MPIfG Working Paper 98/2, September 1998

Interdependence and Democratic Legitimation

by Fritz W. Scharpf 

Fritz W. Scharpf is director at the Max Planck Institute for the Study of Societies, Cologne

Introduction

The premise of the Bellagio Project on Democracy has been that, in recent decades, Western democracies have come to suffer a decline of political "trust" or "confidence" in, or popular "satisfaction" with, the "performance" of their representative institutions, and that this decline needs to be taken seriously as a potential threat to the viability of democratic government (Putnam 1998). The terms used also suggest that the project starts from an implicit principal-agent model in which citizens-as-principals have come to be dissatisfied with the performance of their political agents. If we assume that this is empirically true, and that the change does reflect a deterioration of perceived performance, rather than the rising (or increasingly conflicting) expectations of citizen-principals, there still are two fundamentally different working hypotheses from which one might begin the search for an explanation. Growing dissatisfaction could be caused by factors that have reduced the fidelity of agents -- i.e., their willingness to act in the interest of their principals. But it also could be caused by factors that have constrained the objective capacity of agents to achieve the outcomes expected by principals. Whereas the project as a whole is exploring the first of these working hypotheses, my own paper will focus on a particular type of capacity constraints: growing international economic interdependence. In doing so, I will not review the empirical evidence regarding changes in the levels of popular satisfaction, except to note the high degree of variance among countries (Newton, 1998; Katzenstein 1998). Instead, I will examine the analytical and normative arguments that could link economic internationalization to citizen satisfaction, and ultimately to the democratic legitimacy of national political systems. I will argue that one should indeed expect such links to exist, but that their effect on legitimacy will be strongly mediated by the characteristics of national political discourses.

Interdependence as a Challenge to Democratic Legitimacy

From the Athenian city state to the modern nation state, democratic self-government has been defined by reference to the territorially-based constituencies of local, regional, and national governments. It is true that self-governing associations with a geographically dispersed membership and with internal structures and procedures of a democratic character do exist. Examples that come to mind are professional associations, labor unions, some clubs, and perhaps some non-governmental organizations like Amnesty International or Greenpeace. However, the governing powers which such associations are able to exercise over their membership are either very limited, depending essentially on voluntary compliance, or they are exercised "in the shadow of the state", on whose laws and enforcement machinery they must rely when voluntary compliance is not forthcoming. The monopoly of legitimate coercion, at any rate, on which the problem-solving capacity of democratic self-government continues to depend, has only been achieved within territorially defined units.

But if democratic self-government is defined by reference to territorially limited constituencies, it must be vulnerable to increasing military, economic, technical, ecological and communicative interdependence among territorial units. Under such conditions, choices within a given unit will create, and suffer from, external effects. Spill-outs may reduce the effectiveness of domestic choices, and spill-ins may produce domestic outcomes that have not been chosen internally. In the following sections, I will explore the reasons why the lack of congruence between the constituencies of democratic governments and the populations that are affected by governing decisions may be considered a major problem for democratic legitimacy.

Input- and Output-Oriented Democratic Legitimacy

Democracy is a concept with a variety of meanings, but when we speak of "democratic legitimacy," the reference must be to arguments that can be used to justify the exercise of governing authority -- i.e., of the authority to adopt collectively binding decisions, to implement these with resources taken from the members of the collectivity, and ultimately by resort to the state's monopoly of legitimate coercion. Legitimating arguments, then, must be arguments that are able to establish a moral duty to obey these collectively binding decisions even if they conflict with individual preferences.[1] In the modern period, the concept of democracy has become the major foundation of such legitimating arguments. Their basic appeal was most succinctly expressed, in Abraham Lincoln's Gettysburg Address, by reference to the triple identity of the governed ("government of the people"), the governors ("government by the people") and the beneficiaries of government ("government for the people"). But why should Lincoln's formula create a moral duty to obey acts of government?

Leaving aside for the moment the first element that defines the collectivity that is to be self-governing ("government of the people"), the formula points to two analytically distinct dimensions of democratic legitimation, one input-oriented, the other one output-oriented (Scharpf 1970). In the input dimension, "government by the people" implies that collectively binding decisions should originate from the authentic expression of the preferences of the constituency in question. Government, in other words, is meant to be self-government, and compliance can be expected because the laws are self-determined, rather than imposed by an exogenous will. In the output dimension, "government for the people" implies that collectively binding decisions should serve the common interest of the constituency. Obedience is justified because collective fate control is increased when the powers of government can be employed to deal with those problems that the members of the collectivity cannot solve either individually, or through market interactions, or through voluntary cooperation.

However, by using the singular term for describing the plural originators and beneficiaries of democratic government, both of Lincoln's criteria avoid the critical question of how the exercise of governing authority, and the duty to obey its commands, should be legitimated if "the people" is not considered either an organic unity or an aggregate of homogeneous individuals, but an association of individuals and groups whose preferences may diverge and whose interests may conflict with each other. Within a purely input-oriented frame of reference, there are two possible responses to this problem. The first postulates that government should be consensual, based on the widest possible agreement among the individuals and groups affected, whereas the other one seeks to justify decisions based on the expressed preferences of a majority of the membership (Lijphart 1984; 1991).

From an input perspective, consensual democracy has first-rate credentials, resting ultimately on the Roman-law maxim of volenti non fit iniuria (meaning that if you have consented you cannot claim damages). Its weakness lies in the output dimension since, in the face of divergent or conflicting preferences, the search for consensus may prevent the adoption of any effective solution. Hence output-oriented concepts tend to favor majoritarian democracy because of its greater problem-solving efficiency (Buchanan/ Tullock 1962), but must then try to assure that majoritarian policies will indeed serve the public interest. I will return to that point shortly.

Within a purely input-oriented frame of reference, the justification of majority rule is more demanding than is often assumed.[2] As I have tried to show elsewhere (Scharpf 1997a, chapter 7), it must ultimately presuppose preferences of the majority that will somehow include the welfare of the minority as an argument -- an assumption that must rule out not only the possibility of hostile majorities (think of Nazi Germany or Bosnia) but also the "bloody-minded" pursuit of rational self-interest. At bottom, therefore, notions of democracy that rely exclusively on the "will of the people" as a source of political legitimacy must assume conditions of a strong collective identity, and a pervasive sense of common fate and common destiny, that overrides concerns based on divergent preferences and interests. Only if these Rousseauian assumptions are fulfilled, is it indeed possible to treat the preferences of the majority as a true expression of the volonté générale which the minority, being mistaken, would be wrong to oppose.

