The Rise of the Regulatory State in Europ T G. MAJONE Abstract Privatization and deregulation have created the conditions for the rise of the regulatory state to replace the dirigiste state of the past. Reliance on regulation—rather than public ownership, planning or centralised administration—characterises the methods of the regulatory state. This study examines the growth of regulation in Europe, at the national and Community levels. It stresses the advantages of this mode of policy making, but also recognises its problems. It is suggested that political accountability can be ensured by a variety of substantive and procedural controls, among which judicial review is especially important. Executive oversight and coordination may be improved by using new tools of public management like the regulatory budget or the regulatory clearing house. Regulation and the Redrawing of the Borders of the State A paradoxical consequence of the international debate about privatization and deregulation has been to focus the attention of European policy makers and scholars on regulation as a distinctive mode of state intervention in the economy and society. In the words of a legal scholar, regulation has become the new border between the state and the economy, and the battleground for ideas on how the economy should be run.1 A political scientist observes that regulation is a pervasive and widely accepted phenomenon in all advanced countries.2 According to an economist, the regulation issue—what, how, and at what level of government to regulate—is the core of the compromise between the European Community and its member states that made the Internal Market programme possible.3 This consensus on the significance and distinctiveness of regulation is a relatively new phenomenon. Until recently, European scholars devoted little attention to the special features of regulation that distinguish it from other modes of policy making. Thus, while the American deregulation Rise of the Regulatory State in Europe 193 movement was preceded and prepared by decades of intensive research on the law, economics and politics of the regulatory process, in Europe the terms 'deregulation' and 'privatization' gained sudden currency—even in Great Britain the words were scarcely heard of before 1978—with hardly any intellectual preparation. The academic and political debate in the United States has been greatly facilitated by the fact that the meaning of regulation is fairly clear within the framework of American public policy and administration (see below). By contrast, European scholars tend to identify regulation with the whole realm of legislation, governance, and social control. Such a broad use of the term makes the study of regulation almost coextensive with the study of law, economics, political science, and sociology. If it is true that reduc-tionism is a necessary condition of scientific progress,4 it is not surprising that the analysis of regulation as a particular type of policy making is still in its infancy in Europe. The relative neglect of regulatory analysis in the past corresponded to the low visibility of regulatory activities. While in the United States the j tradition of regulation by means of independent agencies combining legislative, administrative, and judicial functions goes back to the Interstate Commerce Act of 1887—and even earlier in states like Wisconsin, Massachusetts and New York—the tendency in Europe has been to treat regulatory issues as either purely administrative, and so tasks for central departments or ministries, or as judicial, and so matters for determination by courts or court-like tribunals. In Britain, for example, tribunals like the Railway Commission (created in 1873), proved so common that by 1933 the regulation of British public utilities was viewed by some as considerably impaired by this reliance on the quasi-judicial method and by the resulting failure to develop the administrative commission. What was absent was a powerful agency that applied a special expertise, employed its own secretariat and regulated (in the sense of imposing a planned structure on an industry or social issue). Regulators, instead of instituting action, responded to the competing proposals of private interests.5 Even these timid beginnings of an autonomous regulatory function were forgotten in the era of nationalisations and municipalisations. In most countries of Europe, public ownership of key industries such as railways, telecommunications, electricity, gas, water and other natural monopolies was supposed to protect the public interest against powerful private interests. In this respect, nationalisations and municipalisations were the functional equivalent of American-style regulation. Indeed, at 194 G. Majone Rise of the Regulatory State in Europe 195 one level, this equivalence is so close that it is possible to establish a one-to-one correspondence between typical forms of 'regulatory failure' and certain well-known problems of public ownership, as shown by Table 5.1.6 Table 5.1. Comparing Two Types of Government Failure Failures of Economic Regulation_Failures of Nationalised Industries__ Capture of regulators by regulated firms Capture of public managers by politicians and trade unions Overcapitalisation (so-called Averch-Johnson Overmanning effect) Anticompetitive regulation Public monopolies Vague objectives ('regulate in the public Ambiguous and inconsistent goals given to interest') public managers Poor coordination among different regulators Poor coordination among different public enterprises Insufficient political accountability of No effective control over public enterprises by independent regulatory agencies Parliament, the courts, or the sponsoring minister Because of these analogies, it has been argued that there is no great difference between public monopolies, like the European Post and Telecommunications Ministries, and privately owned but publicly regulated monopolies like the American Telephone and Telegraph Company before deregulation. However, this argument overlooks one important point: the purpose of public ownership was not simply to regulate prices, conditions of entry and quality of service, but also to pursue many other goals including economic development, technical innovation, employment, regional income redistribution, and national security. While nationalisations and other traditional forms of direct state intervention were thus justified by appealing to a variety of often conflicting goals, regulation has a single normative justification: improving the efficiency of the economy by correcting specific forms of market failure such as monopoly, imperfect information, and negative externalities. Because regulation is more narrowly targeted, and because many of the traditional goals of nationalisations have become obsolete or can be pursued more efficiently by other means, privatisations tend to strengthen, rather than weaken, the regulatory capacity of the state. The British experience, for all its imperfections, is quite instructive in this respect. Paralleling the sale process of industries such as British Telecommunications and British Gas has been the development of a whole new regulatory structure.7 This structure rests on a body of economic law involving specific obligations and license conditions placed on the privatized industries, and on a new breed of regulatory agencies—the regulatory offices (ROs). The ROs combine several functions: they regulate prices; they ensure that the privatized firms comply with the terms of their licenses; they act as a channel for consumer complaints and as promoters of competition in the industry they regulate. Detected instances of monopoly abuse are referred to the Office of Fair Trading and to the Monopolies and Mergers Commission (MMC). Thus, privatization has led to a considerable widening of the scope of agencies to promote