Self-regulation is a term which is increasingly important in both political and academic debate, yet is one which gives rise to serious problems when subjected to any analysis (for particularly useful recent accounts see Black 1996 and 2001; Moran 2003, ch. 4). It could even be said that, as I shall argue below, there is no such thing in the real world as self-regulation; nor is there any such thing as its perceived opposite, command, and control regulation. These two concepts are used in two ways, both of which raise serious concerns. The first (potentially the more useful) is as two heuristic devices; ways of contrasting different regulatory regimes and identifying different elements in a cocktail of mixed regulatory techniques. The second is as political slogans; self-regulation is good because it benefits business (or bad because it does not benefit anyone else); command and control is bad because it imposes the dead hand of the state on the vibrancy of business and markets (the point about the political role of the concept of self-regulation has been well made by Moran 2003, p.67). These are, however, slogans, not sociological description of any identifiable system. Indeed, the importance of politics is central in discussions of self-regulation; on the one hand the latter appeals to advocates of lifting regulatory burdens, but on the other, moves to self-regulation may be resisted by regulators reluctant to lose powers of control or oversight. In the first part of this article I shall deal with the concept of self-regulation and suggest that to speak instead of co-regulation avoids many of the difficulties; in the second I shall outline the use of both these concepts in the development of the new Audio–Visual Media Services Directive (Directive 2007/65).

Self-regulation, Co-regulation and Institutional Design

The Limits of Self-regulation and Command and Control

Why do the concepts of self-regulation and command and control regulation not correspond to actual practice? Firstly, self-regulation at a national level is often presented as the antithesis of law, yet in fact it is rare indeed for a regulatory system involving major conflicts between values to be unaccompanied by direct forms of legal rules and sanctions. This is what distinguishes self-regulation in areas such as broadcasting content from “technical” self-regulation, for example setting industry standards which may be disputed but do not raise major moral and social concerns. Thus the press in the UK is often claimed to be an example of an extreme form of self-regulation, yet the UK press is subject to laws of defamation which (even after recent liberalisation) are still amongst the most restrictive in the Western World, and we even retain criminal liability for blasphemy. The Internet is claimed as a classic area for self-regulation; yet we have some of the strongest, and most vigorously enforced, criminal legislation of all concerned with child pornography on it. Moreover, there is a strong tendency for supposed self-regulation to evolve close to direct state regulation when there is a crisis; financial services is a good example. Here highly informal “club regulation” progressively turned into a form of self-regulation by so-called “self-regulatory organisations” subject to detailed oversight by the Securities and Investment Board, and then into direct public regulation by the Financial Services Authority. Much of this can be understood as a combination of political responses to successive crises in confidence as scandals affecting the public broke in this sector and of the economic needs of more rational, open and globalised financial markets (Moran 2003, ch. 4). In all these examples, self-regulation is much better seen not as a pervading regulatory approach, but as part of a shifting set of regulatory techniques, the mix depending on external political, economic, and social factors.

As regards command and control, this characterisation of regulation assumes that a public authority gaining its legitimacy from the political process issues orders to companies or individuals requiring them to meet public policy goals; the implication is that these orders are obeyed. This view has been subjected to a barrage of criticism in recent years, especially from researchers in sociology and in law (the literature is huge, but for a particularly important example see Hawkins 2002). Much sociological work has shown the difficulty for regulators in gaining knowledge and understanding of the systems they are regulating; law is a blunt tool for intervening in other complex systems and cultures, including the economy and indeed the media. Indeed, this insight has formed the basis for the development of the important theory of autopoiesis and its application to regulatory systems (e.g., Teubner 1987; Black 1996). The conceptual problem may be resolved by a blurring in practice between regulatory law and the systems it is regulating. Thus even where regulation ostensibly takes the form of command and control, the reality has been shown to involve extensive negotiation between regulators and those they regulate, either through consultation when rules are made or even more importantly through selective enforcement in which rules are enforced not as binding orders but as the basis for negotiation to achieve reasonable results (Hawkins 2002). Moreover, sociological work has shown that more effective compliance can often be achieved not by issuing orders but by creating structures in which economic incentives are created for compliance, or cultures are encouraged in which public policy goals are internalised (Ayres and Braithwaite 1992; Gunningham and Grabosky 1998).