But in light of the totalitarian potential of the Rousseauian tradition (Talmon 1955), and of pervasive misgivings about the cognitive and normative shortcomings of "populist democracy" (Sartori 1965; Riker 1982), modern democratic theory rarely derives legitimacy primarily from the belief that "the people can do no wrong." Where input-elements dominate, theorists take care to restrict the domain of "participative democracy" to the micro-level of local or shop-floor decisions (Lindner 1990), to emphasize procedural safeguards against the dangers of "direct democracy" (Luthardt 1994), or to insist that policy inputs should arise from public debates that have the qualities of truth-oriented deliberations and discourses (Manin 1987; Dryzek 1990; Schmalz-Bruns 1995; Habermas 1996). In effect, the ideal of "deliberative democracy" may also be understood as a concept that builds a bridge between input- and output-oriented legitimating arguments by insisting on specific input procedures that will favor qualitatively acceptable outputs by regulating "the flow of discursive option- and will-formation in such a way that their fallible results enjoy the presumption of being reasonable" (Habermas 1996, 301). I will return to this point below.

In any case, however, input-oriented justifications of majority rule are complemented everywhere by output-oriented criteria with a negative and a positive thrust. In addition to requiring that government should be capable of achieving effective solutions to collective-action problems, output-oriented criteria must also define what governments should not be allowed to do, in order to be considered "government for the people." The emphasis here is on institutional arrangements that are meant to provide protection against the danger that governing powers of the majority will be abused to the detriment of minorities or individuals, and to assure that these powers will only be used to further the common interests of the members of the constituency, rather than the special interests of office holders and their clienteles. These institutional arrangements include constitutional guarantees of individual rights, an independent judiciary and other forms of "checks and balances" as well as the mechanisms of representative democracy which, on the one hand, provide opportunities for public debate, reflection and criticism that are thought to discriminate against self-serving policy choices (Habermas 1962; Elster 1986) and which, on the other hand, are meant to assure the direct or indirect accountability of office holders to the general electorate.[3]

Countries differ greatly in the extent to which their institutions emphasize the negative requirements of output-oriented legitimacy by creating veto positions and electoral vulnerabilities that make it more difficult to coordinate and employ the policy resources available to government as a whole in coherent and effective policy choices (Tsebelis 1995). These differences between, say, the concentration of powers in the British "Westminster Model," and the dispersion of powers in the present German constitution (with coalition governments, opposition veto in the federal chamber, an activist constitutional court, and an independent central bank) are rooted in historical experiences and path-dependent courses of institutional evolution, and there is no reason to expect convergence (Pierson 1997).

With regard to the positive dimension of problem-solving effectiveness, by contrast, democratic theory has generally built on the foundations of the sovereign "Westphalian" state. It is taken for granted, therefore, that the democratic state, like its absolutist predecessor, is potentially omnipotent within its own territory and able to control its external boundaries. There are physical constraints on the internally available resources, and boundaries may be violated by military invasions from abroad. Within these limits, however, the democratic state is as capable as its non-democratic predecessors and competitors of taxing its residents, of regulating their actions with the force of law, of requisitioning their property and their services, and of requiring them to risk their lives. Any limits on these capabilities are thought to be self-imposed -- either by constitutional norm or by political choice. In principle, therefore, the state has the means to achieve all normatively acceptable and politically consented domestic purposes, and governors are rightly held politically accountable for failing to do so.

Interdependence and the Loss of Congruence

If this now seems an unrealistic idealization, in light of growing international interdependence, it is one that was closely approximated in the most important relationship between the state and the capitalist economy not so long ago. In the first three decades following the Great Depression and the Second World War, Western democracies had finally learned to control the cyclical crises of the economy, and they were able to meet the aspirations of their citizens for full employment, rising mass incomes, rising levels of education, lower inequality, and social security in times of unemployment and sickness, and in old age. Only recently, however, has it become evident how much this "Great Transformation" (Polanyi 1957) depended on the fact that, after the rampant protectionism of the 1930s and the Second World War, capitalist democracies were for a while able to control their economic boundaries: Goods and services that did not conform to domestic regulations or that threatened the survival of domestic producers could be excluded, the outflow of capital could be prevented, and immigration was tightly controlled. Under these conditions, national political processes were able to choose among a wide range of options, and while the Scandinavian welfare states differed greatly from the German social-market economy or from post-New-Deal America, all were equally economically viable and legitimated by broad political support.

During these "golden decades", interdependence was slow in increasing under American-led international regimes of "embedded liberalism" (Ruggie 1982), and it was not until the 1980s that pre-1914 levels of international integration in product and capital markets were again reached and surpassed (Hirst/ Thompson 1995; Bairoch 1997).[4] Since then, however, the nation state has again lost control over its economic boundaries. This is most obvious within the European Union, where the completion of the internal market for goods, services, and capital is now being topped by the creation of a monetary union. But beyond Europe as well, repeated rounds of GATT and WTO negotiations have drastically reduced tariffs and non-tariff barriers to trade in goods and services, and the explosive increase of transnational money flows has eliminated any chance of maintaining protected national capital markets.

Consumers are thus free to buy goods and services regardless of their location of production; firms are free to produce at any location without endangering their access to the home market; capital is free to take advantage of profitable opportunities for investment or speculative transactions around the globe and around the clock; and workers are free to choose their places of work at least within the European Union. Governments however, being held accountable for the economic and social welfare of their constituents, must be concerned about the potential loss of jobs if demand for nationally produced goods and services should decline, if firms should relocate their production to other countries, if capital owners should prefer investment opportunities offered elsewhere, if highly-skilled workers decide to emigrate, and if taxpayers or their taxable resources should leave the national territory.

The Impact of Regulatory Competition

As a consequence of the potential mobility of economic actors and factors there is now a much greater degree of interdependence not only between the formerly compartmentalized national economies, but also between national policy choices that have an effect on the economy. If one government cuts its social security contributions, that reduces the international competitiveness of products from other countries that have not done so; and if one country cuts its rate of corporate taxation, that will create incentives for firms to relocate their company headquarters. Hence it is indeed wrong to think that only firms are in competition with each other. Economic interdependence creates a constellation in which nation states find themselves competing with each other for market shares in product markets, for investment capital, and for taxable revenues, and in which that competition constrains their choices among macroeconomic, regulatory and tax policy options.

