensive construction of federal jurisdiction and the gradual disappearance of constitutional limits to the central government's powers. Co-operative federalism is a pragmatical method for the flexible accomplishment of tasks (Ehringhaus L194/55). Compare also Smiley (L757/320), who stresses the piecemeal approach to the interpretation of constitutional responsibility provisions. The tasks are no longer separated. They are considered to be jointly allocated to central and state governments (compare Connally L789/619, Elazar L198/20, Martin L513/37s, Richardson L659/65). Co-operative federalism emphasizes «joint responsibility of governments for the welfare of those they jointly govern» (Ehringhaus L194/21). It reflects shifting power relationships instead of rigid independence (Birch L358/20). According to the strict construction of the federal constitution, the central government can intervene only when constitutionally mandated to do so. Further, federal law usually precedes state law (see Frenkel L264b/879ss) Therefore a broad construction primarily results in an expansion of central jurisdiction. The corresponding possibility of extended member state competence (as was originally the case with new federalism, no.341) remains mostly theoretical. Co-operative federalism thus has a centralizing tendency. This is stressed by many authors. It has been considered to be a justification for national usurpation (Baker L45/10) or, in the context of the problem of French culture in Canada, one of the century's most monumental trickeries (Léger L535/319s, L753/289).
353 Those federalists who emphasize horizontal collaboration arrive at somewhat different conclusions. They see less an expansion of central powers than a possibility of preventing or containing this through self-co-ordination (for a critical view: Frey L267/378). For them collaboration with the central government may help to give protection from further or excessive centralization. It is also centralization, but this centralization is milder because it better preserves member state participation and jurisdiction than would be the case with the traditional inclusion of additional tasks in the catalogue of central government responsibilities.
D.3423 Creative Federalism
354 Hyphenated federalisms seem to be like drugs: continuous use weakens their effect. New epithets replace old ones. Thus in the United States the term creative federalism supplanted co-operative federalism. The term was introduced into political parlance in 1962 by Governor Rockefeller from New York (L789/704). It gained greater acceptance in 1964 when President Johnson used it in Michigan in May (L162/83) and then in his famous Great Society address in June (L162/29). Creative federalism is not significantly different from co-operative federalism, but is rather a variant of the latter where the relative value of the constitutional allocation of functions becomes especially evident (compare Shuman L746/12 and Wright L746/77). The Great Society programmes had as their main target the overcoming of the social handicaps of the poor and mainly coloured urban population groups. Creative federalism, therefore, is mainly concerned with big city problems (slums and high infrastructure costs in the centres as compared with rich suburbs which do not share these costs, see L789/704) and associated federal/local relations (Ehringhaus L194/57).
355 It goes without saying that creative federalism has not escaped the criticism of being merely another political catchword (thus L449/178, L789/847). Its definition in political statements does sometimes indeed appear to be imprecise as when Vice-President Humphrey (L789/959) understood it to mean «effective and timely co-operation between all levels aiding, comforting, propping up weak, obsolete, ineffective state and local government.» However, creative federalism is not an empty formula. Its aims and its methods are obvious: social redistribution to be achieved through extensive federal welfare programmes which largely circumvent the traditional allocation of responsibilities and introduce a strong element of social and structural change.
356 Thus both creative and co-operative federalism are characterized by two elements: As to intergovernmental relations they stress co-operation and co-operative institutions in the place of autonomy. In some countries, direct federal/local relations are beginning to play a greater role. As to the allocation of responsibilities, the constitutional limitations on central government activities have largely lost their force, mainly because of the separation of the spending power from the constitutional distribution substantive powers and the possibility of imposing intergovernmental conditions on grants. In the United States there is an additional element of reshaping the rules of local political life to the advantage of underprivileged groups. This gives creative and co-operative federalism a special pluralist flavour (see Ehringhaus L194/190 and, for Germany, Hempel L352/221).
357 To a certain degree, increased collaboration and blurred jurisdictional limits are an almost inevitable consequence of the modern reinforcement of the interdependence that exists between levels of government in welfare states (compare Mathews L519/xx). Even without any stated programmatical goals, to-day's federalism is more co-operative than yesterday's. But, as has been shown, co-operative and creative federalism have further aims. The evaluation of these further aims is not easy because it depends in the final analysis on ideology.
358 Co-operative federalism (like creative federalism) is problem oriented. No problem must remain untackled just because the allocation of functions excessively delays solutions. Collaboration must be flexible so that all the demands the federation is faced with can be dealt with (Deuerlein L165/298, Martin L513/40, Wright L746/77). If a member state does not act it is the central government's right and duty to intervene (Shuman L746/12). At the same time, co-operative federalism makes possible a concentrated use of public means for the solution of problems, something especially necessary in times of limited financial and human resources (see the Troeger Report L438/20s). This problem oriented approach has obvious idealistic traits. The guiding star is social justice (e.g. Pearson L753/172s). With the introduction of new forms of participation for those who benefit from welfare programmes a democratization goal also becomes apparent (Ehringhaus L194/193).
359 On the other hand, co-operative federalism's main motivation is a technocratic one (thus Hempel L352/221). This is evident from its problem orientation and its emphasis on the concentration of resources. The idealistic traits are mingled with technocratic elements insofar as co-operative federalism seems to be based on an assumption that there are objectively right measures which can be discovered by the application of good intentions. These are deemed to be a matter of course in central administrative agencies and more likely to be found with the central parliament than with state ones. As to the democratization goal -which in this context is mainly a United States problem- it is necessary to remember that we are dealing with forms of participation which, being centrally established and based on idealistic motives, compete with traditional forms of local democracy. Finally, co-operative federalism's executive and administrative orientation is also technocratic in its tendency.
360 The characteristic mark of technocracy is the shift from private sector influence -citizens, political parties, lobbies- to public sector autonomy, that is to say to the authorities themselves and to their experts (compare Beer L58/31).
361 Co-operative federalism has numerous disadvantages. Without anticipating a detailed discussion of the functions of federalism (no.404ss), the conflicts between traditional and co-operative federalism may be briefly mentioned here: In the context of consumer orientation (no.501ss), co-operative federalism favours nationally defined interests to the disadvantage of local ones. As to efficiency (no.476ss), it entails a reduction in local control and a diffusion of political responsibility (see also L352/221, L454/12, L698/30s, L757/327s). Constitutional adaptation is very unsystematic or piecemeal (Smiley L757/320). Vertical division of powers (nos.412ss and 623ss) is bypassed by co-operative federalism's institutions (see Beer L58/33s). This is, to my mind, one of its more questionable aspects. The member units' innovative capacity (no.434ss) is attenuated because of the lesser challenge to their initiative. The nationalization of policies weakens federalism's mechanisms for conflict resolution (no.551ss) and integration (no.570ss), although this at least assists in resolving other conflicts and, sometimes, in integrating other groups. The concentration of financial and human resources may conflict with the crisis stability which results from federalism's many redundancies (no.541ss). Finally, we should consider whether a kind of co-operative federalism that makes collaboration too much of a value in itself may not some day endanger federalism's function of legitimizing government (no.592ss).
362 In parliamentary federations, co-operative federalism poses additional problems. Parliament's powers are in any case considerably weakened because of the executive's control over the voting of their majority parties in two or three party systems. Increased co-operation between federal and state executives adds to this reduction in parliamentary powers (compare Frenkel L264b/1483, Smiley L757/327s).
363 To say that co-operative federalism is technocratic and neglects some essential functions of federalism is not passing judgment. It is a statement of fact. In the final analysis we deal with ideology. The question whether it is national or local democracy that has a better claim to be the basis of government is perpetual and is always being renewed with fresh arguments. It is the question of federalism. The treatise presented here attempts to present criteria for evaluation. But the answer really depends mostly on the individual's own values and preconceptions. It cannot be scientifically proved. As to co-operative federalism, we may only assert with certainty that its programmes generally favour national over local democracy and executive authorities over legislative authorities and popular will. Under my conception of federalism, co-operative federalism is an inappropriate way to reform federalism. I do not disapprove of collaboration where it is appropriate. But I do object to its being elevated to programme status.
D.4 A Brief Survey of Federal History
364 Modern federation had its origin in the United States Constitution of 17 September 1787. But the idea of federalism is much older. The eighteenth century political thinkers were wont to look to Greek antiquity for their examples (e.g. Madison and Hamilton L334/18). However, the city leagues of Greece were not the ancient world's only forms of political federalism (thus Montesquieu L551/9/1ss).
365 The classic nineteenth century exposition of ancient Greek unions was the work of Freeman (L251). He described the Achaean, the Aetolic, the Boeotic, the Lykian, and several other leagues. Freeman sought to prove that true federations were already to be found among the Greeks (L251/202, but for a contrary view see Giovannini L294/93). He did not consider the so-called amphictyonic leagues, i.e. those organized around a common place of worship, as political federations (L251/97s). His examples were city state leagues with worldly goals, for instance defence. This, incidentally, is also the picture we get from reading Aristotle (L34/2/2). According to this reasoning the independent city states came first. They later formed associations for limited ends with common authorities.
366 But the Greek city leagues which are best documented date from the comparatively late Hellenistic period (Giovannini L294/9s). Modern comparative research also suggests that the basic pattern was not an aggregation of independent political communities but the gradual consolidation of tribal units as they changed from a nomadic to a settled style of life. The common organization of the tribal cult in a central place remained as the essential characteristic binding the community together in an increasingly loose form (compare Voyenne L837/29s). Such amphictyonic unions could also comprise common interests of a more or less secular nature (e.g. mutual assistance, intermarriage, trade, etc.).
367 It is interesting to note that such amphictyonic leagues existed all over the ancient world, as for instance in Israel (Bright L89/129ss, Noth L583/37ss, Schechter L710/5, Zuber L883/100s), Greece (Alföldi L18/8, Freeman L251/97ss) and Italy (Alföldi L18/10ss, Freeman L251/558). A conspicuous and still unexplained fact in this context is the ubiquity of the number six and, more especially, number twelve in the organization of the association's units. (Bright and Noth trace this back to the necessity for monthly or bimonthly change in the care of the common shrine; L89/147, L583/85s, compare Alföldi L18/168.)
368 We find similar leagues in more recent times in communities that had attained a comparable level of civilization as, for instance, the mounted herdsmen of Northern Asia in the eighteenth century (Alföldi L18/11), the tribally based feudal order of Germany under Henry I (Dennewitz L161/29s, Jerusalem L385/9), and the famous Iroquois League of the Five respectively Six Nations which remained intact right up to the twentieth century (Calhoun L119/72, Parker L602/45, L603/61).
369 If this theory of the history of ancient leagues is correct, it would provide better support for Sidjanski's contention, based on his perception of the constitutional history of Austria, Brazil, and Mexico that federal unions also result from disaggregation (L747/10, no.674). But the state of research does not yet permit a final answer to the problem of understanding the development of the ancient leagues.
370 Rome was also a member of an amphictyonic association, the Latin league. After having vanquished another such union, the Etruscan one, Rome then assumed control over the Latin league (compare Alföldi L18). But antiquity did not develop a theoretical concept of federalism (Friedrich L274/67), and there is no direct succession leading from the old Greek and Italian leagues to medieval ones (Freeman L251/557s, Mogi L550/24). The Roman Empire which medieval lawyers encountered in the Digests (the code of Justinian) was a universal and hierarchic realm without federal traits.
371 This hierarchical model was to be grafted on to the completely different structure of the medieval empire (compare Eulau L224/2s): «summa potestas» was ascribed to the emperor, the princes were to be «praeses provinciarum», the diet was compared to the senate, the free imperial cities were given the same status as the «municipia». Even the many leagues of the late Middle Ages were regarded as comparable to the simple «universitates» of doctors, artisans etc. Up to the end of the sixteenth century, the Corpus Iuris remained for jurists trained in Roman Law the only source for the legal construction of the Holy Roman Empire of the Germanic Nation. This was so even though the post-glossator Bartolus had laid, in the fourteenth century, the basis for a more realistic theory of government with his differentiation of «universitates quae superiorem recognoscunt» and «universitates quae superiorem non recognoscunt», that is associations which are subordinate to a higher authority and those which are not (see Mogi L550/25).
