MPIfG Working Paper 96/2, July 1996 


Economic Integration, Democracy and the Welfare State


von Fritz W. Scharpf mail


Plenary Lecture, 8th International Conference on Social Economics, University of Geneva, July 13, 1996 Revised and updated version: Cologne, August 10, 1996

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




The present difficulties of democratic welfare states in Europe are often ascribed to economic "globalization", that is to the world-wide integration of markets for capital, goods and services which has eliminated national control over boundary-crossing economic transactions, and which therefore exposes national producers to world-wide competition. In fact, however, globalization in the strict sense of the word is, at the most, realized only for speculative capital transactions. It is true that in all other areas of economic activity the intensity of international competition has increased as well. Nevertheless, even under the liberal regimes of GATT and WTO, governments have not abdicated their capacity for boundary control, and the freedom of world trade is still constrained by a wide variety of tariffs, quotas, "voluntary" restrictions and non-tariff barriers.


All this is different within the European Union, where national governments have delegated control over external trade relationships to the Union, and where they have in fact abdicated boundary control over economic transactions within the internal European market. As a consequence, the competitive pressure on European welfare states originates mainly from the "regulatory competition" within the Union, and the frustrations of democratic governance in Europe mainly result from the fact that the range of feasible policy choices has been reduced at the national level while policy making at the European level still lacks democratic legitimacy. I will begin with an examination of the last point.





The public concern about the European "democratic deficit" is of recent origin. It gained attention in the controversies leading up to the Maastricht referenda in Denmark and France, and it was then elevated to the status of a fundamental constitutional objection in the Maastricht judgment of the federal constitutional court in Germany. From the perspective of democratic theory, however, this debate is deficient in so far as its focus is on the democratic deficit of the European Union, rather than on the democratic deficit in Europe. It is assumed, in other words, that democracy continues to flourish at the national level, and that the deficit would disappear if only national sovereignty could be defended against the further expansion of European competencies. In my opinion, this view suffers from an incomplete conceptualization of democracy.


Democracy aims at collective self-determination. It must thus be understood as a two-dimensional concept, relating to the inputs and to the outputs of the political system at the same time. On the input side, self-determination requires that political choices should be derived, directly or indirectly, from the authentic preferences of citizens and that, for that reason, governments must be held accountable to the governed. On the output side, however, self-determination implies effective fate control. Democracy would be an empty ritual if the political choices of governments would not be able to achieve a high degree of effectiveness in achieving the goals, and avoiding the dangers, that citizens collectively care about. Thus, input-oriented authenticity, and output-oriented effectiveness are equally essential elements of democratic self-determination.


The current debate is deficient since it focuses exclusively on the weakness of the input structures at the level of the European Union. These are obvious enough. The European Parliament is directly elected, but its influence on European legislation is very limited. More importantly, even if the legal competencies of the European Parliament were enlarged, this would not help much, since the structural preconditions on which authentic democratic processes depend are also lacking: there are as yet no European political parties, no European political leaders and no European-wide media of political communication. As a consequence, we also have no European-wide controversies and debates on political issues and policy choices, and we have no Europe-wide competition for government offices, that could assure democratic accountability.


All this is true enough and would, by itself, be sufficient to deny that majority decisions at the European level could claim democratic legitimacy in the same sense in which majority decisions in a national parliament can do so. But the current debate is also conducted at a deeper level where "communitarian" arguments sometimes shade over into ethnic nationalism. It can indeed be shown, on analytical grounds, that majority rule is generally not welfare efficient if individual voters are assumed to be rational self-interest maximizers. Thus, to be normatively acceptable in welfare-theoretic terms, the majority principle must presuppose that voters are, at least in part, oriented toward a notion of the "common good" - which logically depends on a demarcation defining who is, and who is not, to be included in the respective "community".


But that demarcation cannot be defined arbitrarily or in purely technocratic terms. To take an example that Joseph Weiler (1995; 1996) has used: If Denmark were somehow joined with Germany, that would hardly persuade the Danes that they should accept majority decisions in the larger state as being democratically legitimate.[1] In other words, the capacity of the majority rule to create legitimacy depends itself on a pre-existing sense of community - of common history or common destiny, and of common identity - which cannot be created by mere fiat. At this point, however, arguments denying the existence of a "European demos" often come dangerously close to asserting that ethnic and linguistic homogeneity should be considered a necessary precondition of democratic legitimacy - and implying fatalistically that the European democratic deficit could never be overcome.


That is of course nonsense. Collective identities are malleable, as we have again observed this summer at the European football championship where individual players were just about killing themselves for the success of a "team" which, before this championship, had not even existed. Another example is the popular opposition, in Germany, to all attempts to change the territorial boundaries of Länder that, only fifty years ago, had been artificially created by allied military governments without any concern for the historical identities of the territories that were thrown together. Thus, just as playing together can create teams, living under a common government, and participating in common political processes, can create political identities. For that reason, the institutional structure of the Union is by no means irrelevant for the future evolution of a common identity.


Nevertheless, not all artificially created teams achieve efficiency in cooperation, and as the separation of Czechoslovakia and the difficulties of Canada and Belgium remind us, not all institutionally induced political identities hang together. At any rate, nobody claims that the European Union in its present composition should, in the eyes of the citizens of its member states, have already attained a degree of political identity that is at all comparable to that of long-established nation states. Thus it remains indeed correct to point out that the majority principle by itself cannot convey democratic legitimacy, and that even if the European Parliament had all the usual powers of national parliaments, its vote could not legitimate policy choices that run counter to the interests or the deeply held preferences of citizens in some of the member states. And the same would of course be even more true of majority votes in the Council of Ministers.