From the perspective of democratic legitimacy, therefore, economic interdependence between self-governing territories raises two problems: On the one hand, the growing importance of external effects undermines the congruence between the "people" that is being governed, and the "people" that is supposed to govern -- choices that may be legitimated in one country (e.g., the interest-rate policy of the German Bundesbank) may have a direct impact on the economy of another country (e.g., unemployment in France) where this choice was not, and would not be, democratically legitimated. In effect, this reduces the ability of all governments to achieve the purposes, and to solve the problems, that have high salience for their citizens. On the other hand, the competition for mobile factors of production and taxable assets imposes a redistributive bias on national policy choices that will shift burdens from mobile actors and the owners of mobile capital onto immobile actors and of the owners of immobile assets. Again, there is no reason to expect that these policy shifts would be legitimated by corresponding shifts in the authentic preferences of citizens in the competing countries.

At the normative level, both of these changes are widely interpreted as a loss of democratic legitimacy, and at the empirical level they may generate growing dissatisfaction with the government of the day, and perhaps a more general disaffection with the democratic political system as such, reflected in political abstention, alienation or growing support for system-critical movements and radical political parties. The conclusion therefore seems to be that increasing economic interdependence is indeed likely to generate problems for democratic legitimacy at the national level. But before I examine this conclusion more closely, it is necessary to check if international or supranational solutions might avoid, or at least alleviate, the problems faced at the national level.

Supranational Remedies?

If government within a territorially limited unit is considered ineffective in the output dimension as well as unresponsive in the input dimension, territorial enlargement and functional centralization would seem to provide the logical solution to the problems of interjurisdictional interdependence. This at any rate is the standard prescription of fiscal federalism (Oates 1977); it is the logic behind the long-standing recommendations to overcome the deficiencies of joint-decision making in German federalism by merging several Länder to create larger units with fewer externalities (Scharpf 1988); and it is of course the logic driving European political integration.

In general, centralization is justified in output-oriented terms. But this is only plausible if it is also assumed that decisions at the higher level are taken under majoritarian or hierarchical rules and cannot be blocked by the veto of constituent governments. From an input-oriented perspective, by contrast, centralization appears problematic even within the nation state, where competencies would be exercised by a central government with clear democratic legitimacy. If citizen preferences differ, and if policies must be uniform, centralization will necessarily reduce the goodness of fit between preferences and policy choices (Buchanan/ Tullock 1962, chapter 6).[5] It is at the supranational level, however, where the centralizing solutions that are justified by output-oriented arguments become truly problematic. In the following sections, I will discuss these problems by reference to the European Union.

The Preconditions of Majority Rule

Governing systems that are able to overrule dissenting interests need to be legitimated, and before it is meaningful to talk about either input- or output-oriented legitimation arguments with regard to the European polity, it is now necessary to discuss a precondition which is usually taken for granted in a national context: the definition of the constituency that is ready to be governed by majority rule. In Lincoln's triad, this refers to "government of the people" that I skipped above - and that Giovanni Sartori (1965, 26) found to pose "insoluble problems of interpretation". The difficulties are well illustrated by the examples of some countries (like Canada, Belgium, the former Czechoslovakia, or Bosnia) where ethnic, linguistic or religious divisions seem to undermine the legitimacy of majority rule, and of some other countries (like Switzerland, the Netherlands, or the United States) where cleavages of a similar nature do not have nearly the same delegitimating effects. Regardless of the main criterion of sameness or difference, however, it seems obvious that a "we identity" (Elias 1987) that is shared by the members of the constituency is a logically necessary precondition of democratic legitimacy.

In the input dimension, we-identity is necessary to justify "my trust in the benevolence (and perhaps even solidarity) of my fellow citizens" (Offe 1998, 17) --which implies that the welfare of the minority must also be an argument in the preference function of the majority. In the output dimension, identity is necessary to define the membership in the community whose "common interests" are thought to justify governmental action even if it should entail individual sacrifices. In neither dimension is it possible to name a single set of necessary and sufficient criteria for what constitutes an effective we identity -- common language, culture, religion, history or institutions play important but differing roles. There is also no reason to assume that only one specific type of collectivity may be invested with the characteristics of collective identity. Individuals may identify with different units of reference -- religious, partisan, territorial, local, regional, national, European, etc. -- in different contexts or for different purposes. Moreover, collective identifications may differ greatly in their intensities, and thus may permit rather different levels of sacrifices and involuntary redistribution to be legitimated.

But while the willingness to accept sacrifices for the purpose of solidaristic redistribution seems remarkably high at the level of established nation states (Hicks/ Swank 1992), it also seems clear that no political unit above the national level has as yet developed a we identity of comparable intensity. This is true even of the European Union, which has gone further than any other supranational or international organization toward establishing institutions that resemble constitutional democracies at the national level. But even if further institutional reforms invested the directly elected European Parliament with the full range of competencies of a national parliament, there is no reason to think that its majority decisions could legitimate salient sacrifices imposed on a dissenting minority. The reason is succinctly expressed by Joseph Weiler (1996: 523) in an article that criticizes the exclusive focus on ethnic identities, but nevertheless acknowledges that

"democracy does not exist in a vacuum. It is premised on the existence of a polity with members -- the Demos -- by whom and for whom democratic discourse with its many variants takes place. The authority and legitimacy of a majority to compel a minority exists only within political boundaries defined by a Demos. Simply put, if there is no Demos, there can be no operating democracy."

To drive the point home, Weiler then constructs a counterfactual:

"... imagine an Anschluss between Germany and Denmark. Try and tell the Danes that they should not worry, since they will have full representation in the Bundestag. Their shrieks of grief will be shrill not simply because they will be condemned, as Danes, to permanent minorityship (that may be true of the German Greens too), but because the way nationality, in this way of thinking, enmeshes with democracy is that even majority rule is only legitimate within a Demos, when Danes rule Danes."

Turning to Europe, he then concludes that "it is a matter of empirical observation that there is no European Demos -- not a people not a nation." It is hard to see how this conclusion could be denied, and with each territorial expansion, the hope that the multiple peoples of Europe would soon develop a common political identity, and a common space of political communication and policy-oriented discourse, has receded further into the future. This is not meant to discourage efforts that could advance political integration.[6] For the time being, however, the European Union cannot yet rely on the foundation of a collective identity that would be strong enough to legitimate majority rule.