372 The realities of the medieval state corresponded more to old German law institutions and actual power relationships than to the rigid Roman law categories of the learned doctors. In addition, the medieval church developed not only the theory of papal autocracy but also a federal theology based on the «civitas dei» which influenced political thinking. The Christian Family, called thus because of its dynastic pillars, in its external role had to defend the Christian faith against the infidels while internally it had to safeguard the independence of its members (compare Bülck L104/3, Deuerlein L165/33, Reynold L654/228s).
373 With this as a background, several attempts were made to devise a theoretical division of responsibilities between the emperor and the princes. Thus the abbot Engelbert von Admont and Dante Alighieri, who both lived at the turn of the thirteenth to the fourteenth century, attributed to the emperor the realization of general well-being or the pursuit of common and higher interests. The states were to take care of the welfare of individual nations or of special interests (Deuerlein L165/23, Mogi L550/24). The Protestant university of Herborn was a stronghold of federal theology. This was the university Johannes Althusius joined in 1586. His contribution to federal doctrine was a theory of complete and incomplete federations.
374 There are different assessments of Althusius's role as a theoretician of federalism. For Djordjevi?, for instance, his grasp of federalism was much deeper than that of Montesquieu, Madison, or others (L175/3). For Riley, on the other hand, he was a late theoretician of the medieval state rather than a precursor of a federal doctrine, because Althusius's complete federation corresponded to full incorporation of one state into another whereas his incomplete federation involved any contractual relationship between states (L637/6-3/36ss).
375 Medieval history reveals a multitude of temporary and restricted alliances among states, with a few instances of relatively close and stable associations of states such as the German Empire (Jerusalem L385/9s), the Swiss Confederation (Hilty L360/8ss), the unions of Horodlo and Lublin between Poland and Lithuania (Wagner L383/15), and, later, the Netherlands in the sixteenth century (Boogman L78/5ss, Riker L660/9s, Voyenne L837/92s). At the end of the Middle Ages, in 1462, we then find King Georg Podiebrad of Bohemia's grand design for a European union, with common authorities and a common budget, which anticipated the still more encompassing federation project which Sully submitted some hundred and fifty years later to the French King Henry IV (Bernier L63/203).
376 Post-medieval legal science could no longer escape the task of elaborating a serviceable theory of federal associations of states. Besold was the first to formulate, in his 1614 thesis at the university of Tübingen, a new form of government with inferior states - «res publicae subalternae» or «imperia subalterna» - which did not enjoy full sovereignty and which were subject to a fully sovereign superior state. These inferior states had their own structure of government which the sovereign could not change. In their own territory they had the same rights as were enjoyed by the sovereign in the empire as a whole (L638/39*). In 1648 the peace treaty of Osnabrück resulted in the definitive legal recognition of the autonomy and participatory rights of the estates of the realm (article VIII para.1s, von Puttkamer L639a/27s). In 1661, Ludolph Hugo introduced the concept of a state made up of states as a means of describing the actual form of territorial organization in Germany (Deuerlein L165/39s). At about the same time, in 1667, Samuel von Pufendorf published his famous analysis of the German imperial constitution which showed that the empire, if evaluated in the terms of traditional concepts, would have to be called «an irregular body, similar to a monstrosity» (L639/6/9, L639/7/8, L638/28s*). Pufendorf distinguished ordinary states, several regular forms of associations of states which retained the sovereignty of the individual states, and, finally, irregular forms of association between the two other kinds of forms of government (Bresslau L638/30s*). (Other German federal plans and pacts are described by von Puttkamer L639a/43ss.)
377 The political impact of the German ideas was surpassed by Montesquieu. He described the federal republic as a constitution which combined all the internal virtues of the republic with the external force of monarchy. «This form of government is an arrangement whereby several political entities agree to become citizens of the greater state they wish to create. It is a society of societies that may be enlarged by new associates uniting themselves to it» (L551/9/1). Although the corresponding section of the «Esprit des Lois» was very short its influence was great (Voyenne L837/132), especially for the development of North American federalism. Imboden considers the federal idea to have been even more far-reaching in its effects than the doctrine of the separation of powers which made the book famous (L387/29, but compare Nelson L576/9ss).
378 Montesquieu proposed a primarily organizational justification for federalism (safeguarding individual liberty also played a certain role, L551/9/1). In addition to this the framers of the American Constitution were influenced by the personalistic ideas of the social compact (nos.051ss, 282; Schechter L711/1ss). Since Hooker's work in 1594, these had greatly affected English Protestant thinking (L367/1/10/1, Jellinek L382/206s). To the English monarchomachs and covenanters the agreement of the ruled was a necessary prerequisite for legitimate governance. As a consequence, this personalist element was introduced into the compact of union between the states. This was the case with the confederation of the New England colonies from 1643 to 1684 (Bennett L59/39s). But this confederation, like the first American Federal Constitution of 1777 (Bryce L96/20s), united only the colonies and not their populations.
379 In the debates which preceded the 1787 Constitution (compare L232-35), modern federation was for the first time conceived as a form of government uniting both member states and their populations. The Union had its own population which it could deal with directly and by which it could itself be influenced without state government involvement. To this double basis of governance corresponded the new Constitution's preamble: «We the People of the United States,...». The American founding fathers had thus invented the modern, democratic federation. This form of government was soon to conquer large parts of the Old as well as the New World. But with increasing consolidation and centralization, it was also to accentuate conflict between «big» national and «small» local democracy.
380 The American founding fathers were quite aware that they had created something new (compare Diamond L169/133). Patrick Henry for instance illustrated this quite graphically (L516/75): «We may be amused if we please, by a treatise of political anatomy. In the brain it is national: the stamina are federal - some limbs are federal, others national. The senators are voted for by the state legislatures, so far it is federal. Individuals choose the members of the first branch; here it is national. It is federal in its conferring general powers; but national in retaining them. It is not to be supported by the states - the pockets of individuals are to be searched for its maintenance. What signifies it to me, that you have the most curious anatomical description of it in its creation. To all the common purpose of legislation, it is a great consolidation of government.» Or Madison (L334/39): «The proposed Constitution... is, in strictness, neither national nor a federal Constitution, but a composition of both.»
381 The influence of the American discussion about human rights profoundly influenced the French Revolution. For federalism, this was not the case. Some French politicians harboured federal ideas, it is true, as for instance Mirabeau (L328/180) and some early Jacobins (Billaud de Varenne 1791, L837/205ss). (Curiously enough, the Girondins, who later were to be decried as federalists by the Jacobins when the term had become an invective, did not establish such a federal programme, Voyenne L837/234ss.) In some provinces federalist movements came into being, first on 1 August 1789 in Lons-le-Saulnier, followed by actual oaths of confederation on 2 November 1789 in Besançon and on 13 December 1789 at Montélimar (L837/185ss). But the revolution took other turns. In Paris the absolutist core of Rousseau's acceptance of democracy triumphed and it swept everything in its way in the provinces too (compare Carlyle L123/230, Nelson L576/20).
382 The Revolution practically conferred taboo status on the French conception of the unitary and indivisible republic. Even if the occasional statesman such as Clemenceau (L742/46) has favoured federalism, this taboo has remained to our own times. A later revolution, the Parisian commune of 1871, was federalist in essence (L32/266). But its suppression was a decisive defeat for federalism. Only since the beginning of the Mitterand Presidency in 1981 have there been some real, albeit rather timid, indications of a departure from extreme unitarianism.
383 Even if France was not to become a federation, French thinkers played a major role in developing the idea of federalism. Thus the liberal Benjamin Constant (L147/315s), thus the theoretician of anarchy Pierre Joseph Proudhon (L634/361s,507s, L635/150, Hahn L328/257s,271s), and thus especially Alexis de Tocqueville (L806s). French foreign policy also favoured the creation of federations in the rest of Europe, for obvious strategic reasons. Thus NapoleonI sponsored the German Union of the Rhine (L161/52,113) and the Swiss Mediation Constitution (L359/554ss).
384 Much later, after the Second World War, it was again France that among the victorious powers most favoured non-centralizing elements in the structure of the Federal Republic of Germany (Laufer L454/19). Less successful were France's attempts in 1946 to transform its colonial empire into a federal union or, in 1958, into a «Communauté» (L540/4).
385 The close relationship between federalism and idealistic anarchism has been well illustrated by Jean Paul (after Krüger L833/21/106): «Ideally, the small federal states that always give free laws to themselves would divide into federal villages. And these in turn would fall into federal individuals who at any minute might adopt a new legal code.»
386 In the nineteenth century the American example influenced German and Swiss federal developments primarily via the writings of Tocqueville. This has been described in some detail by Mogi (L550, compare Bluntschli L74/398). In Central and South America the concept of federation was more directly taken over from the United States (L408/33ss, L485/290ss, L608/viss, L830/151). The history of Canadian government which in 1867 adopted the first federal constitution of the Westminster type was more independent from the United States.
387 By the end of the nineteenth century, the following federations had been formed and they still exist to-day (even if in some of the Latin American ones federalism is very much attenuated, L173/163, L272/26ss, L830/150ss): Argentina, Brazil, Canada, Germany, Mexico, Switzerland, the United States of America, and Venezuela. Shortly afterwards, in 1901, the Australian colonies federated (after an experiment with confederation from 1885 to 1895 which should have included New Zealand and Fiji, L176a/113ss, L343/57). With the exception of some short-lived Central and South American federations, the federations founded in the nineteenth century proved to be stable. The same observation holds true for one of the two that came into being between the two world wars: Austria (the other one, the Soviet Union, partially lives on in Russia).
388 If Riker called the twentieth century the true era of federalism (L660/2, similarly Brugmans L94/13), if Elazar discerned a «federal revolution» (L202/2, L204/15, 205b/3ss,49ss), and if McWhinney perceives federalism to be a popular cure for the most diverse kinds of political ills (L533/92), these statements are concerned with the period after the Second World War. Indeed, since 1945 federations have multiplied. But most have disappeared again after a very brief existence. Of the new federations, those which still exist are Belgium, the Comores, India, Malaysia, Nigeria, Pakistan, Tanzania, the United Arab Emirates and, in a very much reduced state, Yugoslavia.
| 389 Statistical data (1986) | Area (km2) | Population (mio.) | ||
| Argentina | 2'777'815 | 27.86 | ||
| Australia | 7'682'300 | 14.42 | ||
| Austria | 83'853 | 7.46 | ||
| Brazil | 8'511'965 | 123.00 | ||
| Canada | 9'976'140 | 23.90 | ||
| Comores | 1'862 | 0.30 | ||
| Czechoslovakia | 127'877 | 15.18 | ||
| F.R. of Germany | 248'624 | 61.44 | ||
| India | 3'159'530 | 683.00 | ||
| Malaysia | 333'507 | 13.30 | ||
| Mexico | 1'967'183 | 69.40 | ||
| Nigeria | 923'773 | 81.00 | ||
| Pakistan | 796'095 | 80.20 | ||
| Soviet Union | 22'400'000 | 264.50 | ||
| Switzerland | 41'288 | 6.35 | ||
| Tanzania | 939'701 | 17.60 | ||
| United Arab Emirates | 92'100 | 1.04 | ||
| U.S. of America | 9'160'454 | 226.50 | ||
| Venezuela | 912'050 | 14.54 | ||
| Yugoslavia | 255'804 | 22.30 | ||
| 51.82% | 70'391'921 | 1753.29 | 39.95% | |
| Globe | 100.00% | 135'830'000 | 4387.72 | 100%.00 |
390 Never before were there so many newly independent nations as after the Second World War. Because many of these were former colonies composed of different, mutually suspicious population groups living in geographically separated territories, it appeared natural to the colonial powers to grant independence to the colonies through a federal form of government. This may be one of the main reasons for the popularity of the federal formula after World War II. Belgium proceeded in this way with the Congo, the Netherlands with Indonesia, and France with French West Africa, to cite but a few examples. Proudhon (L634/537) had once called the English the most federal of all races, and the United Kingdom created a large number of federations - not surprisingly, perhaps, since it had the biggest empire to dismantle (compare Macy L502/694, Riker L660/25s, Simandjuntak L752/290ss). Many of the new federations quickly changed their form of government or were stillborn.