competition. Now the MMC has a direct role in the regulation of utilities, while prior to privatization it did not have the competence to examine the potentially anti-competitive practices of the nationalised industries. Regulation of the competitive behaviour of the privatized industries is further strengthened by the availability of the competition law of the European Community (EC) which offers considerably more powerful remedies than are available under British law.8 Similarly, in America newly deregulated industries have lost their pre-existing statutory immunity from anti-trust laws, and despite major changes in the telecommunications sector, important segments of the industry remain regulated.9 In sum, neither privatization nor deregulation have meant a return to laissez-faire or an end to all regulation. Privatization changes the role of the state from a producer of goods and services to that of an umpire whose function is to ensure that economic actors play by the agreed rules of the game. Deregulation often means less restrictive or rigid regulation: a search for ways of achieving the relevant regulatory objectives by less burdensome methods of government intervention, as when command-and-control methods are replaced by economic incentives. Thus, neither American deregulations nor European privatizations can be interpreted as a retreat of the state, but rather as a redefinition of its functions. What is observed in practice is a redrawing of the borders of the public sphere in a way that excludes certain fields better left to private activity, while at the same time strengthening and even expanding the state's regulatory capacity in other fields like competition or environmental and consumer protection. Normative and Positive Theories of Regulation In order to explain the sudden growth of administrative regulation, as well as the lateness of its arrival on the European political stage, it is important to specify more precisely the basic characteristics of this mode of policy 196 G. Majone making. As noted above, within the framework of American public policy and administration, regulation has acquired a meaning which is often imperfectly understood in Europe. To use Philip Selznick's formulation,10 regulation refers to sustained and focused control exercised by a public agency over activities that are socially valued. The reference to sustained and focused control by an agency suggests that regulation is not achieved simply by passing a law, but requires detailed knowledge of, and intimate involvement with, the regulated activity. This requirement will necessitate, sooner or later, the creation of specialised agencies entrusted with factfinding, rule-making, and enforcement. The emphasis on socially valued activities excludes, for example, much of what goes on in the criminal justice system: the detection and punishment of illegal behaviour is not, per se, regulation in the sense in which the term is used here. On the other hand, market activities can be 'regulated' only in societies that consider such activities worthwhile in themselves and hence in need of protection as well as control. If the first part of Selznick's definition reminds us of the different institutional traditions of the United States and Europe, and specifically of the lack of a tradition of independent regulatory bodies in Europe, the second part reminds us of important ideological differences in the past. As discussed above, nationalisations may be viewed, in important respects, as the functional equivalent of American-style regulation. The fact remains that nationalisations were not undertaken in order to support the market by improving its efficiency, but rather to replace market criteria by political and administrative ones. The first wave of nationalisations coincided with the first worldwide depression of the capitalist economy (1873-96) which shattered popular and elite support of the market for almost one century. The late Peter Jenkins 1 exaggerated only slightly when he wrote in 1988 that only now, for the first time in this century, the governing classes of Europe no longer assume that socialism in some form is what history has in store. By contrast, the American rejection of nationalisation as an economically and politically viable option expressed a widely held belief that the market works well under normal circumstances and should be interfered with only when it does not function properly. The normative theory of regulation expresses this belief in analytic terms. According to this theory, public regulation of economic activity is justified only when the market is incapable of producing a social (Pareto) optimum. This occurs when some form of market failure is present: monopoly power, negative externalities, inadequate or asymmetrically distributed information, public goods. Regulation should be used to increase the efficiency of the economy by Rise of the Regulatory State in Europe 197 correcting market failures, but not for other purposes, however legitimate, por example, income redistribution should be achieved through social policy, not through regulation. An important recent development is the recognition that market failures provide only a prima facie case for intervention, since the costs of public intervention may exceed the benefits. Hence 'regulatory failure' must also be considered (see Table 5.1). A frequent criticism of the market-failure approach is precisely that it is a normative, not a positive, theory. It provides a basis for identifying situations where the government ought to do something, tempered by considerations of regulatory failure. Many political scientists and economists argue that analysts should focus their attention not on normative issues but on describing the consequences of government programmes and the nature of the political processes that produce such programmes. Normative analysis, it is said, is irrelevant since policy outcomes depend, not on norms or ideas of the public interest, but on factors such as the rules of the political process, the incentives facing various participants in the process, and the changing configurations of power and interests in society. According to the positive theory, as formulated by George Stigler among others, regulation is not instituted in the public interest, that is, for the protection and benefit of the public at large or some large subclass of the public, but is acquired by an industry and is designed and operated primarily for its benefit.12 How else could one explain the price and entry regulation of basically competitive industries such as airlines, trucking, banking and telephone services, or the anti-competitive licensing of so many professions and trades? The positive theory has greatly enriched our understanding of the regulatory process, but has not made the normative theory obsolete. As always in the social sciences, the distinction between normative and positive is much less sharp than positivists used to think.13 In a useful survey paper on 'Regulation in Theory and Practice', Joskow and Noll call 'normative analysis as a positive theory' (or NPT) the theory which regards market failure as the motivating reason for the introduction of public regulation.14 The characterisation is appropriate, since the normative theory is often successful in explaining the origin and development of many regulatory policies. As Peltzman15 writes: If there is an empirical basis for the NPT's continuing attraction for economists, it is probably its apparent success as an entry theory. Consider Hotelling's classic statement in 1938 of the natural monopoly version of the NPT. In this purely theoretical piece, railroads and utilities are presumed, without much evidence, to be the main real-world examples of natural monopoly. They also occupied most of the regulatory (including public ownership) effort when Hotelling wrote. The 198 G. Majone