The paradoxical logic of collective action

It is commonly assumed that if all the members of a group agree on some common interest, then there would be a tendency for the group to seek to further this interest, i.e., the group would also act in a self-interested or group-interested manner. Mancur Olson's important contribution to political economy has been his proof that this familiar assumption is basically wrong (Olson Reference Olson1971). This proof depends on the argument that the members of the group will not provide as much of the collective good as it would be in their common interest to provide. Such a tendency towards sub optimality is due to the fact that by definition a collective good, such as economic integration, is available to all members of a group once it has been provided by any member. Since an individual member gets only part of the benefit of any expenditure he makes to obtain more of the collective good, he will discontinue his purchase of it before the optimal amount for the group as a whole has been obtained.

In some small groups (like the original Six of the 1951 Paris and 1957 Rome Treaties) one or two members may find that they would be better off if the collective good were provided – even if they had to pay the entire cost of providing it themselves – than they would be if it were not provided at all. In such situations there is a presumption that the collective good will be produced. Thus in small groups, where each member gets a substantial portion of the total gain simply because there are few others in the group, a collective good can often be produced by the voluntary, self-interested action of the members of the group. Actually, the greatest likelihood that a collective good will be provided occurs in the case of small groups of members of unequal size (again, the situation of the original Six); for the greater the interest of any single member in the collective good, the greater the likelihood that this member will get such a significant portion of the total benefit from the provision of the good that she will gain from seeing that the good is produced, even if she has to pay the entire cost herself. Even in the smallest groups, however, the members of the group will not provide as much of the good as it would be in their common interest to provide. This less than optimal provision of the collective good is again due to the fact that, by definition, a collective good is such that other members of the group cannot be kept from consuming it once any group member has provided it.

The logic of collective action operates not only in groups of individuals but also in the case of states. An interesting example of this is provided by the well-known historian Heinrich August Winkler in the first volume of his Geschichte des Westens (Winkler Reference Winkler2012: 760). According to the 1821 ‘federal war constitution’ (Bundeskriegsverfassung) of the German confederation, no state was allowed to participate with more than three army corps to the confederate army. This rule was dictated by the desire to avoid even ‘the appearance of the supremacy of one member state over the other members of the Bund’. The Prussian state was a member of the Bund, but important Prussian territories – East and West Prussia, and Posen – were not part of the German confederation. For this reason the Prussian state was allowed to have nine, rather than three, army corps. Because of this, the other states of the Bund could assume that in case of a foreign threat Prussia would mobilize its entire military might. As a consequence they kept their contribution to the common defence well below the limit allowed by the federal constitution.

In sum, the size of the group is a key element in the logic of collective action, and ‘the larger the group, the further it will fall short of providing an optimal amount of a collective good’ (Olson Reference Olson1971: 35; italics in the original). Olson concludes that if larger groups/organizations do in fact exist this is not because of the collective good they provide but because of the power of coercion they may enjoy (as in the case of the state) or because of selective (positive or negative) incentives that voluntary organizations may provide. A selective incentive is one that applies selectively to the individuals (or subgroups) depending on whether they do or do not contribute to the provision of the collective good. A well-known example of selective incentive in the context of the EC/EU is the Common Agriculture Policy, a positive incentive offered to France for its support of European integration.

Unfortunately, the larger the group the more difficult it becomes to provide suitable selective incentives to all the members of the group. The availability of such incentives is limited, in case of the EU, not only by the size of the group but also by the socio-economic heterogeneity of its members. An even more basic problem in organizing and maintaining socially and economically heterogeneous groups is due to the fact that the members are less likely to agree on the exact nature of whatever collective good is at issue or on how much of it is worth ‘buying’. In the EU context we have the traditional cleavage between the countries that wish to limit integration to the economic sphere and those that support also political integration. Consensus on such matters is especially difficult because the defining characteristic of collective goods – that they go to everyone in the group if they are provided at all – entails that all the members of the group have to accept whatever level and type of the good is provided. It follows that political entrepreneurs who attempt to organize collective action will be more likely to succeed if they strive to organize relatively homogeneous groups. Thus some of the most serious problems the European Commission is experiencing today, particularly in the area of policy harmonization (see, e.g., Piris Reference Piris2011: 8–12), originated in the ‘big bang’ enlargement of 2004–07—an enlargement which the Commission, paradoxically, eagerly supported.

The applicability of Mancur Olson's analysis to the case of European integration could be demonstrated by a number of other examples. It is therefore rather surprising that the concepts developed in the present section seem to have been generally ignored by political scientists, legal scholars and even by economists writing about the EU. Given the importance of the limits of collective action for understanding the causes of the present crisis of EU-style integration it may be useful to list here Mancur Olson's key propositions, as summarized by Cornes and Sandler (Reference Cornes and Sandler1996: 324–6):

1. (1) Large groups may not provide themselves a collective good.

2. (2) The larger the group, the lower the collective provision level.

3. (3) Large, well-endowed members will bear a disproportionate burden of collective provision, so that the small will exploit the large.

4. (4) Asymmetric groups (including both well-endowed and poorer members) are more apt to contain at least one member or subgroup whose benefits from collective action will exceed the associated costs, even if these costs are solely born by the individual member or the subgroup.

5. (5) Collective action can be promoted, up to a point, by selective incentives and/or institutional design.

Olson's Logic of Collective Action was first published in 1965. Since the mid-1970s, neo-institutional economists have provided important insights into other dimensions of the logic of collective action. Key to the understanding of these broader aspects of collective action is the notion of transaction costs.