This limitation is recognized by recent theories emphasizing the non-majoritarian sources of legitimacy of European policy choices (Majone 1994; 1994a). Instead of democratic accountability, what is supposed to matter is the authority of law and, more generally, the authority of technical expertise. The European Union, in this view, should thus be understood not as an imperfectly democratic form of government, but as a "regulatory state" exercising powers that are similar to those exercised by the independent regulatory agencies in the United States, by constitutional courts, or by independent central banks. However, as Giandomenico Majone, the author who has most contributed to its explication, has pointed out, the potential reach of non-democratic legitimacy is definitely limited. It presupposes a normative consensus on the validity of certain norms, or the desirability of certain outcomes, and it also presupposes the existence of professional standards - and professional systems of discourse (Majone 1989) - by which it is possible to judge the interpretation and application of such norms, and the technical means employed for the attainment of consensual goals. This, Majone assumes to be true for European regulations on product safety, environmental standards, and competition rules,[2] and for the role of the European central bank in a future Monetary Union.


As a normative argument, this is persuasive, even if there may be disputes over the empirical classification of some policy areas. For present purposes, however, Majone´s contributions have an even more important implication in that they point out not only that there are non-democratic sources of legitimacy but also policy areas that differ in their need for legitimation - a point which is not much in need of discussion in the national context of democratic constitutionalism, but which is of great importance for the European Union and all other institutions of transnational governance.


In light of the Roman maxim "volenti non fit iniuria", voluntary agreements among the parties affected have intrinsic legitimacy. What needs extrinsic legitimation is the imposition of involuntary losses and the compulsion to act, or to abstain from action, contrary to one's interests or normative preferences. Thus the European Community was not in great need of political legitimation as long as its actions and regulations were perceived as serving everybody's interests - and the fact that it was constructed as a negotiation system in which decisions could not be imposed over the veto of national governments made sure that those decisions that were taken in the fields of trade liberalization and agriculture did not violate any of the interests that were politically influential at the national level (Weiler 1981).


That changed as the Community tried to push integration beyond the level of a customs union, and thus needed to confront the problem of non-tariff barriers to trade - that is of national health, safety and environmental regulations that differed from one country to another. Since the goal of a common European market could not be achieved as long as each country was preventing imports that did not conform to its own regulations, the Community tried to harmonize these regulations. But even though all countries were interested in the larger market, they disagreed over whose national regulations should become the European standard. In game-theoretic parlance, these constellations could be classified as "Battle-of-the-Sexes" problems which are indeed difficult to resolve under the unanimity rule.


Thus, when the single-market program was launched in 1986, it was clear that these deadlocks over the harmonization of national product standards needed to be broken. They were in fact broken by a two-pronged strategy: On the one hand, directives could now be adopted by qualified-majority voting in the Council of Ministers, and on the other hand, these Council directives would only specify basic principles while the details of the technical standards were to be defined by expert bodies. Products that did conform to these standards could then be freely marketed throughout the Community. In combination, these procedural innovations did in fact allow the Single-Market program to succeed by the end of 1992.


But with the move to qualified-majority voting, discussion about the European democratic deficit also assumed greater salience. This was perhaps not justified for the Single-Market program itself which could, as I suggested, be justified on the basis of generally convergent national interests in Battle-of-the-Sexes type constellations, or which alternatively could be legitimated on the grounds developed by Giandomenico Majone.


However, as Majone also emphasizes, this justification could not carry over into areas where interests do not converge, and where policy choices are crucially affected by distributive conflict or by ideological or normative conflict. Here, near-unanimous agreement is unlikely to be achieved and the authority of technical expertise will not suffice to settle conflicts authoritatively. If they are to be taken at all, therefore, such decisions must ultimately depend on democratic legitimation - which is not, or at least not yet, available at the European level.


In other words, it is incorrect to discuss the democratic deficit as if it were a general problem. It does not affect European decision-making across the board, but only certain types of policy areas in which conflicts of interest and of ideology are endemic. Under the present institutional ground rules, requiring nearly unanimous decisions, such policies could not be adopted by the European Union, and if the ground rules were changed to allow majority voting, such decisions would not be accepted as legitimate. They should be, and are in fact, left to the member states.


But does that mean that all is well as it is, since the policies that the Union can and does adopt can generally be legitimated, while policies that could not be so legitimated are practically ruled out by the existing institutional structure? That is the conclusion one might reach if only the authenticity dimension of democratic self-determination is being considered - and it is approximately the conclusion that was in fact reached by the German Constitutional Court. The conclusion changes, however, when the effectiveness dimension is considered as well. Here, all depends on how effectively the nation state is still able to deal with those policy problems that should not, or could not, be dealt with at the European level.


In Majone's analysis, the policy areas that definitely depend on democratic legitimation, and that therefore should be left to the member states of the Union, include all redistributive policies affecting capital and labor and, more generally, all policies relating to the welfare state. What Majone does not take into account, however, is the fact that the future viability of national welfare states is directly challenged by European economic integration which drastically reduces the effectiveness of democratic self-determination at the national level. This, I suggest, is presently the core problem confronting democratic governance in Europe.