correspondence between the NPT and the real-world allocation of regulatory effort seems striking. Now consider the postwar expansion of regulation. In terms of the resources involved, the biggest single chunk is probably accounted for by environmental regulation, where the externalities aspect of the NPT scores another success. In Europe, too, as we shall see in the following pages, much of the recent growth of regulation can be explained in normative, public-interest, terms. But even when regulation is best explained by the political or economic power of groups seeking selfish ends, those who attempt to justify it must appeal to the merits of the case. Legislators, administrators, scholars, and the public at large wish to know whether the regulation is justified. All of them seek standards against which to judge the success of a policy and the merits of specific programmes initiated within the framework of that policy. The Growth of Regulation in Europe Administrative regulation—economic and social regulation by means of agencies operating outside the line of hierarchical control or oversight by the central administration—is rapidly becoming the new frontier of public policy and public administration throughout the industrialised world. The absence of an efficient regulatory framework is increasingly seen as a major obstacle to modernisation. Thus, as a 1993 issue of The Economist points out, one of the serious problems of privatization in Russia is that there is no regulatory system to control trading in vouchers or shares. This 'regulatory black hole' has already claimed many victims among uninformed investors, and creates an irresistible temptation for any swindler.16 The growth of administrative regulation in Europe has greatly accelerated during the last two decades. In France, for example, the expression 'autorite administrative independante' was used for the first time by the law of 6 January 1978 creating the Commission Nationale de l'lnforma-tique et des Libertes, but several independent regulatory agencies already existed prior to that date: l:he Commission de Controle des Banques created in 1941 and transformed into the Commission Bancaire by the law of 24 January 1984; the Commission des Operations de Bourse (1967), whose powers have been significantly extended by the law of 2 August 1984; the Commission des Infractions Fiscales (1977); the Commission des Sondages (1977); the Mediateur (1973), the only single-headed regulatory agency created so far in France. Today there are almost 20 independent Rise of the Regulatory State in Europe 199 agencies including, in addition to those already mentioned, the Commission d'Acces aux Documents Administratifs (1978), the Commission de la Securite des Consommateurs (1983), the Conseil de la Concurrence (1986), and the Commission de Controle des Assurances (1989).17 In Britain, too, the 1970s have been a period of significant innovation, especially in the area of social regulation. The Independent Broadcasting Authority (1972), the Civil Aviation Authority (1972), the Health and Safety Commission (1974), the Equal Opportunities Commission (1976), and the Commission for Racial Equality (1976) are only some of the regulatory bodies created in this period.18 Despite the hostility of Conservative governments toward any kind of 'quangos', regulatory agencies were set up in the 1980s and early 1990s, partly because it was realised that in many cases privatization would only mean the replacement of public by private monopolies unless the newly privatised companies were subjected to public regulation of profits, prices, and entry and service conditions. Hence the rise of a new breed of regulatory agencies, the regulatory offices: Office of Telecommunications (1984), Office of Gas Supply (1986), Office of Water Services (1989), Office of Electricity Regulation (1990). Parallel, if slower, institutional developments are taking place in all other European countries, and the reasons given for the rise of the independent agencies are strikingly similar from country to country. These functional explanations are also strongly reminiscent of the arguments of earlier American writers. Thus it is said that agencies are justified by the need of expertise in highly complex or technical matters, combined with a rule-making or adjudicative function that is inappropriate for a government department or a court; that an agency structure may favour public participation, while the opportunity for consultations by means of public hearings is often denied to government departments because of the conventions under which they operate; that agencies' separateness from government is useful whenever it is hoped to free government administration from partisan politics and party political influence. Agencies are also said to provide greater continuity and stability than cabinets because they are one step removed from election returns; and the exercise of a policy-making function by an administrative agency should provide flexibility not only in policy formulation but also in the application of policy to particular circumstances. Finally, it is argued that independent agencies can protect citizens from bureaucratic arrogance and reticence, and are able to focus public attention on controversial issues, thus enriching public debate.19 200 G. Majone Rise of the Regulatory State in Europe 201 The growth of administrative regulation in Europe owes much to these newly articulated perceptions of a mismatch between existing institutional capacities and the growing complexity of policy problems: policing financial markets in an increasingly interdependent world economy; controlling the risks of new products and new technologies; protecting the health and economic interests of consumers without impeding the free flow of goods services and people across national boundaries; reducing environmental pollution. It is sufficient to mention problems such as these to realise how significant is the supranational dimension of the new economic and social regulation. Hence the important role of the European Community (now Union) in complementing the regulatory capacities of the member states. The European Community as Regulator Apart from competition rules and measures necessary to the integration of national markets, few regulatory policies or programmes are explicitly mentioned in the Treaty of Rome. Transport and energy policies which could have given rise to significant regulatory activities, have remained until lately largely undeveloped. On the other hand, agriculture, fisheries, regional development, social programmes and aid to developing countries, which together account for more than 80 per cent of the Community budget (see Table 5.2), are mostly distributive or redistributive rather than regulatory in nature. This budget of almost ECU 47 billion represents less than 1.3 per cent of the gross domestic product (GDP) of the Community and less than 4 per cent of the central government spending of member states. Given such limited resources, how can one explain the continuous growth of Community regulation, even in the absence of explicit legal mandates? Take the case of environmental protection, an area not even mentioned by the Treaty of Rome. In the two decades from 1967 to 1987, when the Single European Act finally recognised the authority of the Community to legislate in this area, almost 200 directives, regulations, and decisions were introduced by the Commission. Moreover, the rate of growth of environmental regulation appears to have been largely unaffected by the political vicissitudes, budgetary crises, and recurrent waves of Euro-pessimism of the 1970s and early 1980s. From the single directive on preventing risks by testing of 1969 (L68/19.3.69) we pass to 10 directives/ decisions in 1975,13 in 1980, 20 in 1982, 23 in 1984, 24 in 1985 and 17 in the six months immediately preceding