The Limits of Supranational Legitimacy

If Europe cannot yet be a majoritarian democracy, that places severe limitations on its capability to act in the face of politically salient disagreement. As it is, the European Union is relying, for the legitimation of its policy output, on a combination of hierarchical and consensual decision processes. Hierarchical authority is most clearly exemplified by the competencies of the European Central Bank which, under the rules adopted in the Maastricht Treaty, was constructed to be even more independent from political inputs and political accountability than is the German Bundesbank (which it will replace as the author of monetary-policy choices for the members of the European Monetary Union). But whereas the formal independence of the ECB, and the scope of its hierarchical authority, were legitimated by the explicit and highly politicized decisions of national governments and parliaments, much less political attention had accompanied the expansion of the judicial authority of the European Court of Justice (Weiler 1982). Nevertheless, the authority of the Court, together with the active use of independent enforcement powers granted to the European Commission, have been used to define and enlarge the reach of "negative integration" -- meaning the legal rules that restrict the capacity of national governments to interfere with the free movement of goods, services, capital and workers throughout the internal European market (Scharpf 1996). The most important extension was achieved through the application of European competition law to service-public areas such as telecommunications, air, road and rail transport, and energy supply which before had been considered exempt from full market competition in practically all European countries (S. Schmidt 1998). But even though this may have strained the authority of the law to its limits, the legitimacy of judicial law-making has not been seriously undermined.[7]

By contrast, European policy processes of "positive integration" -- meaning the active regulation of the economy -- depend on broad political agreement. It would be wrong to equate this with the classical model of intergovernmental negotiations. The European Parliament is rapidly approaching the point where its veto cannot be overruled in most important fields of European legislation, and the practical importance of the European Commission's monopoly of legislative initiative (in addition to its unilateral enforcement powers) can hardly be overestimated. Nevertheless, national governments represented in the Council of Ministers are ultimately decisive for the adoption of European legislation, and even though decisions by qualified majority are possible in an increasing number of policy areas, the requirements are set so high that small groups of governments with similar interests cannot be overruled. In fact most decisions in the Council are adopted by broad consensus.

But if European legislation thus avoids the threat to political legitimacy that would arise if substantial interests could be overruled by self-interested majorities, one of two consequences is likely to occur: Either policy choices will be blocked by disagreements among national governments, or the burden of legitimating European policy solutions is shifted back to the political systems of member states. In both cases, the outcome will add to the difficulties of democratic legitimation at the national level -- either because problem-solving deficits will persist, or because policy choices must be accepted that may, on the input side, not conform to the authentic preferences of national constituencies or that may, on the output side, not be optimal solutions if judged by criteria of the national interest. Since there are indeed areas where EU policy-making processes are highly effective, and others where the problem-solving capacity of the Union is very low (Scharpf 1997b), both types of legitimacy problems must in fact be dealt with at the national level. I will begin with an examination of the input-oriented problems that will arise at the national level precisely when European policy processes are successful in producing effective policy outputs.

International Problem-Solving and National Preferences

In order to clarify the implications of internationally agreed policy solutions for national democracy, I will refer to a highly simplified model of intergovernmental negotiations (Figure 1). Assume three countries, A, B, and C, facing a problem that none of them could solve nationally, but that could be solved through international cooperation requiring the agreement of all three countries. While all of them dislike the status quo (located at SQ = 0), each prefers a different one of three feasible coordinated solutions, located (in unidimensional and interval-scaled utility space) at points A=1, B=3, and C=5, respectively. Assume also that these "ideal points" are determined, ex ante and in strictly input-oriented procedures, by the citizens (i.e., the median voter) in each country. However, if we should further assume that negotiators from each country are strictly bound by these expressed preferences, it is clear that none of the cooperative solutions could find the agreement of all three countries, and that the undesired status quo would continue. Thus, if negotiations are to serve any purpose at all, the negotiating governments must be allowed to agree to solutions that do not match the ex ante preferences of their constituents -- provided that the solution chosen must increase the welfare of the country (i.e., reduce the distance from the country's ideal point) in comparison to the status quo.

Figure 1: Negotiations in Single-Issue Space

If only the ex-ante positions of each country are considered, the only generally acceptable solution would be located at point A, which represents the lowest-common-denominator outcome. It satisfies the preferences of the most "conservative" country A and is still preferred to the status quo by B and C. The famous Coase Theorem tells us, however, that negotiations could do better. In the absence of transaction costs, they should be able to achieve an overall welfare maximum (Coase 1960). If distances from each country's ideal point are interpreted as welfare losses, Figure 2 shows that this welfare maximum (i.e. the minimum of aggregate losses) is located at point B, rather than at point A. But since solution B would, by itself, be less attractive to country A than the status quo, its veto would need to be bought off by side payments -- say, one unit each from countries B and C, which these could well afford to pay from the gains which they will achieve if the agreed-on solution is located at B, rather than at A.

Figure 2: Welfare Losses in Negotiated Agreements

So far, so good. For each country, this outcome is the best that it can reasonably expect to reach in a world in which solutions cannot be unilaterally imposed but do depend on the voluntary agreement of all parties involved.[8] In that sense, the output-oriented legitimacy of the negotiated outcome would be fully assured. But what about input-legitimacy? In order to appreciate the difficulties here it is useful to consider the preconditions that must be created in order to make Coasian outcomes possible in the real as distinguished from the model world.

In order to achieve the welfare-maximizing outcome, the parties must somehow overcome the "Negotiators' Dilemma" (Lax/Sebenius 1986; Scharpf 1997a, chapter 6) which arises from the simultaneous presence of common interests (in finding the best overall solution) and competitive interests (in maximizing the distributive share of one's own side). This implies that all aspects of the constellation -- available policy options, their likely effects, and the valuation of these effects by all parties involved -- would have to become transparent to all of them. Moreover, the parties would need to agree on a normative rule for distributing the costs and the gains of cooperation. These are extremely demanding preconditions, depending to a large degree on the development of mutual trust, or at least mutual understanding, among the negotiators directly involved.[9] If they are not met, the Negotiators' Dilemma will induce self-serving negotiating strategies that will produce inferior outcomes or frustrate agreement altogether.

With regard to input-oriented legitimacy, this analysis seems to lead to two dismal conclusions. First, it is clear that negotiations cannot reach their optimal outcome (i.e., the outcome maximizing total welfare for the group of countries as a whole) without systematically departing from ex-ante citizen preferences in most or all countries. Second, and more important here, the specific reasons for these departures cannot be fully communicated to the constituencies in each country. If it is assumed that mutual understanding and mutual trust among negotiators is an essential precondition for optimal negotiated solutions, it follows that intergovernmental negotiations are unlikely to succeed in the glare of total publicity, and if that is so, there will be an inevitable communication gap between the international and national levels of this two-level game (Putnam 1988).