391 Some consolidated themselves into unitary states; this was true of Burma (L180/2974), Cameroon (L771a/261), Congo (L770/1612s), the Ethiopian/Erithrean federation (L459/Vs), the Federation of South Arabia (L205b/196), the German Democratic Republic (L771a/510), Ghana (L384/136, L633/51s), Indonesia (L421/195s), Libya (L205b/195s), and Uganda (L180/229, L205b/175s, L583a/43).
392 Others broke up into their constituent parts. This was the case with East Africa (L249/16), the Federation of Arab Republics (L205b/196ss), the French Communauté (L539/16), French West Africa (L528/148), the Ghana-Guinea-Mali Union (L230/3), the Iraq/Jordan Federation (L205b/196), the Mali Federation (L230/2s, L528/148), Rhodesia and Nyasaland (L249/82, L528/132s), the United Arab Republic (L205b/195s, L771a/429), the United Arab States (L205b/196, L771/422, Neue Zürcher Zeitung 28.6.1977), partially the USSR and Yugoslavia, the Federation of Arab Republics (L205b/196ss), and the Federation of the West Indies (L223/19, L249/115s, L834/78). In the case of Malaysia, Singapore became a separate state leaving the rest of the federation intact.
393 The advocates of federal solutions for problems of Third World national integration (e.g. Agyeman L14/120ss, Leibholz L533/7, L343/54) often fail to realize -as was the case with the British colonial administrators- that cultural or territorial diversity does not by itself make federations viable. On the contrary, it may very often be a threat to survival. Any federal system makes certain demands on the political culture, and a federation usually comes into being only if specific environmental conditions are met (compare no.664ss). The cultural elements especially are often lacking in the Third World, in particular the willingness to share control and to tolerate competing centres of power. The political culture is based on other principles (see Frenkel L261/260ss, Nordlinger L582/31ss). Still, it must not be excluded that federalism might help to solve some particularly difficult conflict as for instance the conflict between Israel and neighbouring Arab countries (for a comprehensive discussion of federalist alternatives see Elazar/Sharkansky L205a/243ss).
394 Modern Cyprus represents a curiosity. After the downfall of the pluralist constitution in 1974, there was an agreement in 1977 between the leaders of the Greek and Turkish communities on the creation of a bi-communal federation (L622/205s). This agreement has not yet been implemented, but the Turkish population group considered itself until recently to be a member state of a federation with its own government, postage stamps, Turkish currency, etc. In 1984, they have opted for independence. The Greeks live in a unitary state. Internationally the Greek-Cypriot government is usually recognized as the only Cypriot authority despite its somewhat questionable constitutional basis.
395 A comparatively recent development is Senegambia which was founded on 1 January 1982 as a confederation with a common army, a partially common executive but with separate economic regions (Neue Zürcher Zeitung 31.12.1981). Here a former British (Gambia) and a former French (Senegal) colonial territory with ethnically related population groups joined in a federal structure after the Senegalese army helped to suppress a coup d'état in Gambia.
396 Whether the United Kingdom is also federally organized is a question worthy of scrutiny. Mason (L516/12) describes eighteenth century England as a federal empire, pointing to the distribution of responsibilities between central and local government at that time (compare L251/59, L385/11s). Also a number of federal institutions have been associated with the special legal and parliamentary status of Scotland, Northern Ireland, and, to a minor degree, Wales. Mill (L547/375) explained the far-reaching legal and administrative separation of Scotland from England with words that would equally apply to a federal system: «A people having that unbounded toleration which is characteristic of this country for every description of anomaly, as long as those whose interest it concerns do not feel aggrieved by it, afforded an exceptionally advantageous field for trying this difficult experiment.» But following the consolidation of the doctrine of parliamentary sovereignty at the end of the eighteenth century, British federalism had no chance of developing, except for the special case of Northern Ireland (for a relevant proposal see Burrows/Denton L117/16ss). When the question of some legislative and executive autonomy for Scotland recently became an issue, its advocates went to great pains to avoid the term federalism. They talked about devolution instead (compare L599).
397 Three of Europe's federations were communist: Czechoslovakia, the Soviet Union, and Yugoslavia. Yugoslavia coulc even claim to have the most federalist of federal constitutions, one which provided for single member state veto. We are therefore faced with the problem of the Marxist approach to federalism (as to the history of Soviet federalism see Kis L422/37ss, Lepjoschkin L476/130ss). The relationship is ambiguous. Marx was clearly anti-federalist (Lenin L472/97, Mogi L550/847). Engels had in mind a highly decentralized unitary state with great local autonomy (L550/849). According to Lenin (L472/47s), Marxists «are as a matter of course enemies of federation and decentralization for the simple reason that capitalism for its development needs as large and as centralized states as possible... But as long as diverse nations form a unified state, marxists shall on no account defend federalism or decentralization... Still, it would be unpardonable to forget that with our defence of centralism we only mean democratic centralism.... Democratic centralism does not only not exclude local autonomy or regional autonomy characterized by... specific circumstances. On the contrary, it requires both.» «The right to federate is nonsense because federation is a treaty between two parties.... As to autonomy, marxists do not defend a 'right to autonomy' but autonomy itself as a general, universal principle of a democratic state with colourful national composition and highly diverse geographical and other conditions» (L472/97). But Lenin also considered national self-determination to be essential (L473/58ss) and thereby arrived at the recognition of federal or confederal forms of government. Thus, in 1919, he prophesied a world-wide Soviet federal republic (L63/207). And for Russia he called for a confederation essentially restricted to defence and foreign policy (L790/2s, for Lenin's ambivalent attitude compare also L422/28ss). There his position was opposed to Stalin's, who preferred a policy of annexation. Stalin also recognized, in principle, the theory of national self-determination (L766/261). But he easily found a justification for its qualification: «There are cases where the right to self-determination gets to be at variance with another, higher right, the right of the working class risen to power to consolidate this power. In such cases -this must be said quite clearly- self-determination cannot and must not be an obstacle to the working class' right to dictatorship. The first right has to take second place. That is why we were forced in 1920 to march on Warsaw in the interests of the defence of the working class risen to power» (L766/220).
398 The modern left in Western Europe is much influenced by the «petty-bourgeois» (Lenin L550/851) anarchic ideals of Proudhon and Bakunin. Therefore, federalism as a principle (as opposed to its existing manifestations) is advocated by both bourgeois representatives of personalism (e.g. Alexandre Marc, Ferdinand Kinsky, Denis de Rougemont) and exponents of the new left (e.g. Kappeler L782/563, Zwerin L884/142), based on the same Proudhonian foundations.
399 As to future developments of political federalism, we must start from the realization that the period of colonial emancipation will very soon be a thing of the past. It is an open question whether and when there might again be conditions under which the many miniature states that profit by the geopolitical situation might seek to federate in some way or other. It appears that there may be three other circumstances of special importance to future forms of federalism: the evolution of transnational intercourse, the need to have additional pluralist interests participate in government, and the demand for more local autonomy in highly centralized states.
400 The dream of a united Europe is old. In its federal variant we may trace it back to the fifteenth century (no.375, Bernier L63/203). It is little known that Benjamin Franklin also thought about a European federation after the pattern of the American Constitution of 1787 (L234/131). But it was mainly in the nineteenth and in the early twentieth century that proposals for a federal Europe became common (compare Bakunin L47/15ss, Bernier L63/208, Bülck L104/21, Fouéré L248/149ss, Proudhon L364/335, de Rougemont L675/94, L694/78ss). In the fifties of this century the European Communities (EC, now the European Union) were founded and with them an entity that was dedicated to become the nucleus of a future pervasive union. The European Union with its regulations for certain functional areas represent a form of partial federation (thus Klein L425/36s, compare Scheuner L833/21/123). And certainly the political aspirations are of a federalist nature. EU law constitutes a new level between national and international law (Bernier L63/237). It may be argued that this kind of unification, having mainly functional goals, lacks the common, constitutional ideology that was typical of last century's federations (thus Imboden L387/67). Again, we must not preclude the possibility that the EU might be evolving into a fully fledged union. This was certainly the idea of the founders. Direct elections to the European Parliament and a common currency, the Euro, may herald such a future. A constitutional convention is actually preparing the legal groundwork for it. But it is just as well possible that the soon the be enlarged EU, suffocating in bureaucracy and in paper, has arrived at the end of its tether and that other communities will open up the way, if at all, to more genuine European federalism.
401 Plans for world federation have aims which are even more remote and uncertain. Duchacek (L180/48) calls them daydreams so long as a union closer and more effective than the United Nations is not necessary for the defence of common interests against other planets (compare L180/350, Beres L60/75,86s, Secrétan L740/60, Streit L785a/238ss). It is nevertheless possible that some world-wide functional associations will undergo the same transformation from transnational into supranational institutions as has already happened at the regional level.
402 The second possible development mentioned above (no.399), the need to allow additional pluralist interests participate in government, is dealt with elsewhere (nos.203ss and 589). Undoubtedly this development greatly influences United States federalism.
403 The third tendency is the demand for more local autonomy in highly centralized states. This phenomenon can be encountered in all large unitary states, such as France, the United Kingdom, Italy, and Spain. The institutions designed to meet these demands resemble federal institutions more and more (e.g. autonomous regions in Italy - L558, L681/129ss, L852 - and in Spain, Constitution article 143ss). One is tempted to elaborate some kind of federal theory of convergence: Whereas most federations have become more centralized during recent decades, unitary states have become more decentralized. This is a somewhat simplistic view. In particular, it does not do justice to recent developments in the federal allocation of functions, where the pendulum has again swung in the direction of localism. This can probably be attributed to the same renaissance of localism (no.704) that gives impetus to demands for autonomy in unitary states. But these thoughts show that a neat differentiation of federations and unitary states may become increasingly difficult in the future, because federalism will be less tied to confederation or to federation as the only forms of political implementation. It is quite possible that the twentieth century's contribution to federal practice - apart from transnational superstates - will be the part-regionalized autonomist state, i.e. a state where some, but not all, regions enjoy the kinds of autonomy and participation that are usually reserved for member states in federations. Examples of such forms of government are Åland in Finland, the Aosta valley in Italy, and the Basque province (as well as other provincial statutes) in Spain.
E. Functions of Federalism
404 Federalism is a principle for the organization of political decision making. Its characteristics are on the one hand the autonomy of territorial groups and on the other hand their participation within the framework of an encompassing structure (no.210). If we look now at the functions this principle is deemed to apply to, we must first consider local as against national performance of tasks. Most arguments for federalism are addressed to this aspect. But this is not sufficient. The special quality of federalism is its two-sided nature. It is national as well as local.
405 There is something else, also. If we think of the goals of federalism we usually have a federation in mind. In all classical federations we encounter a three-tier structure. The member states are themselves federally organized. The member state, therefore, has three attributes. It is local in comparison with the national centre. Seen thus those considerations apply that have to do with the functions of locality. It is, secondly, central in relation to local government, which renders the corresponding set of central functions relevant. Thirdly the member state is a middle level within a whole, which poses the problem of the specific functions of such a middle level for a federal system. This last question is usually neglected by students of federalism. Empirical studies are lacking even more in this respect than for the effects of local autonomy. I shall nevertheless try to deal with the functions of the middle level also.
406 My exposition of the functions of federalism corresponds to the model outlined above (no.24ss). First (no.416ss), we deal with functions that have to do with the system's capacity to learn. Then there are the functions that stabilize the system (no.540ss). Finally we will examine assigned functions (no.614ss) such as the promotion of individual freedom, civic participation, etc.