passage of the Single European Act. Table 5.2 The EC Budget by Major Categories of Expenditure (1990, ECU) -Total Administration-Operations Agricultural market guarantee 26 Guidance (agricultural structures) 2 Fisheries Regional development and transport 5 Operations in the social sector 3 Energy, technology, research, nuclear safeguards, information market and innovation Repayments and aid to member states Co-operation with developing countries Other expenditures Operations—Total Commission—Total Other institutions Grand Total 1,529,765,860 .452,000,000 ,073,475,000 376,100,000 209,700,000 672,885,000 763,478,000 335,091,812 503,590,000 000,000,000 386,319,812 916,085,672 847,661,982 763,747,654 Source: The Community Budget: The Facts in Figures—1990 edn., Luxembourg: Office for Official Publications of the European Communities, p. 62. The case of environmental regulation is particularly striking, partly because of the political salience of environmental issues, but it is by no means unique. The volume and depth of Community regulation in the areas of consumer product safety, medical drug testing, banking and financial services and, of course, competition law is hardly less impressive. In fact, the hundreds of regulatory measures proposed by the Commission's White Paper on the completion of the internal market20 represent only the acceleration of a trend set in motion decades ago. The continuous growth of supranational regulation is not easily explained by traditional theories of Community policy making. At most, such theories suggest that the serious implementation gap that exists in the European Community may make it easier for the member states, and their representatives in the Council, to accept Commission proposals which they have not serious intention of applying. The main limitation of this argument is that it fails to differentiate between areas where policy development has been slow and uncertain (for example, transport, energy or research) and areas such as environmental protection where significant policy development has taken place even in the absence of a clear legal basis. Moreover, existing theories of Union policy making do not usually draw any clear distinction between regulatory and other types of policies. Now, an important characteristic of regulatory policy making is the limited influence of budgetary limitations on the activities of regulators. The 202 G. Majone size of non-regulatory, direct-expenditure programmes is constrained by budgetary appropriations and, ultimately, by the size of government tax revenues. In contrast, the real costs of most regulatory programmes are borne directly by the firms and individuals who have to comply with them Compared with these costs, the resources needed to produce the regulations are trivial. It is difficult to overstate the significance of this structural difference between regulatory policies and policies involving the direct expenditure of public funds. The distinction is particularly important for the analysis of Community policy making, since not only the economic, but also the political and administrative costs of enforcing EC regulations are borne by the member states.21 As already noted, the financial resources of the Community go, for the most part, to the Common Agricultural Policy and to a handful of distributive programmes. The remaining resources are insufficient to support large-scale initiatives in areas such as industrial policy, energy, research, or technological innovation. Given this constraint, the only way for the Commission to increase its role was to expand the scope of its regulatory activities. Another important element in an explanation of the growth of Community regulation is the interest of multinational, export-oriented industries in avoiding inconsistent and progressively more stringent regulations in various EC and non-EC countries. Community regulation can eliminate or at least reduce this risk. A similar phenomenon has been observed in the United States, where certain industries, faced with the danger of a significant loss of markets through state and local legislation, have strongly supported federal regulation ('preemptive federalism'). For example, the American car industry, which during the early 1960s had successfully opposed federal emission standards for motor vehicles, abruptly reversed its position in mid-1965: provided that the federal standards would be set by a regulatory agency, and provided that they would preempt any state standards more stringent than California's, the industry would support federal legislation. Analogous reasons explain the preference for Community solutions of some powerful and well-organised European industries. Consider, for example, the 'Sixth Amendment' of Directive 67/548 on the classification, packaging, and labelling of dangerous substances. This amending Directive does not prevent member states from including more substances within the scope of national regulations than are required by the Directive itself. In fact, the British Health and Safety Commission proposed to go further than the Directive by bringing intermediate products within the Rise of the Regulatory State in Europe 203 scope of national regulation. This, however, was opposed by the chemical industry, represented by the Chemical Industries Association (CIA) which argued that national regulation should not impose greater burdens on British industry than the Directive placed on its competitors. The CIA view eventually prevailed. Similarly, German negotiators pressed for a European-wide scheme that would also provide the framework for an acceptable regulatory programme at home, wanted a full and explicit statement of their obligations to be defined at the EC level. Moreover, with more than 50 per cent of Germany's chemical trade going to other EC countries, German businessmen and government officials wished to avoid the commercial obstacles that would arise from divergent national regulations.22 The European chemical industry had another reason for supporting Community regulation. In 1976 the United States, without consulting their commercial partners, enacted the Toxic Substances Control Act (TSCA). The new regulation represented a serious threat for European exports to the lucrative American market. A European response to TSCA was clearly needed, and the Community was the logical forum for fashioning such a response. An EC-wide system of testing new chemical substances could serve as a model for negotiating standardised requirements covering the major chemical markets. In fact, the 1979 Directive has enabled the Community to speak with one voice in discussions with the United States and other Organization For Economic Co-operation and Development (OECD) countries, and has strengthened the position of the European chemical industry in ensuring that the new American regulation does not create obstacles to its exports. There is little doubt that the ability of the Commission to enter into discussions with the USA has been greatly enhanced by the Directive, and it is unlikely that each European country on its own could do so effectively.23 Explaining Regulatory Policy Making in the EC In the preceding section I have considered three variables that are clearly important for explaining the growth of EC regulation: the tightness and rigidity of the Community budget; the desire of the Commission to increase its influence by expanding its competencies; and the preference of multinational firms for dealing with a uniform set of rules rather than with 12 different national regulations. However, these variables are not sufficient to explain the willingness of the member states to surrender important regulatory powers to supranational institutions, nor the ability 204 G. Majone of the Commission to introduce significant innovations with respect to the policies of the member states.24 As already