The European Community has from its beginning been committed to economic and political integration, but it has been much more successful in the former than in the latter. This success is owed, first, to the European Court of Justice which has given constitutional force to the Treaty provisions prohibiting member states from interfering with the freedom of interstate economic transactions and mobility (Weiler 1981); and it is owed, secondly, to the success of Jacques Delor's single-market program which promised, and did in fact achieve, the completion of the European internal market by the end of 1992.


As a consequence, and to cut a long story short (Scharpf 1996), we now have, within the European Union, an integrated economic area within which legal and administrative restrictions against the free movement of goods and of capital have been completely removed, where restrictions on the free competition among service providers are rapidly being eliminated, and where legal restrictions against the mobility of workers and students are in some regards even lower than they are in mature federal states like the United States of America.[3] Moreover, these rules of "negative integration" are vigorously enforced against any remaining national restraints on trade and any national distortions of free competition not only by the European Commission and the European Court of Justice, but also by the national courts in ordinary administrative and civil proceedings.


For practical purposes, therefore, capital is free to move to locations offering the highest rate of return, and firms are free to move their production and their research and development activities to any place within the Union without jeopardizing in any way their access to their home market (which is not assured for locations outside of the EU). As a consequence, the capacity of national governments to protect domestic firms against competitors producing under different regulatory regimes abroad has been eliminated, and their capacity to tax and to regulate domestic capital and business firms is now limited by the fear of capital flight and the relocation of production. Hence all national governments in the European Union are now forced to compete against each other in order to attract, or retain, mobile capital and firms. In that sense, their situation is now comparable to that of subnational governments in the unified national markets of federal states like the United States, Canada, the Federal Republic of Germany or Switzerland.


In these federal states, regional units of government - states, provinces, Länder or Kantone - have long been confronted with the fact that their own territorial jurisdiction is more narrowly circumscribed than the (effectively national) boundaries of the market, and that they are constitutionally required to respect the freedom of border-crossing movements and economic transactions. It is to be expected, therefore, that the traditional allocation of governing functions in federal states, between the center and the regions, will reflect the fact that capital and business have always been mobile within the boundaries of the national market. From this, important lessons might be drawn for the assignment of governing functions in the European Union (CEPR 1993).


The most clear-cut solution is found in the Federal Republic of Germany, where regulatory competition among the Länder in areas that might affect the location choices of capital and firms is almost totally eliminated. All taxes, including those accruing to the Länder, are fixed by uniform federal legislation; the same is true of all social welfare levies and benefits, including those that are paid out of local-government budgets. Similarly, all health and safety regulations, and all environmental regulations that might affect business enterprises are uniform across the nation, and so are corporation law, labor law, collective bargaining law and, by and large, the collective-bargaining agreements negotiated between nation-wide unions and employers's associations. Thus, the main economic function of Länder governments is the promotion of business by providing public infrastructure including education and training, industry-oriented research facilities and the services of technology-transfer institutions, while financial subsidies to business firms are again regulated by national legislation.


In international comparison, the German degree of regulatory centralization is unusual. Nevertheless, all federal systems have been confronted with the same choice of either moving responsibility for policies affecting capital incomes and the production costs of mobile firms "upward" to the national level, or avoiding such policies altogether. An instructive example is provided by the history of child-labor legislation in the United States in the early decades of this century. During that period, the states of the union were prohibited from interfering with interstate commerce, while regulation of the conditions of production was considered to be beyond the reach of the federal power. Under these conditions, those states that attempted to limit the employment of children in industrial production found that their industries were out-competed by imports from states that still allowed child labor. As a consequence, child labor continued unabated even in the most "progressive" states until the New Deal "constitutional revolution" finally permitted regulation at the federal level after 1937 (Graebner 1977).


In fact, therefore, social policy and the welfare state have remained weak when they were left to subnational governments;[4] and where they are well developed in federal states, they have either become "nationalized" or, where they remain formally at the subnational level (as is true in Canada), they have come to depend so heavily on national subsidies that effective political responsibility rests with the central government as well.


Thus, the lesson to be learned from the experience of federal states seems to point in exactly the opposite direction from the one suggested by Professor Majone: At the level of member states, welfare state policies that clearly would have majoritarian legitimacy are increasingly frustrated by economic competition in the integrated European market; as a consequence, national governments are losing democratic legitimacy in the output-oriented or effectiveness dimension. It would seem to follow that, if effective self-determination is to be maintained, the responsibility for redistributive policies must be shifted from the national to the European level. Europe, in other words, should follow the path taken by federal nation states by developing the "Social Charter" into a comprehensive European welfare state - or at least by harmonizing national social policies in order to eliminate the ruinous competition between national welfare regimes (Leibfried 1992; see also Leibfried/ Pierson 1995).





If this course were taken, however, political responsibility would have to be moved to a level where democratic legitimacy is not (yet) available, and where decisions could not be taken by majority vote in the European Parliament and by a democratically legitimated European government, but would continue to depend on the agreement of national governments in the Council of Ministers and in the European Council. Many of these decisions would still have to be taken unanimously, and even where voting is by qualified majority, the opposition of a few countries can block European action.


It is true of course that the European Commission has an important role as agenda setter, and it is also true that the involvement of the European Parliament has been significantly increased during the last decade. But that does not change the character of the European Union as a negotiation system - it only increases the number of "veto players" and the range of interests that must be taken into account, and it thus makes European policy negotiations even more complex.