407 The ambivalence of the goals of federalism (Diamond L169/130) is obvious in the writings of classical authors. Thus Montesquieu (L551/9/1) said about confederation: «Composed of small republics it enjoys their good government in the interior and, as to the outside world, it has through the force of association all the advantages of the great monarchies». Similar thoughts may be found in Rousseau L679/5, Tocqueville L806/232, and amongst modern authors Duchacek L180/199s and Kafka L833/21/108s. If today the values of localism are in the foreground of arguments for federalism, this may have to do with the fact that increasing centralization of federations and a tendency on the part of scientists towards hierarchical thinking provoke the emphasizing of autonomy. The following selection of catalogues of goals may serve as an introductory survey.
408 Aubert (L782/1081): «1.Federalism guarantees - to the extent that a system may guarantee anything.. - the proximity of power. ... 2.Federalism serves to vulgarize power. 3.It assures the division or separation of power. 4.It preserves power as the opposite to anarchy».
409 Friedrich (L277/57s): «Federalism 1) increases the opportunities for dissenting minorities to make their views known to other citizens and policy makers; 2) multiplies the opportunities for citizens to participate in political life; 3) enhances consensus in political discussion in the sense that solutions are sought that will reduce the size, resentments, and coercion of defeated minorities; 4) greatly improves the chances of the peaceful resolution of conflicts; 5) aids the solving of urgent policy questions by providing an opportunity for experimenting with solutions on a limited scale; and 6) enhances confidence in and loyalty to a constitutional polity.»
410 Journal «Zeit» (cited in L457/85): Advantages of the federal system: «It makes for vertical division of powers and thereby furthers rule of law; it provides a multitude of economical, political, and cultural centres and favours a balanced regional structure of the Federal Republic of Germany; it facilitates a diverse political composition of federal and state governments thereby integrating the pluralist country as a whole; and, finally, it increases demand for political cadres thus creating a broad basis for testing young political talent.»
411 Laufer (L454/33s): «The federal organization is more adequate to democracy.... Federation furthers the separation of powers and intensifies rule of law.... Federalism makes multiformity possible and strengthens competition».
412 Thieme (L800/148): «If today we look for the meaning of the federal system we find seven different and partially related viewpoints being proffered: 1. Regional influence on the solution of political problems, 2. administrative influence on the decision of overarching questions taking place already in the stage of political planning, 3. an element of division of powers in a situation where the division of powers between Parliament and government has been largely overridden, 4. experimentation and competition, 5. safeguarding of minorities, 6. unburdening of central institutions, and 7. realization of the subsidiarity principle.»
413 Once I myself proposed a catalogue of justifications for federalism for the use of the Swiss experts' commission on the total revision of the federal constitution. I was asked to make do with one half page. The result was the following text (L577/1975/52) which, to judge from their final report, in no way influenced the experts' centralizing bias:
414 «Federalism is no statement of law. Federalism is no confession of faith. Federalism, for Switzerland, is an organizational principle of political sagacity. Federalism, for Switzerland, is federation. Switzerland is a federation because this organization allows: Switzerland has, on the other hand, to put up with: 1. to keep the distance narrow between citizens and polity; 1. occasional cleavage between costs and benefits (territorial spillovers; 2. to distribute power amongst the many without its becoming anonymous (guaranteeing participation and stability); 2. overtaxing of political cadres (cumulation of roles); 3. to tailor public services to different local demands (efficiency as consumer orientation); 3. relative significance only of the federal constitution's equality clause; 4. to minimize administrative costs because of short distances for information processing and democratic control (efficiency as cost minimization); 4. neglect of economies through large scale consolidation (economies of scale); 5. the regions to pursue their development goals with vigour; 5. sluggishness in determining and carrying out national priorities; 6. different cultures, languages, and temperaments to live together without destructive conflicts. 6. renunciation of coherent 'national' politics.» 415 For the purposes of this study I do not attempt to trace federalism back to some specific ideological or religious position. It is obvious that such positions would emphasize certain functions to the disadvantage of others. The history of federalism (no.364ss) provides us with a colourful palette of such ideological starting points for federal solutions. And today it is not only conservatives who don the cloak of federalism but also progressives (for the USA see, for instance, L181/9s). As to religion it may well be that the attempt to justify federalism with some commandment of God (thus Lang L451/101) or with the Christian faith as such (L451/110ss, Reynold L654/114) is relevant in a given situation because of the integrative power of religion. But I doubt that such a situation now exists. And for a scholarly analysis the ideological point of departure is too restrictive.
E.1 Federalism and the Political System's Learning Capacity
416 We shall first examine the functions of federalism which are related to the maintenance of the political system's learning capacity. Then we will consider those functions that have to do with the perception of demands and the quality of solutions.
E.11 Maintenance of the Political System's Learning Capacity
417 Federalism maintains the political system's learning capacity by not overloading a single centre with too many problems (no.418ss), by forcing the system to learn continuously through division of powers and competition (no.424ss), and by furthering local experimentation (no.434ss). Furthermore, federalism provides for decision-making rules (thereby freeing the system's capacities for new learning processes). These rules will not be considered here. They are an essential element of given constitutional arrangements which are outlined in the German companion volume (L264b). To put it in a nutshell: their characteristic is that they create a balancing or optimizing framework for small scale and large scale democracy. The justification for this framework, on the other hand, is the subject of the following analysis of the functions of federalism.
E.111 Relief of Decision Centres
418 Overload of decision centres has two aspects. One has to do with the demands which are placed on the system by the outside world. The other aspect is the system's capacity to deal with a given number of demands.
419 There is reason to believe that, as compared with a unitary state of similar size, a federation will have to confront a lesser sum total of demands. This is because most problems will only be perceived as problems through the working of the political process. Let us suppose, for instance, that in a federation there are several population groups with markedly different attitudes to world politics (to some extent this is true for Canada and Switzerland). Then we may expect foreign policy to be more passive than elsewhere because this will help to diminish potential internal conflicts about foreign policy directions (compare Lorwin L491/42s). The more passive is foreign policy the less likely it is that there will be problem overload of the federal system stemming from this sector of politics. (In some extreme cases, if a federation is so weak internally that it kindles foreign appetites for interference or annexation, the result may however be the opposite.)
420 This is what Wheare (L858/186) makes allusion to when he states: «Federalism and a spirited foreign policy go ill together» (see also Hahn L328/273s, Tripp L813/69).421 We may also propose the thesis (which requires to be empirically validated) that most of the mechanisms which generate political demands are nationally organized and oriented. This is especially true for the mass media and large-scale interest groups. Problems that are generally considered to be outside the scope of this level are not treated by these mechanisms -the grid is too wide-meshed- and they are thus not perceived as problems. In a unitary state they may find their way into the political arena more easily, thereby heightening demand. Scharpf's findings (L705/19s) point in the same direction. The multitude of interrelated decision centres in a federal system make it much more difficult to carry particular demands through all of them, thereby limiting the generation of too many such demands.
422 Nevertheless, some qualities of federalism counteract this relief function. For one thing, federal systems tend to be confronted with a considerable number of demands which are energetically supported by regional authorities. This may for instance be the case for regional economic development (it would be interesting to have a comparative study of this question). And then there is the additional task of harmonizing the different policies among the different levels of government. This is a typically federal political problem of co-ordination (see Simeon for Canada L753/295).
423 Let us now address the second aspect mentioned above (no.418), the ability to deal with a given number of demands. Here we may talk of federal systems having a clear advantage. The existence of several autonomous decision centres makes for a division of labour which prevents the piling up and overload of problems at one national centre (Bryce L96/353). Two conditions have to be met if this relief is to be fully realized. It is necessary that the system be a federal one, that is a non-centralized as distinguished from a decentralized one (for the difference see no.226). It is also desirable that not too many policy sectors are organized as joint tasks for more than one level of government (see L264b/1043ss), adding demands to the load of each of them (compare also Ostrom L593/229, Scharpf L702/81s, L705/22).
E.112 Competition and Division of Powers
424 The existence of various and largely autonomous power bases is characteristic of a federal system. Such a situation entails competition and makes for a form of vertical division of powers which is at the centre of most theoretical justifications of federalism (as to the general doctrine of division of powers see Frenkel L253/8ss). For the purpose of the present analysis, the functions of division of powers and competition are treated as different aspects of federalism. Here we deal with maintaining the political system's learning capacity. Later we shall look at the processing of demands (no.501ss) and the support of liberty (no.625ss).
425 Competition takes place horizontally between the member states. To the extent that this maintains learning we shall deal with it in the context of innovation (no.434ss).
426 In the foreground of discussion is the vertical division of powers between the central government and the member states. We start with the observation that centralized organizations are in danger of being unable correctly to perceive and process local information. «This is one of the critical properties of all large-scale systems. The authorities must act in an atmosphere of partial ignorance, considerable misinformation, and consequently of great uncertainty» (Easton L188/415). Every bureaucracy has a tendency to base its decisions on data which are in the final analysis self-engendered rather than taken from the outside world (compare Deutsch L166/215). And as Ostrom points out (L595/252), central agencies rarely surround themselves with critics, thereby restricting the search for alternatives and augmenting the probability of errors (also L593/214).
427 It was Tullock who especially drew attention to this fact (cited by Ostrom in L595/251): «In his Politics of Bureaucracy, Gordon Tullock has demonstrated how a hierarchically-oriented structure functions as a transmitter of information and develops selective biases at each link in the communication chain. This process results from two assumptions Tullock makes about centralized structures: that superiors have a controlling influence over the career advancement of subordinates and that subordinates as selfinterested individuals will act so as to advance their career potentials. Thus, Tullock argues, at each link in the hierarchical chain of command subordinates will advance information reflecting favorably, and repress information reflecting adversely, upon their career potential - thus causing a selective biasing or distortion of information.» Or Chesterfield: «... they thought I informed, because I pleased them» (L131/221).428 The federal structure counteracts this tendency by creating hierarchically autonomous centres of decision making (Truman L817/45). A member state government is, as a rule, not too dependent on the goodwill of the central government. The kind of information it adds to decision making is independent of the information available to the central bureaucracy. «Thus, the diverse structures and processes of government decision-making contribute to a due process of inquiry, and can correct for the distortion of information and proneness to error that inhere in highly integrated, large-scale bureaucracies. This is only possible when the diverse structures can function independently of the central executive authority» (Ostrom L595/253). To this we may add that such data cannot be neglected, because they emanate from authorities which have their own power base and are conflict-efficient (Scharpf L705/52) because their consent is necessary or highly desirable in many areas.
429 Competition between different political parties, stemming from differences in the composition of central and state legislatures, goes in this direction (Laufer L457/9s), as does the linguistically motivated special status Québec claims for herself in Canada (Macmahon L501/17). But it must be said that the conventional view generally has little comprehension for such considerations, preferring frictionless proceedings (Frenkel L255/730). The former Canadian Prime Minister Trudeau appears to be one of the few national politicians who, when in government, commented favourably on such conflicts (L816/14s): «It may not be recognized, for example, that the system is working well when a government whose programs are affected by another government's legislation, complains about that legislation and brings about the removal of a harmful or wasteful provision. ... Federalism in Canada will never be free of disagreement. ... For it has been said that a crucial benefit of federalism is the unlikelihood that the governments of the constituent parts and the central government will all decide at the same time to do the same foolish thing to people.»430 This function of federalism applies not only to federations proper but also to subsystems like political parties, associations, etc. Thus Fetscher, Hesse, and Tiemann consider the federal principle to be the essential factor for democratic relaxation of party hierarchies in the Federal Republic of Germany (L239/166, L804/316). It may well be that the organization of such subsystems is better evidence for the reality of federalism in a federation than the latter's written constitution (see no.682ss).
431 A threat to the federalism function of division of powers may be seen in the practices of so-called co-operative federalism (no.351ss), which in their tendency replace conflict by co-operation. If such a maxim should become the accepted standard in the politicians' and the public's imagination, competition will be weakened in importance and legitimacy. The cult of an ideology of collaboration, coupled with the fact that the goals of such collaboration will tend to be defined by central interests according to national criteria, will bring to co-operative federations a markedly hierarchical structure with all the concomitant results of rigidity and of distortion of information mentioned above. To this there will be added an element of displacement of politics by administration, as co-operative federalism is mainly geared to the collaboration of administrations (L264b/1552, compare also Grawert L304/158, Hempel L352/221, Laufer L454/12).
