Why and how to regionalise the Common Fisheries Policy

The concept of regionalisation, as it has been employed in connection with the Common Fisheries Policy (CFP), is both ambiguous and multidimensional in the sense that it can have different meanings to different people and subsumes several discussions under one heading. Similarly, the perceived benefits of regionalisation can vary. In this article we develop a conceptual framework, which allows structuring of different, perceived benefits of regionalisation according to various objectives of CFP governance, as well as disentangling the different sub-themes that the discussion of regionalisation subsumes: the questions of ‘what’, ‘where’ and ‘whom’. Eventually, we present a suite of five different models of regionalisation— ‘archetypes’—that we believe are representations of important perspectives on what regionalisation means in practice and might facilitate a structured discussion of where the European Union should be heading in relation to fisheries governance: ‘Nationalisation’, ‘Regional Fisheries Management Organisations’, ‘Regional Fisheries Co-Management Organisations’, ‘Regional Marine Management Organisations’, and ‘Cooperative Member State Councils’.


Introduction
The discussion of regionalisation, as it has in later years unfolded in relation to the Common Fisheries Policy (CFP) of the European Union (EU), is both complex and politically sensitive, as are the decisions potentially to be taken. Nevertheless, having observed and, to some extent, taken part in the discussions of recent, we feel that the lack of a clear (and perhaps common) conceptual framework for discussing regionalisation further complicates the matter. Consequently, this article offers a framework of concepts and possible regionalisation models, which we believe resonate in the ongoing discussion of regionalisation as a way forward for the CFP.
To us regionalisation is strongly associated with decentralisation of CFP governance, as it is also the case for Symes earlier in this issue (Symes 2012). A main point in this regard is that the discussion of regionalisation must-due to the present highly centralised structure of the CFP-basically evolve around the need to make the management regime less centralised, rather than about making the management regime more centralised by strengthening a regional structure above for instance the level of individual countries. Interestingly, it can be argued that the initial establishment of the CFP is an example of the latter version of regionalisation: centralisation of fisheries management authorities. Thus, as pointed out by Symes (2012), the CFP was originally basically designed and set up with the North Sea in mind. However, with numerous enlargements of the EU the CFP expanded from a policy for one regional sea to the continental wide policy, which it was never designed to be. Nevertheless, decentralisation and regionalisation is not the same. Regionalisation includes a territorial component that decentralisation does not necessarily and, in that sense, regionalisation is a very particular instance of decentralisation. For a more thorough discussion of decentralisation, regions, regionalism and regionalisation in the context of the CFP, please refer to Symes (2012) article within this special issue. Similarly, for those interested in a more general introduction to conceptions of decentralisation, which is beyond the scope of this article and the special issue as such, we refer to Dubois and Fattore (2009), who provide an overview of the many typologies that have been employed to categorise different types of decentralisation in the context of public administration research and as such provide a good entry point to decentralisation research.
The present article consists of three parts before concluding in a brief discussion. After briefly introducing the background for discussing regionalisation of the CFP, the first part is a systematic overview of different reasons for why the CFP should be regionalised. The assertion is that regionalisation to be perceived as preferable must be seen as able to contribute to what can be perceived as basic objectives of CFP governance. The second part outlines and discusses the three main problem dimensions to address when designing a regionalised governance system for the CFP. In the third section, we outline a handful of what we refer to as 'archetypes' of regionalisation to flesh out possible ways that regionalisation of the CFP can be designed vis-à-vis the different problem dimensions.
The current article is kept on a relatively conceptual level. However, to draw more specific implications for policy, the following article of this issue (Hegland et al. 2012, In Press) applies the framework from this article to the discussion of regionalisation and reform of the CFP as it has played out in recent years, as well as reports on how the archetypes were received by stakeholders confronted with them by means of a survey.