It is this systematic gap which poses the most serious threat to democratic legitimacy. Where it exists, the opposition in each country cannot only claim that the outcome achieved does not conform to ex ante citizen preferences, but also that the national interest was sold short by incompetent or illoyal negotiators.[10] Governments, on the other hand, arguing that this was the best that could be obtained under the circumstances, would have to refer to inside information about feasible options, and to informed guesses about the true preferences and outside options of the other governments -- none of which could be fully scrutinized and verified in public or parliamentary debates.

Moreover, given the difficulties of re-negotiation, governments can no longer afford to respond to criticisms and suggestions raised in public debate, even if the outcomes need to be approved by parliament, or by referendum. Instead, agreements must be presented as a fait accompli that confronts the democratic sovereign with a take-it-or-leave-it proposition whose rejection will cause the collapse of the cooperative international effort. Also, given the joint responsibility of all negotiating governments, no single government could in truth be held politically accountable for the ultimate outcome.[11] If that is so, intergovernmental negotiations will indeed disable the institutional mechanisms that connect government action to the expressed preferences of constituents or to the scrutiny of parliaments, political parties, and public debate.[12]

It is true that the need to discipline domestic preference formation has always been an obstacle to demands for a "democratic foreign policy" -- except under hegemonic conditions when one government is able to impose domestically generated preferences on its external partners. What is new is that with increasing transnational interdependence the same compulsion is now manifest in ever larger areas of what used to be purely domestic policy choices. In effect therefore, the more policy choices are moved from the national level to the level of intergovernmental negotiations, the more the institutions that are meant to assure input-oriented influence and accountability are losing their effectiveness.[13]

Legitimate Democracy without Omnipotence

Thus we seem to face a veritable dilemma. As interdependence increases, the nation state finds its former range of policy options exogenously constrained, and some previously legitimated policy choices will now become less effective, more costly, or downright unfeasible -- which must be counted as a loss of democratic self-determination even if new options are also added to the policy repertoire. It is true, however, that constraints do not rule out choice, and that it may indeed be possible to achieve the former (or newly consented) policy objectives through the choice of new policy instruments (Scharpf 1999). In that case, output-oriented legitimacy may still be maintained. But the new policy instruments must be adopted either in domestic choice processes that are extremely sensitive to international constraints or in processes of negotiations at the international level. In both cases, the increase in output effectiveness seems to have a high price in terms of input-oriented legitimacy.

The Inevitable Corruption of Input Legitimacy?

In the input dimension, conventional notions of "popular sovereignty," and expectations that governments should carry out the "will of the people," are directly challenged by the increasing importance of external economic and institutional constraints. Ever more frequently, policy choices that would be both domestically popular and economically feasible must be avoided out of respect for the legal constraints of GATT rules and European law, or as a consequence of decisions by the WTO, the European Commission, or the European Court of Justice. Conversely, policy choices that would be both legally permissible and domestically popular must be ruled out because they could have disastrous consequences for the international competitiveness of national producers, for the confidence of investors, or for the stability of the national currency in global money markets.

As external legal and economic constraints multiply under conditions of growing international interdependence, the role of experts and of specialized knowledge will increase to an extent that may render the role of authentic but untutored popular preferences and demands practically insignificant. All this is even more true if solutions can only be achieved through international negotiations. In short, popular approval and popular demand are becoming less and less sufficient for assuring, or even for justifying, corresponding policy choices. As a consequence, input-oriented legitimating arguments will become less plausible, and government at the national level must increasingly depend on output-oriented legitimation arguments alone.

In fact, much of this is happening already. As more and more domestic policy areas have become internationally interdependent, governments are increasingly tempted to invoke Bismarck's Primat der Außenpolitik -- meaning that foreign policy should override domestic political considerations -- to immunize policy choices with an international dimension against the demands and criticisms of domestic public opinion, political parties, parliaments and other democratic input processes. To the extent that they succeed, the remaining legitimating arguments take on a paternalistic and technocratic character, insisting that under difficult circumstances and in a dangerous environment the government is doing the best it can to promote and defend the national interest, and that any demands for more direct participation and control could only make a difficult job even more difficult. When that is accepted, partisan controversies and political attention at the national level are likely to be diverted in two directions - personalities and scandals on the one side, and policy outcomes (rather than policy choices) on the other side. Elections will then be either about candidates and their personal qualities and deficiencies, or they will be about the performance of the stock market, the level of unemployment, the rate of inflation, the size of the public-sector deficit, or even natural disasters like floods and earthquakes, without regard for the question of whether the government was in any way responsible for these outcomes. In other words, input-oriented politics in general, and political accountability in particular, will lose their connection to, and their disciplining effect on, policy choices.

Toward Internationally Embedded Policy Discourses?

The question is whether this could be otherwise. A positive answer does require a reconsideration of the role that input-oriented mechanisms could and should play in the democratic process. I begin by returning to the discussion of intergovernmental negotiations. The input-oriented objections presented above are fully compelling only when they are raised against negotiated solutions for problems that could just as well have been dealt with at the national level. In German federalism, we have indeed identified instances of Überverflechtung in which the practice of joint-decision making went far beyond the "objective" need for coordination in the face of important interdependencies (Scharpf 1988), and the same may be true in some European policy areas as well. But as economic interdependencies increase, these instances will become rarer, and the input-oriented critique of intergovernmental negotiations will be weakened. With regard to problems that cannot be solved at home, within the boundaries and with the means of the nation state, the relevant criterion for judging solutions cannot be conformity to the solipsistic preferences of citizens of that state. For any country that is not a hegemon, the interests of necessary partners in a cooperative solution must be considered as well. Preferences formulated within a national frame of reference are relevant for defining the "ideal points" of a country's negotiators. But it is not reasonable to expect that negotiated outcomes should conform to any one of these national aspirations. Instead, the most that one could legitimately ask for, within a national frame of reference, is that the outcome should be better than the "best alternative to negotiated agreement" (BATNA), and that it should come as close to nationally defined aspirations as is possible, given the bargaining constellation and the BATNA positions of the other countries.

For democratic theory, this implies that it can no longer treat popular preferences as being exogenously given. In order to be normatively relevant, they must relate to policy outcomes that could be considered feasible within the international context in which the choice must be made. As a consequence, "democratic decisionism" (Greven 1998) and the assumptions of omnipotence associated with "popular sovereignty" are no longer theoretically viable options. Within the context of input-oriented theories, these requirements are met by concepts of "discursive" or "deliberative democracy" which insist on procedures of "will formation" that are supposed to lead to "reasonable" conclusions (Habermas 1996). However, in trying to avoid the pitfalls of unrefined populism, Habermas and others tend to insist on extremely demanding "procedural" preconditions that would assure a very high degree of moral and intellectual sophistication in public debates. In the tradition of "critical theory," these demands are not meant to be practicable - and if they could be approximated, political discourses would be restricted to a small elite of philosopher-kings.

