Abstract
This article examines the interplay between relational and structural factors in establishing the legitimacy of social governance processes at the European Union (EU) level, specifically the peer review process in the Social Open Method of Coordination (OMC). Through a multi-level governance lens, the article uses network analysis to evaluate the roles of different actors within the Social OMC governance structure and links that with the ideas of input and throughput legitimacy. The findings suggest that the peer reviews have limited capacity in improving input or throughput legitimacy. Input legitimacy has improved somewhat to include non-traditional actors, but the governance structures in place do not provide a clear role for these new actors and it remains a mostly government-driven process. While the Social OMC does open up new avenues for transparency, information sharing and improving procedural throughputs, the process does not have a clear vision in how to translate this transparency into significant increases in legitimacy.
The peer review process in the Social Open Method of Coordination (Social OMC) is a non-binding policy tool that aims to foster dialogue, coordination and mutual learning between European Union (EU) member states on various social policy issues. However, this focus on the coordinating and facilitating functions of the Social OMC underplays the effects of its policy instruments on EU legitimacy. As an actor-driven tool, the process can serve a legitimating function by opening up the policy process to discussion and debate involving different state and non-state actors, thus bolstering EU-level governance processes in general, and aiming to improve practices in social policy more specifically. Given the actor-driven nature of the process, analysis of the effectiveness of the peer reviews as a legitimating tool must take into account not only the underlying structures driving the process but also the relationships between actors that determine how the process works in practice and how it works over multiple political levels.
This article utilizes a multi-level governance (MLG) framework to examine the configuration of the Social OMC peer review networks in establishing power and coordination in social policy and its effects on legitimacy at the EU level. The Social OMC process as a whole is vast and complex, and as such this article focuses only on the peer review process. This allows for a deeper examination of the relationship between legitimacy and power in non-binding policy instruments. The research provides deeper institutional understanding of non-binding governance processes and the EU’s role in social policy, making three theoretical or analytical contributions to the debate surrounding legitimacy and power. First, it provides nuance to the interplay between different types of legitimacy by expanding on the importance of throughput legitimacy in non-binding processes. Second, it provides a suitable case for deepening the conceptualization of MLG in a way that encompasses both structural and relational factors. Finally, it applies these theoretical considerations of legitimacy in multi-level systems in a relatively understudied tool of the EU’s Social OMC.
The central questions of this article are as follows:
In what ways are Social OMC peer review processes shaped by political structures and actor relations?
How do these structural and relational factors affect the legitimacy of this governance process?
The article will utilize qualitative methods and network analysis of the peer reviews to examine the structural and relational factors that shape the peer review process, along with the potential effect of these factors on legitimacy. The first section will briefly outline the theoretical and methodological underpinnings of the concepts explored in this article and develop a framework of analysis for policy-making in multi-level contexts that takes into account both structural and relational factors. Then, the article will look at the structural base and the role of actors in the Social OMC peer review process, and how these factors shape MLG and legitimacy. This will lead into an examination of the substance of the peer reviews and the ways in which the peer reviews enhance legitimacy. Finally, this article will examine the implications of these findings on non-binding EU forms of governance and the benefits and trade-offs these forms have in terms of creating EU-level legitimacy, a key concern in the literature (Kröger and Friedrich, 2011). The research shows that the Social OMC has allowed for a more flexible structural approach to governance, but has not significantly altered the hierarchical relations between actors in the process, which in turn limits the legitimacy of the process in significant ways.
Theoretical approach
MLG
MLG developed as a way to understand EU integration processes that stretched up to supra-national levels while also incorporating regional levels and new actors into the process. Since its EU origins, it has been expanded into new areas and contexts (Conzelmann, 2008), but it is typically viewed as a dichotomy between federal-type structures with clearly defined jurisdictions, little overlap and stable relationships between levels (Type I) and systems with overlapping jurisdictions, fluid, function-specific and sometimes unclear connections and ever-changing relationships between the levels (Type II) (Bache and Flinders, 2004; Hooghe and Marks, 2001). However, use of the term has become muddied in the 20 years since it was coined as authors tried to deal with both increasing issue and institutional complexity. The concept itself has evolved from its original basic conceptualization towards more functional, comparative and normative uses, including consideration of legitimacy (Stephenson, 2013).