Background
Before turning our attention to regionalisation more specifically, it seems appropriate briefly to introduce the current set-up of the CFP, which will be the point of departure of regionalisation, as well as introduce the different shortcomings that the current CFP is suffering from (see also Raakjaer and Hegland 2012, as well as Symes 2012, for more details).
In short, the core governance system of the CFP operates across three politicoadministrative levels: the member state level, the (embryonic) intermediary level of regional EU seas (or areas covered by the Regional Advisory Councils (RAC)), a and the EU level. Due to the fact that the protection of living aquatic resources is recognised as one of only a handful of issues under the exclusive compentence of the EU, the central EU level is in a position to take a very wide range of decisions relating to fisheries management.
In addition, the policy system is embedded in the binding principle of non-discrimination, equating to a standard set of regulations or 'harmonisation' for its 'common pond'-the combined waters of the member states (Symes 1997). Furthermore, the EU allocates fishing opportunities based on a firmly rooted principle of 'relative stability' , which may be the most path dependent element of the CFP (Hegland and Raakjaer 2008). The implication hereof is that the CFP governance system is predisposed to exerting a high degree of micro-management from the central EU level.
In consequence, the situation of the CFP remains one charachterised by a top-down, command-and-control approach to management. Simply put, management is to a large extent being shaped in what is a continous process of draft legislation being prepared by the Commission of the European Communities (Commission), legislative acts being jointly decided by the Council of the European Union (Council) and the European Parliament acting together through the codecision procedure, and adopted legislative acts being implemented by the EU member states. b Council Regulation No 2371/2002, known as the basic regulation, provides the Council with the necessary authority to govern access to fishing zones and resources and the sustainable pursuit of fishing activities, including limiting catches; limiting fishing effort; adopting technical measures; adopting multiannual recovery plans; and adopting multiannual management plans. Unfortunately, as discussed in the introduction to this issue , the set-up may have provided a stable and strong institutional basis for EU-wide fisheries management but has not to the same extent provided sustainable fisheries.
In an analysis of the problems facing the CFP, Raakjaer (2009) listed the following features of fisheries and fisheries management in the EU: For many years overfishing has been evident, resulting in a critical resource situation. A fragmented fishing industry has led to a fragmented interest structure in the EU fishing industry. Lack of commitment has existed within the Council to ensure sustainable fishing. Persistent lack of political will has been evident in the Council and the member states to reform the CFP. Member states have emphasised domestic interests. A strong tendency to apply off-the-peg approaches (one size fits all) has existed. Inconsistency has been evident between structural policy elements and conservation elements within the CFP. The management regime building on total allowable catches (TAC) set for single species is not effective in multispecies demersal fisheries. Problems of 'implementation drift' and lack of enforcement exist in the member states. A clash has been present between the ways administrators and fishermen view the goals and means of the management regime. Attempts to introduce elements of 'New modes of governance' have not been successful in the fisheries domain. The type of co-management introduced has not led to responsible behaviour.
Within the area of fisheries management, there has been a growing recognition that effective management relies on allowing rules to differ between different fisheries, as pointed out by Symes (2007). However, the monolithic structure of the CFP complicates this in Europe, where even matters of details are decided at the central EU level. Following from its status as an exclusive competence area of the EU, the protection of living aquatic resources is exempted from the principle of subsidiarity, which otherwise entails that decisions in the EU should be taken at the lowest level appropriate. However, that it is not mandatory to follow the principle does not mean that it is not allowed or preferable.
Thus, understanding the present structural failures of the CFP closely relates to the mismatch in scale levels, particularly the lack of ability to find the 'right fit' of scales levels for governance intervention. The problem of scale and governance has been addressed by several influential scholars. For example, Ostrom (1990) invokes the concept of 'nested enterprises' , Kooiman et al. (2008) refer to 'orders of governance ' , and Symes (2007) uses the term 'scales of governance'. They all introduce different levels of rules or orders of governance that can be thought of as different spheres of influence encased in one another. However, matching the political boundaries to manageable areas or ecologically appropriate scale levels poses particular challenges in relation to fisheries and other marine management concerns. Specifically vis-à-vis the CFP, the EU's Marine Strategy Framework Directive (MSFD) and the Integrated Maritime Policy (IMP) both present new considerations in this regard (van Hoof et al. 2012, In Press).
Not surprisingly, in light of the above, regionalisation has increasingly become part of the discussions on how to reform fisheries management in the EU.

Why regionalise? Outlining the objectives of CFP governance
Discussing regionalisation of the CFP is only relevant insofar as CFP governance could potentially in some way gain from being regionalised. Consequently, in this section, we develop and outline a relatively simple typology of objectives (or qualities) of CFP governance that regionalisation could in principle be perceived as contributing to. The specific benefits from regionalisation that people invoke can thus be expected to relate to one or more of the objectives in our typology and will, furthermore, affect how they imagine regionalisation should be put into practice, which is the topic of the subsequent sections.

The three basic objectives of CFP governance
Our starting point is an understanding of 'governance' as consisting, on one hand, of a 'governance system'-defined as the system or network of public and private bodies having an impact on the content, implementation, enforcement, or interpretation, et cetera, of policies and measures adopted-and, on the other hand, the 'policies and measures' themselves-defined as the actual rules and regulations, et cetera, to which citizens are subjected.
Accordingly, in the context of the CFP, we argue that any perception of regionalisation as a favourable option will be motivated by a belief that either the governance system of the CFP or the policies and measures of the CFP-or both simultaneously-will benefit in one way or another from regionalisation.
Such distinction between the system itself and its products in the shape of policies and measures is related to Scharpf's (1997:19) notion of democracy having to be 'understood as a two-dimensional concept, relating to the inputs and to the outputs of the political system at the same time.' Scharpf's distinction highlights that a well-functioning democracy, which means one that is able to deliver 'collective self-determination' , is contingent on both the system's ability to take up the preferences of citizens and balance the differences of interests through the decision-making system (which Scharpf refers to as 'input-oriented authenticity') along with the ability to transform preferences into actual outcomes that are effective in achieving stated goals (which Scharpf refers to as 'outputoriented effectiveness'). In our typology, we operate with these qualities of a democracy as two of the basic objectives of CFP governance and hence also major categories of reasons for wanting to move towards regionalisation.
However, to these two we add a third objective that also relates to the governance system, namely efficiency. In other words: how efficient the system is in the process of taking up preferences and transforming those into effective policies and measures. Where 'effectiveness' is about getting the actual product (in this case a policy or a measure) in the best possible shape vis-à-vis the needs, 'efficiency' is about optimising the production process, so to speak.
Consequently, our typology includes three basic objectives of CFP governance, which regionalisation can potentially contribute to: 1) ability to take up and balance preferences, 2) efficiency of the system, and 3) effectiveness of the policies and measures. c In the following sections we take a closer look at these basic objectives and sub-divide each in more specific objectives by drawing more directly from literature specifically on fisheries management.
Objective 1: increasing the ability to take up and balance preferences The objective of being able to take up and balance preferences relates broadly to the legitimacy of the governance system. Arguments for regionalisation that invoke this objective are consequently concerned with how regionalisation can assist in bringing about a system where the dispersal of authority, power, responsibility or related measures of influence d across actors and institutions is (perceived as) more legitimate/fair or just than an alternative distribution. Where Scharpf labels this 'input legitimacy' , the more common term in the fisheries management literature is 'process (or procedural) legitimacy' , which has traditionally been employed to refer to the legitimacy that fisheries management derives from being the product of a process or system perceived as fair and just (Jentoft 1989, Jentoft 1993, Jentoft and McCay 1995Raakjaer Nielsen and Mathiesen 2003).
Literature on fisheries management has argued that participation of user groups positively influences the perception of the legitimacy of fisheries management and that this in turn will improve the level of compliance (Jentoft 1993), which is widely considered a sine qua non of successful fisheries management. e In relation to this, however, Jentoft introduces an important distinction by emphasising the need: To distinguish between internal and external legitimacy, and to recognize that these two types of legitimacy may be in conflict. For those directly involved, userparticipation in decision making may improve legitimacy, but for the general public and other groups that are outside the process, user-participation may well be seen as a step in the wrong direction (Jentoft 2000:145).
On a general level Jentoft's observation highlights that there is no universal yardstick on how good a system is at taking up and balancing preferences. For some, one specific way of organising the system might be perceived as legitimate, to others that same specific way may be viewed as illegitimate. At the same time Jentoft points us in the direction of what might be one of the defining cleavages on our particular issue, namely the different perspectives of users as opposed to a wider set of stakeholders or citizens.
Consequently, arguments for regionalisation that invoke the objective of improving the take-up and balancing of preferences, which in turn 'produces' process legitimacy, can be sub-divided on the basis of whether they are driven by a concern for internal or external legitimacy. And these two perspectives will likely often translate into different perceptions of how regionalisation should look in practice.
Objective 2: making the system more efficient As mentioned, this objective is concerned with whether the policy process can be organised more efficiently-preferably, of course, without reducing its ability to take up or balance preferences or the effectiveness of its outputs. Arguments invoking this objective will advocate that regionalisation contribute to creating a situation where the same (or higher) level of goal realisation can be achieved with the use of less resources by reorganising the system and putting the resources in play in a more efficient way.
In our understanding this covers two interrelated but nonetheless distinct sub-objectives: efficient use of financial resources, and efficient use of human resources. The two are interrelated insofar as financial resources are to some extent a means by which human resources can be acquired.
In relation to the efficient use of financial resources a main issue is the balance between the public costs of managing the fisheries sector vis-à-vis the public benefit of having a fisheries sector. It has been noted that in several member states the value of landings is now lower than the costs on the public budget related to carrying out fishing activities (Commission 2009). Although the public benefit of having a fisheries sector can be difficult to make up and that the value of landing is not a definitive measure for the public benefit, f this discrepancy clearly puts the issue of the efficient use of financial resources at the centre of attention for the reform of the CFP.
In principle, financial resources are a means by which human resources can be acquired. However, when we talk about the efficient use of human resources in the context of the governance system of the CFP, we primarily think of human resources that cannot be acquired in a simple way merely by spending financial resources. It is more than 'just' a financial issue. Examples of this could be highly qualified scientific expertise or elected decision-makers. Consequently, when using scarce human resources, it is vital that these resources are put to work in the most efficient way.