If we start instead from the policy-oriented "discourses" that are in fact going on within existing Western democracies, they are indeed largely elite affairs - conducted by politicians, spokespersons for interest groups, prominent experts and journalists under the filtering, amplifying and distorting conditions of the media. Discussion takes place in many specialized policy communities with specialized publics of interested non-elites. At the same time, specialized discussions are linked into the more general political discourse carried on among policy generalists in governments, parliaments, political parties, associations and the media, on issues that could potentially catch the attention of the wider public and affect the outcomes of general elections. It is in these interwoven patterns of communications among specialists, generalists, and communicators that problem definitions are proposed and rejected, that policy options are presented, criticized and justified, that political performance is being evaluated, and that political trust and, ultimately, legitimacy is constantly being generated, eroded or destroyed.

These communications will surely not approximate ideal debates among philosopher-kings. What matters, however, is that they are conducted in public, and that they allow statements to be supported and contradicted in ways that may catch the attention of unspecified non-elites. The importance of these two conditions - publicness and contestation - can hardly be overstated. Publicness works as a powerful censorship mechanism (Elster 1986), allowing only public-regarding communications to be made. It simply would not do to publicly justify a political demand or a policy proposal in terms of what it would do for yourself or your own group. That does not rule out self-serving communications. In public debates, however, self-interest is forced to masquerade as public interest - at which point the possibility of contestation allows competing interests or public-interested critics to challenge such claims.

From the perspective of democratic theory, public discourses may serve two critical mediating roles in the relationship between governors and the governed (V. Schmidt 1997; 1998). On the one hand, they greatly reduce the information costs of non-elites. Reasonably interested members of the public will have a chance to sort out the pros and cons of policy proposals and to form an opinion of government performance either in terms of their own self-interest or in terms of the public interest - which then may enter into their electoral choices. For the governors in turn, public discourse provides a sounding board for trying out problem definitions and policy solutions, and an early warning system for issues that might achieve electoral salience - which is critical for the mechanism of "anticipated reactions" that links policy choices to voter reactions (Scharpf 1997a, chapter 8).

What matters for input-oriented democracy is the quality of these public discourses. They may perform an orienting and a legitimating function if they communicate the wider framework of ongoing policy controversies, the definition of the situation and the political aspirations in light of which problems and options can be meaningfully considered. For the individual citizen, such discourses provide the context within which it is possible to make sense of what is happening, and to respond to specific policy choices by approval, unconcern, or active opposition. For the public as a whole, orienting discourses provide the stimuli in response to which the electoral expression of political support or opposition can indeed claim, and bestow, a maximum of democratic legitimacy.

But in what way would a realistic reformulation of "discursive democracy" provide a more promising perspective on the legitimacy deficits associated with increasing international interdependence? The answer lies in the connection between the orienting and the legitimating function. In order to maintain legitimacy even under conditions of international interdependence, national policy discourses must provide orientations that are free of the suggestions of omnipotence that still infect not only conventional notions of popular sovereignty but also the mutual recriminations between governments and oppositions -- where governments claim exclusive credit for everything that seems to go well, while the opposition blames the government for everything that seems unsatisfactory. Instead, orienting discourses should provide a realistic picture of the country's present place and future options in an institutionally and economically integrating world; they should reassess policy goals with a view to their feasibility under international economic and institutional constraints; and they should emphasize the search for policy instruments that are still viable under these constraints.

When it is made clear, moreover, that important national goals can no longer be achieved through purely national action, the possibility of pursuing them through internationally coordinated or supranational action will be understood as a gain, rather than a loss, of collective fate control. If that is acknowledged, the national interest can no longer be defined in solipsistic terms, and policy options must be discussed in light of the relevant decision rules and actor constellations at the international level, and with an empathetic understanding of the preferences, worldviews and capabilities of the other countries involved. As a consequence, the information and communication gap discussed above will be greatly reduced. National policy discourse will be able to shadow more closely the real choice situations that governments must deal with at the international level, and governments will have less opportunity to escape from political accountability by the mere reference to external necessities and constraints.

Is this an impossible ideal? I think not. Small European democracies have long had much more open economies than the larger European states, let alone Japan and the United States. As a consequence, they have never been able to control their policy environment or to indulge in omnipotence fantasies (Katzenstein 1984; 1985). Nevertheless, they have done very well economically over the past decades, and some of them are also now much more successful in coping with the challenges of economic interdependence and systems competition than their larger, previously more self-sufficient neighbors. They also seem to have higher levels of public trust or political satisfaction than is reported for the larger countries that have only more recently felt the full thrust of international economic interdependence (Katzenstein 1998).

It seems plausible, therefore, that the secret of the economic and political success of small and open countries, like Switzerland, Austria, Denmark or the Netherlands, lies precisely in their ability to conduct policy discourses that are based on a realistic understanding of their own capabilities and constraints, and to focus debates on those policy alternatives that could be feasible and effective in an international policy environment that is characterized by high degrees of institutional integration, economic interdependence and regulatory competition (Visser/ Hemerijck 1997). Under these conditions, public opinion will not proceed from solipsistic definitions of policy problems and policy goals, and when that is assured, the existence of international constraints, and the need for international cooperation will not be experienced as a delegitimating disappointment, but will have been taken into account all along. For these countries, democratic legitimacy no longer presupposes omnipotence and is not challenged by the realization of their interdependence.

There is no reason why the larger democracies should not also come to live with international interdependence. But the lesson they need to learn from the successful small and open countries is that orienting discourses do require political leadership. They cannot merely reflect untutored popular opinions and preferences, but must impose the discipline of Freud's "reality principle" on policy-oriented public debate. If that is not achieved, effective international problem solving will remain domestically vulnerable to populist appeals to wishful thinking, nostalgia for past national grandeur, resentment of foreign influences, or xenophobia. It is the responsibility of policy elites to communicate the extent to which international involvement, cooperation, and trustworthiness have become a precondition for the effective pursuit of the national interest. If they succeed, policy discourses even in the larger countries should be able to maintain the tenuous linkage between the perceptions and preferences of non-elites and policy choices that are effective under the constraints of an increasingly interdependent international environment -- and then there would be no reason to fear that international interdependence would undermine democratic legitimacy.