However, ideas of structure and relations between actors can further be separated out from Hooghe and Marks’ approach. Governance structures may change without altering the actors involved, and while the number of actors may increase in either Type I or Type II of MLG, this does not necessarily mean that the structure underlying these relations will change as actors may be given more or less autonomy within existing structures (Sørensen, 2006: 99). Some work has been done, mainly from a policy perspective, in separating out structural and relational factors in order to deepen conceptualizations of governance (e.g. Curry, 2015; Howlett et al., 2009; Offe, 2006; Treib et al., 2007; Weaver and Rockman, 1993), and this work separates Hooghe and Marks’ dichotomy into a two-axis framework of analysis, as shown in Figure 1.
A framework for understanding multi-level governance processes. (Curry, 2015: 34) reproduced with permission of Palgrave Macmillan.
This framework differentiates between structural factors (rigid vs flexible structural design) and relational factors (hierarchical vs heterarchical relations). Structures, in this context, refer to institutionalized processes that affect governance. The dichotomy presented on this axis refers to the flexibility of the political structure in allowing actors to deal with issues as they see fit. This flexibility may arise through the use of policy instruments, differences in policy implementation or other factors, while rigidity usually results from a strict adherence to statutory and legal procedures. The flexibility afforded these processes in turn affects their connection to legitimating functions, inasmuch as long-term legitimacy is harder to develop given shifting and uncertain rules of the game. These structures can develop and change over time and range from highly formalized and rule-driven (e.g. constitutional edicts) to semi-rigid structures (policy instruments) to informal, but still institutionalized ideas (constitutional conventions or traditions). Hay (2009) further differentiates between structures that can be shaped by the actors within the structure and those that do ‘not include the opportunity to reconfigure the rules governing the operation of the system itself’ (p. 266). The Social OMC, as an actor-driven process, clearly fits into the former type. The hierarchy/heterarchy relational axis refers to the spectrum of ultimate accountability and responsibility over decisions and the level of autonomy from oversight by other actors. This is developed through the relations exhibited between the different actors involved in the process. Heterarchy arises in situations where actors are able to operate freely, autonomously and in the way they best see fit within the bounds of their power. Hierarchy exists in situations where actors clearly control (or are controlled by) other actors or when they operate at the discretion of another actor. Hierarchy does not need to be blatantly exercised to exist, as a shadow of hierarchy may still be present if there is the possibility of this control to be exercised through various formal or informal means (Héritier and Lehmkuhl, 2008; Smismans, 2008).
This approach does not position the Social OMC and its tools (including the peer reviews) as an alternative mode to MLG, as some authors contend (Stephenson, 2013: 823). Instead, it shows that ‘heterarchical’ and ‘decentred’ processes (Hodson and Maher, 2001: 719) such as the Social OMC do not conflict with an MLG framework. Indeed, the Social OMC has significant multi-level components as a mode of governance whose aim is to coordinate between actors at all levels. Michalski (2012) draws a clear link between the Social OMC and interactions with other governmental levels in federal and unitary, centralized and decentralized states. The author notes how the relative multi-level nature of member states has a significant effect on how they engage with the Social OMC, as well as how they involve sub-national authorities or other stakeholders (such as parliaments) (Michalski, 2012: 413–416).