Objective 3: providing for more effective policies and measures
Whereas the first two objectives of our typology highlight respectively the legitimacy of the governance system and its efficiency, the third objective-more effective policies and measures-focuses attention on the outputs from the system. Scharpf uses the concept of 'output-legitimacy' to refer to the legitimacy that a policy or a measure can derive from being perceived as effective in reaching its goals. Output-legitimacy is thus not the same as effective policies or measures but rather a result of them. Jentoft (1989), focussing on the legitimacy of management measures among fishermen, argues that there are primarily two elements at play in relation to this kind of legitimacy, which he labels 'content legitimacy': 1) that a measure-to be perceived as legitimate-has to 'coincide with the way the fishermen themselves define their problems' (Jentoft 1989:139) and 2) that the measure needs to be equitable. Content legitimacy is 'in the eye of the beholder' , so to speak.
So fishermen prefer policies and measures that coincide with how they perceive the problems, but so will other stakeholder groups; they will also have different perceptions of what the problems and thereby appropriate solutions are-and equity among fishermen may not necessarily be one of them.
Currently, the stated objective of the CFP is to 'Ensure exploitation of living aquatic resources that provides sustainable economic, environmental and social conditions' (Council 2002: Article 2.1). Of course, the perception of whether CFP measures are appropriate to achieve the goals will to a significant degree depend on how the individual weighs these different, specific fisheries management objectives in relation to each other.
The three objectives of the CFP reflect what Charles (1992) presents as the three conflicting fisheries management world views or paradigms: 1) the 'conservation paradigm' , focussing on conservation and resource maintenance; 2) the 'rationalisation paradigm' , focussing on economic performance and productivity; and 3) the 'social/community paradigm' , focussing on community welfare and equity. It should be noted, though, that the paradigms are theoretical constructs, which will usually not be found as pure orthodoxies in real life but rather as blended preferences.
Anyhow, in relation to our conceptual framework the implication of the co-existence of conflicting fisheries management worldviews has the implication that arguments for regionalisation that depart from the objective of more effective policies and measures might very well depart in conflicting ideas of what fisheries management is all about and therefore be fundamentally different in their implications for regionalisation in practice.

Putting the typology of CFP governance objectives together
Based on the above sections we can now 'assemble' our typology. As evident from Figure 1 below, the typology is based on the notion that there are three basic objectives of CFP governance and a range of more specific sub-objectives. Our assertion is that any specific argument for of how the CFP would benefit from moving towards regionalisation will relate to one or more of these objectives or sub-objectives, Thus, our typology of objectives is effectively also a typology of the different benefits that regionalisation can be perceived as having.
Having now identified different objectives that regionalisation can in theory contribute to; we turn our attention to the different dimensions of the regionalisation discussion.