suggested, available theories of policy making in the Community do not explain why the member states would be willing to delegate regulatory powers beyond the level required by an integrated market; nor can they explain the policy entrepreneurship of the EC Commission. This is because such theories stress the dominant role of the member states in all stages of the policy process, from initiation (which comes from the heads of state or governments in the European Council) to formal adoption (the prerogative of the Council of Ministers), to implementation (in the hands of the national administrations). A model capable of explaining the above mentioned phenomena must come to grips with two issues that have been overlooked by the traditional theories: first, problems of 'regulatory failure' in an international context, which limit the usefulness of purely inter-governmental solutions; and, second, the fact that regulation, as a very specialised type of policy making, requires a high level of technical and administrative discretion. To start with the first issue, market failures with international impacts, such as transboundary pollution, could be managed in a co-operative fashion without the necessity of delegating powers to a supranational level, provided that national regulators were willing and able to take into account the international repercussions of their choices; that they had sufficient knowledge of one another's intentions; and that the costs of organising and monitoring policy co-ordination were not too high. These conditions are seldom, if ever, satisfied in practice. Experience shows that it is quite difficult to verify whether or not inter-governmental agreements are being properly kept. Because regulators lack information that only regulated firms have, and because governments are reluctant, for political reasons, to impose excessive costs on industry, bargaining is an essential feature of the process of regulatory enforcement. Regardless of what the law says, the process of regulation is not simply one where the regulators command and the regulated obey. A 'market' is created in which bureaucrats and those subject to regulation bargain over the precise obligations of the latter.25 Because bargaining is so pervasive, it may be impossible for an outside observer to determine whether or not an international regulation has been, in fact, violated. When it is difficult to observe whether governments are making an honest effort to enforce a co-operative agreement, the agreement is not credible. For example, where pollution has international effects and fines impose significant competitive disadvantages on firms that compete inter- T Rise of the Regulatory State in Europe 205 nationally, firms are likely to believe that national regulators will be unwilling to prosecute them as rigorously if they determine the level of enforcement unilaterally rather than under supranational supervision. Hence the transfer of regulatory powers to a supranational authority like the European Commission, by making more stringent regulation credible, may improve the behaviour of regulated firms. Also, because the Commission is involved in the regulation of numerous firms throughout the Community, it has much more to gain by being tough in any individual case than a national regulator: weak enforcement would destroy its credibility in the eyes of more firms. Thus it may be more willing to enforce sanctions than a member state would be.26 In fact, the Commission has consistently taken a stricter pro-competition stance than national authorities such as the British Monopolies and Mergers Commission, the German Bundeskartellamt, or the French Conseil de la Concurrence. In short, the low credibility of inter-governmental agreements explains the willingness of member states to delegate regulatory powers to a supranational authority. At the same time, however, governments attempt to limit the discretion of the Commission by making it dependent on the information and knowledge provided by national bureaucrats and experts. We must now explain how the Commission often manages to overcome these limitations. The offices of the Commission responsible for a particular policy area form the central node in a vast 'issue network' that includes not only experts from the national administrations, but independent experts (also from non-EU countries), academics, public-interest advocates like environmentalists and leaders of consumer movements, representatives of economic and professional organisations and of regional bodies. Commission officials listen to everybody—both in advisory committees, which they normally chair, and in informal consultations—but are free to choose whose ideas and proposals to adopt. They operate less as technical experts alongside other technical experts, than as policy entrepreneurs, that is, as 'advocates who are willing to invest their resources—time, energy, reputation, money—to promote a position in return for anticipated future gain in the form of material, purposive, or solidary benefits'.27 In his study of policy innovations in America, Kingdon identifies three main characteristics of successful policy entrepreneurs: first, the person must have some claim to be taken seriously, either as an expert, as a leader of a powerful interest group, or as an authoritative decision maker; second, the person must be known for his political connections or negotiating skills; third, and probably most important, successful entrepreneurs 206 G. Majone Rise of the Regulatory State in Europe 207 are persistent.28 Because of the way they are recruited, the structure of their career incentives, and the crucial role of the Commission in policy initiation, Commission officials usually display the qualities of a successful policy entrepreneur to a degree unmatched by national civil servants. Actually the Commission officials' typical motivational structure is quite different from that of the average national government official. While the staff of the national governments is often recruited from persons who tend to be—compared with their peers who choose an industrial career—solid, correct, security-oriented, conservative, risk-averse and often somewhat narrow-minded, the Commission recruits its staff from people who are highly motivated, risk oriented, polyglot, cosmopolitan, open-minded and innovative... From the beginnings in the 1960s and up to the present, it has indeed been officials of a special type who chose to leave the relative security of their national administrations to go to Brussels to do there a well-paid but extremely challenging job... The structural conditions of recruitment and career favour a tendency to support new ideas and to pursue a strategy of innovative regulation which attempts to go beyond everything which can presently be found in the Member States. 