Co-regulation

As a result it is clear that there is a continuum between different regulatory regimes with varying degrees of public and private input; what is apparent in practice is a cocktail of different techniques dependent on context. This is perhaps why we are hearing much more of the concept of co-regulation which suggests a more sophisticated and balanced approach than the two alternatives I have suggested above. The definitions of this concept also vary enormously.

One example is that of the UK Government in its White Paper that formed the basis for the reform of communications regulation through the Communications Act 2003 (Department for Trade and Industry and Department for Culture, Media and Sport 2000, para. 8.11.1). This distinguished co-regulation from self-regulation on the basis that the former had an active involvement by the regulator to ensure that an acceptable and effective solution was achieved, for example through setting objectives or providing support for sanctions whilst leaving space for self-regulatory initiatives by industry. A recent study undertaken for the European Commission identified criteria for co-regulation (Hans-Bredow Institute/EMR 2006, para. 2.4). They are that a system is established to serve public policy goals (excluding mere industry self-promotion, for example), that there is some sort of connection between the non-state regulatory system and the state (though not necessarily a statutory one; contract will suffice), that some discretionary power is left to the non-state system, but that the state uses regulatory resources to guarantee the fulfilment of the regulatory goals. The design of the system, and the balance of responsibility between private actors and the state, will vary in different co-regulatory systems, as will the allocation of standard-setting and enforcement within the system, but there will always be some sort of public–private mix.

It should be added that the conditions for establishing an acceptable co-regulatory system may be demanding. One illustrative example is that of the criteria drawn up by the UK communications regulator, the Office of Communications (Ofcom) for promoting successful co- and self-regulation (co-regulation in my terms) (Ofcom 2004b). These include requirements, for example, of accessibility of co-regulatory systems to members of the public, near universal participation in them by the industry, independence from interference by interested parties, auditing, and review by Ofcom and appropriate appeals mechanisms. A further example from Ofcom is that of the delegation of the regulation of broadcast advertising to the Advertising Standards Authority (Ofcom 2004a). This is often portrayed as the useful adoption of a previously established self-regulatory regime on the part of the public regulatory authority. However, viewed as a whole, the system is hardly one of simple self-regulation for it involves the retention of backstop sanctioning powers for Ofcom, and this regulator also monitors the effectiveness of the system. Importantly, Ofcom also retains control of the regulation of advertising scheduling, and this has been crucial in the most important recent controversial issue of policy, that of junk food advertising to children. Here Ofcom was able to introduce what were seen by the industry as highly restrictive rules on a matter of substantial public concern. In my terms this system is clearly one of co-regulation, not of self-regulation.

The Role of Normative Principles

This discussion has one particularly important implication for practical institutional design; that definitions of self-regulation or co-regulation are of limited use, and indeed may be undesirably constraining. We shall see in the latter part of this article that this is precisely what was evident in the early stages in the drafting of the Audio-Visual Media Services Directive. Thus self-regulation and command and control regulation merely represent the extremes of a continuum of different regulatory models; extremes that are unlikely to be found in real life. Moreover, any regulatory regime is not characterised by a single model but will be a cocktail of various techniques with interventions by both private and public actors. Does this mean that there are no principles which can be used to categorise and propose different regulatory mixes? Of course, there are such principles, but narrow definitions are unlikely to be of assistance in finding them. What are much more important are the normative considerations we should bring to bear on designing regulatory regimes; answering the question of how the regimes can maintain their legitimacy. Assessing what these might be involves us in looking at the reasons why we might decide to prefer different mixes of regulatory models.

At base there are two different reasons for favouring a regime which is closer to the self-regulation extreme, which themselves draw on very different approaches to regulation (see Cafaggi and Muir Watt 2007). The first is an argument of principle; it favours a more self-regulatory approach because this is more consistent with private autonomy. Thus regulatory techniques which favour autonomous action by those regulated and interfere as little as possible with market operation are desirable, because their adoption limits state interference in an autonomous private sphere and so promotes freedom. A further, economic, reason for emphasizing this recognition of the importance of autonomy is that such an approach is more compatible with the open operation of markets, and is thus likely to encourage the attainment of economically efficient results through market operations. On this view the main actor in self-regulation is the industry, not the consumer, and self-regulation is essentially a “bottom-up” phenomenon with its base in the recognition of individual freedoms. This approach has, of course, a strong political appeal to those wishing to lift regulatory burdens.