Problem dimensions of regionalisation
As mentioned in the introduction, we believe that part of the complexity of the regionalisation discussion stems from the fact that the discussion subsumes three intertwined sub-discussions or problem dimensions. These discussions pertain to the questions of 'what' , 'where' and 'whom'-issues which can only be separated analytically. Consequently, this section disentangles the discussion into the three important problem dimensions of regionalisation of the CFP: 1) what kind of authority should be regionalised, 2) where should regionalisation take place on the politico-administrative scale, and 3) who should be the recipients of authority at any given level of the scale?
The first problem dimension of the discussion relates to the question of determining which authorities that should be placed at the various politico-administrative levels. Although all kinds of authority are in play in this regard, the power to take decisions is of special importance.
Generally, developed fisheries management systems include a hierarchy of decisions ranging from the most general to the most specific. In the terminology of Ostrom (1990) 'constitutional rules' represent the outermost sphere, as they dictate the structure of governance and the overall political organisation of the system, 'collective-choice rules' are nested within the constitutional rules and concern policy-making and management decisions, and, lastly, 'operational rules' involve the daily decisions of managing fisheries, as they involve monitoring, enforcement, and other actions on the ground level. Operational rules are the innermost sphere and, thus, are nested in both collective-choice and constitutional rules. Other scholars employ different terminologies. However, a main message that we draw from this is that for each of the 'nested enterprises' , 'orders of governance' or 'scales of governance' it is in principle possible to determine which decisions are most appropriately taken there. However, in practice this is a highly complex and politicised discussion, not least in the context of an organisation such as the EU where there are marked differences between the member states when it comes to both the national fisheries systems as well as the culture of public administration.
Without going into the actual discussion of what specific decisions under the CFP that are most appropriately taken at what politico-administrative level, it might be useful in brief to outline the current hierarchy of the regulatory framework of the CFP. Ignoring that the EU is part of a system of international agreements, the top level of the CFP regulatory hierarchy is the Lisbon Treaty, which outlines the basics of the mode of cooperation in the EU as such and the area of fisheries in specific ('constitutional rules'). As an example, it is the Lisbon Treaty that stipulates that the area of fisheries is a co-decision area, where the Parliament and the Council decides in cooperation. The second layer of the hierarchy is the so-called Basic Regulation, which generally deals with a mixture of 'constitutional' and 'collective choice' rules and is reviewed every ten years. It is, for instance, in the Basic Regulation that the system of total allowable catches is outlined as well as rules on access to waters and so on. Likewise, overall strategies for fisheries management (such as 'maximum sustainable yield' (MSY) as a target for stock exploitation) are contained in the Basic Regulation. The third regulatory layer consists of more specific regulations or decisions for different components of the CFP, in relation to conservation the most important being the Technical Measures Regulation, which is negotiated yearly and outlines detailed provisions on how to fish in practice in the EU by stipulating days-at-sea, mesh sizes, allowed gear types, and so on, for various fisheries and fleet segments ('operational rules'). Finally, the member states have the authority to decide on for instance how to allocate fishing rights ('collective choice'), as well as on how in practice to implement the decisions taken at EU level ('operational rules').
So regionalisation is, we would argue, about allocating the right to take particular decisions (and other authorities) to the right levels. Recalling the notion of 'nested systems' , there is nothing that says that elements of for instance 'constitutional rules' (e.g. how to include stakeholders) cannot be decided at lower levels, as long as they respect the 'constitutional rules' decided at higher levels (e.g. democracy and sustainable fishing). In fact, this may in the light of the principle of subsidiarity be preferable.
Finally, it should be recalled that vis-à-vis the discussion above, there is a separate discussion to be taken in relation to the extent to which fisheries management decisions should be dealt with in separation from broader marine management or not.