Input, output and throughput legitimacy
As MLG systems will involve multiple and discrete actors with differing levels of power, along with a variety of structural arrangements, the legitimacy of governance processes also becomes more complex. Legitimacy may come in three forms: input, output and throughput, which refer broadly to participation, performance and process, respectively (Schmidt, 2013). Input legitimacy, where more actors are allowed to take part in political decision-making, opens up participation to make policy accountable to more stakeholders (including citizens). However, if this opening up of the process favours unelected or unrepresentative stakeholders, it can have a negative effect on perceptions of legitimacy by moving it further from the citizenry. In output-based legitimacy, the emphasis is on the legitimacy of the performance (Scharpf, 1999). Output-based legitimacy is improved if the interests or goals of more stakeholders are met by a certain policy outcome. However, output legitimacy can be exclusionary and limit the role of stakeholders in identifying issues and developing approaches to deal with them. Finally, throughput legitimacy looks at consultation and the efficacy, accountability and transparency of governance processes in crafting policy. Although trade-offs and complementarities exist between input- and output-based processes (increasing one may hamper the other, or limitations in one may be offset by strengths in the other), throughput legitimacy often operates on a parallel level. Increasing throughput legitimacy does not necessarily offset limitations in input or output legitimacy, but reductions in throughput legitimacy may act to undermine the other two types of legitimacy (Schmidt, 2013). While the different types of legitimacy have been extensively analysed in EU literature (e.g. Holzhacker, 2007; Lindgren and Persson, 2010; Meunier, 2003; Moravcsik, 2002), the interplay between these three types of legitimacy remains an area that can be more thoroughly explored.
Non-binding forms of EU governance that aim to operate transparently and broaden participation in order to improve outcomes, such as the Social OMC, carry the promise of addressing real or perceived shortfalls in all three types of legitimacy, and these approaches tie into academic discourse on MLG. Different MLG processes may minimize the role of hierarchy in decision-making, and by allowing more actors into the process, more people and groups gain a voice in the decision-making process, thus (theoretically) improving input legitimacy. In addition, horizontal links between actors help to improve their influence over the process, and deliberation and transparency (throughput legitimacy) can be increased through relational or structural factors as well. Finally, these governance processes can provide a supplementary form of engagement above and beyond traditional channels of political participation, which in turn should improve policy output (Sørensen and Torfing, 2009: 244). However, this raises questions of what actors are considered legitimate and whether ‘too many cooks’ or overly influential actors (through resources, political clout or other means) can crowd out other voices. The ‘multi-levelness’ of the process may hamper democratic legitimacy of all types by clouding visibility of decision-making and moving the policy process further away from representative institutions (Papadopoulos, 2010; Suškevičs, 2012), producing ‘a complex structure of interlocking institutions and procedures, designed to generate consensus and obscure asymmetries in power and influence’ (Brzezinski, 1997: 27). This legitimacy gap can become pronounced when actors in the policy process may be several steps away from being directly accountable to the public. Although the main concern of the peer review and the Social OMC as a whole is not legitimacy, it inevitably adds to that debate.
Given the above-mentioned framework for MLG and the nature of the Social OMC, two hypotheses can be developed about the role of structural and relational factors in the peer review process and how these may act to affect the perceived legitimacy of the process. As the level of analysis focuses on the EU level, only effects on input and throughput legitimacy will be assessed:
H1. Peer reviews will affect actor relations, provide a new avenue for participation of actors and a shift in how these actors interact, allowing for movement towards a more heterarchical approach to policy-making. This in turn will affect perceptions of input legitimacy.
H2. The introduction of new structures of engagement will allow for more clarity in social policy processes at an institutional level while also increasing the flexibility and choice in how these policies are addressed. This in turn will affect perceptions of throughput legitimacy.
Peer reviews and the social open method of coordination: a background
While other authors have written on the Social OMC in general (see, for example, Borrás and Jacobsson, 2004; Borrás and Radaelli, 2010; De la Porte, 2007, 2011; Tholoniat, 2010; Visser, 2009; Zeitlin and Pochet, 2005), this work, along with work by Kröger (2007), Büchs (2008) and Duina and Raunio (2007), broadens the perceptions of the Social OMC beyond its traditional coordinating and facilitating role to thoroughly consider its role as a legitimating tool. It expands on earlier research using novel methodological techniques (network analysis) to assess legitimacy and examines procedural implications of the Social OMC on throughput legitimacy. Finally, it examines legitimacy in the Social OMC through the lens of structural and relational factors, allowing for a more nuanced examination of the Social OMC process that takes into account both political institutions and agency. The European Commission as an unelected body lacks direct input legitimacy and has proven eager to try to engage civil society in new ways to deal with policy under its control (Kohler-Koch, 2010).