The second reason for adopting a self-regulatory model is a very different one. This is that the use of such a regime does not represent in principle an acknowledgement of necessary limits to regulation but instead a way of making regulation “smarter” through increasing its flexibility and, importantly, enabling it to take on board the pooled knowledge of those regulated and others with an interest such as the citizen and consumer. It is thus a form of “top-down” delegation of its prerogatives by the state in order to achieve better governance. A corollary of this view is that such delegation is partial and revocable and remains subject to public oversight and to overriding public service and public interest requirements. On this view, self-regulation is essentially fragile and may be inadequate to cope with crises of trust, so requiring a continuing role for oversight by the regulatory state.

Both these arguments, that based on private autonomy and that on “smarter” regulation, favour a greater self-regulatory element in the regime, but each will have different implications as to the degree of public oversight to which it is made subject. They do have in common an assumption that self-regulation can achieve legitimacy through the participation of those subject to regulation in the actual regulatory process, either as a recognition of private autonomy or as a means of accessing knowledge necessary for successful regulation.

On the other hand, there may be compelling arguments against any self-regulatory elements, and in favour of regulation carried out more directly by public agencies. Again, these arguments may be drawn from a number of different sources. Thus one of these is competition law; although self-regulation is usually presented as a market-friendly technique, it can carry with it the danger of encouraging cartel-like agreements to close markets; the early history of financial markets shows this clearly with “club-like” restrictive practices which effectively closed markets to outsiders (Moran 1989; 2003 ch. 3). Partly as a result of business pressure, these had to be swept away and replaced by more rule-based regulation as part of the process of market liberalisation. Co-regulation may be a means of lessening this danger through permitting some public oversight of the process so that self-regulation does not degenerate into a cartel. This model is particularly important given that it acts as a counterweight to the assumption that self-regulation is inherently pro-market.

The second normative restriction arguing against a more self-regulatory model is that of how we can avoid the problems of legitimacy associated with self-regulation. Such a regime is often seen as replacing the public interest with the self-interest of participants and marginalizing both some sort of overall public interest and the interests of particular groups such as consumers and employees who play no part in the self-regulatory arrangements. The fear is not simply that self-regulation encourages cartels, but that it represents an abandonment of the legitimate interests of the state in acting as a guarantor of the public interest and the interests of under-represented groups; on this view, self-regulation is essentially undemocratic. By contrast, public regulation by the state can at least claim to be based on some sort of democratic mandate; government departments are at least subject to some form of democratic scrutiny.

Co-regulation and Legitimacy

Can co-regulation combine these different sources of legitimacy? Co-regulation might be seen as offering either the best or the worst of both worlds, either a system in which private and public interests are effectively reconciled, or one in which neither is respected and any values are subjected to unprincipled bargaining between the state and private interests. Which is achieved will depend strongly on the context and is essentially an empirical question. However, it is important to stress that, to assess legitimacy of a regime properly, we have to examine co-regulation from the viewpoint of the extent to which it respects process values as well as from that of its efficiency and effectiveness in meeting public policy goals. The most important will be firstly “input values”; values of participation both by those active in the industry being regulated and by outside groups such as consumer groups and those representing other interests (there is no shortage of these in the broadcasting area). Secondly, there will be “output” values of accountability that will be important here; to what extent are co-regulatory arrangements required to be transparent and to offer explanations of their decisions both to those directly affected and to others with an interest? There is also a third issue; to what extent do co-regulatory arrangements preserve their independence and avoid being subject to regulatory capture? This is often a very difficult issue for examination as capture may come not just from the regulated industry but from different parts of it or indeed from other sources; thus one can imagine an apparently co-regulatory regime being subject to capture by government where industry actors are particularly weak.

The extent of different forms of capture is essentially an empirical question and is difficult to predict from abstract institutional design; again transparency is the key and one will need to examine appointments procedures, decision-making procedures, duties to give reasons and provision for access to information on the operation of the regime to assess this fully. Indeed, these are indeed the very issues which Ofcom highlighted in its guidance on promoting effective co- and self-regulation mentioned above (Ofcom 2004b). Thus the guidance stresses the need for the independence of a co-regulatory system to be guaranteed by its having independent members comprising half to two-thirds of its governing body. There should be independent appeals mechanisms, and oversight through auditing and review of co-regulatory arrangements by Ofcom itself; Ofcom would also approve the co-regulatory body’s governance and funding arrangements and any codes or guidelines which it published. There should also be “effective arrangements for wide public consultation on significant issues (e.g., about substantive changes in codes or procedures)” (10). All this is far from the idea of self-regulation as giving unfettered power to industry bodies immune from public oversight.