Regionalising to where?
As mentioned earlier, the governance system of the CFP operates over several politicoadministrative levels, most prominently the member state level, the intermediary level of regional EU seas (or the RAC areas), and the EU level. Moreover, the EU is embedded in a global international level as signatory to a number of treaties, conventions, and declarations dealing with fisheries policy and management. Likewise, at the other end of the scale, a number of member states embed within them subnational politico-administrative levels with relevance for fisheries management.
In any effort to regionalise a decision must be made as to which level or component of the organisation or system that a particular authority could most appropriately be allocated. In relation to the CFP, it is important to bear in mind that the politico-administrative scale has its counterpart in a bio-geophysical scale reflecting the biological system of the sea. One scale level of the natural system could be a fjord or a bay, and on the other end of the spectrum would be the oceans or ultimately the global marine ecosystem. In between we have the regional seas, of which the North Sea and the Baltic Sea are examples. The scale levels of the natural system are not, however, neatly reflected by corresponding levels of policy-making/management on the politico-administrative scale, which is one of the challenges of operating the current governance system of the CFP. Thus, creating a better match between scale levels of the governance system and the scale levels of the natural system may call for regionalisation.
A particular challenge in relation to this problem dimension relates to the fact that the EU legal framework does not allow politico-administrative structures with decision-making powers between the EU level and the member states-in effect these powers can be held be either by the EU or by the member states. Of course, however, this does not rule out regionalisation. In principle, we would argue, there are at least three ways to get around this problem. The least ambitious is the one already applied: Establishing the RACs did represent some sort of regionalisation of the CFP, and since the RACs are exclusively advisory bodies that did not constitute a problem in a legal sense. However, insofar that regionalisation presupposes more authority at the regional level than merely that of a right to be consulted, a possibility would be to create an informal structure that is in a legal sense not decision-making but might be so de facto. Finally, the most advanced solution, but also legally most tricky, is to resort to indirect regionalisation, where authority is awarded to the EU member states on the condition that they exercise that authority jointly with the other relevant member states in the region. This resembles to a wide degree one of the visions presented in the Commission's Green Paper: Another option to be carefully considered would be to rely wherever possible on specific regional management solutions implemented by Member States, subject to Community standards and control. [. . .] In most cases this delegation would need to be organised at the level of marine regions [. . .]. Member States would therefore have to work together to develop the setups required. (Commission 2009:10f ) Regionalising to whom?
Even if a clear vision might exist as to what authorities are best placed at what politicoadministrative level, as discussed in the sections above, the issue remains to decide to whom management authority should be awarded. A key issue in this regard is the question of how private interests-as opposed to public bodies that are traditionally in one way or the other linked up to representative democracy-should be involved in the process at any given politico-administrative level.
According to democratic theory, those affected by a decision should be given the opportunity to participate in the decision-making process related to it (Dahl 1989), and today most fisheries management systems worldwide include elements of involvement of users and stakeholders in the development of management measures (beyond the involvement coming from representative democracy in which users and stakeholders are of course also voters). Often these arrangements are of a sort that is generally known under the label 'co-management' , which Jentoft (2003:3) defines as 'a collaborative and participatory process of regulatory decision-making between representatives of usergroups, government agencies, research institutions, and other stakeholders'. Sen and Raakjaer Nielsen (1996) have provided empirical evidence on a variety of different co-management arrangements based on user and stakeholder involvement in fisheries management. Not surprisingly, they observed a large degree of variation across fisheries and regions and even within specific countries. Their conclusion is that the proper design principles depend upon the context and conditions under which the co-management arrangement must operate. Sen and Raakjaer Nielsen (1996) observed moreover that the development of co-management arrangements often evolves gradually through a process of muddling through, noting that the process is always dynamic; a finding also supported by Jentoft and McCay (1995).
In order to facilitate the discussion of different options for user and stakeholder involvement in fisheries management in the EU, we have outlined a revised and modified version of the typology of fisheries co-management regimes originally presented by Sen and Raakjaer Nielsen (1996). Inspired also by Raakjaer (2009), we propose the existence of five levels of stakeholder-involvement, g which form a continuum range over various visions of co-management from little or no involvement to ultimately self-management: 1) 'Top-down hierarchical management by the state'; where mechanisms for dialogue with users and stakeholders might exist, but only minimal exchange of information takes place and EU/national governments decide what information to share. 2) 'Co-management by consultation'; where extensive formal mechanisms for consultation (and feedback on use of recommendations) with users and stakeholders exist, but all decisions are taken by EU/national governments. 3) 'Co-management by partnership'; where EU/national governments, users, and stakeholders cooperate as decision-making partners in various aspects of management. 4) 'Co-management by delegation'; where EU/national governments have devolved de facto decision-making power to users and stakeholders in relation to various aspects of fisheries management. 5) 'Industry self-management with reversal of the burden of proof'; where the government has devolved wide-ranging management authority to users and stakeholders, who must demonstrate to EU/national governments that management decisions are in accordance with the given mandate.
Traditionally, stakeholder involvement within the CFP has primarily been variations of the two top categories, in other words the least ambitious. Even the 2002 CFP reform, which introduced the Regional Advisory Councils (RAC), only consolidated the 'co-management by consultation' approach in the CFP as the means to obtain inputs from the regional and local level into the CFP decision-making.
However, the question of who are the specific stakeholders that should be involved is also contentious. In the EU it is commonly accepted that those dependent on fishing for their livelihood ought to be well-represented in the management process. Nevertheless, as part of a broader trend worldwide, the EU is encouraging broader stakeholder representation. Even though industry upholds a dominating position, conservationists and consumers are today represented in EU consultative bodies. There are many interests related to fisheries issues and this might, at least in principle, call for representation of a broad set of stakeholder organisations. However, in a system such as the EU where it is already a challenge to balance industry interests from different countries with each other this serves to further complicate matters and constitutes yet another choice to be made when designing regionalisation.
Merging what, where and whom: a selection of archetypes of regionalised governance of the common fisheries policy The following section outlines five qualitatively different models of regionalisation of the CFP on the basis of our discussion so far. In other words, each of the following models reflects an integrated perception of how the questions of who, where, and what could be answered 'in practice' and they also contribute to the objectives of CFP governance in varying ways. By referring to these models as 'archetypes' we emphasise that they are-rather than detailed ready-to-apply systems-rough skeletons drafted primarily with the intention to create an illustrative suite of potential models, which could form a point-of-departure for discussions.
We do not claim that this is a complete list of possible models-far from it. A full list of hypothetical models based on our three problem dimensions including sub-dimensions would be almost unimaginably long. Simply consider that the variation of stakeholder involvement (five options) multiplied by the issue of whether regionalisation should be about fisheries management or marine management (two options) results in ten different models.
Consequently, in deciding on our selection of archetypes we have not resorted to any such automatism but rather included five models that based on our pre-knowledge of the CFP we believe capture important perceptions of ways to go forward. It should be mentioned, however, that we have prioritised presenting and discussing models that represent a significant and qualitative change from the current system as opposed to models that represent variations of the current system.

Archetype 1: nationalisation
The first alternative to the current governance system in our selection of regionalisation archetypes is the Nationalisation model. The Nationalisation model represents a qualitatively different configuration to the current system and also stands in contrast to the subsequent models outlined, where an intermediary level of governance between the EU and the member states is actively sought strengthened. This is not the case under the Nationalisation model, which puts the member states at the centre. In that way it represents a perspective on regionalisation that responds to the question of 'where' in a fundamentally different way than any of the other models in the selection-and this is its trademark.