432 On the other hand, federal competition may also appear as a pretext. In the nineteenth century Constantin Frantz observed (L487/850): «Most often what is presented to us as federalism has no intrinsic positive meaning, being but resistance to centralism.» The German Hesse writes (L356/9): «It is well known that today quarrels between the Federal Government and the Länder are, as a rule, no longer true federal conflicts, but conflicts between different political tendencies in the public realm clothing themselves in the garments of federalism to have a standing in constitutional jurisdiction.»
433 Nevertheless, in summing up we may say that federalism's function in enlarging information through competition and division of powers is enhancing democracy by furthering both the weight and the breadth of the information flowing from the periphery to the centre.
E.113 Innovation
♦ 434 «A sound justification for regional segmentation of political power is connected with the diversity of initiatives stemming from a variety of decision centres.» (Hempel L352/218). «The pressure on the several levels of government to show results furthers activity and imagination» (Tiemann L804/316).435 This innovative function of federalism is one that is very often emphasized. In this context, the American Supreme Court Justice Brandeis designated the states as «laboratories of democracy» (L548/3). There are many examples. In the United States, at the beginning of the century, Wisconsin was leading in the fields of social insurance, vocational training and budgetary procedures (L114/329). New types of insurance (no-fault insurance) were first introduced in some states (L30/4). Grants-in-aid for the fight against cystic fibrosis were innovated by Connecticut (L22/232). New techniques for sewage treatment are usually developed in localities of 15'000 or fewer inhabitants (L201/293s). A new generation of American progressives has withdrawn from the federal political arena because they see more opportunities for social change at the local and state levels (Steele, The Guardian Weekly 3.9.1978). In Switzerland the institutions of direct democracy, such as constitutional or legislative referenda, popular initiatives, and proportional elections, were first introduced in the cantons before they became part of the Federal Constitution (L243/209). The Canton of Glaris was first in continental Europe to adopt a law on factories. The first Swiss ombudsman, to take a recent example, started his work in 1971 in the city of Zurich, the Cantons of Zurich and Basle-Country following in 1978 and 1984, respectively. Numerous other illustrations from these and other federations might be cited.
436 It would be misleading, however, to ascribe innovative effects uniformly to federalism. It is also often the case that federalism prevents, or is said to prevent, nationwide innovation. Thus Neidhart, in Switzerland, perceives unco-operative cantonalism which, according to him, plays a major role in rendering changes in the Swiss system of government difficult (L575/37, L782/612). In a general way it may well be said that the many powers of veto in a federal system slow down change. They are intrinsically conservative (Dicey L170/173s, Friedrich L275/10). This is not negative in itself. We may, with McWhinney (L533/7), also consider it to be the political realization of the need for wide national consensus before attempting large-scale experimentation. But the fact of curbing innovation remains. Porter (L753/281) puts forth the hypothesis that member states are more inclined to conservatism than the federation as a whole because they are more lacking in means as well as more exposed to the influence of lobbies. This still has to be tested. I have some doubts. It may be true that there are other lobbies at work. But looking at federations in general I think that national lobbies and media are at least as influential as, if not generally more influential than, their member state counterparts. The disparities in the willingness to innovate in certain fields might also be due to the greater distance between federal parliaments and electors and the more bureaucratic or intellectual quality of central decision making. This hypothesis is also worth serious consideration.
437 We should consider the possibility that restraints on national innovation are not so much a characteristic of federations as the result of any highly complex governmental structure. They are, therefore, also to be found in other countries.438 Scharpf writes (L703/183) that plural systems with relatively independent subsystems on the one hand increase the probability of initiatives. But, on the other hand, these initiatives are usually solutions for the subsystem's own problems whereas a fragmented polity has a tendency to neglect overarching issues. This observation is certainly valid for federal pluralism. «Compared to this, centralized systems appear to be less sensitive for their subsystems' problems, ... Decentralization, on the other side, furthers those innovations for which consensus might be arrived at within individual subsystems but not with the system as a whole» (L703/183).
439 In this area also the practices of co-operative federalism (no.346ss) probably have a rather negative result. The member states governments are tempted not to solve their problems but to eschew them, pointing to the advantages of waiting for the central government's decisions in the matter. Or the member states may attempt to introduce common solutions even in fields where this is not necessary. This is especially the case in the Federal Republic of Germany, where there is constant media pressure to unify anything that smacks of diversity. Thieme notes that «the various intergovernmental Länder-conferences... have therefore found their main occupation in unification much more than in experimentation, to rather prevent individual solutions than to further them» (L800/151s). A similar situation obtains in the United States. We are confronted with more than the «largely removable side-effects of a methodical/administrative kind» (Ehringhaus L194/179).
440 A further barrier to innovation may be erected by intergovernmental competition. Legislation on environmental protection, labour codes, new taxes etc. is not very probable if it entails the threat of an exodus to other member states of those adversely affected (Wright L746/75s). In such cases the central government may further member state experimentation by restricting the possibilities for interstate «dumping». The American system of unemployment insurance is an illustration of this (Freund L746/47).
441 These hindrances may be the reason why some authors, who expect perhaps too great a degree of diversity in federations, are rather disappointed at the actual number of experiments (compare Macmahon L501/3s,143ss, Miller L548/3s). During the 1975 election campaign the Austrian Chancellor Kreisky went as far as to describe the practice of the member state parliaments as «petrified» and to deny the existence of any real member state politics (L587/23).442 Baker (L45/10) calls for more detailed research into this whole question of federalism and innovation. I cannot but concur.
E.12 Solving Problems in Pluricentral Organizations
443 Disputes on the advantages and disadvantages of federalism are at the same time discussions about this system's efficiency, that is to say about its ability to solve problems in an adequate way. There is, then, a tendency to place fragmentation, overlapping and inefficiency on one and the same footing. If there exists a kind of political rhetoric that tends to correlate federalism with goodness itself, there is also, in everyday life, a set of arguments at least as politically influential which suggests that the alleged weaknesses of federal performance may be overcome by methods of so-called modern management. It is said that eighteenth or nineteenth century federalism has not been designed to deal with the problems of the twentieth century. «As the United States moves into the 1970s, it is clear that our federal system is not working» (Reuss L594/1). «Our present federal structure is a relic from the horse-and-buggy past - a structure fundamentally intractable to successful grappling with today's complex social and economic issues» (Humphrey L594/1).
444 These critics usually ask for a more clearly structured, hierarchical government organization, free of redundancies, along patterns developed in the business world. There is a certain irony in the fact that such proposals are put forth at a time when economic evolution favours the loosening up of big centralized hierarchies to render the individual parts more autonomous; or, in other words, when business begins to see virtues in federalism (compare Elazar L202/21, L637/6-4/10, Porter L625/84s).445 Efficiency is one of those words that, whilst indicating a direction, do not have a uniformly accepted meaning. If we were to equate efficient with functional the whole of the present analysis of functions of federalism would also involve the analysis of efficiency. But we are nearer to the general acceptance of the term if we think especially of the cost and quality of solutions when we deal with efficiency in relation to federalism. In talking about costs, we would stress production aspects whereas quality is concerned with the demand for public services in terms of preferences and the satisfaction of wants.
446 Finally, we should not overlook a further source of criticism which might well be the most important one. It is what I call intellectual æsthetics (no.531ss). These matters are now considered in turn.
E.121 Costs of Federalist Solutions
♦ 447 «Federal systems in general tend to be extravagant in the use of governmental manpower» (Sawer L698/50). «Trouble, expense, and delay due to the complexity of a double system of legislation and administration» (L99/341). - «Experience presents us with counter-evidence: if we sum up we arrive at the conclusion that administration organized under the principle of subsidiarity will cost less and will be more economical than centralized administration» (Marcic L509/431).448 The positions are clear and contradictory. We might expect a wealth of empirical studies to support them. This is not the case. As with nearly all aspects of federalism very little research has been undertaken (Ostrom L590/33). In theory, some costs which have to do with territorial disparities between providers and beneficiaries of public goods («spillovers»), or with the impossibility of providing such goods for too small a population («economies of scale»), militate for a high degree of centralization. Costs of decision and control, on the other hand, point rather in the opposite direction (Frey L267/371). The optimal degree of centralization under the cost criterion has yet to be determined. In order to establish it, we would have to confront centralization not only with non-centralization, but also with decentralization (see no.225). And given the prevailing centralizing bias we would also have to ascertain whether economies of scale are not, in some instances, accompanied by diseconomies of scale (Walsh L844/16s).
449 According to a European Community inquiry (L437/34) it appears that the proportion of the gross national income used for public expenditures at all levels of government tends to be somewhat lower in federations than in unitary states. But the countries thus analysed are not representative enough, and the difference is too small, to make clear-cut assertions admissible (also Beer L57/89s). - On the other hand, it appears that local government expenditures in the Western World have expanded faster since the Second World War than central government ones (Gould L798a/226ss). This may be a consequence of the increased delegation of responsbilities to the local level and, possibly, a larger growth of some service oriented governmental functions as e.g. health and education (local) if compared to others like defense (national, L798a/228).450 I am inclined to think that the question of costs is not a decisive one. Certainly not if we identify costs with money. The political costs or advantages connected with other functions of federalism are more important. However, both theoretical considerations about the costs of information processing (empirical studies are especially lacking here) as well as the analysis of advantages and disadvantages of size lead us to perceive merits with non-centralized or partially centralized forms of organization. The hypothesis that central agencies will more efficiently allocate scarce means (Neidhart L575/37) cannot be upheld, at least not from a financial viewpoint.
451 I shall deal more specifically with the costs of information, of mobility, of scale, and of externalities. It goes without saying that all my considerations on functions of federalism could be clothed in the form of reflections about costs. This would be a consistently economical approach. Such an approach is for instance at the basis of the theory of Albert Breton and Anthony Scott (see L87/31ss, L307/44ss). These authors distinguish between costs of signalling, of mobility, of administration, and of co-ordination. Yet such a limitation on the provision of public goods and services cannot but very clumsily do justice to the political and psychological aspects of, for example, the function of integration. I prefer a less pure but, in my view, more meaningful approach.
E.1211 Costs of Information
452 Processing information involves costs. These will be minimal where the distance between the origin of data and their final use is small. This happens to be one of the justifications for the subsidiarity principle (no.231ss). Each step of transmission of information engenders two kinds of costs: there are the actual costs for the necessary intermediary agencies (salaries, administrative infrastructure, etc.; compare Proudhon L635/99s). But there are in addition the costs arising from the distortion of information which any hierarchy is prone to and which may lead to expensive misjudgment. The modern American theory of bureaucracy deals especially with these questions; allusions to this have already been made above (no.426ss, see also Deutsch L166/215, Walsh L844/17s). Examples will be given below (no.485ss).
453 On the other hand, federations may have to face additional costs of information. This may for instance be the case if for some territorially far-reaching decision there is distortion of information due to the fact that only local data are taken into consideration. This is typically the case with spillover problems (no.472ss). But we may also think in this context of the costs for consensus and co-ordination due to the veto positions of additional units of government, like member states or local governments, which render agreement more difficult (for Canada see Sproule-Jones L763/123ss) or necessitate additional flows of information.
454 Still another viewpoint is that of the individual citizen. A federal system provides him with multiple access points for his demands (Landau L449/190) and, at least to some degree, shorter distances. The other or cost side of the coin is a more demanding standard of civic education necessary for the utilization of such a system (Ostrom L591/231).
E.1212 Costs of Mobility
455 Below (no.503) I shall discuss the so-called Tiebout hypothesis. It is based on the theory that citizens or individual firms may, in a federation with different service structures in its member states, show their preferences by changing their domicile. This, for the individual, is an advantage. For member states and other territorial units the corresponding competition brings costs because some measures which entail the danger of an exodus of tax-payers, such as increases in taxation, are thereby constrained. This is also one of the reasons why some tasks such as income redistribution and economic stabilization policy can only be exercised by member states to a limited degree (L264b/901a).