The OMC was launched at the Lisbon Council in its current form in March 2000 (European Commission Communication COM/199/347 to the Counciile, the European Parliament, the Economic and Social Committee and the Committee of the Regions, 1999) and initially aimed to identify and promote ‘the most effective social issues’ within a European context (Eurostat, 2012). Efforts to strengthen and consolidate the Social OMC process were undertaken in 2005–2006 and culminated in a new Commission Communication in 2008 that aimed to improve mainstreaming, horizontal coordination and analytical tools in developing a more coherent idea of social policy at the EU level (European Commission Communication COM/2008/0418 to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, 2008). Some research has examined the role of the Social OMC in EU governance (Zeitlin and Pochet, 2005) and mostly concluded that the Social OMC lacks the clear vision and authority necessary to effect policy change (Kröger, 2009; Radulova, 2007; Song, 2011). Others see the Social OMC as a ‘transformational’ form of government in which binding and non-binding initiatives interact (Dawson, 2011). While both of these are potentially valid arguments, this article contends that highlighting the Social OMC’s role as simply a coordinating and facilitating instrument underplays important questions about the role of its policy instruments in affecting EU legitimacy.
Social policy is an area where the EU does not have any formal jurisdiction, so the main policy tools at its disposal are non-binding, including policy indicators, joint reporting, shared objectives and sharing of good practice. It is on the latter that this work focuses, in the form of the peer review process. Peer reviews were established in 2004 under the EU Social Inclusion Programme and broadened in 2005 to include pensions, health and long-term care. The peer reviews are a sub-programme of PROGRESS, the EU Community Programme for Employment and Social Solidarity (European Parliament and Council Decision No 1672/2006/EC, 2006). The overall aims of the peer reviews are as follows: To contribute to a better understanding of the Member States’ policies, as laid down in their National Reports on Strategies for Social Protection and Social Inclusion and of their impact; To improve the efficiency and effectiveness of the policies and the strategies for social inclusion, pensions, healthcare and long term care in present and future Member States and at EU level, by learning from the experiences in the Member States; and To facilitate the transfer of key components of policies or of institutional arrangements, which have proved effective in their original context and are relevant to other contexts. (DG Employment, Social Affairs and Inclusion, 2010: 5) Ensure greater involvement in peer reviews of officials at local and regional levels. (European Commission Communication COM/2008/0418 to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, 2008)
The peer reviews are hosted by one member state and typically attended by seven or eight other member states, as well as EU-level stakeholder networks (non-governmental organizations (NGOs)) and the European Commission. Usually, 30–40 people are involved, including government representatives, independent experts, NGO representatives and EU officials. The peer reviews provide a forum to discuss good practice examples in that country and in other participating nations, as well as national-level issues, how these policy examples contribute to wider EU goals and how (or whether) they can be applied in other country contexts. While most peer reviews of policies are undertaken ex post, the process also allows for ex ante evaluation and information sharing if a country is looking to undertake future reforms.
The peer reviews should be well integrated with other aspects of the Social OMC process. The literature on EU forms of governance such as the Social OMC has stressed the importance of peer review processes (and other forms of sharing and reporting) in ensuring accountability and connectedness for all levels in designing and implementing policy (Sabel and Zeitlin, 2008). Other analyses of the peer review process have shown somewhat mixed results from the instrument. They have been shown to provide policy learning from best practices and to be effective in information sharing. However, this did not translate into engagement in the policy debate at the national level and did not often result in long-term interaction with other participants (Public Policy and Management Institute, 2012). However, given the fact that participants are drawn from both EU and member state levels, there is potential for development of networks across levels (Agostini et al., 2013: 20).
Methodology
Network analysis
Governance as a concept can be understood as a series of networks, with EU governance in particular explainable as both governance by networks and governance in networks (Börzel and Heard-Lauréote, 2009), and network analysis provides a way of analysing these connections (Henning, 2009). Network analysis focuses on the relationships between actors as a way of explaining different phenomena, creating a way to clearly, numerically establish the roles, connections and overall nature of actor relations in fitting into and perhaps shaping structural processes (Krackhardt and Raider, 2001; Wasserman and Faust, 1999; Wellman and Berkowitz, 1997).