Political transaction costs

Transaction costs are the costs of operating an economic, political, or social system. Thus the costs of using the market system have been classified as: (1) costs of preparing contracts, i.e., search and information costs, narrowly defined; (2) costs of concluding contracts: bargaining and other decision-making costs; and (3) costs of monitoring and enforcing the contractual obligations. Such costs are clearly different from the costs of producing goods and services – the only costs considered by neoclassical economics – but they are as real and significant as production costs. Following Williamson (Reference Williamson1985), we can think of transaction costs, whether they arise in economic, political, or social systems, as the equivalent of friction in physical systems.

The idea of studying political processes in the transaction-cost mode seems to have originated with Douglass North (Reference North1990), who, however, focused on a particular kind of transaction cost, namely failure of ‘instrumental rationality’ for participants in the political process. Furubotn and Richter (Reference Furubotn and Richter2000: 47) define political transaction costs as the costs of supplying public goods by collective action. Specifically, these are: (1) the costs of setting up, maintaining, and changing a system's formal and informal political organization; and (2) the costs of running a polity. Actually, most forms of economic transaction costs appear also in political processes; but political processes are likely to be even more beset by transaction costs than are economic processes and relationships. This observation suggests that a transaction-cost politics perspective should take advantage of the main results of transaction-cost economics, but move beyond them if it is to provide new insights. This is precisely the research programme sketched by the distinguished Princeton economist Avinash Dixit in The Making of Economic Policy (Dixit Reference Dixit1996).

The most significant transaction costs of EU-style integration are best understood as costs of collective ownership, i.e., by thinking of the member states as the ‘owners’ of the EU – in the same sense in which the members of a producer or consumer cooperative are, in fact, its owners. Ownership, as the term is conventionally used, has two essential attributes: the exercise of formal control and the receipt of residual gains. Formal control does not necessarily mean effective control. Thus, in business corporations, cooperatives, and mutual enterprises, formal control generally involves only the right to elect the firm's board of directors and to vote directly on a small set of fundamental issues, such as mergers or the dissolution of the firm (Hansmann Reference Hansmann1996). The costs of ownership, and of collective ownership in particular, have received only limited attention by the new institutional economics of organization, while they have been either ignored or not recognized as such by writers on European integration. The costs of collective ownership may be broken down into three categories: the owners' costs of decision-making; the costs of delegation (or agency costs); and the costs of risk bearing.

To grasp intuitively the nature of decision-making costs in cases of collective ownership, think of a worker-owned firm. When many persons share ownership of a firm there are likely to be differences of opinion concerning the firm's policies and programmes. Serious problems arise especially when the outcome of the collective decision will affect different owners differently. Thus, if a worker-owned firm must shut down one of its two plants, the workers at the two plants are likely to have very different preferences about which plant should be chosen. When the interests of the individual owners differ, mechanisms for collective choice, such as voting, engender costs, known as ‘costs of collective decision-making’.

The costs of collective decision-making are logically distinct from the agency costs discussed below. The latter costs arise in the context of a principal(s)–agent relation and include the costs of monitoring the agent and the costs resulting from the failure to monitor the agent with perfect effectiveness. On the other hand, the costs of collective decision-making can be significant even in modest-sized organizations, in which there are no hired managers and hence no significant agency costs. The costs associated with collective choice mechanisms are of two broad types. First, there are the costs resulting from decisions whose outcomes fail to maximize the aggregate welfare of the owners themselves as a group, favouring instead the interests of some subgroup. During the current euro crisis, for example, a number of key collective decisions were deemed to favour the interests of the creditor countries rather than the interests of the other members of the Eurozone, not to mention the interests all the members of the EU. Second, there are the costs of the decision-making process itself.

Examples of the second type of costs of collective decision-making are bargaining and influence costs. Bargaining costs are the costs involved in negotiations among different parties. They include not only the opportunity costs of bargainers' time, but also resources expended trying to improve one's bargaining position, the costs of monitoring and enforcing the agreement, and the losses from failure to reach an agreement. A most important source of bargaining costs at the supranational level is the opportunistic behaviour of bargainers mainly interested in national (rather than ‘European’) benefits and costs. Influence costs are the costs incurred in attempts to bias the decisions of the central institutions in a self-interested fashion; in attempts to counter such influence activities by others; and by the resulting degradation of the quality of decisions. Students of economic organizations have suggested that influence costs are one important reason why many, if not most, merged organizations cannot do everything the separate components did, and more (Williamson Reference Williamson1985; Milgrom and Roberts Reference Milgrom and Roberts1992). Judging from the extensive literature on interest groups in the EU, influence costs are at least as pervasive, and potentially even more serious, than bargaining costs in their distorting consequences. Influence costs are particularly significant when decisions are taken at the level of the European Council or of the EU Council of Ministers.

One of the main causes of high decision-making costs in the EU is the need to reach a collective decision satisfying different, even contradictory, national interests. As a result, many important issues might remain unresolved or are even completely neglected. Agreement on EMU, for example, was celebrated as a turning point in European integration. However, the Maastricht Treaty left unanswered a number of fundamental policy questions concerning monetary union. In order to make political agreement possible, the question of measures to coordinate economic policies, or to provide compensatory budgetary transfers, was sidestepped. Also, issues of external monetary policy, unitary external representation of the monetary union, exchange-rate policy, and political accountability were left unsettled. Even the basic question, whether it made economic sense to adopt a ‘one-size-fits-all’ monetary policy for structurally very different economies, was never properly discussed. As a result, EMU turned out to be a high-risk project with no easy exit option when things go wrong.