456 Even for the individual citizen the federal structure may bring about mobility costs (apart from the obvious costs of moving according to the Tiebout hypothesis). It may well be that member states render interstate immigration or movements of goods more difficult (sometimes also emigration, as formerly in Switzerland, C62). This is usually connected with linguistic or economic protection. With the exception of Canada (where article 6 of the 1982 Constitution addresses this very problem) these impediments to mobility are usually without importance in modern federations. They are certainly not more important than in unitary states.
E.1213 Costs of Scale
♦ 457 Proudhon said (L634/469s): «I have pointed out ... that general government costs increase in direct and geometric connection with centralization. In the Canton of Vaud the average tax yield per head of the population is fr. 15.77. Together with the federal contribution which amounts to fr. 6.89 per head, we arrive at a total of fr. 22.66. This same average corresponds to fr. 30.-- in Belgium and fr. 54.-- in France. On the other hand there are 25 cantonal governments as well as a federal one in Switzerland for a population of 2'392'760 inhabitants.» - «Though the subject is largely unexplored, a recent pioneer study suggests that the economists, who are undecided as to whether a large nation is significantly cheaper to administer than a small nation, would agree that a federal state might be less able to achieve economies of scale than a unitary state» (Hicks L358/45).458 It may be admitted as a fact that the village where I live is unlikely to be able to shoot a satellite into orbit. For the United States this is no longer a problem. It is rather obvious, too, that a town must be of a certain size before it is in a position to support an ensemble of first-class opera singers or an art gallery (paradoxically, American studies have come to the conclusion that, on the other hand, such institutions are visited in inverse proportion to a city's size: L156a/117). Very often a central computer might serve without significantly higher costs a very much larger area than it is actually used for. All this is to say that a certain minimal area and, therefore, a certain degree of centralization are necessary prerequisites for the provision of some public services or, at least, for their cost-efficient provision. Centralization makes for economies of scale. This is a quite popular and politically strong argument against federal fragmentation.
459 The existence of economies of scale is evident. Diseconomies of scale also exist: costs that rise disproportionally with size. Economies of scale are usually the result of a better exploitation of fixed facilities. But in the context of the local government amalgamations which have been fashionable during recent years in both federations and unitary states, these economies often had a tendency to be offset by diseconomies. Hawkins (L346/212s), on the basis of his empirical studies, said in respect of the U.S.A.: «Yet what cannot be dismissed is the consistency of the conclusions - that while diseconomies of scale appear, economies of scale do not». Elazar (L203/72) pointed out «that in cities of more than a million, the average per person costs were 51% greater than those of cities with populations between 100'000 and 250'000 and were 100% greater than the costs of services rendered in cities having fewer than 25'000.» In 1968, Hirsch arrived at the conclusion that for many services efficiency gains were highest for a size of population in the range of 50'000 to 150'000, with the exception of specific services which require a large amount of capital like water supply, public transport, and sewage treatment (L595/119).
460 The English local government reorganization of 1974 was also designed to result in high efficiency gains. The subsequent experiences were particularly painful for the populations concerned. According to Richards (L658/96) the new local governments very often raised taxes by about 50 per cent, in some extreme cases by nearly 100 per cent.
461 Knaub (L428/162ss) has analysed the French cost/population ratio in different sized communities for the year 1968:
Current expenditures per inhabitant Personnel expenditures per inhabitant Infrastructure costs for each additional inhabitant Population Cost Population Cost Population Cost 5000-10'000 245F Below 2000 64F Below 10'000 17'300F 10'000-20'000 276F 2000-5000 86F 10'000-20'000 24'100F 20'000-50'000 335F 5000-20'000 124F 20'000-50'000 24'400F 50'000-100'000 362F 20'000-50'000 165F 50'000-100'000 33400F 100'000-300'000 417F 50'000-100'000 191F 100'000-200'000 36'700F 300'000+ 388F 100'000 236F 200'000+ 37'000F 462 Thurner's study (L802/41ss) on the relationship between the number of public employees and population size in local governments in Upper Austria points to the same conclusion.463 What are the reasons for such rapidly mounting diseconomies of scale? Costs for the treatment of information as mentioned above (no.452ss) probably constitute one set of reasons. But then we are also confronted with personnel expenditures and increases in the demand for public services.
464 There are several inducements for personnel costs to rise disproportionately. Thus in the Federal Republic of Germany, this may result from local government amalgamation followed by pension payments for redundant, often part-time, functionaries. Still more important is professionalization of the public service. Functions that heretofore were honorary posts, paid with a song and a whistle, now command a salary that is likely to be higher than the sum of the cost economies. In addition, full-time employees need offices and other facilities. Personnel employed before the reorganization will need to be better paid for additional responsibilities (after all, their unions' agreement to the amalgamation plans need to be recognized in some tangible way). Notwithstanding anything that might be contemplated at the outset, it would be naive to believe that personnel who become superfluous under the new arrangements will be dismissed. On the contrary, Parkinson's law will make for new jobs, especially in relation to staff functions. In addition, the problems of information treatment and raised expectations about services will encourage the growth of bureaucracies. «Overstaffing is enemy number one with local government reformers ('in England'), who point out that since 1955 town hall staff have almost doubled in numbers... After the recent reorganisations of councils into bizarre-sounding new areas, the staff, whose function remained the same, went up by six per cent and is now ten per cent up on the old total. 'Where there are cutbacks, it tends to be the services, not the staff, that get the chop'» (Sale, Punch 26.5.76).
465 According to Hawkins (L346/213) «a 1974 State of California study found it would cost fifty-four percent more to have the State Division of Forestry provide service to rural areas presently receiving fire services from special districts. The California Local Government Reform Task Force found that special district sewage treatment facilities were just as efficient, at comparable levels of quality, as city and county operations. A study conducted by the Environmental Protection Agency found that private garbage collection was only fifty-seven percent as costly as equivalent services provided by the public sector. And a private fire company in Scottsdale, Arizona, not only reduced local fire insurance rates but also reduced per capita fire costs by forty-seven percent. According to the evidence, large size does not appear to be a critical factor. In fact, size is often positively associated with decreasing levels of effectiveness.» «On the contrary, there are now bits of positive evidence that the quality of service declines (below a minimum level) with increasing size. Using a variety of measures of student performance in California, for example, Niskanen and Levy find that 'School district size has a consistent negative relation to student performance'» (Wildavsky L861/119).466 While such studies as are available suggest that economies of scale do exist (Metwally L543/42), they are to be found primarily in connection with services that require capital rather than manpower. It is thus probable that big computers can be more efficiently exploited in areas with large populations than in less populous areas. A public welfare agency, on the other hand, will incur additional costs as it grows, because of the need to employ additional personnel for direction and control. This tendency is accentuated by the absence of a competing market in the public sector. Public enterprises are comparatively inattentive to management costs that rise disproportionally with increased size. This may lead to the point where the management costs engendered by each additional official are higher than his productivity so that diseconomies of scale rather than economies result (compare Grewal L308/11ss, Ostrom L592/59s).
467 Public services hardly ever require either much capital and no manpower or considerable manpower and no capital. Public services need personnel as well as expensive equipment. If the organization of such services required a mutually exclusive choice between different patterns of organization, we would have no choice but to neglect benefits of scale in order not to increase personnel costs, or to neglect rising personnel costs in order to keep capital investment within sensible limits. But in reality we have more choice. Special districts or single-purpose authorities, contracting out, and other organizational arrangements make it possible even for small agencies to achieve most of the benefits of scale through co-operation with others without losing the cost advantages of small size (see Hawkins L346/212). Evidently these hybrid forms also have their disadvantages. Public control of joint institutions is more difficult and their operational performance might be difficult to monitor. Both situations may result in increased costs.
468 If we assume that public services are to be financed from a local government's own funds (i.e. without grants-in-aid) we may presume that smaller governments keep house more economically than bigger ones. I have the impression that, in Switzerland, the connection between the cost of a new service and the level of taxation plays a greater role in local politics than in the Cantons, where it again seems to be a more important consideration than in national politics. There are several possible reasons. There is the fact that in a large budget an individual item of expenditure has a comparatively small influence. Then there is the psychological inability to visualize big sums, or to oversee and co-ordinate a large scale bureaucracy with all its services. Increasing professionalization of centralized politics (full-time politicians, large staffs of administrators and interest groups etc.) encourages the level of demand in the political arena (compare Birgersson L71/265s, Friedman L273/343). Finally, in large groups it becomes much easier for free riders to profit from public services without participating in the costs of their provision (L430/60).
469 It seems to be a consequence of local government amalgamation in various European countries that the different patterns of public services in the old jurisdictions are not only maintained in the original areas but are extended to the whole of the expanded jurisdiction. This obviously results in an overall increase of expenditures (Frenkel L263/24).470 In an abstract way and without relation to actual public services within a given polity, it is difficult to conclude whether scale produces economies or diseconomies. But since public discussion seems to assume that there are economies of large scale, it is especially in this direction that a caveat is necessary. In addition, we should not forget that the above considerations about costs are primarily related to the production side of public services. They shed little light on the problems of how these services are made available to the public and do not say much about optimal political organization, especially when we recognize that for each public service the optimal area will be different. In a general way and returning to the question of federalism, we may say that a federal system, with its three or more different sizes of government, provides for a certain balance of economies and diseconomies of scale (thus Ostrom L593/229).
471 «Naturally, there are trade-offs between size and number. From the federal viewpoint, the curse is taken off size by number: the larger the number of units involved, the larger each unit can be without decreasing effective delivery of services. The value of size, then, depends on number: the smaller the number of units, the more large size contributes to ineffective delivery of services» (Wildavsky L861/107).
E.1214 Spillovers
472 Under economic theories of federalism, spillovers or spatial externalities constitute an important cost factor. Spillovers are the result of public as well as other goods being offered or produced in one area, but also benefiting (or burdening) the inhabitants of other areas who do not participate in their costs (or who do not enjoy corresponding advantages, compare Beer L57/63ss, Pawlowsky L607/17s). Some examples are state universities, theatres, and local industries. State universities in Switzerland are financed by one canton and attended by students from other cantons which, until recently, did not contribute to the costs. Theatres often involve a heavy financial burden for their cities. But the theatregoers who profit from their existence may very well come from other localities which contribute nothing to the deficits. The opposite situation may arise in the case of industries that pollute a wide area, while tax and other benefits accrue only to their own local governments.
473 If we look at spillovers from a purely financial viewpoint without taking into account considerations of justice, we may ask whether in the context of the national economy their existence has a tendency to lead to oversupply or to undersupply of public goods. Economists are not unanimous on this subject (see Frey L266/96, Pawlowsky L607/59). At least until recently the advocates of the undersupply thesis were in the majority. They base their case on the assumption that member states only produce a public good until its marginal costs coincide with the marginal benefits for the states' own inhabitants (L266/96s). Against this Williams and Frey demonstrate that spillovers, in terms of the national aggregate may result in excessive expenditures if, in addition to the reactions of governments, private investment and labour force mobility induced by infrastructure expenditures are also taken into account. Because individual regions utilize infrastructure expenditure as an instrument for regional economic development there will be an oversupply in infrastructure if all governments adopt the same policy (Frey L266/97, Pawlowsky L607/59s,64).
474 Of course it would be ideal if, for each public good, the spatial boundaries of service provision could be circumscribed in such a way as to have no benefits or costs transgressing them. Spillovers would then be internalized. Such a solution, consisting in creating a special district for each public good, is at most a theoretical possibility (compare public choice theory, no.516ss, Pawlowsky L607/122). Some externalities will always remain. They may be small scale, as for instance local border effects as described by Tullock (L820/27). Or they may be large scale, as illustrated by economic stabilization or redistribution effects. Perfect internalization presupposes unitary government (in strict logic it is only feasible under world government, L820/27). Considerations such as these cause Riker to claim: «Hence in all cases a uniform policy is cheaper from the point of view of external costs than a non-uniform policy» (L660/147). «Decisions made by constituent units are invariably minority decisions that impose high external costs on the national majority. This assertion is wholly irrefutable on the level of theory» (L660/151). The flaw in this reasoning is its use of an analytical instrument for discrete relationships as a recipe for government organization which is always multi-dimensional. Or as Pawlowsky describes what is in his eyes the basic weakness of spillover analysis (L607/75): «It is an attempt to solve a macro-economic problem with a micro-economic tool.»