29 Because of this tendency to favour innovative regulatory solutions, even national experts may find the Community a more receptive forum for their ideas than their own administration. A 1989 directive on the safety of machinery (89/392/EEC) offers a striking example of this phenomenon. The crucially important technical annex of the directive was drafted by a British labour inspector who originally sought to reform the British regulatory approach. Having failed to persuade the policy makers of his own country, he brought his innovative ideas about risk assessment to Brussels, where they were welcomed by Commission officials and eventually became European law.30 Moreover, what is known about the modus operandi of the advisory committees suggests that debates there follow substantive rather than national lines. A good deal of copinage technocratique develops between Commission officials and national experts interested in discovering pragmatic solutions rather than defending political positions. By the time a Commission proposal reaches the political level, first in COREPER (the committee of permanent representatives of the Member States) and then in the Council of Ministers, all the technical details have been worked out and modifications usually leave the essentials untouched. The Council may, of course, delay a decision or reject the proposal outright, but these options are becoming increasingly problematic under the qualified majority rule and the 'co-operation procedure' between the European Parlia- ment and the Council introduced by the Single European Act. Fitting together all the variables introduced in this and the preceding section— budget constraints, bureaucratic and economic interests, the poor credibility of purely intergovernmental arrangements and, last but not least, the highly technical nature of most regulatory policy-making—we begin to understand not only the origin and growth of Community regulation, but also its increasingly innovative character. The Problem of Political Accountability We just noted the significance of administrative and technical discretion for EC policy-making. Of course, regulatory discretion has important institutional and political implications also at the national level. One obvious consequence is the creation of specialised agencies such as the French autorites administrates independantes and the British regulatory offices mentioned above. Such agencies are independent of the central administration, and hence of civil service rules, and often combine legislative, judicial, and executive powers—rule making, adjudication, and enforcement, in the terminology of American administrative law—in more or less narrowly defined areas of policy making. As already suggested, such institutional arrangements represent an important departure from European constitutional and administrative traditions. The implicit, and in some cases explicit, model is the American independent regulatory commission (IRC). The IRCs were created by Congress to ensure agency independence from presidential control and short-term political considerations. Although they cover an extremely wide range of administrative activities—from the control of prices, routes and service conditions of railway companies by the Interstate Commerce Commission created in 1887 to the licensing of nuclear power plants by the Nuclear Regulatory Commission created in 1975—all IRCs share some organisational characteristics that are meant to protect their decisional autonomy: they are multi-headed, having five or seven members; they are bipartisan; members are appointed by the president with the consent of the Senate and serve for fixed, staggered terms. Unlike the single-headed line agencies, the IRCs operate outside the presidential hierarchy in making their policy decisions. As the US Supreme Court asserted in Humphrey's Executor vs. United States (1935) commissioners can be removed from office only for official misbehaviour, not for disagreement with presidential policy. In the course of their century-old history, IRCs have been often criticised I 208 G. Majone accountability, and for an alleged tendency to be captured by private interests. Not surprisingly, the same criticisms are heard now in Europe. Here, regulatory agencies are still seen as 'constitutional anomalies which do not fit well into the framework of controls, checks and balances', 1 even as challenges to basic principles of democracy and of the Rechtsstaat.32 To be sure, it is no easy task to fit the new institutions into the constitutional framework of countries where the diffraction of state power is seen as a direct challenge to parliamentary sovereignty and to the principle of a rigid separation of powers. Expressed in traditional terms the dilemma is: either the regulatory agencies are part of the state administration, and then they cannot be independent; or else they are independent, but in this case to whom are they accountable? It is impossible to escape this dilemma without questioning the relevance of traditional notions such as the constitutional axiom of the tripartite separation of powers, or the political principle that governmental policy making ought to be subject to control only by persons accountable to the electorate. It is certainly not a coincidence that similar issues are being raised in the ongoing debate about the proper scope of judicial review and judicial policy making. The rise of judicial review in Europe shows that the triad of government powers is no longer considered an inviolable principle. At the same time, courts find their policy making role enlarged by the public perception of them as guarantors of the substantive ideals of democracy when electoral accountability in all spheres of government seems to be waning.33 What connects the discourse about administrative regulation with that about judicial review and policy making is the issue of the role of non-majoritarian institutions in democratic societies. Again, it is no mere coincidence that the same country has developed both the most advanced system of judicial review and the most extensive network of regulatory institutions. The American experience shows that a highly complex and specialised activity like regulation can be monitored and kept politically accountable only by a combination of control instruments: legislative and executive oversight, strict procedural requirements, public participation and, most importantly, substantive judicial review. Measured against these standards, regulation in Europe is seen to be highly discretionary, suffering from weak accountability to Parliament, weak judicial review, absence of procedural safeguards, and insufficient public participation.34 The issue of the political accountability of regulators, who are neither elected nor directly responsible to elected officials, is particularly visible at the EU level precisely because of the central importance of regulatory Rise of the Regulatory State in Europe 209 policy-making in the Community system. However, the remedies should not compromise the effectiveness of the supranational institutions. The comparative advantage of EU regulation lies mainly in the relative insulation of Community regulators from the short-run political considerations and pressures which tend to dominate national policy-making. As was noted above, the Commission has more to gain by being tough in any individual case than a national regulator. This is because the Commission is involved in the regulation of firms throughout the Community, so that weak enforcement would destroy its credibility in the eyes of more firms. For the same reason, the Commission is less likely to be captured by a particular firm or industry than a national regulator. In the language of James Madison, the insulation of the Commission from day-to-day politics is an important safeguard against national and sectoral 'factionalism'. In fact, as I have shown elsewhere,35 many of the arguments developed along Madisonian lines by the American advocates of an 'independent fourth branch of government'—the regulatory branch—apply, mutatis mutandis, in the context of the European Union and its member states. These writers acknowledge that government by judges and technocratic experts raises serious issues for democratic theory, but point out that government by elected officials, too, suffers from defects. For example, in seeking re-election legislators engage in advertising and position taking rather than in serious policy making, or they design laws with numerous opportunities to aid particular constituencies. Thus, re-election pressures have negative consequences for the quality of legislation. On the other hand, pro-regulation scholars ask, if the courts require the regulatory process to be open to public inputs and scrutiny and to act on the basis of competent analyses, are the regulators necessarily less accountable than elected politicians?36 The procedural remedies