Limits to Co- and Self-regulation

A related question is that of enforceability; are enforcement and sanctioning weaker at the more self-regulatory end of the continuum? The argument that they are is often combined with the assumption that self-regulation is associated with “soft law,” with no legal enforceability, whilst state regulation is associated with “hard law,” assumed to be directly legally enforceable. Of course, it would be a mistake to assume that state regulatory bodies operate only through “hard law,” or indeed that “hard law” is always unproblematically enforced. The sociological literature referred to earlier makes it clear one must not assume this, and that “soft law” sanctions may be sociologically much stronger. The effectiveness of enforcement is an empirical matter, not one to be determined a priori by the choice of regulatory regime.

It would also be wrong to suggest that a co-regulatory regime cannot make law enforceable by formal processes. One way in which this could occur is through, for example, a private group such as a professional organisation developing rules which can ultimately be enforced by public authorities; the example of the ultimate role of Ofcom in relation to advertising content rules developed by the Advertising Standards Authority is a case in point. Here the licence conditions under which the broadcasters operate require them to comply with directions made by the Authority; failure to do so could attract sanctions from the regulator including fines and, ultimately, withdrawal of the licence (Ofcom 2004a). Similarly, in its guidelines on promoting effective co- and self-regulation discussed above, Ofcom emphasizes the need for effective and credible sanctions on the part of a co-regulatory body (Ofcom 2004b). These will be graduated, with the most far-reaching sanctions, such as the removal of the ability of a company to function by withdrawing access to telecommunications networks, lying with Ofcom itself.

A striking example of linking a process normally described as self-regulatory into the formal legal process is that of section 12(4) of the Human Rights Act 1998, which requires a court, in deciding whether to grant relief which might affect freedom of expression, to pay particular regard to “any relevant privacy code.” This refers in particular to the code produced by the self-regulatory Press Complaints Commission and so represents a way of anchoring the Code into the court system; a publication which has breached the Code is more likely to find itself subject to legal constraints protecting the right to privacy.

This is not to deny that in particular contexts there may be strong arguments against using self- or co-regulation and soft law. Examples would be where there are strong economic incentives for firms to evade rules, or where markets are highly internationalised, and building cultures of compliance is difficult. On this basis it could be argued that the promotion of public service broadcasting, in an internationalising media environment, is not appropriate for a regime with large elements of self-regulation, despite the fact that this has characterised the approach of the UK Government and of Ofcom under the Communications Act 2003 (Prosser 2006). Similarly, a less self-regulatory approach will be preferable where it is intended to provide major rights for third parties who are unable to participate directly in industry processes; once more this is fully recognised in the Ofcom guidelines. Again, though, the protection of such rights is one element to be built into the cocktail of a regulatory regime, not an argument for completely rejecting any elements of co-regulation.

A related issue is the relationship between co-regulation and international obligations, including both those protecting rights and those implementing Community obligations (see Hans-Bredow Institute/EMR 2006, ch 5). Thus the first question is adequacy of self-or co-regulatory arrangements to provide an effective remedy under the European Convention on Human Rights. The European Court of Human Rights held in the Peck case that the UK Press Complaints Commission did not provide an effective remedy under Article 13 for an infringement of Article 8 of the Convention in a case of a serious breach of privacy because it could not award compensation nor prohibit publication.Footnote 1 This case involved facts before the Human Rights Act came into effect, and the Act does now provide the courts with the power to award damages for breach of a Convention right, but this does not affect the general question of the acceptability of self-regulatory regimes to protect basic rights. The case also, of course, raises the bigger issue, largely outside the scope of this article, about the applicability of constitutional guarantees to non-state bodies, an issue on which the UK courts have not had an impressive record so far.Footnote 2 However, once more, these limitations should not condemn the use of co-regulation as part of a cocktail of regulatory techniques. What has happened in the UK is that, alongside the “self-regulatory” regime for the press administered by the Press Complaints Commission, the courts have developed rights to privacy which can be enforced by legal means, including the award of damages for breach. This has its dangers, notably of untidiness and of a two-tier system permitting better remedies for those who can afford expensive litigation; however, it does demonstrate that self and co-regulation are not incompatible with a system which provides legal remedies for breaches of human rights. Similarly, it seems from the case-law of the European Court of Human Rights that co-regulatory measures where there is state delegation or oversight will not fail the requirement in Article 10 that measures restricting freedom of expression must be “prescribed by law” (Hans-Bredow Institute/EMR 2006 151; Barthold v Germany (1985) EHRR 383).