In the Council, interests that have influence on the national ministries involved cannot be overruled; the Commission uses a wide network of committees of experts and interest organizations to draft its own initiatives; and the European Parliament prides itself on representing those broad-based concerns that have difficulty in organizing effective pressure groups. In short: In the European negotiation system it is unlikely that major interests will be ignored. This should increase the legitimacy of the outcomes that can be agreed upon - but it also reduces the possibility of reaching agreement.


In fact, therefore, the European Union is capable of effective action only in areas in which the major interests affected are either convergent or complementary. Such areas do exist, and they mainly coincide with those fields of regulatory policy that have been discussed by Professor Majone. But social policy and the welfare state are not among them.


One reason lies in the differences in the level of economic development among the Member States of the Union. In its present composition, the Union includes some of the most highly developed national economies in the world, and some that have only recently been at the level of threshold economies. If we take GDP per capita as a measure of the productivity of an economy, Denmark reached over 28 000 US Dollars in 1994, while in Portugal it was below 8 800 - a difference that is larger than existing disparities among the subnational regions in most federal states. Thus, in order to compete in the European internal market, Portuguese firms must be able to pay not only much lower wages than are paid in Denmark, but non-wage labor costs, taxes and the costs of environmental and other regulations must be much lower as well. Given the much lower levels of disposable income, such costs also could not be shifted from firms to households. In short: the economically less developed Member States simply could not afford to burden their firms, their workers or their consumers with the same level of welfare costs or, for that matter, environmental costs that citizens in the highly developed Member States have come to demand and to accept.


As a consequence, the share of GDP that is committed to the total welfare budget varies between 17.6 % in Portugal, 33 % in the Netherlands, and 40 % in Sweden. It is clear that if these differences were to be harmonized at the lower level, the welfare systems of the more advanced welfare states would be destroyed - and if they were to be harmonized at the higher level, the economies of the less developed member states would be wiped out, just as the East German economy was wiped out when West German standards were imposed.


But even if the North-South conflict within the Union could somehow be neutralized, the harmonization of European welfare states would be extremely difficult and probably impossible. In fact, the more important difficulties are not economic but structural and institutional. To begin with structural differences on the spending side, Italy uses 50 % of its welfare budget for old age pensions, while pensions amount to less than 30 % of total expenditures in Germany and to less than 25 % in Ireland. By contrast, health care takes 30 % of the welfare budget in Germany, but only 20 % in Britain and 18.5 % in Denmark. Denmark on the other hand spends 10.3 % on family allowances, which amount to only 5.5 % in the Netherlands and 3.6 % in Italy (BMA 1995, 14).


Similarly, on the financing side, general tax revenues pay for 87 % of the welfare budget in Denmark, but only for 18 % in France. Employers contributions amount to 53 % of total expenditures in France and to only 24 % in the Netherlands (and less than 8 % in Denmark). Conversely, workers contribute 48 % of the welfare budget in the Netherlands, 29 % in France, 16 % in Italy and merely 5 % in Denmark (BMA 1995, 13).


In the institutional dimension, finally, European welfare states have evolved in highly heterogeneous forms (Esping-Andersen 1990). Take the difference between the universalist welfare systems of the Scandinavian countries on the one hand, and the corporatist systems in Continental Europe. The former have emphasized social services, the latter social transfer payments. The former have tax-based basic pensions combined with contribution-based and "funded" supplemental pensions, while the latter have pay-as-you-go systems that are financed through non-wage labor costs. The Scandinavian countries and Britain have national health services while most Continental countries have insurance-based health care systems in which private physicians are paid on a fee-for-service basis. The list could be extended, and the differences would appear even greater if industrial-relations systems (Crouch 1993) were included in our definition of welfare-state institutions.


Under these conditions, any attempt at European harmonization would require fundamental structural and institutional changes in most of the existing national systems, and we should expect fierce conflicts over which of the institutional models should be adopted at the European level. In the countries that lose out in this battle, it would be necessary to destroy, or to fundamentally reorganize, large and powerful organizations from which hundreds of thousands of employees derive their livelihood and on whose services and transfer payments large parts of the electorate have come to depend. In other words, the political difficulties of harmonizing the institutional structures of mature welfare states would be so overwhelming that it is perfectly obvious why nobody, neither governments nor opposition parties, nor employers associations or trade unions, are presently demanding that the harmonization of social policy should be put on the European agenda.


Instead, we see that in all Member States, governments and unions are struggling with the consequences of economic competition in the internal market. Everywhere, unions accept wage cuts, and governments deregulate and cut taxes on business and on capital incomes, while public services, welfare benefits and pension payments are being reduced. In this, Britain has taken the lead, France has faced the challenge last Winter; the German government formulated the most severe austerity program of the postwar period in April of 1996, and Sweden has followed suit a few weeks later. There may still be ideological differences between governments that welcome the external pressure to reduce the level of redistribution and welfare regulation, and governments that reluctantly give in to economic compulsion. The effect, nevertheless, is the same:


With the completion of the European internal market, all Member States are now competing against each other to attract or retain internationally mobile factors of production - primarily investment capital, producing firms, research and development, and other highly productive services. As a consequence, all see themselves compelled to reduce the regulatory and tax burdens on capital and firms (Sinn 1993; 1994).(5) By the same token, they must increase the tax burdens on workers and consumers, and they must resist and reduce the claims of those groups - the young, the sick, the unemployed and the old - that depend on public services and welfare transfers.