Bibliography

Bairoch, Paul, 1997: Globalization Myths and Realities. One Century of External Trade and Foreign Investment. In: Robert Boyer/ Daniel Drache (eds.), States against Markets. The Limits of Globalization. London: Routledge, 173-192.

Brennan, Geoffrey, 1989: Politics with Romance. Towards a Theory of Democratic Socialism. In: Alan Hamlin/ Philip Pettit (eds.), The Good Polity. Normative Analysis of the State. Oxford: Basil Blackwell, 49-66.

Buchanan, James M./ Gordon Tullock, 1962: The Calculus of Consent. Logical Foundations of Constitutional Democracy. Ann Arbor: University of Michigan Press.

Coase, Ronald H., 1960: The Problem of Social Cost. In: Journal of Law and Economics 3, 1-44.

Cooke, Jacob E., (ed.), 1961: The Federalist. Cleveland: Meridian Books.

Dryzek, John S., 1990: Discursive Democracy. Politics, Policy, and Political Science. Cambridge: Cambridge University Press.

Elias, Norbert, 1987: Wandlungen der Wir-Ich-Balance. In: Norbert Elias, Die Gesellschaft der Individuen. Frankfurt/M.: Suhrkamp, 207-315.

Elster, Jon, 1986: The Market and the Forum: Three Varieties of Political Theory. In: Jon Elster/ Aanund Hylland (eds.), Foundations of Social Choice Theory. Cambridge: Cambrige University Press, 103-132.

Greven, Michael Th., 1998: Output-Legitimation. "Der Zweck heiligt die Mittel" in der Demokratie nicht. In: Michael Buckmiller/ Joachim Perels, eds., Opposition als Triebkraft der Demokratie. Bilanz und Perspektiven der zweiten Republik. Jürgen Seifert zum 70. Geburtstag. Hannover: Offizin-Verlag, 477-492.

Habermas, Jürgen, 1962: Strukturwandel der Öffentlichkeit. Untersuchungen zu einer Kategorie der bürgerlichen Gesellschaft. Neuwied: Luchterhand.

Habermas, Jürgen, 1996: Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy. Translated by William Rehg. Cambridge: MIT Press.

Hicks, Alexander M./ Duane H. Swank, 1992: Politics, Institutions, and Welfare Spending in Industrialized Democracies, 1960-1982. In: Americal Political Science Review 86, 658-674.

Hirst, Paul/ Grahame Thompson, 1995: Globalization and the Future of the Welfare State. In: Economy and Society 24, 408-442.

Kaiser, Karl (1971): Transnational Relations as a Threat to the Democratic Process. In: Robert O. Keohane/ Joseph S. Nye (eds.), Transnational Relations and World Politics. Cambridge, MA: Harvard University Press, 356 -370.

Katzenstein, Peter J., 1984: Corporatism and Change. Austria, Switzerland, and the Politics of Industry. Ithaca: Cornell University Press.

Katzenstein, Peter J., (1985): Small States in World Markets. Industrial Policy in Europe. Ithaca: Cornell University Press.

Katzenstein, Peter K., 1998: Trust and International Relations. Paper prepared for the Bellagio Conference on "Public Trust and Governance in the Triateral Democracies," June 29-July 3, 1998.

Kirchgässner, Gebhard, 1992: Towards a Theory of Low-cost Decisions. In: European Journal of Political Economy 8, 305-320.

Lax, David A./ James K. Sebenius, 1986: The Manager as Negotiator. Bargaining for Cooperation and for Competitive Gain. New York: Free Press.

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

Lijphart, Arend, 1991: Majority Rule in Theory and Practice. The Tenacity of a Flawed Paradigm. In: International Social Science Journal 129, 483-493.

Lindner, Clausjohann, 1990: Kritik der Theorie der partizipatorischen Demokratie. Opladen: Westdeutscher Verlag.

Luthardt, Wolfgang, 1994: Direkte Demokratie. Ein Vergleich in Westeuropa. Baden-Baden: Nomos.

Manin, Bernard, 1987: On Legitimacy and Political Deliberation. In: Political Theory 15, 338-368.

Moravcsik, Andrew, 1993: Preferences and Power in the European Community. A Liberal Intergovernmentalist Approach. In: Journal of Common Market Studies 31, 473-524.

Moravcsik, Andrew, 1994: Why the European Community Strengthens the State. Domestic Politics and International Cooperation. Working Paper 52. Cambridge, MA: Center for European Studies, Harvard University.

Newton, Kenneth, 1998: Trust in People and Confidence in Institutions: Trends and Patterns in the Trilateral Countries. Paper prepared for the Bellagio Conference on "Public Trust and Governance in the Triateral Democracies, June 29-July 3, 1998.

Oates, Wallace E., ed., 1977: The Political Economy of Fiscal Federalism. Lexington: Heath. Offe, Claus, 1998: How Can We Trust Our Fellow Citizens. Manuscript. Berlin: Social Science Department, Humboldt University.

Pierson, Paul, 1997: Path Dependence, Increasing Returns, and the Study of Politics. Center for European Studies. Program for the Study of Germany and Europe. Working Paper Series No. 7.7. Cambridge, MA: Harvard University.

Polanyi, Karl, 1957: The Great Transformation. The Political and Economic Origins of Our Time. Boston: Beacon Press.

Putnam, Robert D., 1988: Diplomacy and Domestic Politics. The Logic of Two-Level Games. In: International Organization 42, 429-460.

Putnam, Robert D., 1998: What's Ailing the Trilateral Democracies: The Case for Declining Trust. Paper prepared for the Bellagio Conference on "Public Trust and Governance in the Triateral Democracies," June 29-July 3, 1998.

Riker, William H., 1982: Liberalism against Populism. A Confrontation between the Theory of Democracy and the Theory of Social Choice. San Francisco: W. H. Freeman.

Ruggie, John Gerard, 1982: International Regimes, Transactions, and Change. Embedded Liberalism in the Postware Economic Order. In: International Organization 36, 379-415.

Sartori, Giovanni, 1965: Democratic Theory. New York: Praeger.

Scharpf, Fritz W., 1970: Demokratietheorie zwischen Utopie und Anpassung. Konstanz: Universitätsverlag.

Scharpf, Fritz W., 1988: The Joint-Decision Trap. Lessons from German Federalism and European Integration. In: Public Administration 66, 239-287.

Scharpf, Fritz W., 1996: Negative and Positive Integration in the Political Economy of European Welfare States. In: Gary Marks et al., Governance in the European Union. London: Sage, 15-39.

Scharpf, Fritz W., 1997a: Games Real Actors Play. Actor-Centered Institutionalism in Policy Research. Boulder: Westview.