suggested here are also relevant to the problem of the 'democratic deficit' of the EU. For example, it is well known that the Treaty of Rome does not structure the executive power of the Community in a single way, applicable to all instances of legislation needing further execution. Instead, it has been left to the Council, in its capacity as legislative decision maker, to organise, case by case, the executive process.37 This ad hoc approach is the very negation of the idea of transparency which plays such a large role in the current discussion of regulation. The adoption of something like an Administrative Procedures Act for the European Union could do more to make public accountability possible than the wholesale transfer of traditional party politics to Brussels. Any progress along such lines at Union level would have positive spillovers 210 G. Majone Rise of the Regulatory State in Europe 211 for the member states where, as we saw, the accountability of regulators is still an open issue. The fact that regulation is relatively more important at EU than at national level, makes the Union an ideal laboratory for the study of the problems of the regulatory state. This can be seen also by examining the issue of co-ordination and executive control—the second key issue of regulatory policy-making. Co-ordination and Control An important characteristic of regulatory policy making is the absence of a regulatory budget procedure. Because, as was shown above, the size of regulatory programmes is not significantly constrained by legislative appropriations and by the level of tax revenues, as in the case of non-regulatory programmes, no mechanism exists for regulation that requires policy makers throughout the government to solve the two-level budget problem—how much to spend during a given period and then how to allocate this total amount among alternative uses—which is addressed by any government in its direct expenditure activities. The result is both economic inefficiency and inadequate political oversight. These defects of the regulatory process are, again, particularly evident in the case of EU rule-making. Thus budgetary discipline is even weaker than at the national level since the burden of implementing Union regulations is carried by the governments of the member states. Also, because of the absence of a central political authority, regulatory issues are dealt with sector by sector, with little attempt to achieve overall policy coherence. Even within the same sector it would be difficult to maintain that regulatory priorities are set in a way that explicitly takes into consideration either the urgency of the problem or the benefits and costs of different proposals. The piecemeal procedure of the Commission in proposing new regulatory measures has resulted in directives in areas where harmonisation is a low priority, while neglecting other areas which need a considerable amount of harmonisation. Since the lack of budgetary discipline is a basic reason for the structural defects of the regulatory process, one should attempt to create control mechanisms similar to those traditionally used in the case of direct public expenditures. This simple idea has led several analysts of the American regulatory process to propose the introduction of a regulatory budget. In its basic outline the regulatory budget would be established by Congress and the President for each agency, perhaps by starting with a budget constraint on total private expenditures mandated by regulation, and then allocating the budget among the different agencies. By setting a budget constraint on mandated private expenditures, the regulatory budget would clarify the real costs to the economy of adopting a regulation and encourage cost effectiveness. The knowledge that agencies would be competing against each other would lead them to propose their 'best' regulations in order to win presidential and congressional approval. Simultaneous consideration of all new regulations would permit an assessment of their joint impact on particular industries and the economy as a whole.38 Serious technical difficulties (e.g., estimating the full social cost of regulations, especially when the regulations restrict outputs or behaviour rather than merely requiring outlays for compliance) have to be resolved before the regulatory budget could actually be implemented. Nevertheless, because the budget is such a useful analogy for highlighting the defects in the current regulatory process, a promising approach consists in developing methods of regulatory oversight and control that incorporate budget concepts in a workable fashion. One possible model suggested by the analogy with the budgetary process, deserves special investigation: a regulatory clearing-house. In the EU context, such a clearing house should be located at a sufficiently high level in the Union bureaucracy, possibly in the office of the President of the Commission. Directorates-General would be asked to submit annually draft regulatory programmes to the clearing house for review. When disagreements or serious inconsistencies arise, the President or a 'working committee on regulation' would be asked to intervene. By extending centralised control over the regulatory agenda of the Directorates-General, this review process would help the Commission shape a consistent set of regulatory measures to submit to the Council and the Parliament. The usefulness of the procedure as a tool of managerial control could be increased by co-ordinating the regulatory review with the normal budgetary review, thus linking the level of budgetary appropriations to the cost-effectiveness of the various regulatory programmes. One key function of such a clearing-house system, in addition to providing for greater coherence, would be to flesh out the concept of subsidiarity: only through systematic review of all proposals put forward by the various Directorates-General will the Commission be able to determine when action by the Union is necessary. Obviously the idea of a regulatory clearing-house system would be useful also for the member states, as one way of reducing the negative consequences of the diffraction of state power which is one of the significant characteristics of the regulatory state. 212 G. Majone Rise of the Regulatory State in Europe 213 Conclusion: The Paradoxes of Privatization, Deregulation and Re-regulation At the beginning of this analysis we noted that European scholars and policy makers began to recognise regulation as a distinct mode of policy making only after deregulation became a popular theme of political discourse. This is only one of several paradoxes, real or presumed, that seem to characterise the development of regulation in Europe during the past two or three decades. Thus, the privatization and/or deregulation of potentially competitive industries have not meant the end of all regulation; on the contrary, they have created the conditions for the rise of a regulatory state to replace the dirigiste state of the past. Where competitive conditions did not yet exist, as in the case of telecommunications, only public regulation could ensure that privatization did not simply mean the replacement of public monopolies by private ones. Often, deregulation is only a first step towards re-regulation, that is, regulation by other means—economic incentives instead of administrative rules, statutory instead of self-regulation—or at different levels of government—for example, at Community rather than national level. This paradoxical combination of deregulation and re-regulation is what is usually meant by regulatory reform. On the other hand, the experience of countries such as Britain shows that old habits of secretiveness and ministerial interference, characteristic of the management of nationalised industries, continue to persist even after privatization. Serious flaws in the design of