Nevertheless, for all we know from public opinion polls, mass demonstrations, protest strikes and election results, the welfare state continues to have broad political support in Europe (Borre/ Scarbrough 1995). Thus, present welfare cutbacks are not generally interpreted as the outcome of democratically legitimated political choices. Instead, each government is adopting and defending them under the external compulsion of economic necessity.


This necessity, however, has not just occurred by accident; it was brought about by the political decision to complete the European market through the Single European Act of 1986 that was ratified, practically without opposition, in all national parliaments. The ensuing difficulties were easy enough to foresee, and they were in fact correctly predicted by many. We must assume, therefore, that these difficulties were thought to be overbalanced by the economic benefits of the internal market (or, conversely, by the alleged "costs of non-Europe"). At any rate, the European market is now integrated, its economic consequences are unlikely to be reversible, and there seems to be no one that is seriously advocating that Europe should return to a regime of protected national markets.


If that is so, the dilemma that has been brought about by the political choices of democratically accountable national governments and parliaments must be squarely confronted. With the completion of the internal market, national governments are no longer able to continue the social and welfare state policies that their citizens have come to take for granted and continue to demand. At the national level, therefore, democratic self-determination has lost its effectiveness. At the same time, the attempt to harmonize social policy at the European level would lack authentic democratic legitimacy and, within the existing institutional structures of the Union, would be politically unfeasible.


At this point, we might well join Jean-Marie Guéhenno (1993) in declaring the "end of democracy" in Europe. If we are not willing to do so just yet, we also need to once more re-examine our concept of democracy. We have defined self-determination as authenticity plus effectiveness, meaning that democratic governments should be effective in realizing the undistorted preferences of citizens. But if we were to apply this definition to individual self-determination, it would describe the expectations of a fairly immature person. What is missing is what Sigmund Freud called the "reality principle" through which infantile fantasies of omnipotence are tempered with a realistic assessment of one's own possibilities and limitations.


In the political world, it is true, the ideas of popular sovereignty and of parliamentary sovereignty are easily associated with the notion of omnipotence, especially in unitary nation states like Britain where policy choices are constrained neither by a constitutional court, nor by federal structures, nor by an independent central bank. Even there, of course, there have been external constraints in foreign and military relations - but these areas have always been either ignored or considered anomalies in democratic theory. On the other hand, while constraints on the exercise of political power are well known in countries with a bill of rights, constitutional separation of powers and federal structures, these institutional "checks and balances" could be understood as being self-imposed by the democratic sovereign in order to better structure and control the actions of its agents. Thus, there is again no theoretical conflict with the dual postulates of democratic self-determination.


The economic constraints under which national governments are operating in the internal market are of a different nature. They are not generally accepted as being normatively legitimated - even though there is a body of neo-liberal and Hayekian theory arguing the case for an "economic constitutionalism" that presupposes regulatory competition among territorially limited jurisdictions (Mestmäcker 1994; Streit 1996). Nor are these constraints quasi-natural givens. They had not existed twenty or even ten years ago; they were brought about by the cumulative effects of political choices; and they could still be changed, even though at very high cost, by concerted political action.


In other words, these are constraints that limit the freedom of political choices in ways which are neither normatively beyond challenge nor empirically inevitable, and which moreover operate not at the fringes of the political domain, but in the very core of the modern welfare state. This condition is also different from the one which has always characterized democracy at the level of local and regional governments, which was often construed to exist by delegation of the democratic sovereign defined at the national level, and where in any case it was possible to assume that problems that could not effectively be dealt with at local and regional levels could be left to national political processes of superior democratic legitimacy.


Instead we have here a situation in which the freedom of choice of democratic processes at the national level is being massively constrained in an area of central political concern, while at the higher, European, level where action might be effective, democratic legitimacy is weaker or non-existent. In short: We have reached the point where it is imperative, for the survival of democratic governance, that normative political theory as well as political practice should come to grips with the conditions of "democracy without omnipotence".


These are difficult conditions which force us to acknowledge how much even "realist" theories of democracy have depended on the assumption of omnipotence to support the comfortable belief that the political practice of Western democracies could be taken as a pragmatically sufficient approximation of democratic ideals. Democratic authenticity presupposes that, in principle, every citizen should have the ability to participate in, or least attend to, public discourses over policy choices. That does not seem altogether impossible as long as such discourses will primarily focus on values, or policy goals, or on the desirability of policy outcomes - rather than on issues of feasibility. Even more important: Only if governments are thought to be omnipotent does it make sense to base political accountability on the judgment of voters who, on the basis of their own experience, are only able to judge the desirability of (actual or promised) policy outcomes, rather than the technical quality of policy choices in the light of feasible alternatives. In other words, the more choices are constrained, and the more issues of feasibility increase in salience, the more difficult is it to argue that the empowerment of ordinary citizens, and be it only in the attenuated role of voters in general elections, will be automatically conducive to good government and, hence, to effective self-determination.