Network information on the Social OMC peer reviews was collected from participant lists and synthesis reports of all peer reviews between 2004 and 2012 in order to develop actor configurations (input legitimacy). Peer reviews are voluntary, so the number of peer reviews by year and by country varied, but all member states have taken part in the process. Participant lists were encoded into the network analysis software UCINET (Borgatti et al., 2002) to analyse the networks present in the peer reviews, and actors were considered to be directly connected if they attended the same peer review. Networks of individual peer reviews were analysed, looking at actors’ relations with other actors in each specific event. The overall peer review network over the 9-year period was also analysed, which assessed individual actors’ connections to all events over time. Finally, individual actor (ego) networks were examined to see how individual actors in the process connected with other actors in all events. These three types of network analysis provided evidence of (a) actor connections within specific peer reviews, (b) overall actor connections to all peer reviews and (c) individual actor connections with other actors. This provides evidence of shifts over time as well as more in-depth analysis of individual peer reviews and participants. UCINET was used to check for general measures of cohesion (density of networks), centrality (both direct and indirect connections and distance between actors) and core/periphery-ness of actors, as well as measures to establish whether any groupings of actors were formed (factions and cliques). In order to gauge transparency and deliberation in the process (throughput legitimacy), all synthesis reports were reviewed to determine key themes of the peer reviews, as well as a general idea of issues and approaches to social policy that were discussed. The network analysis and the qualitative data were then linked to provide a clearer indication of participation patterns within the peer review process, information developed during the peer reviews and implications of these factors on input and throughput legitimacy.
Findings 1
Relational factors
Three key indicators in the peer review process can be used to assess the nature of relational factors: governmental and non-governmental participation in the process, iterative interaction of actors over time and the type of actor involvement promoted in the peer reviews themselves. Strong participation by both governmental and non-governmental actors in a manner that facilitates ongoing, repeated interactions and relationships will help to improve the input legitimacy of the process. This can be used to ascertain the validity of Hypothesis 1.
There were a total of 71 peer reviews between 2004 and 2012, for an average of 8 peer reviews a year, with a low of 7 in 2005 and a high of 10 in 2009. In a 9-year period, all 28 EU member states (including Croatia, which joined in 2013) attended peer reviews, as did Norway and Serbia. Looking at the overall peer review network in terms of governmental participation, 20 of 28 member states hosted peer reviews, with France and Germany hosting the most (each hosting eight), followed closely by Belgium (seven) and Spain (five). All countries apart from Slovakia and Croatia attended at least 15 peer reviews (Slovakia attended 10 and Croatia attended 13), with Finland and Luxembourg attending the most (22 each). In terms of non-governmental participation, 33 NGOs representing a variety of interests attended peer reviews, with 14 of these organizations attending multiple peer reviews (the European Anti-Poverty Network attended the most, at 20 peer reviews).
Looking at specific governmental actors shows that France and Germany were the most central actors to the process overall, but there is some fluctuation when other factors are taken into account. When looking at direct connections to other member states and NGOs, Belgium proved to be the most central actor in forging these direct links, with Germany, Spain and the United Kingdom also establishing numerous direct connections. Combining hosting and participation into a ranking of connections, Germany proves to be the most central country to the process, followed by France and the United Kingdom. In general, there was a trend for the EU-15 to be more actively involved in the peer review process than the new (2004, 2007 and 2013) accession countries, and the process was very much member state driven with little input from other lower levels of government. In terms of non-governmental actors, they were most active in the area of homelessness, and the European Anti-Poverty Network was the most central participant, followed by the European Federation of National Organizations Working with the Homeless (FEANTSA), AGE Platform Europe, Eurocities, the European Social Network and Eurochild. However, overall non-governmental actors were more fragmented in terms of participation, and notably, those central to the process tended to be large, umbrella organizations. The centrality of central state actors and large, umbrella NGOs points to the continued dominance of traditional actors in the process.