Delegation, or agency, costs are the second category of costs of collective ownership. One of the most significant features of the institutional architecture of the EC/EU is the extensive delegation of powers to the supranational institutions. According to Miles Kahler, the member states ‘have delegated more important and extensive functions to European institutions than has been the case with the members of other international or regional institutions. The Commission, for example, surpasses even the strongest secretariat of an international organization’ (Kahler Reference Kahler1995: 85). Agency costs are particularly significant in the case of multiple owners, or principals, all of whom have some power to influence the actions of the agent. What in the literature is known as ‘common agency’ is a situation where multiple principals try to influence the agent. In the usual case, i.e., when the principals have different preferences and priorities, the delegation produces results that are significantly inferior to what could be achieved by one truly unified principal. In part this is because the results that can be achieved when each principal offers a particular incentive to the agent tend to cancel each other out (for a formal proof of this and related results concerning multi-principal agencies, see Dixit Reference Dixit1996: 157–71).

The costs of risk bearing have been deliberately ignored in the EU before the seriousness of the euro crisis had to be officially admitted. As I argue elsewhere (Majone Reference Majone2014), the political culture of European leaders, until recently, has been one of total optimism. The traditional approach to European integration (generally known as the Monnet strategy of fait accompli) implies that the success of a collective decision is determined by the decision-makers themselves – by the very fact that they agreed on the decision – rather than by those who will be affected by the eventual outcomes. This emphasis on the process of decision-making rather than on actual results excludes a priori the possibility of failure. For this reason the question of feasibility has been systematically ignored by integrationist leaders. Thus, there is no indication that the feasibility of the goal of the 1970 Werner Plan – monetary union by 1980 – was ever seriously considered. The so-called ‘bicycle theory’ of European integration – according to which integration must keep moving forward, especially in a crisis, for the bicycle (that the EU is seen to be) not to fall – provides the conceptual justification of the strategy of fait accompli. As noted above, the costs of risk bearing are today recognized and accepted (however reluctantly) by all EU leaders. It is, however, important to realize that, in the words of an economist at Goldman Sachs, Dirk Schumacher, ‘as long as the main political legitimacy rests on the national level, there are limits to how much international risk-sharing is possible’ (Wall Street Journal, 20–22 December 2013, p. 4).

A strong institutional leadership could reduce, if not annul, many of the costs mentioned above. However, the idea that leadership may be provided by a single country or institution is incompatible with the principle of ‘collective leadership’, hence with the assumption of the essential equality of all the member states. To understand the problem of leadership in the context of EU-style integration, it may be helpful to keep in mind that there are different kinds of leadership.

Three kinds of leadership

John Plamenatz (Reference Plamenatz1973: 83–90), distinguishes three activities, all of them leadership in the broad sense of the word, which are important in the sphere of governance and politics: management; government; and leadership in the strict sense. Managers provide leadership by directing the work of others. Government, as distinct from management, consists, not in directing the work of others but in making rules for their guidance and applying those rules to them. Finally, leadership in the strict (political) sense consists in promoting or defining some goal that the people (or a section of them) share, or that the promoter hopes to get them to share. This kind of leadership consists, not in ruling or managing others, but in speaking for them. Speaking for people in the sense relevant here means, above all, giving expression to aims, beliefs, and feelings that people are willing to endorse, though they may never have thought about them until they were adopted by some trusted leader. To be a leader in this sense is to have a large say in defining goals, principles, and attitudes which are, or are supposed to be, shared.

Plamenatz's taxonomy allows us to identify situations that require leadership in the political sense of the term. In his classic study, Leadership in Administration, the well-known American sociologist Philip Selznick points out that when leadership (in Plamenatz's strict sense) fails, it is more often by default than by positive error. One type of default is the failure to set goals. Selznick writes:

Much administrative analysis, he adds, takes the goal of the organization as given, whereas in many crucial instances this is precisely what is problematic. It is therefore important to emphasize the leader's responsibility to define the mission of the enterprise. A related type of default of leadership occurs when goals, however neatly formulated, enjoy only a superficial acceptance and do not really influence the total structure of the enterprise and its specific activities. It follows that the task of building special values and a distinctive competence into the organization is a prime function of leadership. Selznick's analysis of leadership relies heavily on the distinction between ‘organization’ and ‘institution’. While the term ‘organization’ refers to an ‘expendable tool’, an ‘institution’ is a responsive, adaptive organism, heavily dependent on its social or political embedding, and on its contacts with the outside world. Keeping this distinction in mind, it becomes clear why the default of leadership shows itself in an acute form when organizational survival – or even organizational achievement in terms of resources, procedures, and stability – is confounded with institutional success.