A related point is that of the extent to which a form of co- or self-regulation may be adequate as a means of implementing an obligation under European Community law. To summarize very briefly, in the implementation of a directive, the principle that Member States can choose the form and method of implementation (Art. 249(3) EC) would seem to permit the use of co-regulation, although national traditions may vary considerably both in the extent to which co-regulation is recognised and what it means. This suggests a considerable role for subsidiarity to respect national traditions. The freedom is limited in that if a directive makes specific provisions relating to its implementation, these must be followed; moreover, the transposition must always be in a clear and precise manner, and must be fully effective. Transposition through the use only of internal regulatory orders or administrative practices is inadequate, at least where the directive is intended to create rights for individuals. Other issues include compatibility with Art 49 of the EC Treaty on freedom to provide services and also with the competition law provisions of the Treaty. It is thus clear that the use of co-regulation in Community law may raise issues of some complexity, but is not in principle unacceptable (for a much more detailed analysis of the relevant caselaw see Hans-Bredow Institute/EMR 2006, pp.152–73). This now makes it necessary to examine the steps which have been taken by the Commission to develop the concepts of self- and co-regulation, and their role in the new Audio-Visual Media Services Directive (Directive 2007/65).

The Audio-Visual Media Services Directive

Background

My argument so far, then, has been that narrow definitions of self-regulation are not helpful, and that co-regulation is a much more useful concept for two reasons; it makes it clear that there is a continuum between “pure” industry self-regulation and “pure” command and control regulation, and it is compatible with systems which are cocktails of different types of regulatory approach. What is more important than to construct definitions is to apply the relevant normative concerns relating to regulatory legitimacy. Let me now say something about this in the context of EU broadcasting regulation (for more on co-regulation in the European context see European Audiovisual Observatory 2003; Hans-Bredow Institute/EMR 2006).

There has been considerable interest generally in the potential use of self-and co-regulatory techniques in EC law recently. Thus the Commission’s 2001 White Paper on European Governance proposed that implementing measures might be prepared within a framework of co-regulation. This was characterised by combining binding legislative and regulatory action with actions taken by the actors most concerned, drawing on their practical expertise. The exact shape of co-regulation would vary from sector to sector. The conditions where co-regulation would be appropriate would be where legislation sets a framework of overall objectives, basic rights, enforcement and appeal mechanisms, and conditions for monitoring compliance. Co-regulation was only suited to cases where fundamental rights or major political choices were not called into question, and must not be used where rules need to apply in a uniform way in every Member State. The participating organisations must be representative, accountable, and capable of following open procedures, and the arrangements must be compatible with the European competition rules (European Commission 2001, p.21).

In stressing the flexibility to tailor co-regulation to the circumstances in different sectors, and in emphasising quite demanding conditions for legitimacy, this seems compatible with the approach suggested in this article. The only potential problem relates to the exclusion of co-regulation where fundamental rights are called into question; rights issues are of course pervasive in broadcasting, so if this were taken literally it would suggest little scope for co-regulatory techniques. What it seems to mean, however, is that self-and co-regulation may be part of the cocktail with other protections for basic rights alongside them, something which already applies in the press sector in relation to confidentiality and privacy rights.

A more detailed account of the roles of self- and co-regulation can be found in the 2003 Interinstitutional Agreement on Better Law Making (European Parliament, Council and Commission 2003, paras. 18–23). The Agreement recognised the need to use alternative forms of regulation in the form of co- and self-regulation, whilst repeating that they were not appropriate where fundamental rights or important political options are at stake or in situations where the rules must be applied in uniform fashion in all Member States. However, it also included highly restrictive definitions of co- and self-regulation; note that they are definitions, not statements of relevant normative concerns in designing the system.

Thus Co-regulation means the mechanism whereby a Community legislative act entrusts the attainment of the objectives defined by the legislative authority to parties which are recognised in the field (such as economic operators, the social partners, non-governmental organisations, or associations).

This is very much a “top-down” view with minimal discretion for the co-regulatory body, and appears not to cover the “bottom-up” development of private regulation with state back-up powers, or the contracting-out of regulatory functions to such bodies. The narrowness of this view is reinforced by related provisions in the Agreement stating that the legislative act involved must comply with the principle of proportionality and must define the measures to be taken should there be non-compliance with its provisions. Voluntary agreements by the parties affected were to be assessed by the Commission for compliance with Community law and the basic legislative act, with further scrutiny on request by the Parliament and Council. It is difficult to see how such cumbersome processes would provide much extra flexibility compared to direct forms of public regulation.