Scharpf, Fritz W., 1997b: Introduction. The Problem-Solving Capacity of Multi-Level Governance. In: Journal of European Public Policy 4, 520-538.

Scharpf, Fritz W., 1999: Governing in Europe. Democratic and Efficient? Oxford: Oxford University Press, forthcoming.

Schmalz-Bruns, Rainer, 1995: Reflexive Demokratie. Die demokratische Transformation moderner Politik. Baden-Baden: Nomos.

Schmidt, Susanne K., 1998: Commission Activism. Subsuming Telecommunications and Electricity under European Competition Law. In: Journal of European Public Policy 5, 169-184.

Schmidt, Vivien A., 1990: Democratizing France. The Political and Administrative History of Decentralization. Cambridge: Cambridge University Press.

Schmidt, Vivien A., 1997: Discourse and (Dis-)Integration in Europe. The Cases of France, Great Britain, and Germany. In: Daedalus 126, 229-241.

Schmidt, Vivien A., 1998: Democracy and Discourse in an Integrating Europe and a Globalizing World. Working Paper. Berkeley: Center for German and European Studies, University of California.

Talmon, Jacob L., 1955: The Origins of Totalitarian Democracy. London: Secker and Warburg.

Tsebelis, George, 1995: Decision Making in Political Systems. Comparison of Presidentialism, Parliamentarism, Multicameralism, and Multipartism. In: British Journal of Political Science 25, 289-325.

Visser, Jelle/ Anton Hemerijck, 1997: A Dutch Miracle. Job Growth, Welfare Reform, and Corporatism in the Netherlands. Amsterdam: University of Amsterdam Press.

Weiler, Joseph H. H., 1982: The Community System. The Dual Character of Supranationalism. In: Yearbook of European Law 1, 267-306.

Weiler, Joseph H. H., 1996: European Neo-Constitutionalism. In Search of Foundations for the European Constitutional Order. In: Political Studies 44, 517-533.

Notes

1   The presumption is, of course, that governments which, in the absence of legitimacy, could not count on voluntary compliance but would have to rely purely on the exercise, or the threat, of superior force, would only be able to reach comparatively low levels of governing efficiency.

2   The general assumption is well stated by Michael Greven who postulates: "The crucial idea, from which the legitimacy of government in democracies derives, is the possibility of participation of all citizens, based on the mutual recognition of their civic and political equality... Precisely because and if this is true, democratic theory implies that the outcome of political will-formation has a claim to recognition and legitimacy even among those whose arguments failed in the discussion and whose preferences were defeated in a vote." (Greven 1998, 480; my translation).

But why should the mere chance of equal participation have legitimating force? At bottom, the argument seems to rest on the logic of the duel: you have no reason to complain if you have fought and lost in a fair fight. Under the conditions of modern mass democracies, this logic may indeed be relevant for candidates for political office. It is harder to see why it should persuade individual citizens.

3   Representation and accountability based on general elections have been thought to counteract the dangers of self-interested majorities since the Federalist Papers (Cooke 1961). The argument can be restated in in rational-choice terms, if it is first assumed that, for any individual voter, the vote is a "low-cost decision" (Kirchgässner 1992) -- meaning that the probable effect on individual self-interest is so low that it is reasonable to think that at least some voters will be motivated by public-interest considerations (Brennan 1989). If that is granted, the anticipation of a public-interest oriented swing vote creates strong incentives for office holders to select policies that can be publicly defended as serving common interests, rather than special interests (Scharpf 1997, chapter 8).

4   These data are often cited to suggest that, since "globalization" is nothing new, there should also be no reason to be concerned about its political impact. But that argument forgets that international capitalism before 1914 and again in the 1920s was characterized by deep economic crises. Before 1914, political democracy was underdeveloped in most countries, and the level of political aspirations -- and hence the potential impact of economic performance on political trust --was much lower than it is now. In the interwar period, however, the crises of international capitalism had serious, and in the case of Germany catastrophic, consequences for the viability of political democracy.

5   Both the assumption and the conclusion can be questioned: In the Jacobine tradition of French democracy, centralization is considered desirable precisely because it imposes uniformity and hence civic equality (V. Schmidt 1990). Conversely, centrally imposed policy might at least in theory also provide for differentiated solutions that would fit the differing conditions or preferences of subgroups or regions within the larger constituency or territory. An example might be the Spanish constitution which grants differing degrees of autonomy to different regional units. In general, however, the empirical association between centralization and rule uniformity seems to be quite strong.

6   A plausible proposal was recently promoted by Jacques Delors. It would require European parties to nominate their own candidates for the office of President of the European Commission in the European elections. This would not only force governments to nominate the winning candidate for confirmation by the European Parliament, but it would also focus public attention in all member states on the competition for a highly visible European office, and it would put European issues on the agenda of the European election campaign, which so far is dominated by purely national concerns. Moreover, this option could be realized without any change in the treaties. So far, however, European parties do not seem to be responding.

7   However, in the face of growing political unease, the Court and the Commission themselves have recently become more sensitive to the limits of negative integration (Scharpf 1999).

8   The negotiated solution would not necessarily be the one preferred by the median voter in a larger country (A+B+C), since negotiations tend to equalize the bargaining powers of the participating countries regardless of differences in the size of their populations.

9   The difficulties of reaching agreement are reduced, and the approximation of Coasian outcomes is facilitated, if negotiations are "embedded" in stable network constellations, and conditions can be improved even more through institutional arrangements that increase transparency, that provide for the good services of an "agenda setter," and that may generate mutual trust through the evolution of normative "regimes" (Scharpf 1997, chapter 6).

10   In the early years of the Federal Republic of Germany, when Adenauer had to defend the disappointing outcome of negotiations over the "occupation statute," he was attacked by the leader of the opposition as being "the Chancellor of the Allies."

11   Exactly the same criticism is directed at the effect of interstate and federal-state negotiations in the joint-decision system of German federalism (Scharpf 1988).

12   The same argument, without democratic-theory pretensions, supports the proposition that European integration is strengthening national governments in relation to other national and subnational political actors (Moravcsik 1993; 1994).

13   For an early recognition of the problem, see Kaiser (1971).

Copyright © 1998 Fritz W. Scharpf

No part of this publication may be reproduced or transmitted without permission in writing from the author.
Jegliche Vervielfältigung und Verbreitung, auch auszugsweise, bedarf der Zustimmung des Autors.

MPI für Gesellschaftsforschung, Paulstr. 3, 50676 Köln, Germany