But neither would it be plausible to argue (and Majone certainly does not so argue) that technical expertise, by itself, could provide a sufficient basis of legitimacy for the full range of policy choices for which governments must assume responsibility. Even if issues of feasibility increase in importance, conflicts over desirability retain their salience. Ideally, therefore, democratic publics as well as governments and opposition parties would have to debate over operative preferences in the light of what are feasible aspirations under the prevailing external constraints, and voters would have to respond to government performance in the light of the same criteria. The implication is that if democratic authenticity is to be maintained in the absence of assumed omnipotence, the informational demands on democratic processes must increase substantially. Alternatively, "realist" theories of democracy would in fact have to conclude that effective policy can only be made behind a public smokescreen of symbolic politics focusing on personalities, scandals and trivialities - in which case the prevention of "predatory rule" (Levi 1988) by self-serving governors would indeed be as unlikely as some Public-Choice theorists have claimed (Riker 1982).


Yet none of this would make much of a difference one way or another if the completion of the European internal market would in fact imply the frustration of all goals and aspirations that had in the past been associated with the welfare state in European democracies. Even if democratic self-determination cannot be equated with omnipotence or wish fulfillment, it does imply that in areas of high political salience meaningful political choices must remain available at those levels of a multi-level political system where authentic democratic participation is possible.


At the national level, therefore, it is important to explore whether and how it is still possible to pursue the aspirations of the welfare state under the economic conditions of the European internal market. Similarly, it is necessary to explore what contribution might be expected from European policy processes under conditions where the European Union cannot be a majoritarian democratic polity but will remain, for the time being, a complex negotiation system.


In either case, it will be necessary to respecify the original goals and strategies in the light of what the nation state is still able to do, and in terms of what the European Union in its present institutional form is able, or unable, to achieve. But if that is done, I am fully convinced that the range of significant choices that are still available is much wider than is suggested by the political hysteria about cost cutting and welfare cutbacks that is presently sweeping all European countries.


There is no possibility for me to spell out the concrete implications of this invocation of the reality principle. Nevertheless, a few examples are necessary to illustrate its potential. I begin with some strategic options that are still available at the national level:

  • First, if the terms of trade between capital and labor have irreversibly shifted in the favor of capital, political parties and unions that are still committed to egalitarian goals would finally need to shift their attention from wage and tax policy to the distribution of capital assets. It would be useful therefore to go back to earlier versions of Rudolf Meidner's plans for the Swedish "wage earner funds" - that is before they were overburdened with all the anti-capitalistic ambitions of the social-democratic Left of the 1970s. Acceptance by firms could be secured if, in collective bargaining agreements, a part of the "normal" wage settlement were used to "buy" equity in the firm, title to which would then be transferred to investment funds. In order to avoid disincentives to mobility, and to spread the risks of bankruptcy, individual workers could then become shareholders in these investment funds, rather than in their own firms. Over the medium term, then, income from work would be supplemented by income from capital.

  • Second, if wage and tax cutbacks seem unacceptable as a competitive strategy, unions and governments must concentrate their efforts on helping to raise business profitability in other ways. That implies industrial-policy and industrial-relations efforts that will increase productivity and the capacity to innovate, rather than protect existing jobs and firms against the market. It is important to note, however, that such efforts are only likely to succeed if each country is building on the comparative advantages of its existing institutional structures, rather than trying to import the latest fads from abroad. Thus, if Britain should find its competitive advantage in deregulated and highly flexible labor markets, that is no reason for Germany to dismantle its own institutional advantages that depend on trustful cooperation between management, banks and organized labor.

  • Third, if financing welfare expenditures through taxes on business and non-wage labor costs leads to capital flight, disinvestment and job losses, a larger part of the welfare burden must be shifted to taxes on consumption. As I have said before, this is largely true in Denmark, where 87 % of the welfare budget is financed from general tax revenues - including high levels of the value added tax. Significantly, therefore, Denmark also seems to be one of the few countries in which the welfare state is hardly affected by the present agitation over international competitiveness. One implication is of course that the European Union should not impose upper limits on VAT rates, and that the attempt to switch from the country-of-destination principle to the country-of-origin principle should finally be abandoned.

    An even more radical solution would generalize the Swiss model of health care finance. Here the role of the state would be reduced to requiring everybody to take out private insurance against typical social risks, and to subsidizing the premium payments of low-income groups. Since this would shift the major burden to private consumption, the overall size and cost of the welfare state would cease to be a major public or economic issue.

    In both instances it is clear that the problems of transition from one system to another could be enormous, and they would vary greatly from one country to another, depending on the particular staring situation. This is another reminder that European harmonization could not succeed in the social policy field, and that individual countries must be allowed to work out their own, path-dependent solutions.

  • Finally, if the welfare state is bankrupted by high and rising rates of unemployment, governments could attempt to increase the "employment content" of welfare expenditures - for instance by subsidizing low-wage employment instead of financing full-time unemployment.

In a very limited form, this is achieved by the "Earned Income Tax Credit" in the United States. The most general, and in many ways the most attractive solution would be the "negative income tax" which would provide a basic income for everybody while maintaining incentives to seek paid employment. Given the financial constraints under which all governments are presently operating, however, it is unlikely that a generous scheme of this kind would be adopted in the near future. More attractive from a fiscal and labor-market point of view would be a targeted scheme in which a degressive income subsidy would be paid to all workers in jobs paying wages at or below the poverty level.


Such a program would encourage the evolution of a low-wage labor market - which presently does not exist in European welfare states, and in which a considerable part of the American "employment miracle" has taken place. If employment is to increase significantly, these low-paying job opportunities for persons with low levels of skill cannot be ignored any longer. Income subsidies would permit their utilization without allowing the emergence of a large population of the "working poor" who are compelled to work full time to achieve incomes that remain below the poverty level. If successful, such income subsidies could pay for themselves in terms of reduced outlays for full-time unemployment. The major opposition therefore comes from unions that are still committed to the principle that only "good jobs" should be allowed in European welfare states - even if that implies accepting high levels of unemployment.