The EU may be the clearest illustration of Selznick's thesis that, for the bare continuity of organizational existence, leadership in the strict sense is often dispensable. The greatest challenges to institutional leadership, on the other hand, arise when goals are not well defined, and when fluid situations require constant adaptation. Many of the present problems of European governance can be traced back to the scarcity of institutional leadership relative to the scope of the powers transferred to the supranational level. The fact that the need of leadership grows with the scope of institutional commitments should be as obvious as the point that the need of democratic legitimation grows with the scope of competences transferred to the supranational level. In fact, both needs are seldom recognized explicitly and hence deserve to be emphasized here. The 1957 Treaty of Rome did not require much institutional leadership since its most important articles were concerned with negative integration, i.e., the removal of obstacles to the free movement of people and goods (Majone Reference Majone2009: 138–43). Given these limited objectives, the kinds of leadership discussed by Plamenatz under the labels of management, and of government as rule-making, were sufficient. The default of institutional leadership became increasingly apparent as successive treaties kept expanding supranational competences without a statement of mission more specific than ‘ever closer union’; without robust popular support; and even without a corresponding increase in institutional resources. Indeed, the strategy of fait accompli, traditionally used to expand the activities of the EC/EU, represents the clearest admission of the absence of institutional leadership at the European level. This strategy consists in pushing ahead with ambitious integration projects without worrying too much about either feasibility or democratic legitimacy. The rationale behind this strategy, as Pascal Lamy clearly understood long ago, was the realization that the people were not ready to accept the steady expansion of supranational powers. Thus, instead of trying to rally the people in support of some clearly defined goals, Jean Monnet and his followers preferred, in Lamy's words, ‘to get on without telling [the people] too much about what was happening’ (cited in Ross Reference Ross1995: 194). Such lack of trust is incompatible, not only with the notion of institutional leadership but also with the ideology of collective leadership.

The European Commission as would-be leader

A key element of the ideology behind the post-Second World War model of European integration was the basic equality and equal dignity of all members of the would-be community of nations, from the smallest to the largest: no leader but a ‘collective leadership’. This principle of formal and (to the extent possible) substantive equality of the member states has inspired all the European treaties and also the day-to-day practice. It is also reflected in the design and modus operandi of EU institutions. Indeed, an important, if tacit, responsibility of the Commission is to ensure that the interests of the smaller member states are sufficiently taken into consideration – which explains why the strongest supporters of the Commission have always been the smaller countries. Clear evidence of the importance the smaller member states attach to being directly represented in this body is the promise made by the European Council, after the failure of the first Irish referendum on the Lisbon Treaty (June 2008) to abandon the planned reduction of the number of Commissioners – a promise meant to facilitate the success of the second referendum. The problem is that a return to the old rule of one Commissioner per member state would result in too high a number of Commissioners for the institution to work efficiently. With a membership of 28 states (with more to come in the future) at very different levels of socio-economic development and with quite different needs and priorities, it is increasingly difficult to follow the traditional ‘one-size-fits-all’ approach (‘harmonization’) to policymaking.

As a consequence of this diversity, in a number of important policy areas the Council does not receive enough proposals because the Commission itself cannot agree on what is needed or desirable. The problem is aggravated by the fact that the member states, old and new, increasingly see ‘their’ Commissioner as their own representative in the Commission. To avoid these and other difficulties, decisions are made by consensus, which means that with so many members the Commission can hardly take decisions (Piris Reference Piris2011). Under such conditions this institution is unable not only to use its monopoly of legislative and policy initiative but even to exercise the weaker forms of leadership mentioned in a preceding section.

Thus the Commission, which many euro-enthusiasts used to see as the kernel of the future government of a politically united Europe, in fact looks more and more like an international bureaucracy and less and less like a proto-government. Even the European Council – the most likely candidate to provide leadership at the supranational level – is only able to achieve what the member states want it to achieve, with agreements hammered out, often bilaterally, beyond its walls (Peterson and Shackleton Reference Peterson, Shackleton, Peterson and Shackleton2012). There are good reasons to believe that, short of a radical transformation, political leadership will always be an extremely scarce commodity in the EU. The problem is that in most circumstances a leaderless group will provide a less than optimal supply of a collective good – even if every member of the group desires that good. In a later section of this chapter we shall see how recent attempts to overcome the crisis of monetary union have undermined the principle of equality both within the euro group and between this group and the other members of the EU. This development is closely related to the steady loss of influence of the European institution, which, in the intentions of the founding fathers, should have provided all three kinds of leadership mentioned by Plamenatz. Indeed, the Commission is, or was meant to be, ‘a distinct hybrid: the European Union's largest administrator and main policy manager, as well as a source of political and policy direction’ (Peterson Reference Peterson, Peterson and Shackleton2012: 96).

Walter Hallstein, the first president of the Commission from 1958 to 1967 stressed particularly the political function of this hybrid institution, often referring to himself as the equivalent of a European prime minister. Also Jacques Delors envisioned an economically and politically united Europe, with the Commission as the federal executive. Both Hallstein's and Delors's activism illustrate the pitfalls and risks of such analogies. Hallstein's attempt to turn the Commission into an embryonic government of the European Economic Community put him on a collision course with French president de Gaulle, who bitterly opposed any extension of the Commission's powers. The showdown came in early 1965, when Hallstein introduced proposals to link completion of the financial arrangement of the Common Agricultural Policy with greater executive authority for the Commission and budgetary powers for the European Parliament. The ‘Empty Chair’ crisis, which paralyzed the Community for seven months, was eventually resolved at a foreign ministers’ meeting in Luxembourg on 28 and 29 January 1966, where agreement was reached to adopt an interim financial regulation for the CAP, deferring the question of additional powers for the Commission and the European Parliament. Both sides approved the famous Luxembourg Compromise, which maintained the principle of majority voting, but acknowledged that ‘when very important issues are at stake, discussions must be continued until unanimous agreement is reached’. The Commission was in a marginalized position during the crisis and in finding a solution to it. In the words of Piers Ludlow (Reference 240Ludlow, Palayret, Wallace and Winand2006: 90), ‘Community crisis and the need for political decisions brought the member states to the fore and minimised the scope for the Commission to use the vast technical expertise it had built up of the EEC's day-to-day operations’. The Commission was not invited to the ministerial discussions among the other five member states, and was even excluded from some of the meetings of the committee of permanent representatives. The regularity and frequency of such encounters during the period of the French boycott ‘did graphically underline how completely the Commission had been pushed to the margins of a dispute it was widely – if rather unfairly – blamed for starting’ (ibid.: 91).