Self-regulation is also narrowly defined as the possibility for economic operators, the social partners, non-governmental organisations or associations to adopt among themselves and for themselves common guidelines at European level (particularly codes of practice or sectoral agreements).

The definition itself does not include any reference to enforcement, sanctioning or the normative considerations set out earlier, although the Commission would notify the Parliament and the Council of self-regulation practices it regarded as compatible with Treaty provisions and as satisfactory in terms of the representativeness of the parties concerned, sectoral and geographical cover, and the added value of the commitments given.

The Initial Negotiation of Directive 2007/65

Despite the restrictiveness of these definitions, it is clear that in broadcasting there is extensive use of co-regulation in national arrangements. I have mentioned above advertising content in the UK, but other examples include such contested areas as protection of minors in Germany and in the Netherlands (these are summarized in Hans-Bredow Institute/EMR 2006, pp.48–50, 75–8, 130–2, 137–9). In general these systems have worked well. It should be noted that they are co- rather than self-regulatory. Thus in the German arrangements, the initial source was the Interstate Treaty on the Protection of Minors in the Media, which established a certification system for non-state bodies, and such certification will only be granted if they are sufficiently independent and have acceptable procedural protections, including provision for hearings and for the giving of reasons. The non-state bodies will classify content and ensure the enforcement of rules. In the Netherlands, the non-state Netherlands Institute for the Classification of Audio-Visual Material was founded after the government had agreed to bear the costs of a body in which all the relevant media organisations would participate; once more, conditions were required for accreditation including independence and the involvement of stakeholders. The Institute developed and applies a classification system, and imposes sanctions, including financial sanctions. These examples are particularly interesting in that they show that co-regulation can be effective in a controversial area, that of the protection of minors, and one in which the right to freedom of expression will have a major and highly contested role. The task faced in developing the new Directive was to develop a system which was compatible with such sophisticated arrangements already in existence, and this was to prove difficult.

The original proposal for the new Audio-Visual Media Service Directive noted that where fundamental rights were touched upon, regulation would be necessary, but there would also be a role for co-regulation and self-regulation. Thus “[c]o-regulation is already widely used with regard to the protection of minors and the proposal should encourage co-regulatory regimes in the fields coordinated by the Directive. Such regimes must be broadly accepted by the main stakeholders and provide for effective enforcement” (European Commission 2005, p. 9). The proposed text included in its recitals the following:

In its Communication to the Council and the European Parliament on Better Regulation for Growth and Jobs in the European Union the Commission stressed that a careful analysis on the appropriate regulatory approach, in particular whether legislation is preferable for the relevant sector and problem, or whether alternatives such as co-regulation or self regulation should be considered. For co-regulation and self-regulation, the Interinstitutional Agreement on Better Law-making provides agreed definitions, criteria and procedures. Experience showed that co- and self-regulation instruments implemented in accordance with different legal traditions of Member States can play an important role in delivering a high level of consumer protection (European Commission 2005, Recital 25).

In the text, however, the only reference was to the encouragement of co-regulatory regimes;

Member States shall encourage co-regulatory regimes in the fields coordinated by this Directive. These regimes shall be such that they are broadly accepted by the main stakeholders and provide for effective enforcement. (European Commission 2005, Art. 3.3).

This reflected a problem with the definitions; the Commission’s lawyers were concerned that the definition of self-regulation in the Interinstitutional Agreement was not appropriate, presumably because of its very restrictive scope; there was also a concern that self-regulation was very undeveloped in some Member States and so could not be prescribed for all of them.

The initial draft created serious potential difficulties of inflexibility. These largely stemmed from the inappropriateness of definitions contained in the Interinstitutional Agreement, designed to improve the Community’s future law making process, for schemes which were already established in other ways and working well. For example, in the UK, the Advertising Standards Authority pointed out that the effect would be to prohibit the use of self-regulation and to permit only a very narrow form of co-regulation in the policy areas covered by the Directive. The highly formal definitions in the Agreement did not recognize the different traditions of self- and co-regulation across different Member States; the approach adopted was also contrary to Art 249 EC leaving to Member States the choice of form and methods for implementing a directive. The definition of co-regulation would restrict industry participation in funding the regulatory system, and ignored the positive impact of practitioner recognition and support for “their” system. The Advertising Standards Authority role in broadcast advertising regulation would fit into neither the definitions of co-regulation nor of self-regulation. As a result, the proposed text “could severely inhibit the continued operation and development of effective advertising self-and co-regulation in the UK and across the EU-25.” (House of Lords European Union Committee 2007, evidence 145–147).