The evidence is more mixed on whether this participation facilitated ongoing connections and iterative interaction between actors. Looking at the networks in specific peer reviews, countries attended peer reviews strategically. Certain groupings of like-minded countries were especially evident on issues such as integration of ethnic minorities and immigrants, homelessness and housing exclusion, and health and long-term care. For instance, new countries of immigration (Czech Republic, Greece and Spain) were the most likely to engage in peer reviews on integrating ethnic minorities and migrants. There was some weak division in attendance based on type and topic of peer review. One group of countries slightly favoured peer reviews based on process 2 – those focused on best practice examples, governance and inclusion issues – while the other favoured those focused on general policy issues. 3 While there were weak actor correlations on topics and approach, the key finding in relation to actor alignment is actually the lack of a discernible pattern between actors over time. While connections in specific peer reviews may be forged, these do not appear to lead to any sustained coordination. In addition, most peer reviews focused on best practices of policies that had already been implemented, rather than seeking insight into new processes, which highlights the primacy of state-led interventions over EU-level coordination or collaboration. On one hand, the exchange of good practice is possible on an individual peer review basis, and case-by-case involvement ensures that member states can engage in areas that are particularly relevant to them. On the other hand, constantly changing participation may lead to shallower engagement and hamper deeper coordination across policy areas (and across the EU as a whole).
Finally, it is also helpful to consider the nature of actor involvement that was promoted in the peer reviews, and this highlights an important multi-level component to the process. Over the 9-year time period studied, there were numerous peer reviews that addressed local- or regional-level policy initiatives, as well as geographic issues affecting social policy provision. While this is understandable due to the federal nature of many EU states, it also points to an underlying understanding of the importance of MLG issues in addressing social policy. However, regional approaches declined in prominence over the assessed time period. In contrast, the perception of the role of non-governmental actors in the policy process changed over time, with a shift from seeing these actors as mere service providers towards a more active engagement with them in the policy process as a whole. This deeper procedural engagement of non-governmental actors can help to legitimate the process (if the actors are representative), but the lack of engagement with regional- and local-level actors undermines this increase in inclusiveness and legitimacy. The prominence of these issues over time indicates that peer review participants are aware of the necessity of involving different types of actors over multiple governmental and policy levels in governance processes, but this awareness was not necessarily met with an actual increase in procedural involvement of these actors. The lack of deep and consistent actor networks, coupled with the lack of significant engagement of non-traditional actors, undermines the peer review process as a structure of input legitimacy.
Structural factors
Measurement of throughput legitimacy focuses on structural aspects of the peer review process. Ideally, processes are legitimized in terms of throughputs by processes that are transparent and open in a way that clearly links up to (and has a potential effect on) the policy process. This transparency and openness can be ascertained in two ways: by examining the themes and ways in which actors engaged with the peer reviews and by examining the information created and disseminated by the peer reviews. These structural factors can confirm or refute the second hypothesis of this work.
The key themes of the peer reviews fit broadly into either policy-specific issues 4 or process-related issues, 5 although many combine policy and process. In all, 61 percent of peer reviews focused on specific policy areas, although there was a move over time towards a greater focus on procedural topics. There was a significant uptick in peer reviews focused on governance, for instance, with half of the peer reviews in this area coming in 2010–2012. This does indicate a maturation of the peer review process as there is a move from specific (and often ungeneralizable) policy issues towards a deeper procedural understanding of social policy problems that can be used in different contexts, which may improve its standing as a throughput legitimating tool. Over time, peer reviews became more self-reflective as well, often referencing previous peer reviews, and issues more often cut across policy areas. Still, while there was general agreement on what issues needed to be addressed as regards social policy, there was little consensus on how these social policy issues should be addressed. This could indicate a shallower type of coordination and a failure to link policy outcomes to the process. Ultimately, member states appear to engage in the peer review process beyond simple dialogue only insofar as it supports national-level change that was already planned, which supports earlier findings (Büchs, 2008).