Jacques Delors took over as president of the European Commission in January 1985. The new president was looking for a key project around which the process of European integration could be relaunched. After rejecting a number of possibilities that might fail to attract the unanimous support of the member states, Delors settled on the completion of the internal market as the central plank of his rélance of the integration process – the White Paper on Completing the Internal Market listed the obstacles to free movement within the Community and identified the types of action necessary to overcome these obstacles. All the approximately 300 measures of the legislative programme were to be adopted no later than by the end of 1992. More than 20 years after the ‘Europe “‘92”’ target, however, the single market is still far from being a reality. The most serious problem area when it comes to completing the single market was, and remains, the services sector. Since in all modern economies the services sector accounts for at least 70 per cent of GDP and 50 per cent of employment, the free movement of goods is no longer sufficient to ensure market integration. Most services are still regulated at the national level, while socio-economic conditions vary so much across the Union – income inequality, as measured by the Gini concentration coefficient, is far greater in the enlarged Union than in the USA – that centralized harmonization of national laws and regulations is likely to reduce aggregate welfare, while ex post harmonization, via regulatory competition, continues to meet strong political opposition from EU institutions and most member states. The aim of Delors's Single Market project was to open the EU internal borders to the free movement of all the factors of production and exchange, as within a nation state. This aim is generally supposed to have been more or less achieved but the truth is rather different: ‘In many areas the Single Market exists in the books but, in practice, multiple barriers and regulatory obstacles fragment intra-EU trade and hamper economic initiative and innovation’ (Piris Reference Piris2011: 15).

Another aim of Delors was to advance political integration in Europe on the back of a programme of market liberalization. This goal has been missed even more patently than the single market for services. ‘Delors wanted a politically integrated Europe. The first step was the single market, once you have a single market you have a single currency, and then the door is open to political integration’. This is how Alexandre Lamfalussy, at the time general manager of the Bank for International Settlements, summarized the grand strategy of the president of the European Commission in an interview with David Marsh in May 2007 (Marsh Reference Marsh2010: 122). In 1988 – almost 20 years after the initial announcement of the Werner Plan for achieving economic and monetary union by 1980 – Delors began a new campaign for EMU, and persuaded the European Council to set up a committee composed of central bankers and economic experts, under his (Delors's) chairmanship. Echoing the proposals of the 1970 Werner plan, the Delors Committee Report recommended a three-stage move to EMU, starting with closer economic and monetary coordination. Unfortunately, the document left many important questions open, supposedly to be settled by the governments later on. As explained by Delors himself to David Marsh in 2007, ‘fundamentally EMU was a political issue that could only be decided by the governments. In the report, I wanted to draw attention to the need for sufficient economic coordination as a precondition for EMU’ (ibid.).

What Delors called the ‘acceleration of history’ in Central and Eastern Europe in 1989 and 1990 made differences of opinion concerning monetary union appear relatively unimportant. Germany's attention was now focused on national reunification rather than European affairs, and this fact may have facilitated final agreement on EMU. At the same time, the geopolitical events of those years seemed to revive interest in European Political Union, but the most visible result of two years of frenetic discussions about the nature, scope, and competences of EPU was the three-pillar ‘temple’ of the Maastricht Treaty. Delors had vainly opposed this arrangement in favour of a ‘tree’ whose different functions would be connected to a common Community ‘trunk’ (Ross Reference Ross1995). In fact, the Commission had very little influence over the negotiations on political union. Maastricht marked the end of the post-1985 rélance of European integration, rather than the big leap forward that Delors had wanted. George Ross has correctly identified the basic flaw in Delors's strategy: the failure of the French leader to realize that the traditional Monnet method of ‘integration by stealth’ could only work as long as Europe did not impact significantly on the daily lives of ordinary Europeans. After monetary union this method could not work anymore.

Delors was followed by Jacques Santer as Commission president. Presiding over an administration that had become inefficient, in some respects even chaotic, the Santer era culminated in the dramatic mass resignation of the entire Commission in March 1999, after the publication of a report of a Committee of Independent Experts, convened by the European Parliament, on charges of fraud, mismanagement, and nepotism. Since the resignation of Santer and his Commissioners, the institution that was supposed to drive integration forward has become increasingly weak. Thus the Commission was marginalized first in the negotiations that led to the Treaty of Nice, and then in the Convention on the Future of Europe that drafted the Constitutional Treaty. Romano Prodi, the new Commission president (1999–2004), proposed without success a constitutional structure in which almost all decisions would be taken by majority vote; closer cooperation among some member states would be excluded; and the Community method would be extended to apply to most policy areas, including the common foreign and security policy. Unsurprisingly, a former member of the Commission later lamented the Prodi Commission's ‘astonishing weakness’ (citation in Peterson Reference Peterson, Peterson and Shackleton2012: 117). As noted above, today the Commission is unable, not only to use its monopoly of legislative and policy initiative, but even to exercise the weaker forms of leadership mentioned in a preceding section.