The list could be extended. That is not my purpose here. I merely hope to suggest that even in the field of social policy, which is said to be most directly exposed to competitive pressures in the internal market, there are still economically feasible choices at the national level that would make a significant difference for the solidaristic and egalitarian goals that have always been associated with the welfare state.


Similarly, it is true that European decision making in the social policy field is impeded by high consensus requirements in the face of widely divergent interests. Nevertheless, it seems possible to respecify some of the goals of social regulation at the European level in such a way that agreement is facilitated, and that European regulation could be employed to reduce or limit the damage done by the ruinous competition among national regulatory and welfare systems in the internal market. Again, I will limit myself to a few suggestions that should merely illustrate the basic approach.

  • First, if conflicts over environmental and other regulations arise primarily from differences in the level of economic development among member states, agreement on harmonization might nevertheless be reached for regulations stipulating dual standards - one defined at the aspiration level of the most highly industrialized and most polluted countries, and one defined at the level which the economically less developed Member States can afford.

  • Second, if conflicts of interest arise primarily from structural and institutional differences among Member States, agreement might nevertheless be reached on the harmonization of quantitative levels of effort. Thus I have argued that in the social policy field not only institutional differences but also the different structures of financing the welfare state and the different emphases on certain types of welfare spending must rule out harmonization at the European level. At the same time, however, differences in total welfare spending are less great and much more regular. By and large, the richer European countries commit larger shares of their GDP to welfare expenditures than do poorer countries.

[click on graphic to enlarge] 

  • While welfare spending is not quite proportional to wealth measured in per-capita GDP, it is nevertheless highly correlated. Thus, if we plot welfare spending as a share of GDP (Eurostat 1995, Table 3.3, 1992 data) against per capita GDP (1994 data), the 12 member states of the European Union were all fairly close to the regression line. Among the countries that did spend more than their "normal" share on welfare were the Netherlands and, remarkably after 17 years of Conservative welfare retrenchment, the United Kingdom. Equally remarkable is the fact that among the lower outliers we find not only Portugal and Ireland, but also (West-)Germany with its allegedly excessive welfare burden.

    The diagram suggests the existence of a latent consensus among the member states of the Union, according to which, regardless of structural and institutional differences, the welfare state should increase in relative importance as countries become more affluent. It also suggests the possibility that this latent consensus could be transformed into an explicit consensus supporting a European convention against ruinous competition in the welfare sector. What would be needed is a binding agreement according to which all countries would avoid welfare cutbacks that would push their total welfare expenditures below a lower threshold which might be defined at, or slightly below, the present location of countries that are lower outliers in the diagram. If such a rule were in force now, Germany could not cut welfare spending to reduce its public-sector deficit, while the Netherlands and Britain would still be able to do so. In other words, the rule would merely limit the extent to which welfare cutbacks could be used to increase international competitiveness, but it would leave countries free to pursue any structural or institutional reforms within these overall quantitative constraints.

  • Finally, agreement on the lower boundary of welfare spending would also not prevent the competitive reduction of employers´contributions. In a sense, this is inevitable: Countries that have traditionally financed large shares of the welfare budget in the form of payroll taxes are not only at a competitive disadvantage internationally, but are also depressing overall employment as a consequence of high total labor costs. It would be counterproductive if they were prevented from shifting some of the welfare burden to other sources of finance.

    What would be desirable instead would be the Europe-wide harmonization, perhaps again graduated relative to per-capita GDP, of taxes on business and capital incomes. However, since national tax systems and systems of tax collection also differ very widely, it would probably be extremely difficult to achieve harmonization at the level of specific taxes. What might nevertheless be feasible is, again, agreement on a quantitative rule defining the minimum share of total revenue that a country should obtain from all types of taxes on business and capital incomes.

    Once more, this would leave countries free to reform their tax systems according to their own preferences, but it would reduce the pressure, created by capital mobility in the European market, to use tax cuts for competitive advantage and to shift all financial burdens from capital onto labor and consumers.

These lists are certainly neither exhaustive nor definitive. But they suggest that even under conditions of a fully integrated internal market in Europe opportunities for significant and effective political choices are still available at the national level. It is true that they do not allow the continuation of solutions that were successful and have found democratic support in the postwar decades. They require major changes of strategy that are particularly painful for labor unions and the parties of the left. But they do not imply the capitulation of democratic politics before the overwhelming force of capitalist markets.


At the same time, the European Union cannot be transformed into a democratic polity within the near future, and it also cannot assume the functions of the national welfare state. But the institutions of the European Union can be used to enable and to assist national political solutions that will defend the rights of citizenship and the values of the welfare state against the forces of the integrated market. When these opportunities are grasped, the future of democracy in Europe seems not nearly as bleak as the present debate often suggests.



  1. The same argument would stand in the way of Europe-wide referenda which have been suggested as a potential remedy against the democratic deficit (Zürn 1995; Grande 1996).

  2. It is also interesting to note that the first serious conflict in which the legitimacy of European decisions is directly being challenged in Germany has arisen over the application of the competition rules of the Treaty, in particular of Art. 92 para. 2, to subsidies paid by the East German Land of Saxony to Volkswagen. 3 For instance, state universities in the United States charge much higher fees for students from other states, while the European Court of Justice has ruled that no fees may be collected from students from another EU member state, even if fees have to be paid by domestic students.