475 Spillover analysis is a valuable instrument if the problem is how to organize regional financial transfers or to define the area of a special district. Spillover considerations play a prominent part in the justification of federal grants-in-aid (L264b/1114). To use them as a basis for large-scale governmental reorganization may be tempting because of the spillover concept's one-dimensionality and apparent clarity. But it is this one-dimensionality which leads us down a blind alley, probably in theory and certainly in reality. «Spillovers are on their own no sufficient justification for the creation of larger areas of government or for the centralization of public services» (Pawlowsky L607/133, compare Bird/Hartle L70/51ss).
E.122 Quality of Federalist solutions
♦ 476 «It seems clear on the basis of both theory and example that uniform national decision-making is invariably more efficient... than is local decision making» (Riker L660/147). «Etzioni warned: City and state administrations, hard as it is to believe, are generally, on every account, more inefficient and ineffectual than are the federal agencies» (L181/8). - «The philosophical argument of federalism is based on the highest possible utility, in the conditions of a particular time and place, in the course towards decentralisation or towards centralisation. This being the philosophical assertion of federalism, federal theory in our own time should be based on pragmatic utilitarianism» (Mogi L550/1059). «The test of federal theory is federal action - superior delivery of services» (Wildavsky L861/106).477 These citations show that, for both critics as well as apologists of federalism, the quality of service delivery is an important aspect. To what degree is it, then, possible to make valid generalizations about a federal system's efficiency as compared to a centralized one?
478 The first problem we encounter is the concept of efficiency itself. It is a popular term. But it is a difficult one if we want to use it as a workable instrument for analysis (L346/212). Scharpf's definition is cost oriented (L704/100, see also Robinson L667/9): «Efficiency, to define it within a goal/instruments model, is either the attaining of a given end with minimal use of resources, or the use of given resources to achieve a maximum degree of goal fulfillment.» This is the aspect I dealt with above (no.447ss). But the concept of efficiency may well be a broader one, such as Lehmbruch's (L465/94s): «We might define it as the capacity of a political system to resolve, within a reasonable time, the problems considered as vital or important by substantial segments of public opinion, including those problems that can be expected to gain such importance in the foreseeable future.» Such a definition, which we will now focus on, emphasizes the quality of solutions.
479 For the sake of clarity, a qualification is called for. We are concerned only with problems that society's decision-making process defines as problems which need to be solved. These we also term problems in everyday linguistic usage. They range from small problems, involving for example the building of a school or the regulation of liquor retailing, to large-scale concerns, such as national defence or the reduction of regional economic disparities. On the other hand, we are not here concerned with system problems, that is problems connected with the existence of a social system as a system (compare no.23). Such problems would involve a system's learning capacity (adaptability, change, efficiency, etc.) or its stability (integration, legitimation, etc.). These are the themes of our concern with the functions of federalism generally.
480 If by efficiency we mean to designate the quality of a solution we use the term as the basis for evaluation. We need yardsticks. These do not exist per se. They result from the same social decision-making process that defines the problems to be solved. Efficiency, then, is the degree to which a solution attains its end and thereby makes the problem disappear, without causing other avoidable problems. Efficiency is the successful management of problems. Efficiency is relative. Absolute efficiency is an unattainable ideal.
481 It is unlikely that we can arrive at a clear-cut judgment about the efficiency of federalism. We may say that the organization and maintenance of this form of political structure in a given society is an indication, by and large, that it is considered to be efficient. But it is obvious that for each of the society's members the assessment will depend on his or her personal conception of the hierarchy of public goals. If someone values an energetic foreign policy highly (e.g. Chenaux-Repond L782/1011) he will rate federalism with its tendency to complicate this differently from someone whose paramount aim is the peaceful living together of groups with markedly different value systems and who therefore prefers weak central government to none at all (e.g. McWhinney L533/6).
482 For the purpose of the following analysis I distinguish between two dimensions of quality: technical adequacy and public satisfaction. Technical adequacy means that a given problem is solved correctly according to the state of technical expertise in the relevant field. A data processing problem, for instance, is solved correctly when the computer used has been chosen, programmed, and fed with information in a way that will satisfy a computer expert that the solution uses the equipment's possibilities according to specifications, economically, and in a manner which does not foreclose future changes. Public satisfaction, on the other hand, obtains where a solution is considered to be the right one by the public, including the beneficiaries, those who bear its costs, and outsiders. In this view the data processing problem mentioned is correctly solved when the citizen is satisfied that a solution is really in his interest, that it does not occasion unnecessary costs, and that it does not invade his privacy more than is strictly necessary. It is obvious that these aspects are not mutually exclusive: technical adequacy per se does not exist. It also requires evaluation. But the differentiation between criteria which are deduced from the properties of a problem itself and those that are founded in social attitudes does elucidate the two essential dimensions of the concept of quality.
483 These two faces of quality may also be found in practical politics. To-day's reform postulates usually demand more efficient, i.e. technically adequate, solutions which should not result in more bureaucracy but, on the contrary, bring about more citizen participation. Proposals to merge communes, with a view to achieve more efficient local government, are nearly always accompanied by statements to the effect that the citizens' preferences would be better heeded or that their participation would be increased under the new regime because of more political accountability or better neighbourhood representation (Schwartz L736/326s).484 Before delving further into the theme another qualification has to be made. Here I treat only the general suitability of federal organization for the efficient solution of problems. I do not go into the question of the efficient allocation of functions amongst the different levels of government in a federation. This is the subject of a chapter in volume II of my German text (L264b/894ss).
E.1221 Quality as Technical Adequacy
485 I dealt above (no.457ss) with economies or benefits of scale. Wherever they are attainable they presuppose larger jurisdictions. They are therefore centralizing. The benefits do not necessarily have to be of a financial nature. They may also be qualitative ones. A university must have a minimum number of students, a hospital must have a minimum number of patients to be able to maintain a certain qualitative level of specialist services. Some political tasks also demand a large or national scale, as for instance defence, interregional economic adjustments, and some aspect of equality.
486 However, this does not answer the question of how these tasks are to be organized. There are different ways for that: centralized, decentralized, part-centralized (e.g. through central skeleton laws), co-operative (e.g. through interstate compacts), or non-centralized (e.g. harmonizing effects of quasi-market mechanisms, using the «invisible hand»). Any federal system utilizes nearly all of these organizational elements. But it is primarily the qualitative problems of local problem solving (modelled to the pattern of subsidiarity, no.231ss) that are the focus of the following considerations.
487 First, territorial differences of all kinds often make it essential to regionally differentiate public policy. Thus, it has often been claimed that federal organization is best suited to the evolution of large-area government because it combines political unification with regional adaptation (Bryce L96/351, Dikshit L173/14). The expediency of having regional adaptation within and by the politically organized regions is based on several factors.
488 Distortion of information in multi-level hierarchies has already been mentioned (no.427). The probability of disparities between a centre's ideas about the effects of some policy and the policy's actual effects in the local context is thus quite high. Attempts to correct deviations through increased control very often bring results that are no longer justifiable in relation to the resources deployed (Ostrom L592/60). Then there is the natural tendency of any centre to view local adaptation as involving failures of central controls which have to be rectified with more controls. This may be an inadequate response if the mistake lies with central planning to begin with (van Gunsteren L320/19).
489 We may also perceive a certain loss of selectivity through increased distance. The farther away a centre of decision and the more encompassing its responsibilities, the more difficult it is for it to deal adequately with small problems or problems of a predominantly local character. A megalopolis like New York is not able to repair pot-holes or to keep pavements clean (L593/223). In Los Angeles, since the amalgamation of school districts, it has become impossible to deal with cockroaches in school buildings (L592/121). «It does seem apparent, though, that ceteris paribus the more localised the benefits generated by the service, the more likely it is that this 'uniformity inefficiencies' argument will support localised political provision» (Walsh L844/20). «Tasks with technologies that require mutual adjustment and/or have environments which are shifting and heterogeneous require an organizational structure which includes small, localized units. To be effective, these units should be endowed with considerable discretion, for they must make judgments about appropriate responses to clients or decide how to cope with complicated environmental situations» (Porter L625/91s). Such autonomous units are to be found in the elements of federal systems.
490 Scharpf's argument uses the same approach (L705/22): «Political tasks and problems may arise locally or be functionally delimited. They may then be dealt with by local or functional decision centres which are narrowly circumscribed but which have a higher problem sensitivity and a swifter responsiveness than is the case with larger centres with pervasive responsibilities.» Tocqueville arrived at similar conclusions (L806/132): «Central power, however enlightened, however wise it may be, will never be able unaided to grasp all the details of a great people's life. It cannot do so because it is humanly impossible to achieve such an aim. If it nevertheless attempts to create on its own so many different things and have them work it will either be content with a result that is very incomplete, or it will exhaust itself in futile efforts.»491 The reasons for quality advantages in local solutions are their lesser proneness to information distortion - usually described as greater familiarity with local situations (L267/370) -, faster information treatment (L682a/15), and the more direct interest in sensible decisions of those who are at the same time their authors and the bearers of the consequences.
492 This last kind of interest was classically formulated in 1839 by the Earl of Durham in his famous report on Canada. Although concerned with the autonomy of a colony, his words are also applicable to problems of federalism (L384/166s): «The colonists may not always know what laws are best for them, or which of their country-men are the fittest for conducting their affairs; but, at least, they have a greater interest in coming to a right judgment on these points, and will take greater pains to do so, than those whose welfare is very remotely and slightly affected by the good or bad legislation of these portions of the Empire. If the colonists make bad laws, and select improper persons to conduct their affairs they will generally be the only, always the greatest sufferers; and, like the people of other countries, they must bear the ills which they bring on themselves, until they choose to apply the remedy.»493 Society's demands have steadily risen during recent decades. This has brought about a corresponding growth in government responsibilities and increased political complexity (see Neidhart L573/39). The adequate response to this problem is not the design of a less complex, i.e. a more centralized political system, as is the current fad of political journalism and many scientists. The proper approach is rather the setting-up of a differentiated governmental structure, as for instance a federalist one. This provides for requisite variety to deal with environmental complexity, thereby reducing the latter and making it amenable to processing (compare Baars L42/1, Huber L782/1160s, Ruffieux L782/1152).
494 Still another advantage of the federalist principle may be seen in its furthering of quasi-market mechanisms (Ostrom L593/230). The citizen sees with his own eyes that public tasks may be tackled differently and sometimes with unequal success. This may result in a quality-improving learning process. I refer here to considerations already presented in the context of federalism and innovation (no.434ss).
495 Finally, it is obvious that the many and mostly part-time public positions and functions which exist in a federalist system result in a much larger part of the population gaining political and public managerial expertise than is the case in unitary states. Cadre selection and preparation rest on a considerably broader base.
496 But local solutions are not uniformly better. They also have specific disadvantages. It is only partially true to say, for instance, that local government is more accessible to the individual citizen than big government, thereby making it more likely that decisions will be technically correct. It may well be that there is less openness in middle-sized communes than in larger communities which are more closely monitored by the media (Kock L430/49, L513/189ss). The elite might find it easier in middle-sized communities to push their private interests to the detriment of the public ones.
497 An additional negative point is the tendency of local solutions to consider only local costs and benefits and to neglect disadvantages for others (as to spillovers see no.472ss) and superior goals (e.g. an expenditure policy in agreement with national economic aims). Frey (L268/35) gives the example of hospitals or mountain resorts where city people pass their holidays. «If it were but the citizens of the localities and Cantons concerned who decide on the service infrastructure the national welfare would be diminished.» Intergovernmental competition may also result in the same disadvantage because self-interested behaviour of each government may sometimes result in the failure of each of them to take the most appropriate measures. Such is the case when no member state may raise its level of taxation because each of them fears that some of the richest taxpayers would in such an eventuality migrate to another one who, thereby, might solve its own revenue problems through a hold-out strategy. A similar illustration is the weakening of public control over large enterprises because of interstate competition in the United States (Stevenson L779/40).