The peer review process also serves an important information and best practices sharing role, identified in other studies of the process (Agostini et al., 2013; Public Policy and Management Institute, 2012). From 2008 to 2012, 998 documents were produced in 71 peer reviews, averaging over 14 documents per peer review. For each peer review, a synthesis report, a short report, a discussion paper and comment papers from all participating member states and other stakeholders are produced. However, there have been significant problems in disseminating this information after the peer review ends, and this dissemination rarely extends beyond the organizations (or ministries) taking part (Public Policy and Management Institute, 2012: 12–13). The sheer volume of reporting leads to concerns of ‘information overload’, which has been experienced in other areas of member state and EU-level policy processes (Brownson et al., 2006; Dunlop, 2010; Mabbett and Schelkle, 2009).
Taken together, these measures reveal a relatively cohesive but indistinct network within the peer reviews. There were many interconnections, but no noticeable cliques or factions developed based on actor or subject groupings, which may indicate that participation is linked more to stakeholder interest in topics rather than an overarching desire to network or more deeply embed this form of governance. The core actors unsurprisingly consisted mostly of member states, along with DG Employment (Directorate-General for Employment; the organizers of the peer reviews) and one NGO – the European Anti-Poverty Network. In addition, the peer review process did not involve local or regional governments in significant ways (Dawson, 2009). This shows that the peer reviews have allowed only limited room for non-traditional (regional and non-governmental) actors and that mainly traditional actors were able to be central to the process.
Legitimating effects of new governance processes in Europe
The Social OMC has two main ways of potentially increasing legitimacy at the EU level (Table 1). First, by involving more stakeholders, it can theoretically improve input legitimacy by giving voice to more groups. Second, the Social OMC operates as a source of information and dialogue, opening the door to more throughput legitimacy by increasing the information available to the public, thus theoretically increasing transparency of the process. This throughput legitimacy also requires that actors take advantage of that information and use it to shape policy outcomes. In both cases, the potential of the Social OMC to increase legitimacy has only been partially realized.
Key findings.
NGOs: non-governmental organizations.
On paper, the peer review process opens up new opportunities for actor involvement, and these opportunities are taken by many central governmental and some non-governmental actors. By involving more actors in the process and creating relatively dense social networks in various policy areas, the process does increase the potential for improved input into the process. However, as shown through the network analysis, the involvement of new actors is limited. There is no clear arrangement to include sub-national actors, and consideration of regional-level approaches has actually lessened over time. However, participation of non-governmental stakeholders has increased, albeit in a way mostly restricted to large, umbrella organizations. Therefore, its success in increasing input legitimacy is partial at best as smaller non-traditional actors are still mostly shut out of the process, involvement requires proactive engagement on NGOs’ parts and hierarchically, member states remain very much in control. Therefore, Hypothesis 1 remains unproven, as there has not been much movement towards a more heterarchical approach to policy-making.
The peer review process also improves the clarity in how social policy issues are addressed at the EU level, which in turn increases the potential for improved throughput legitimacy. Indeed, the sheer quantity of information produced by the peer reviews indicates an opening up of deliberation and transparency of the process, but this considerable amount of data can also lead to information overload and difficulty for actors to process and internalize it. Still, transparency of the process is increased through the widespread dissemination of information. In addition, the increased focus on procedural issues in the peer reviews indicates a more open and flexible approach to social policy, one that does consider new approaches to governance in that policy area. Therefore, Hypothesis 2 is proven, as the peer review process has helped to clarify social policy processes at the EU level and resulted in a more flexible approach to process in crafting social policy.
Taken together, however, a true increase in legitimacy brought about by a non-binding, deliberative form of governance must be one that allows traditional political actors, new ones emerging from civil society, and coalitions among these to contest official proposals against the backdrop of much richer information about the range of arguably feasible choices, and better understanding of the argument about their merits, than traditionally available in domestic debate. (Sabel and Zeitlin, 2010: 8)
As shown, couched in those terms, evidence of overall effectiveness of the peer review process is somewhat mixed as the process has only increased structural clarity and flexibility while hierarchical actor configurations remain largely unchanged. This is illustrated in Figure 2.