Nationalisation
The member states are awarded the responsibility for the conservation of resources in their own Exclusive Economic Zones. Issues relating to shared stocks would be sorted out through a system of bilateral agreements between member states or any other arrangements that the member states themselves deem necessary. The level of involvement of stakeholders would be an issue for the individual member state to decide.
In its pure form the Nationalisation model entails arguably the most radical change from the current system among the presented selection of regionalisation models. In fact Long (2010) argues that nationalisation, in the sense that we outline it above, would require an amendment of the Lisbon Treaty. This model turns EU fisheries management upside down by awarding the authority for resource conservation measures to the member states within their own Exclusive Economic Zones (EEZ)-as opposed to the current system where this is one of only a few exclusive competences of the EU. Management of stocks shared by different states would then, likely, be a matter of setting up bilateral or multilateral agreements among the countries (EU as well as non-EU) in whose EEZs the stock in question inhabits, much like the agreements that are negotiated yearly between for instance Norway and the EU.
The Nationalisation model would per definition eliminate the problem of excessive micromanagement or one-size-fits-all solutions from the EU level, as in principle the EU would not even be in a position to coordinate approaches. Because it will in general be up to the individual member states to decide their national approach, it is difficult to say much more about how such a system would operate in practice at the level of fisheries. For instance, at member state level micromanagement may continue under such a system depending on the national style of management. Nonetheless, stakeholders exclusively concerned with stocks present in the EEZ of only their own member state will likely feel that decisions are being made closer to them and they could, depending on the national style of management, have more direct say in the management of such stocks. In opposition, those concerned with stocks shared with other member states or stocks in other member states' EEZs will have less say, and their interests will have to be defended by their government in negotiations with other states. This would be the situation in many fisheries, as the geopolitical characteristics of EU waters results in a high number of shared stocks.
Archetype 2: regional fisheries management organisations The heart and soul of the Regional Fisheries Management Organisation (RFMO) model h is the establishment of formalised structures with 'institutional personalities' at the regional level. The RFMOs are given authorities to be exercised via delegation of authority to the member states on condition that they exercise their power jointly and without interference from the central EU level within a specified mandate.

Regional Fisheries Management Organisations
Under this model the member states would be given wide authority for fisheries conservation on the condition that the member states with fishing interests in a regional sea area establish a regional fisheries management organisation (RFMO) to deal with fisheries management issues specific to that area. A general framework for regional approaches will be provided by the central EU institutions. The stakeholders' input will continue to be channelled through the RAC. However, the RAC would in most cases advice the RFMO rather than the central EU institutions. The exact extent to which stakeholders' input is given weight in the decision-making process of the RFMO is up to that organisation on a case-by-case basis.
Generally, only member states with fishing interests in the specific sea area covered by the individual regional management organisation would participate in the RFMOor alternatively, in a stricter interpretation than the one reflected in our text above, only member states with coastlines to the region. The geographical coverage of the individual regional management organisation would be the 'regional sea area' , which may have different meanings to different sets of interests. However, generally the term 'regional sea area' when discussed in the context of EU fisheries management would refer to areas such as the North Sea or the Baltic Sea, or marine regions such as those covered by the RACs or alternatively the regions and sub-regions outlined in the MSFD (European Parliament and Council 2008).
The RFMO model presupposes considerable authority placed with the regional organisation-referred to as 'wide authority' above. Although not specifying how much authority this entails, the text above indicates that their authority should at least allow the different regional organisations to develop different approaches to management.
Under this model the RACs would continue to provide stakeholder input to the decision-making process. Basically they would continue operating as they do under the current system with the modification that instead of the Commission being the primary recipient of advice, the RFMOs or the member states in their capacities as part of the RFMO would become the main recipients.
This model potentially relieves the central level of the burden of micro-managing; at the same time, the EU would, as opposed to under the Nationalisation model, maintain a coordinating role as well the ability to set the overarching goals and the frame for the regional approaches. This would potentially increase the system's ability to tailor-make management, as the EU would not have to apply off-the-peg, one-size-fits-all management solutions to the same extent as today. Furthermore, the feeling of distance between the decision-making body and the place where impacts of management or mis-management are felt would be reduced. Delegation of some authorities to a lower level would potentially facilitate more timely management measures as only the most principal decisions would have to go through the lengthy process of joint decision-making between the Council and European Parliament. Ideally, therefore, this model offers a more efficient and legitimate governance system as well as policies and measures more closely corresponding to the needs of the specific region.
However, what this model offers in particular, compared to most of the models in our selection of archetypes, is first and foremost that the model builds to a relatively wide extent on the current structures. The RAC can, as an example, be allowed to continue operating in a relatively unchanged format as an advisory body. As a consequence, this model also offers a governance system where there is a clear differentiation between those governing and those being governed. Based on advisory input from stakeholders, the EU and the member states would decide on how to do management. Coordination with other policy areas would need to be taken care of through the general framework decided upon by the central EU institutions or possibly in cooperation with other regional management organisations charged with other elements of marine management.
In terms of the level of stakeholder involvement, however, this model does not move beyond the weakest form of co-management, namely 'co-management by consultation'. However, the feeling of being heard (and maybe also in reality being listened to) might be increased by getting a closer match between the RACs as advisory bodies and the RFMOs as decision-making bodies. Moreover, it should be noted that there is nothing in the model that precludes that a decision could be taken at regional level to involve stakeholders more in the decision-making process, which would then potentially move the regional system towards more evolved co-management.
However, these advantages are not guaranteed. In contrast, it could also be argued that the governance system will lose efficiency by including yet another decision-making level and that the system loses legitimacy because of even more complex procedures. Moreover, establishing mechanisms for ensuring delivery of the targets/the frame agreed at central level remains a challenge in this regional management organisation model as well as the next two variations, see below.

Regional fisheries co-management organisations
Our second variation on the regional management model theme in our selection of archetypes, the Regional Fisheries Co-Management Organisations (RFCOMO) model, distinguishes itself from the RFMO model described above by specifying stronger direct involvement of stakeholders in the regional decision-making process.