The euro crisis and the problem of the reluctant hegemon

In his classic study of the 1929–39 world depression, Charles Kindleberger writes: ‘for the world economy to be stabilized, there has to be a stabilizer, one stabilizer’ (Kindleberger Reference Kindleberger1973: 305). He goes on to note that in the 1970s leadership was a word with negative connotations, while collective decision-making was regarded as more ‘aesthetic’. In those years the leadership of the USA was beginning to slip, and the American scholar doubted that ‘the rising strength of Europe in an enlarged European Economic Community will be accompanied by an assertion of leadership […] and assumption of responsibility for the stability of the world system’ (ibid.: 307–08). In fact, neither the EEC nor the EU ever tried to assume responsibility for the stability of the world system. The present crisis of EMU has made clear that only Germany could provide the needed leadership – if not for the world system then at least for the Old Continent; not necessarily on a permanent basis but at least in the present predicament and in equally serious crisis situations of the future (see below). The attempt to establish a sort of French–German duopoly or diarchy – in the short Merkozy interlude – did not work, as Kindleberger would have predicted. Actually, this author could call attention to his discussion of the role of France in the great depression of the 1930s, to point out analogies with the present euro crisis:

Today's Germany is neither feeble like the United Kingdom in the inter-war period nor irresponsible like the USA in the same period. However, in spite of its active role in the euro crisis, it is extremely reluctant to play the role of a true leader – for very good reasons having to do with history but also with the ideology of the European integration project. As far as history is concerned, one should always keep in mind Ludwig Dehio's conclusion that the only occasion on which the peoples of Europe seem to display solidarity is when they oppose the hegemonic aspirations of one of the states of the continent (Dehio Reference Dehio1961: 25). This historical legacy explains why the basic equality and equal dignity of all the member states is a key element of the ideology of European integration, as noted above. Given the level of mutual mistrust, especially in crisis situations, it is thus almost unimaginable that the members of the EU, large or small, would be willing to accept a German hegemony, even though Germany, de facto, already conditions the economies of the other member states in important respects.

A common currency was supposed to accelerate the process of economic and political convergence towards a common model of European governance. Paradoxically, one unanticipated consequence of monetary union has been the extent to which the competitive position of the members of the Eurozone have diverged. Italy, Spain, Portugal, Greece, and, more recently, France have lost a significant amount of competitiveness, while Germany, Austria, the Netherlands, and a few other northern countries have gained a significant amount of competitiveness. The other unanticipated consequence of EMU is the general acknowledgment of the infeasibility of political union, now and in the foreseeable future. The possible consequences of the present state of affairs have been drawn by the Hungarian-born American financier George Soros who, in an essay published in September 2012, argued that in order to avoid a definitive split of the Eurozone into creditor and debtor countries, and thus a likely collapse of the EU itself, Germany must resolve a basic dilemma: either assume the role of the ‘benevolent hegemon’ or else leave the Eurozone.

If Germany were to give up the euro, leaving the Eurozone in the hands of the debtor countries, all problems that now appear to be insoluble could be resolved through currency depreciation, improved competitiveness, and a new status of the ECB as lender of last resort. The common market would survive, but the relative position of Germany and of other creditor countries that might wish to leave the Eurozone would change from the winning to the losing side. Both groups of countries could avoid such problems if only Germany were willing to assume the role of the benevolent hegemon. However, this would require the more or less equal treatment of debtor and creditor countries, and a much higher rate of growth, with consequent inflation. These may well be unacceptable conditions for the German leaders, for the Bundesbank, and especially for the German voters. At any rate, both conditions could be satisfied only with significant progress on the road to a political union, in which Germany would accept the responsibilities implied by its leadership role (Soros Reference Soros2012). In a later essay, Soros (Reference Soros2013) added that the present problems are the consequences of the fatal flaw of the European monetary union: in creating the ECB as a fully independent central bank the member states effectively indebted themselves in a currency which they cannot control. As a consequence, when the risk of a Greek default became concrete, the financial markets reacted by reducing the status of all heavily indebted members of the Eurozone to that of developing countries with large debts in a foreign currency.

Also, some German commentators have recently maintained that their country should agree to play a hegemonic role, and not just in the present crisis. One of the clearest statements of this view is an essay by Christoph Schoenberger, ‘Hegemon wider Willen’ [Unwilling hegemon] (Schoenberger Reference Schoenberger2012). The author argues that the crisis of the Eurozone has made clear what people have been reluctant to acknowledge in the past, namely that Germany has become the hegemonic power in Europe. While this leading role provides a unique possibility to influence the future of the Old Continent, it also entails duties and burdens. Unfortunately, German political leaders and public opinion are quite unprepared to meet this challenge. In fact, as the old East–West conflicts and the post-war division of Germany pass into history, awareness of the precarious situation of the country, placed as it is at the heart of Europe, seems to recede. And yet German guidance is precisely what Europe needs in order to face up to the uncertainties and challenges of the twenty-first century, according to Schoenberger. The largest, most populous and economically most powerful member of the EU cannot evade this responsibility. To speak of German hegemony is not to advocate Germany's dominance over the rest of Europe, but only to recognize the fact that there are no strong and democratically legitimated institutions at the European level. Those who criticize the European Commission for its alleged interference in domestic affairs, for example, do not realize that very little can happen in Brussels without the agreement of the governments and the bureaucracies of the member states. Even the European Parliament tends to be ignored by voters who see no connections between national and European politics. The steadily decreasing turnouts in European Parliament elections provide the clearest proof of the people's indifference towards the institution that should represent them at the European level. In fact, the crisis of the Eurozone, with its sequel of summits of the heads of government following one another, has emphasized the intergovernmental features of the Union. On the other hand, the difficulty of finding a solution of the present crisis at the intergovernmental level, under the assumption of the basic equality of all the member states, can only lead to the conclusion that what the EU needs is a hegemonic power – a need usually concealed by the search for consensus which characterizes decision-making in Brussels.