Effects of peer reviews on EU multi-level governance processes.
While the increase in throughput legitimacy can be seen as positive, its usefulness is less assured as it does not act as a complementary force to input legitimacy. In this instance, the throughput legitimating effects of the Social OMC do not affect input legitimacy and instead act as a parallel process that augments but does not connect with traditional input ideas of legitimacy. As Schmidt (2013) puts it, and this article shows, throughput legitimacy does not make up for problems with input legitimacy (p. 18). This is supported by other research on the Social OMC, which shows that any ‘deliberative polyarchy’ developing from the Social OMC is often driven by national-level processes, rather than any facilitation created by the EU process itself (MacPhail, 2010), and there is little indication that the peer review process has allowed for a more heterarchical approach to social policy-making. The process remains member state driven, and even at the national level there is often ‘depoliticization’ of the process that precludes involvement outside of member state elites (Michalski, 2012). It is possible that other aspects of the Social OMC such as the development of indicators may allow for more movement along the hierarchical/heterarchical axis, but the peer review has only created more flexibility in structural approach while maintaining a shadow of hierarchy (Héritier and Lehmkuhl, 2008).
Conclusion
This article uses a modified conceptualization of MLG to highlight the effects of structures and actor relations on legitimacy in the Social OMC peer review process. This adds to the academic debate by clarifying the interplay between a ‘traditional’ idea of input legitimacy and the throughput legitimacy created by governance processes such as the peer reviews. In addition, it provides a more nuanced understanding of the role of actors and institutions in a relatively understudied policy tool used by the EU. The findings show that while the peer reviews have acted to increase throughput legitimacy through moderate structural change, this has not translated into a change in how actors are involved, which has limited the input (and overall) legitimacy of the process. Other research has shown input legitimacy to be closely linked to policy outputs and their legitimacy (Suškevičs, 2012: 235), which further calls into question the effects of the peer review process. These findings have a significant impact on social policy as viewed from the EU level, as limited movement in actor configurations not only affects legitimacy but can also hamper the coordinating effects of the Social OMC and the effectiveness of EU-level social policy processes.
In looking at the peer reviews, the flexibility of the non-binding governance process created by the EU has allowed for the potential introduction of new actors to the process. However, it is less clear to what extent these actors are involved, and involvement of smaller regional and non-governmental actors is still limited. While the peer reviews do have an effect on improving transparency and clarity of the governance process, this transparency has not resulted in a shift in other types of legitimacy or a change in power and actor structures. The peer reviews have created a more flexible structural design, but there has been significantly less change in how new and old actors engage with these structures to affect social policy. This mismatch between input and throughput legitimacy supports a potential ‘diagnostic deficit’ (Schout et al., 2010), where a clear direction for the peer review process, at least as regards its legitimacy, has not been established.
Still, new forms of non-binding governance processes should not be seen as useless. Deliberation under the peer review process improves throughput and can have subtle or overt effects on social policy choice at the member state level, even if these effects are hard to isolate (Sabel and Zeitlin, 2010: 2). In addition, there is some evidence that findings from peer reviews are increasingly being diffused (albeit in a limited fashion) beyond the immediate participants (Public Policy and Management Institute, 2012). The peer reviews allow the EU some discretionary control over aspects of social policy (namely, dialogue) that the member states are willing to cede. Still, this power remains wholly contingent on member state support and lacks influential clout, making it an incomplete solution. While the process has, to a great extent, been institutionalized, it has not necessarily been translated downwards to ‘couple’ with national-level policies (Heidenreich and Bischoff, 2008). In addition, depoliticization of the EU-level process within member states has also created questions of its legitimacy at the national level (Michalski, 2012). This leads to the bigger question of where the Social OMC can be placed in closing the perceived legitimacy gap in the EU as the peer reviews create an intermediate level of legitimacy that only partially increases social policy engagement at the EU level.
Footnotes
Funding
The author(s) received financial support from the Marie- Skłodowska Curie EU H2020 action, grant agreement No. 661479 (Project MLGPIL), Socio-economic Sciences & Humanities, for the research, authorship and/or publication of this article.