institutions to regulate the newly privatized industries can be detected in the choice of a non-participatory model, with none of the public hearings and other procedural characteristics of US regulation; in the creation of a system of agencies linked to particular industries, rather than the pattern of commissions regulating a range of utilities in order to reduce the risk of agency capture; and in the fact that government departments still preserve important regulatory powers, so that the operations of agencies are often dependent on prior decisions of the minister laying down the principles to be applied. The danger, Tony Prosser concludes, is that these powers of direction 'could be abused to exert behind-the-scenes pressure on the regulation in much the same way as pressure was put on the nationalised industries by government, precisely the situation which the privatization programme is supposed to render impossible'. Also the stupendous growth of EU regulation has a certain paradoxical quality. By imposing a tight and rigid budget, the member states no doubt wished to restrict as much as possible the competencies and decisional autonomy of the Commission. Accustomed to think of state power primarily in terms of the power of taxing and spending, national leaders did not apparently realise that regulatory activities cannot be controlled by means of the traditional budget constraints: only a 'regulatory budget' could introduce the necessary discipline. In the absence of a regulatory budget procedure, the rule-making power of the Union has proved well-nigh irresistible. Moreover, the growth has been qualitative as well as quantitative. As noted above, in some areas of economic and social regulation EU directives go beyond the levels achieved by the legislation of the most advanced member states. This is another paradox, at least for theories claiming that member states control all stages of Union policy making. Finally, the terms of the debate about the 'democratic deficit' of the Union are often paradoxical when not simply hypocritical. Problems of political accountability can be perceived most clearly at Union level precisely because regulation is at the core of EU policy making. Yet the frequent criticisms that Union institutions lack direct democratic legitimacy also apply to many national institutions, including courts and independent regulatory agencies. The problem of a 'democratic deficit' concerns all regulatory states, not just the Union. The problem has no simple solution, but it can be mitigated by a variety of substantive and procedural means ranging from judicial review to the 'regulatory budget'. The shift to regulation at the national and supranational level is an attempt to improve the procedural and substantive rationality of public policy in a dramatically changing world. However, the changing role of the state raises new conceptual and practical issues that are still poorly understood, let alone resolved. Notes 1. Cento Veljanovski, 'The Regulation Game', in Veljanovski (ed.), Regulators and the Market (London: Inst, of Economic Affairs, 1991), pp. 3-28. 2. Kenneth Dyson, 'Theories of Regulation and the Case of Germany: A Model of Regulatory Change', in Dyson (ed.), The Politics of German Regulation (Brook-field, VT: Dartmouth, 1992), pp. 1-28. 3. Jacques Pelkmans, 'Regulation and the Single Market: An Economic Perspective', in H. Siebert (ed.), The Completion of the Internal Market (Tübingen: J.C.B. Mohr, 1989), pp. 91-117. 4. Jon Elster, Explaining Technical Change (Cambridge: CUP, 1983). 5. Robert Baldwin and Christopher McCrudden, Regulation and Public Law (London: Weidenfeld, 1987). 214 G. Majone Rise of the Regulatory State in Europe 215 6. Giandomenico Majone and Antonio La Spina, 'Lo Stato Regolatore', Rivista Trimestrale di Scienza dell' Amministrazione 13/3 (1991), pp. 3-62. 7. Cento Veljanovski, Selling the State (London: Weidenfeld, 1987). 8. Ibid., pp. 165-73. 9. Giandomenico Majone, 'Introduction' in G. Majone (ed.), Deregulation or Re-Regulation? (London: Pinter Publishers, 1990), pp. 1-6. 10. Philip Selznick, 'Focusing Organizational Research on Regulation", in R. G. Noll (ed.), Regulatory Policy and the Social Sciences (Berkeley and Los Angeles: Univ. of California Press, 1985), pp. 363-7. 11. Peter Jenkins, Mrs. Thatcher's Revolution (Cambridge, MA: Harvard UP, 1988). 12. George J. Stigler, 'The Theory of Economic Regulation", Bell Journal of Economics and Management Science, 6/2 (1971), pp. 114-41. 13. Giandomenico Majone, Evidence, Argument and Persuasion in the Policy Process (New Haven, CT: Yale UP, 1989). 14. Paul J. Joskow and Robert G. Noll: 'Regulation in Theory and Practice: An Overview', in G. Fromm (ed.), Studies in Public Regulation (Cambridge, MA: MIT Press, 1981), pp. 1-65. 15. Sam Peltzman, 'The Economic Theory of Regulation after a Decade of Deregulation", in Brookings Papers in Economic Activity (Washington, DC: The Brookings Instn., 1989), p. 17. 16. The Economist, 27 Feb. 1993, pp. 89-90. 17. Marie-Jose Guedon, Les Autorites Administratives Independantes (Paris: Librairie Generale de Droit et de Jurisprudence, 1991). 18. Baldwin and McCrudden (note 5). 19. Catherine Teitgen-Colly, 'Les Autorites Administratives Independantes: His-toire d'une Institution", in Claude-Albert Colliard and Gerard Timsit (eds.), Les Autorites Administratives Independantes (Paris: Presses Universitäres de France, 1988); Giulio Vesperini, 'Le Funzioni delle Autoritä Amministrative Indipend-entf, Diritto della Banca e del Mercato Finanziario, 4/4 (1990), pp. 415-19; Baldwin and McCrudden (note 5), pp. 4-9; Guedon (note 17), pp. 16-27. 20. Commission of the European Communities, Completing the Internal Market, COM(85)310, final, (Luxembourg: Office for Official Publications of the European Communities, 1985). 21. Giandomenico Majone, 'Market Integration and Regulation: Europe after 1992', Metroeconomica 43/1-2 (1992), pp. 131-56, and, 'Regulatory Federalism in the European Community", Government and Policy, Vol. 10 (1992), pp. 299-316. 22. Ronald Brickman, Sheila Jasanoff, Thomas Ilgen, Controlling Chemicals (Ithaca, NY: Cornell UP, 1985). 23. Ibid., p. 277. 24. Giandomenico Majone, 'The European Community between Social Policy and Social Regulation', Journal of Common Market Studies 31/2 (1993), pp. 153-70. 25. Alan Peacock (ed.), The Regulation Game (Oxford: Basil Blackwell, 1984). 26. Kristos Gatsios and Paul Seabright, 'Regulation in the European Community', Oxford Review of Economic Policy, 5/2 (1989), pp. 37-60. 27. John W. Kingdon, Agendas, Alternatives and Public Policies (Boston: Little, Brown, 1984), p. 188. 28. Ibid., pp. 189-90. 29. Volker Eichener, Social Dumping or Innovative Regulation!' Working Paper SPS No. 92/28 (Florence: Eur. Univ. Inst., 1992). 30. Ibid., p. 52. 31. Veljanovski, Regulatora and the Market (note 1), p. 16. 32. Teitgen-Colly (note 19), p. 49. 33. Mary L. Volcansek, 'Judges, Courts and Policy-Making in Western Europe", West European Politics 15/3 (July 1992), p. 5. 34. Tony Prosser, 'Regulation of Privatized Enterprises: Institutions and Procedures', in Leigh Hancher and Michael Moran (eds.), Capitalism, Culture and Economic Regulation (Oxford: Clarendon Press, 1989), pp. 135-65; Baldwin and McCrudden (note 5); Veljanovski (note 1). 35. Giandomenico Majone, 'Controlling Regulatory Bureaucracies: Lessons from the American Experience' in Hans Ulrich Derlein, Ita Gerhardt and Fritz W Scharpf (eds.), Systemnationaltta't und Partiellintereste (Baden-Baden: Nomos, 1994), pp. 291-314. 36. Susan Rose-Ackerman, Rethinking the Progressive Agenda (NY: The Free Press, 1992). 37. Koen Lenaerts, 'Some Reflections on the Separation of Powers in the European Community", Common Market Law Review, Vol. 11 (1991). 38. Robert E. Litan and William D. Nordhaus, Reforming Federal Regulation (New Haven, CT: Yale UP, 1983). 39. Prosser (note 34), p. 147.