These criticisms were addressed in part in the revised Council draft of November 2006. This removed the reference to the Interinstitutional Agreement definitions, and instead stated in the Recital 25 that:

Experience in the audiovisual sector has shown that co-and self-regulation instruments implemented in accordance with the different legal traditions of Member States can play an important role in delivering a high level of consumer protection. Without prejudice to Member States’ formal obligations regarding transposition, this Directive encourages the use of such instruments. This neither obliges Member States to set up co- and/or self-regulatory regimes nor disrupts or jeopardizes current co- or self-regulatory initiatives which are already in place within Member States and which are working effectively.

The proposed Art 3 was amended to read:

Member States shall encourage co- and/or self-regulatory regimes in the fields coordinated by this Directive to the extent permitted by their legal systems. These regimes shall be such that they are broadly accepted by the main stakeholders and provide for effective enforcement.

Part of the background was, of course, the study for the Commission of Co-Regulation in the Media Sector referred to above, showing it to be far more extensive than previously thought across a range of Member States, and called either co- or self-regulation interchangeably (Hans-Bredow Institute/EMR 2006).

Parliamentary Amendment and Political Agreement

The Parliament considered the draft with a report submitted by Ruth Hieronymi as rapporteur for the Culture and Education Committee. This welcomed the fact that the Commission was, for the first time, recommending the use of co-and self-regulation in the implementation of a Directive, but proposed amendments “to make clear that each national legislature should decide under what conditions co-regulation and/or self-regulation instruments are used at national level, how the interested parties should be appointed to form the relevant bodies and what penalties the legislature imposes in the event of failure by the appointed self-regulation bodies” (European Parliament 2006). The First Reading text attempted to do this, but in a complex and wordy way; it is nevertheless worth quoting in full to show the complexity of attempting definitions here. The new text amended Recital 25 to remove the reference to the Interinstitutional Agreement and replaced the earlier wording as follows:

In its Communication to the Council and the European Parliament on Better Regulation for Growth and Jobs in the European Union the Commission stressed that a careful analysis on the appropriate regulatory approach, in particular whether legislation is preferable for the relevant sector and problem, or whether alternatives such as co-regulation or self regulation should be considered. Furthermore, experience showed that both co- and self-regulation instruments implemented in accordance with different legal traditions of Member States can play an important role in delivering a high level of consumer protection. Measures aimed at achieving public interest objectives in the emerging audiovisual media services sector will be more effective if they are taken with the active support of the service providers themselves. Thus self regulation constitutes a type of voluntary initiative, which enables the economic operators, social partners, non-governmental organisations or associations to adopt common guidelines amongst themselves and for themselves. Member States should, in accordance with their different legal traditions, recognise the role which effective self-regulation can play as a complement to the legislation and judicial and/or administrative mechanisms in place and its useful contribution to the achievement of the objectives of this Directive. However, while self-regulation might be a complementary method of implementing certain provisions of this Directive, it cannot constitute a substitute for the obligation of the national legislator. Co-regulation gives, in its minimal form, a “legal link” between self-regulation and the national legislator in accordance with the legal traditions of the Member States.

(25A)The generic term “co-regulation” covers regulatory instruments which are based on cooperation between State bodies and self-regulating bodies and vary widely in terms of their designations and structures at national level. The actual form which such instruments take reflects the specific tradition of media regulation in the individual Member States. What co-regulation systems have in common is that tasks and objectives which were originally the preserve of the State are achieved in cooperation with the players affected by regulation. Designated or authorised by the State, it is for the participants themselves to guarantee the achievement of the regulatory objective. In every case the systems are founded on a State legal framework which lays down instructions as to content, organisation and procedures. On this basis, the interested parties create further criteria, rules, and instruments, compliance with which they themselves monitor. Self-regulation as thus defined enables specialist knowledge to be exploited directly for administrative tasks and bureaucratic procedures to be avoided. It is necessary for all, or at least the most influential, players to participate in or recognize the system. Co-regulation operates by combining instructions to the interested parties with opportunities for State intervention should those instructions not be carried out.