498 Again, local solutions may be inadequate if they «are confronted with material problems the extent of which transgresses the limited authority (responsibilities, competences, resources, and goals) of the individual decision centres» (Scharpf L705/22). The reduction of a system's optimizing potential may also be considered a negative aspect, i.e. the lessening of the possibility of combining individual problems which have little to do with one another in such a way as to achieve a result which, from a technical point of view, may not be the best for each of them but which is optimal in a global perspective. Examples of this may be found in the distribution of scarce resources or in foreign policy making.
499 If we attempt a summing up of the aspects of technical adequacy, it appears that local solutions have great advantages which result from close distance. But this, to be meaningful, has to be seen in the context of an overarching system that sets the threshold values as well as the rules of the game and which provides for co-ordinating machinery.
500 A striking parallel to this embedding of a small system in a larger one is provided by Scott's inquiry into corruption, especially in the form of the former «political machines» in U.S. cities (L739/156): «The general line of reasoning developed above suggests that the machine flourishes best at the subnational level,which is where it was confined in the United States. That is, the durability of this political form is maximized where there is an external guarantor of the electoral process, where the machine is part of a larger growing economy that can afford its expensive habits, and where its bosses do not have a monopoly of coercive authority. A large measure of the instability of machines in developing nations may thus derive from their national rather than local character.»
E.1222 Quality as Public Satisfaction
♦ 501 «The State should offer those services that correspond to its citizens' preferences.... If we start from an assumption which is by no means incorrect, i.e. that in a given civil community there are no homogeneous value acceptations to deal with the organization of its economy and its environment, but that there are different groups that... differ in their concepts, it is easy to realize a federal structure's advantages. Federal decentralization is better apt to take into account the different preferences of the community's members than are extremely unitary states with their schematically uniform approach.... Political decisions always entail the submission of a more or less substantial minority to the wishes of the majority which, for them, results in a sacrifice of welfare. A federal structure lessens the numbers of those who find themselves in a minority position» (Frey L268/27).502 Before we may solve a problem we have to know what it is. It is again the social decision-making process that makes society's problems exist, i.e. perceived as such. This leads us to inquire federalism's part in the definition of problems. We may say in this context that the federalist perception is more responsive than a centralist one to the citizens' actual wishes. There are several reasons for this. Negative central aspects are the professionalized creation of demands and the lesser adaptability to local preferences and variations. Both factors have been dealt with above (nos.468, 488). Positive, non-central aspects flow from the more direct contact with the problems at hand (no.507) and with the option to choose among different sets of public services, that is to say with the introduction of some sort of market-mechanism.
503 The latter phenomenon may be illustrated by the opportunity which the citizens of a federation have to choose their place of residence among different member states with differing patterns of policies. «Thus, the public services and the different levels of taxation are valued nearly as in a market-place» (Frey L266/96, compare Ostrom L593/230, Pawlowsky L607/130). The same kind of choice is made by the students of the University of Michigan who prefer to be arrested by the local chief of police rather than by the more adamant sheriff. Or we may point to U.S. lawyers, who are often able to start an action in either a state or a federal court and who will obviously choose the court they consider more sympathetic to their case (L46/49). Tiebout has argued that public authorities at all levels are not very sensitive to citizens' wishes. But in a decentralized system this is not so important as long as there are variations of services offered amongst the various units of government which make it possible for the individual citizen to select that pattern of services and taxes which most closely corresponds to his personal preferences (L803/34, see King L419b/28ss, Pestieau L585/173ss, Walsh L844/18). A federation, in this view, appears as a continuing consumer plebiscite with the consequential gains in welfare as postulated by neo-classical market theory (Winkel L867/57ss). The Tiebout hypothesis sounds plausible. There are cases for which it may serve as a direct explanation, as for instance for migrations resulting from sudden local tax increases. But by and large the applicability of the hypothesis is limited. To «vote with one's feet» may be considerably impeded by the fact that the costs of migration are higher than the advantages expected elsewhere. This is especially so if we also include intangible advantages and disadvantages (Frey L268/28). In addition to that, Buchanan and Goetz have demonstrated (L421/12, L844/19) that individuals have usually but very fuzzy ideas about the cost/benefit effects of their choice of residence.
504 Notwithstanding the actual relevance of the Tiebout hypothesis, for the reasons mentioned we may say that from the viewpoint of public satisfactions there are distinct advantages in locally organized problem perception. But we must take care not to be too dogmatic about this. It may be that local decision-making is oriented too exclusively to the interests of a small group because of a lack of accountability (no.496). It may also be that local majorities are very intolerant when dealing with minorities. «In addition, far from encouraging wide discussion and the extensive canvass of alternatives, politics in the community setting can be oppressive and intolerant.... The pressures for ideological conformity and purity flourish in small and communal settings» (Eisinger L197/147). Sometimes, because of that, minorities will find a better audience for their demands with central politicians than with local ones, because it is often easier to organize themselves and to wield clout at the national level. «Blacks have learned to trust the federal government more than state and local governments because the federal government has been more protective of their rights and responsive to their needs» (Baines L44/117, compare no.512, L46/46s, L533/65).
505 Riker (L660/153ss) uses this argument as a basis for his scathing criticism of U.S. federalism: «The judgment to be passed on federalism in the United States is therefore a judgment on the values of segregation and racial oppression.... Thus, if in the United States one approves of Southern white racists, then one should approve of American federalism. If, on the other hand, one disapproves of the values of the privileged minority, one should disapprove of federalism. Thus, if in the United States one disapproves of racism, one should disapprove of federalism.» These polemics, which cast away the good with the bad and which are somewhat fortuitous in the context of Riker's analysis, do not represent a balanced assessment. The real beneficiaries of federal programmes for underprivileged coloured minorities are often city-hall unions and bureaucracies that work «for» them rather than the minorities themselves (Long L490/24s, compare Derthick L637/6-2/121ss).506 Here we are confronted with the most clear-cut conflict between federalism and democracy. But it is in reality a conflict between two dimensions of democracy: large-scale and small-scale democracy. Before addressing myself in more detail to this problem (no.519ss) I will discuss further federalist solutions and public satisfactions, that is the standards of efficiency themselves.
507 It is rarely the case that a problem is technical to such a degree that the standards for its solution may be directly derived from the matter at hand. The yardstick is itself an object of the process of social decision making. In other words: efficiency is not absolute. In the final analysis it is a value judgment. To the extent that we deal with public services it therefore makes sense to use the appraisal of a service's putative beneficiaries or cost-bearers as a yardstick for the solution's efficiency. This, incidentally, is also the standard most in agreement with democratic theory.
508 Only recently has empirical research started to look into the relationship between consumer satisfaction with public services and organizational arrangements, and this especially in the United States. Elinor Ostrom, for instance, has tested the validity of the hypothesis that amalgamation of police districts in urban agglomerations entails more efficient policing than it is possible for fragmented autonomous departments to achieve (L590/33ss). Several comparable agglomerations with differently organized police forces were analysed in depth. The results of her inquiry point rather in the opposite direction. It has become clear that most of the specialized technical services which can only be organized for a large area, and which are usually the starting point for projected mergers, may also be contracted for by small districts or made otherwise available to them. On the other hand, urban areas with small-scale police organizations show a marked tendency to greater satisfaction of inhabitants with the police services available to them. This is in addition to the fact that these departments are comparatively less costly. One of the reasons for these findings may be the fact that large police organizations invest more of their funds in administration and crime detection than small ones, which operate patrols directed to crime prevention and thereby more visible to citizens (V. Ostrom L593/225s). After statistical analysis of a variety of indicators E. Ostrom concludes (L590/70): «The evidence presented here should not be interpreted to support contentions that all efforts to consolidate police departments are likely to lead to less effective or efficient performance. That some of the relationships appear to be curvilinear suggests that departments in the 10 to 76 range may have higher levels of performance on some variables than either the small or large departments. Consolidation of small departments into a medium-sized department may indeed increase effectiveness. However, the a priori judgment that very small departments are necessarily ineffective and should be eliminated is not supported by this evidence. In fact, the evidence finds the severe problems of law enforcement in the large departments. If future study commissions wish to focus on the most ineffective segment of the police industry, they should shift their focus to the large.»
509 The California Local Government Task Force (L346/214), after polling some 1500 citizens, also arrived at the conclusion that citizen satisfaction with local government is directly related to small size (we should, however, bear in mind the possibility that smaller units of government in California may also be comparatively rich and not burdened with too many social problems). Conclusions similar to that of Elinor Ostrom are to be found in a survey of empirical studies about school organization (L590/73). Also a Harris poll of some 16'000 Americans came to the conclusion (L346/214): «1. Citizens found larger cities to be more disorganized than did their counterparts living in smaller cities, which are serviced by more units of government. 2. Citizens living in smaller units of local government tended to use their local governments more than those in large cities. 3. Citizens in smaller jurisdictions tended to have higher levels of satisfaction than those in large cities.» An inquiry by Rogers (L669/19ss) about public service evaluation by the residents of two similar suburbs concluded: «It became quite obvious that the assumptions supporting metropolitan consolidation were not supported by our findings. Consistently, residents of the small, independent community, Berry Hill, expressed higher degrees of satisfaction with 1) the level of public services provided, and 2) the concern of local officials for the welfare of the community than did residents of Woodbine - the community serviced by a large, metropolitan government» (L669/32).510 It is evident that these American findings must not be applied directly to other countries. But if we consider the critical attitude of the U.S. media towards local efficiency we may nevertheless assume that similar results may be arrived at in countries where criticism is not as harsh as there. In addition, these results are consistent with the theoretical considerations presented above.511 Government organized according the principle of subsidiarity also entails greater citizen efficacy. The individual citizen's impact on policy formulation is relatively significant (Ostrom L593/229, Prewitt L736/223). «Each merger of local units and every centralization of public responsibility will bring about higher political costs for the individual citizen. The probability that his preferences are similar to those of the median voter decreases; the probability that majority decisions deviate from his own wishes increases» (Pawlowsky L607/131).
512 This holds true primarily for individuals who are in agreement with the local, homogeneous majority. The effectiveness of dissidents, on the other hand, will be smaller because of the absence of allies and the lesser importance of political competition. They will presumably find a more heterogeneous, larger governmental unit more responsive (Dahl/Tufte L156a/138, no.504).513 One additional form of citizen effectiveness in a federally structured state is provided by the opportunity for citizens to voice their demands at all levels of government without taking into account the constitutional allocation of tasks. They may thus make one level of government their advocate in relation to another one, where they will directly intervene, or play one government against the other (May L528/6).
514 But to think, as it is sometimes done (e.g. Usteri L826/11), that in a federation the individual citizen's voice has a double weight because of the double majorities needed to carry constitutional referenda - votes aggregated nationally as well as by member states - is an error. Double weights can only be exercised by voters who happen to belong to their local majority but whose states are not the most populous ones. The individual citizens of the large states are disadvantaged by the double majority system.515 If we sum up our findings about efficiency in relation to satisfactions, we arrive at very similar conclusions to those about efficiency as technical adequacy (no.499). There is a considerable array of both theoretical and empirical reasons for a local bias. Local solutions correspond often more closely to public interests than is the case with a more professionalized central approach. This holds true especially for solutions that have primarily local effects and where local adaptation is important. But today it has become difficult to define problems clearly as either local or national. And the public is not only local. There is also a regional one, a national one, etc. This again supports the case for a complex, pluricentral system. Or with Dahl and Tufte: «No single type or size of unit is optimal for achieving the twin goals of citizen effectiveness and system capacity» (L156a/138). «1) Only a complex polity consisting of a number of interrelated political units above and below the level of the nation-state has the capacity to respond adequately to the collective preferences of citizens of nation-states. 2) Therefore units at and below the level of the nation-state should not be wholly autonomous» (L156a/25).
516 The most complex model of federalism is the public choice or marketplace model that is associated with the names of Olson, Tullock and the two Ostroms (especially L596/129ss). Its starting point is that of the production and consumption of public goods. The residents of a certain area form a collective-consumption unit for one or more public goods, as water provision, education, road maintenance, s