Unfortunately, the argument continues, Germany is not mentally prepared to assume the burden of a hegemonic role in Europe, and also its institutions were not designed for such a task. Thus, the Bundestag, under the pressure of public opinion and the influence of recent decisions of the German Constitutional Court, has developed a tendency to limit more and more the freedom of action of the federal government in foreign and European affairs, pretending that the decisions of the executive in this area should depend on explicit parliamentary mandates. What, then, are the alternatives to a German hegemony? The only reasonable alternative, according to Schoenberger, would be a strengthening of the supranational institutions to make them dependent, not on the agreement and directions of the national governments, but on the support of European voters. Unfortunately, such a solution is impossible since the EU has been designed around an inter-governmental core, as the present crisis has amply demonstrated. Thus we are back to the need of a ‘wise’ German hegemony – unless we are prepared to accept the break-up of the Union.

It is refreshing to read such a crisp analysis of an issue that most Germans, in all walks of life, prefer to ignore. Unfortunately, the analysis is far from being either entirely convincing or complete. First, the author starts from the traditional notion of hegemony as preponderance of material resources. Other, more refined, notions of hegemony and hegemonic stability do not assert an automatic link between power and leadership (see below). But even the cruder notion does not seem to fit the facts. While it is certainly true that Germany is larger, more populous, and economically more powerful than any of its European partners, its superiority is hardly as clear as in historically undisputed cases of international hegemony. In the nineteenth century, the Pax Britannica was based not only on Britain's naval supremacy but also on Britain's pioneering role in the first industrial revolution, on the higher productivity of its industry, at least until the 1870s, and on the role of London as the financial centre of the world. The Pax Americana of the 1950s and 1960s was based on the absolute superiority of the USA in production, services, research and development, technological innovation, not to mention military might. Today Germany does not play any comparable role, even on the modest European scale. In particular, this country is hardly a leader in the services sector, which, as we know, accounts for more than 70 per cent of the GDP of all modern economies. In the case of financial services, for example, the European leader is Britain, not Germany.

The more sophisticated notion of hegemony emphasizes the importance of domestic attitudes, political structures, and decision-making processes: ‘Decisions to exercise leadership are necessary to “activate” the posited relationship between power capabilities and outcomes’ (Keohane Reference Keohane1984: 35). Schoenberger is fully aware that this kind of ‘force activation’ is still lacking in Germany; unfortunately, he gives no idea of how this limitation could be overcome, or how hegemonic stability could be maintained once the present crisis of the Eurozone is somehow resolved. The most serious flaw in the analysis of the German scholar, however, is the implicit assumption that, barring a total collapse of the EU, European integration will remain stable at the present level, or even resume its march towards ‘more Europe’ once the crisis is over. In fact, integration is becoming more differentiated, and there are also clear indications of a retrograde movement in several respects. Thus, the goal of a truly integrated European market appears to be less and less feasible; and it can even be argued that the European market is actually less integrated today than twenty or thirty years ago. This is because of the growing importance of services, which, as was noted above, are still largely regulated domestically; but also because growing socio-economic heterogeneity within the Union makes policy harmonization increasingly problematic.

Again, monetary union was originally meant to be a public good, shared by all member states, while it turned out to be a club good (see the following section). Recent proposals of two-speed integration (e.g., by Piris Reference Piris2011 and by Bofinger et al. Reference Bofinger, Habermas and Nida-Ruemelin2012) go in the same direction of differentiated integration. At the end of this process of opting out, making exceptions, and separating avant-gardes and rear guards, we may reach a situation where a hegemon leading the entire membership of the EU is no longer needed: either the aims of European integration are redefined to include a few, limited, and widely shared goals – a return to the spirit of the Rome Treaty with few, concrete commitments, and collective leadership; or else the EU becomes a ‘club of clubs’, with the supranational level mainly responsible for monitoring competition among different territorial and/or functional clubs (Majone Reference Majone2014). As noted above, it seems to be highly unlikely that European integration can remain stable at the present level. In fact, the crisis of monetary union and the remedies so far adopted have already deeply transformed the nature of the EU.

European integration: from collective good to club good

The cost of adhering to the principle of the essential equality of all the member states keeps growing as the membership of the EU expands and more and more competences are transferred to the supranational level. As mentioned in an earlier section, the incentives for group action diminish as group size increases, so that larger groups are less able to act in their common interest than small ones. This is why participation in large voluntary organizations depends, not on the collective benefits but on the individualized incentives these organizations provide. For the EU, however, it is increasingly difficult to provide such incentives since its resources do not grow in proportion to its expanding membership and to the growing scope of its competences.