By contrast, a new Article 3 is proposed in much simpler terms which are almost identical to that in the revised Council draft:

Member States shall encourage self and/or co-regulatory regimes at national level in the fields coordinated by this Directive. These regimes shall be such that they are broadly accepted by the main stakeholders in the Member State concerned and provide for effective enforcement.

Within the verbiage of the new Recital, there is thus a stress on the need for both co- and self-regulatory instruments, whilst attempting to provide new descriptions of them which are somewhat less restrictive and emphasize the role of different traditions in different Member States. There are also some new normative requirements; self-regulation may be complementary to, but must not replace, the obligation of the national regulator, and it is necessary for all, or at least the most influential, players to participate in or recognize the system.

Finally, in May 2007 a political agreement on a common position of the Parliament and Council was reached. The newly modified Commission proposal retained the first paragraph of the revised Recital 25 proposed by the Parliament, but replaced the second with a simpler statement that:

Co-regulation should retain the possibility for State intervention in the event that its objectives are not met. Without prejudice to Member States’ formal obligations regarding transposition, this Directive encourages the use of such instruments. This neither obliges Member States to set up co-and/or self-regulatory regimes nor disrupts or jeopardises current co- or self-regulatory initiatives which are already in place within Member States and which are working effectively.

In the finally adopted Directive 2007/65 the material is collected in Recital 36. Article 3(7) of the regime post-Directive 2007/65 adopts without modification the text proposed by the Parliament. There is thus no longer any attempt to adopt restrictive definitions of co- and self-regulation, and the stress is on openness towards the different traditions of Member States. There is some reference to normative concerns through the requirements that self- and co-regulatory regimes are broadly accepted by the main stakeholders and provide for effective enforcement. This new approach is vastly superior to that originally adopted by the Commission with its potentially highly constraining definitions.

Conclusions

Although this article has been in two very different parts, I hope that it is possible to see common themes in each of them. I have stressed that we should not adopt an approach which in principle favours a particular type of regulation or self-regulation as the two are not as distinct as they seem and in fact, as I have stressed, form a continuum. In many circumstances a more useful concept will be that of co-regulation, and this is happily now coming into favour, including in European Union arrangements for achieving agreed goals. This may in part be a matter of politics; one need not be a full-blooded public choice theorist to appreciate that commitments to full self-regulation may raise hostages to fortune through creating expectations which limit the capacity of public institutions to respond when the inevitable crises of non-compliance arise. Thus co-regulation may appear much more attractive than self-regulation to bodies such as the Commission or regulators which wish to retain some powers of effective crisis management. There are also, however, considerations of principle. Co-regulation is preferable to the use of the term self-regulation not only because it recognises the continuum of different regulatory approaches, but because it is open to the idea of mixed regulatory regimes, with different approaches being undertaken together based on contextual factors such as the different interests being protected by the regime and the most effective means of enforcement. This being so, restrictive definitions (as found in the Interinstitutional Agreement) are not helpful and are likely to cause problems in legislative drafting, as has been vividly illustrated in the process of drafting the new Directive. In particular, whereas the Interinstitutional Agreement is directed at the establishment of better law-making by the Union institutions, and so at the creation of new regimes, it is unsuited to the recognition of the large variety of co-regulatory regimes which have already developed. Differences between them are partly caused by differences in national cultures and traditions, but also by different mixes in the cocktail comprising each regulatory regime. Thus they cannot be tied down by definitions of this kind.

What are much more important than definitions are the sets of normative concerns which can be brought to bear on self- and co-regulatory schemes. These may be formed from a variety of different sources, ranging from national constitutional provisions and principles to “soft law” such as the Ofcom criteria for promoting effective co- and self- regulation. In the new Directive (Directive 2007/65), the Commission has made interesting moves towards using self- and co-regulatory techniques in the future; the dangers of a restrictive approach have largely been overcome due to the changes in drafting. The hard questions will be at Member State level relating to institutional design; the plurality of concepts of self- and co- regulation through various national traditions does not avoid the need for normative principle on issues such as participation and accountability. The draft finally adopted leaves plenty of scope for examination of these questions and for learning from the widespread arrangements for self- and co-regulation already in place. The most important developments, however, are for the future as different regulatory regimes are characterised by varied cocktails of regulatory approaches, but, it is to be hoped, common commitments to normative principles of openness and accountability.