Regional Fisheries Co-Management Organisations
Under this model the member states would be given wide authority for fisheries conservation on the condition that the member states with fishing interests in a regional sea area establish a regional fisheries co-management organisation (RFCOMO) to deal with fisheries management issues specific for that area. A general framework for regional approaches will be provided by the central EU institutions. The RACs would cease to exist; instead stakeholders, scientists, and member states' administrators would work together within the RFCOMO to determine the best strategies for their regional area.
This model delivers stronger ownership of regulations by those subjected to them. Where the RFMO model allowed regional decisions to be made by the member states' representatives alone acting on advice from stakeholders in the RAC, this model presupposes joint decision-making between member states' representatives and stakeholders. Consequently, not only does this model decentralise authority to the regional level with the potential benefits that this entails, it also moves the system from 'co-management by consultation' to 'co-management by partnership' in relation to the specific authorities given to the RFCOMO.
Enlarged with government representatives and possibly scientists, in principle the RACs could be reorganised into RFCOMOs since the stakeholder expertise needed is already present there. In any case, the RAC, as the type of organisation it is presently, would likely cease to exist. As compared with the RFMO model, the RFCOMO model represents a more drastic change from the current system. Not only does the RFCOMO require delegation of authority from the central level to the regional level-as in the case of the RFMO via delegation of authority to the member states on condition that they exercise their power jointly-it also blurs the line between those being governed and those elected to govern. A lack of distinction such as this could pose particular challenges in terms of traditional good governance criteria such as accountability and transparency, as it might become less apparent who is actually doing the governing and in relation to some of those involved it will be difficult to hold them democratically accountable. Moreover, there is also a discussion to be taken in relation to how different stakeholder groups should be balanced. This discussion becomes much more delicate in the RFCOMO than in relation to the other models, where stakeholders are not awarded direct decision-making capabilities.
In terms of the geographical coverage and the interpretation of 'wide authority' the RFCOMO resembles the RFMO. Likewise, besides the potentially stronger buy-in of stakeholders to management measures than what can be expected by the RFMO, the RFCOMO potentially delivers the same benefits and suffers from the same weaknesses, as further expanded above.

Regional marine management organisations
Compared to the RFMO and the RFCOMO, the distinctive feature of the Regional Marine Management Organisation (RMMO) model, which also suggests the setting up of a regional organisation, is that it broadens the perspective of the possible regional organisation from dealing exclusively with management of fisheries to managing several or all activities of a specific regional marine area. Establishing this kind of organisation attempts in particular to tackle one of the coming years' major challenges, namely that of implementing a more holistic approach to environmental management.

Regional Marine Management Organisations
Under this model the member states would set up regional marine management organisations (RMMO) with responsibility for coordinating all matters relating to the regional sea areas. Stakeholders from all sectors would be involved in some form-either as advisors or in a more co-management-like structure. The RACs could continue to operate, but would only be providing advice as one of the affected sectors of the RMMO. Alternatively, the current RACs could be opened for a wider group of stakeholders. A general framework for regional approaches will be provided by the central EU institutions.
The advantage of this model is its holistic approach, which goes hand in hand with the ecosystem approach to management. Potentially, the RMMO would be a forum for coordination of all the interests that claim their right to the marine space in a particular region. Having an integrated organisation for this would be an advantage insofar that integration of policies would be much more manageable since it would only involve one organisation and not several organisations having to interact.
In terms of the actual authority of the RMMOs, the regional organisation would be tasked with 'coordinating' all matters relating to a particular marine region, which might be interpreted as somewhat weaker than the 'wide authority' of the previous two regional management organisation models. Moreover, when discussing the issue of stakeholder involvement, this archetype leaves that an open question.
A particular challenge of this model is that it only to a very limited extent builds on the current system. As suggested by van Hoof, van Leeuwen and van Tatenhove (2012) elsewhere in this issue, a way forward could be to establish regional bodies by merging the RACs with the Regional Sea Conventions, e.g. HELCOM (the Helsinki convention for the Baltic Sea) and OSPAR (Convention for the North Sea). These organisations are presently not part of the EU system but play nonetheless an active role in relation to implementation of the MSFD.

Cooperative member state councils
The last model in our selection of regionalisation archetypes, the Cooperative Member State Council (CMSC) model, is the one that necessitates the least change to the current CFP governance structures. In fact the CMSC model can in principle be set up without having to formally reallocate authority in the governance system at all.

Cooperative Member State Councils
The institutional structure and formal distribution of powers remains largely unchanged. However, the member states with fishing interests in a regional sea area establish mini-councils to deal with fisheries management issues specific to that area. These mini-councils forward their recommendations for formal approval to the overall EU Fisheries Council. The RAC would in most cases advice the mini-council rather than the central EU institutions. The exact extent to which stakeholders' input is given weight in the recommendations of the mini-council is up to that mini-council on a case-by-case basis.
As described, the formal distribution of authority and the institutional structure stays largely as it is in the current system. Consequently, the central EU institutions remain formally in charge of deciding on most fisheries management issues in EU waters. However, for each regional sea the member states with fishing interests there (or in a stricter interpretation: coastlines) set up on a more or less formalised basis mini-councils to provide the central EU Fisheries Council with recommendations on fisheries management issues specific to that area. The RACs would provide the regional mini-councils with stakeholder input and thus the RACs would continue to operate more or less in the current form, now with the member states of the regional mini-council and the Commission as the primary recipients of advice.
To provide significant advantages as compared to the current system, it is a precondition that the recommendations from the mini-councils are, as a general rule, not challenged by or renegotiated in the central EU Fisheries Council, but rather adopted as the position of the Council in the decision-making process. Consequently, this model is highly dependent on the presence of political will in the EU Fisheries Council and the Commission-as well as in the European Parliament-to allow for different approaches and accept that one region might favour a different path than another region.
The primary advantage of this model is the relative ease with which it could be installed. Under this model the EU central level would still apply management from the centre, but the exact nature of that management would de facto (but notably not de jure) be decided at the regional level. This would in practice allow member states sharing an interest in a specific region to develop their own strategies without having to worry about how the adopted approach will be received by all the other EU member states, which by default would opt out of having a say in management in the areas where they do not have a direct interest.
Although empowerment of stakeholders is not a specific aim of this model, it is nevertheless likely that a relationship could develop between the mini-council and the relevant RAC. Potentially this could reduce the feeling of distance between those making the decisions and those subjected to them, perhaps enabling stakeholders to propose measures that better fit the actual situation of the region in question.
In contrast to the various regional management organisation models, as an example, this model does not have the same potential in terms of delivering more timely management by excluding the central level from a number of more technical decisions.