If neither the leadership nor the incentives and sanctions that make collective action possible is available, then one may well doubt whether EU-style integration may still be properly considered to be a collective good. Thus, monetary union, initially conceived as a collective good to be shared by all the member states of the EU, was quickly transformed into a ‘club good’ by the British and Danish de jure opt-outs and by Sweden's de facto opt-out. According to the economic theory of clubs (Buchanan Reference Buchanan1965), a ‘club good’ is a public good from whose benefits individuals may be (or may wish to be) excluded; an association established to provide an excludable public good is a ‘club’. The same definitions apply if, instead of individuals, we consider independent states. Associations of independent states, such as confederations, are typically voluntary, and their members are exclusively entitled to enjoy certain benefits produced by the association. The club goods in question may be collective security, policy coordination, common technical standards – or a monetary union limited to a subset of members. The important point is that as an association of states expands the cost of uniformity in the provision of collective goods can escalate dramatically. The economic theory of clubs predicts an increase in the number of voluntary associations to meet the increased demand of goods more precisely tailored to the different requirements of various subsets of more homogeneous states. Aggregate welfare is maximized when the variety in preferences is matched by a corresponding variety of institutional arrangements.

Buchanan's economic theory of clubs provides a good conceptual foundation for the functional (rather than territorial) approach to supranational governance – an approach advocated by David Mitrany in the 1940s and by Ralph Dahrendorf in the 1970s. The essential principle of a functional organization of supranational activities, according to Mitrany, ‘is that activities would be selected specifically and organized separately, each according to its nature, to the conditions under which it has to operate, and to the needs of the moment’ (citation in Eilstrup-Sangiovanni Reference Eilstrup-Sangiovanni2006: 57). At the same time, Mitrany was sceptical about the advantages of political unions. His main objection to schemes for continental unions was that the closer the union the more inevitably would it be dominated by the more powerful member – a point which has been hardly considered by later writers on European integration but is becoming highly relevant today, as suggested by the above discussion of Germany as potential, if reluctant, hegemon.

The economic theory of clubs emphasizes the advantages of institutional pluralism, and implies that an efficient assignment of tasks between different levels of governance need not coincide with existing national boundaries: there may be significant externalities and a need for coordination between some, but not all, regions within a country or group of countries. The theory also explains why a number of tasks which used to be assigned to central governments are today performed by private, increasingly transnational, organizations. Although there is a strong historical correlation between standardization and the emergence of the sovereign territorial state (Spruyt Reference Spruyt1994), current views on standardization have changed radically as a result of the advance of globalization, the development of technology, and the growing variety and sophistication of technical standards. As Alessandra Casella (Reference Casella, Bhagwati and Hudec1996) has argued, the fact that in today's integrating world economy the relevant community of standards users need not be territorially defined, distinguishes the traditional approach from the contemporary understanding of standards as a special class of club goods. Moreover, as the complexity of a society increases, the number of clubs tends to increase as well. This is because greater diversity of needs and preferences makes it efficient to produce a broader range of club goods. The general implication of Casella's argument is that top-down harmonization may be desirable when the market is relatively small and homogeneous. In a large market, on the other hand, harmonization tends to be brought about by the recognition of similar demands or needs, rather than by a policy imposed from the top. Hence a multiplicity of club goods tends to replace harmonized policies.

A Europe of clubs organized around functional tasks would not exclude the possibility of large projects supported by all the member states – as long as there is clear evidence (through referendums, supermajorities in national parliaments, etc.) of sufficient popular support. This is precisely the point Dahrendorf wished to emphasize with his statement that ‘everyone does what he wants and […] no one must participate in everything’, a situation that ‘though far from ideal is surely much better than avoiding anything that cannot be cooked in a single pot’ (cited in Gillingham Reference Gillingham2003: 91–92). Concretely this meant that there would be common European policies in areas where the member states have a common interest, but not otherwise. This, said Dahrendorf, must become the general rule rather than the exception if we wish to prevent continuous demands for special treatment, destroying in the long run the coherence of the entire system.

The view of Europe as a ‘club of clubs’, far from being new, has deep roots in the history of the Old Continent. As several distinguished historians have argued, the European global dominance of the past was made possible not by centralization but by fragmentation and by the competition stimulated by fragmentation. In turn, political rivalry facilitated the commercial revolution of the Middle Ages by making it possible for the mercantile community to bypass, where necessary, the rules of this or that city or state. The Hanseatic League, which has been viewed by some scholars as a major unifying force in northern Europe, developed from being an association of merchants to a league of towns on the Baltic and the North Sea. One of the tasks of the league was to facilitate the exchange of information between merchants, but the Hansa was not merely an economic association. Like a state, it waged wars, and on occasion it could make or break kings. In fact, the Hansa was an alternative to the sovereign state (Spruyt Reference Spruyt1994: chap. 6). The ‘European miracle’ made possible by a unique combination of competition and cooperation continued in later centuries. The states of early modern Europe were surrounded by actual or potential competitors. Hence the fairly rapid diffusion of policy and institutional innovations throughout the continent in the period preceding the full development of the nation state. In its state-system, writes Eric Jones (Reference Jones1987:115), ‘Europe had a portfolio of competing and colluding polities whose spirit of competition was adapted to diffuse best practice’. But as the national states of Europe became more self-conscious they began to impede the movement of capital and people by closing their borders and enforcing cultural and legal homogeneity on their people. The long century of nationalism that followed was an aberration in European history, and the integration process started after the Second World War was a healthy, indeed a necessary, reaction to the catastrophic consequences of nationalist ideology. In the early stages of the integration process it was not unreasonable to assume that the European Economic Community would evolve, sooner or later, into a politically integrated bloc, perhaps even into something like a federation. That assumption is no longer tenable in a Union of 28 members at vastly different stages of socio-economic development, with different geopolitical concerns, and correspondingly diverse policy priorities. A Europe of clubs seems to be the only solution of the present crisis of European integration that is both feasible and historically justified.