  3. Critics of policy centralization or harmonization point to the fact that Switzerland has been able to maintain a highly decentralized tax and welfare system despite complete market integration (Feld/ Kirchgässner 1995). But the share of GDP committed to total welfare spending is lower in Switzerland than in any other European country at a comparable level of economic development (Eurostat 1995, Table 3.31).

  4. I do not wish to argue that regulatory competition in Europeanized or globalized markets must always lead to deregulation. Like firms, countries may not only be engaged in price (or cost) competition but also in quality competition. That is true where high levels of regulation are advantageous to consumers or investors. Thus international competition has in fact contributed to the upgrading of national regulations in the fields of banking and stock exchanges. But that is no reason for comfort when social policy regulations and taxes that impose costs on consumers and investors are in question.



BMA (1995)
Euro-Atlas. Soziale Sicherheit im Vergleich. Bonn: Bundesministerium für Arbeit und Sozialordnung.

Borre, Ole/ Elinor Scarbrough, eds. (1992)
Beliefs in Government. Volume Three. The Scope of Government. Oxford: Oxford University Press.

CEPR (1993)
Making Sense of Subsidiarity. How Much Centralization for Europe? London: Centre for Economic Policy Research.

Crouch, Colin (1993)
Industrial Relations and European State Traditions. Oxford: Clarendon Press.

Esping-Andersen, Gøsta (1990)
The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.

Eurostat (1995)
Statistische Grundzahlen der Europäischen Union. 32. Ausgabe. Luxemburg: Amt für amtliche Veröffentlichungen der Europäischen Gemeinschaften.

Feld, Lars P./ Gebhard Kirchgässner (1995)
Fiskalischer Wettbewerb in der EU: Wird der Wohlfahrtsstaat zusammenbrechen? In: Wirtschaftsdienst 10, 562-568.

Graebner, William (1977)
Federalism in the Progressive Era: A Structural Interpretation of Reform. In: Journal of American History 64, 331-357.

Grande, Edgar (1996)
Demokratische Legitimation und europäische Integration. Ms. Köln: Max Planck Institute for the Study of Societies.

Guéhenno, Jean-Marie (1993)
La fin de la démocatie. Paris: Flammarion.

Leibfried, Stephan (1992)
Europe's Could-Be Social State: Social Policy in European Integration after 1992. In: William James Adams, Hrsg., Singular Europe - Economy and Polity of the European Community after 1992. Ann Arbor: University of Michigan Press, 97-118.

Leibfried, Stephan/ Paul Pierson, eds. (1995)
European Social Policy: Between Fragmentation and Integration. Washington, D.C: Brookings.

Levi, Margaret (1988)
Of Rule and Revenue. Berkeley: University of California Press.

Majone, Giandomenico (1989)
Evidence, Argument and Persuasion in the Policy Process. New Haven: Yale University Press.

Majone, Giandomenico (1994)
The European Community: An "Independent Fourth Branch of Government"? In: Gert Brüggemeier, Hrsg., Verfassungen für ein ziviles Europa. Baden-Baden: Nomos, 23-43.

Majone, Giandomenico, (1994a)
The Rise of the Regulatory State in Europe. In: West European Politics 17, 77-101.

Mestmäcker, Ernst-Joachim (1994)
Zur Wirtschaftsverfassung in der Europäischen Union. In: Rolf H. Hasse/ Josef Molsberger/ Christian Watrin, eds., Ordnung in Freiheit. Festgabe für Hans Willgerodt zum 70. Geburtstag. Stuttgart: Fischer.

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.

Scharpf, Fritz W. (1996)
Negative and Positive Integration in the Political Economy of European Welfare States. In: Gary Marks/ Fritz W. Scharpf/ Philippe C. Schmitter/ Wolfgang Streeck. Governance in the European Union. London: Sage, 15-39.

Sinn, Hans-Werner (1994)
Wieviel Brüssel braucht Europa? Subsidiarität, Zentralisierung und Fiskalwettbewerb im Lichte der ökonomischen Theorie. In: Staatswissenschaften und Staatspraxis 5, 155-186.

Sinn, Stefan (1993)
The Taming of Leviathan. Competition Among Governments. In: Constitutional Political Economy 3, 177-221.

Streit, Manfred E. (1996)
Competition among Systems, Harmonisation and European Integration. Jena: Max Planck Institute for Research into Economic Systems, Diskussionsbeitrag 01-96.

Weiler, Joseph H.H. (1981)
The Community System. The Dual Character of Supranationalism. In: Yearbook of European Law 1, 257-306.

Weiler, J.H.H. (1995)
Does Europe Need a Constitution? Reflections on Demos, Telos and the German Maastricht Decision. In: European Law Journal 1, 219-258.

Weiler, J.H.H. (1996)
European Neo-constitutionalism: in Search of Foundations for the European Constitutional Order. In: Political Studies 44, 517-533.

Zürn, Michael (1996)
Über den Staat und die Demokratie im europäischen Mehrebenensystem. In: Politische Vierteljahresschrift 37, 27-55.


Copyright © 1996 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, D - 50676 Köln, Germany




MPIfG:  MPIfG Working Paper 96/2

[Zuletzt geändert am 29.03.2007 10:59]