Discussion
In the present article we have outlined a conceptual framework for aspects that need to be taken into consideration when discussing regionalisation of the CFP. The article unfolded in three parts, respectively concerned with: 1) developing a typology of CFP governance objectives, which regionalisation can be seen as contributing to; 2) outlining and discussing the three main problem dimensions to address when designing a regionalised governance system; and 3) the presentation of a representative suite of archetypes of regionalisation.
Our article takes outset in an understanding of three basic objectives of CFP governance, namely 1) the ability to take up and balance preferences in a way that is perceived as legitimate, 2) the ability to do this in an efficient way, and 3) the ability to deliver policies and measures that are effective. We subsequently divided each of these objectives into more-in some cases partly conflicting-sub-objectives.
Further, we argued that overall challenge in relation to designing regionalisation of the CFP can be related to three problem dimensions: the kind of authority that should be decentralised; the politico-administrative scale to where decentralisation should take place; and who should be the recipients of authority at any given scale and these dimensions have been unfolded in details in the framework.
In particular, in relation to the problem of who should be the recipients of authority at any given scale, the article offers a typology of how stakeholders can be involved in CFP governance, ranging from little to no involvement, over what is referred to as variations of co-management, to self-management. The CFP has, as previously described, traditionally been heavily top-down driven and even the latest CFP reform, which introduced RACs, did not move governance beyond a weak version of co-management-what we refer to as 'co-management by consultation' in our typology. A major opportunity related to future regionalisation, consequently, is to let regionalisation be not only about moving authorities between different politico-administrative levels but also moving authority outwards, towards users and stakeholders.
As an example, one could argue for a package solution with an inbuilt hierarchy-a funnel approach to representation, in which the EU level sets clear principles and longterm objectives, standards and frameworks, but where extensive formal mechanisms for consultation (and feedback on use of recommendations) with users and stakeholders exist, although all decisions are taken by EU/member states. The regional level could then develop implementation plans and guidelines tailored to regional conditions, and here stakeholder involvement could be through consultation as for the EU level or through the more ambitious 'co-management by partnership' , where regional member states, stakeholders and users cooperate in developing the implementation plans.
In the last part of the article, we explored ways of going about regionalisation in practice and presented a suite of so-called archetypes of regionalisation: five models that we based on our knowledge of the CFP believe capture important perceptions of ways to go forward towards regionalisation of the CFP's governance system. The five archetypes provide different ways of decentralising management authorities and vary in terms of stakeholder involvement or the type of co-management to become in place. In other words, each of the archetypes reflects integrated perceptions of how the questions of who, where, and what could be answered 'in practice'. As in many other situations, it is fair to say that the devil lies in the detail when it comes to designing regionalisation in practice. Consequently, it is very difficult with any degree of certainty to say how the different archetypes would contribute to the objectives of CFP governance. Anyhow, in the following article of this special issue (Hegland, Ounanian and Raakjaer 2012) more flesh will be put on the different perceptions of regionalisation by reference to both qualitative and quantitative empirical data.
On a general level, however, each of the models represents qualitatively different ways forward. The Nationalisation model represents an end to the attempt to coordinate fisheries management in the EU in an integrated way. The CMSC model, on its side, may represent a quick, pragmatic fix to the current system but does not constitute any great overhaul of the current system. The RFMO model is more ambitious in the sense of actually providing significant change of the current structures, but this approach also presents significant challenges in terms of how to set it up. The same can be said about the RFCOMO, which takes a step further in moving management closer to those affected. Finally, the RMMO may prove to be the most forward looking of our models as it as the only model puts the integrated marine management at the centre of the reform towards regionalisation. But, as discussed, the RMMO model is also a far step from what are the current structures.
Endnotes a It should be noted, however, that the RAC structure is not the only spatial structure applied for the EU waters. Of particular relevance for fisheries are the regions and subregions outlined by the Marine Strategy Framework Directive (European Parliament and Council 2008), an environmental policy initiative; see van Hoof, van Leeuwen and van Tatenhove (2012) in this special issue or, for a more general introduction to marine policies and maritime spatial planning in Europe (de Vivero et al. 2009). b Nevertheless, we should not neglect to recognise an underlying existence of a considerably more complex multi-level governance setting. A setting in which networks of both formal and informal interactions transgress the different institutions at the different level of the system as well as across, where various stakeholder groups work to exert influence on the various EU institutions also outside the remit of the official advisory structure, where national civil servants and politicians meet in formal and informal working groups organised under the Council or the Commission, et cetera. c It should be emphasised that the typology outlined here is relatively simple; particularly in terms of operating with only three basic objectives. The primary reason being that the first objective of our typology with its focus on take-up and balancing of preferences effectively collapses and includes a range of what could be seen as more detailed 'good governance' criteria, such as for instance of governance being transparent, inclusive, participatory, and so on. d It should be acknowledged that authority, power, et cetera, can be both formal and informal. This distinction is of significant importance in the context of the EU, as recent years has evidenced an increase in the use of more informal ways of moving towards increasingly ambitious objectives-often referred to as soft law or new modes of governance-compared to the community method, which the traditional operating mode of the CFP is an example of. e However, it has also been argued that many other factors influence compliance and that the degree of user participation might not be the most important (Raakjaer Nielsen and Mathiesen 2003).
f As an example, in several member states fishing activities are considered part of a cultural heritage and something that attract tourists; likewise there might be regional concerns that are not easy to calculate. g We acknowledge that this typology is a simplification of a very complex situation.
There is a multitude of tasks that can be managed under different types of institutional arrangements and at different stages in the management process. Moreover, we have decided to refer to the typology as one of 'stakeholder involvement' , as the two outliers are hardly co-management variations. h It should be noted that the RFMO archetype is not equivalent to the organisations that are commonly referred to a Regional Fisheries Management Organisations, such as the North-East Atlantic Fisheries Commission (NEAFC) and the International Commission