University of Amsterdam
Graduate School of Social Sciences
Master thesis
The European Union and Social Dialogue:
Orchestrating Transnational Labour Regulation
Political Science. Specialization Track: International Relations
Research Project: Transnational Advocacy
June 23, 2017
Author Supervisor Second reader
B.A. David Tischer Dr. C.M. Roggeband Dr. P. Schleifer
11232978
1. Introduction ...1
2. Theoretical Framework ...3
2.1. Transnational New Governance and Private Standard Setting ... 4
2.2. Orchestration: A regulatory strategy ... 6
2.2.1. Intergovernmental Organizations as orchestrators ... 7
2.2.2. Properties and mechanisms of orchestration ... 8
2.3. The European Union and orchestration ...10
3. Research Design ... 11
3.1. Cases and justification ...12
3.2. Methods ...13
3.2.1. Policy documents ...14
3.2.2. Stakeholder interviews ...14
3.2.3. Document analysis in Atlas.ti 8.0 ...15
3.3. Data ...15
4. Analysis and Results ... 17
4.1. European Social Dialogue ...18
4.2. The European Social Dialogue in action ...20
4.3. The European Commission and the European Social Dialogue ...22
4.3.1. Support ...23
4.3.2. Agenda setting ...24
4.3.3. Intermediary influence ...25
4.4. The regulation process ...26
4.4.1. Categories of regulation ...27
4.4.2. Limitations to transnational regulation ...28
4.4.3. Target influence ...28
4.5. Lack of capacities and organisational self-empowerment ...29
5. The O - I - T relationship revisited ... 31
Documents ...38 Interview questionnaire ...39 Transcripts ...40
BF Business federation
CAQDAS Computer-assisted qualitative data analysis CEEP European Centre of Employers and Enterprises CSR Corporate social responsibility
EC European Commission EEA European Economic Area ESD European Social Dialogue ESP European Social Partners ERN European regulatory network
ETUC European Trade Union Confederation
EU European Union
H&V Harassment and violence at work IGO Intergovernmental Organisation ILO International Labour Organisation IMF International Monetary Fund MNC Multinational company
NGO Non-governmental Organisation O-I-T Orchestrator – Intermediary – Target
TFEU Treaty on the Functioning of the European Union
TU Trade union
UEAPME European Association of Craft, Small and Medium-sized Enterprises UNEP United Nations Environment Programme
Abstract
THE EUROPEAN UNION HAS ESTABLISHED THE EUROPEAN SOCIAL DIALOGUE TO FOSTER TRANSNATIONAL REGULATION OF LABOUR AND SOCIAL POLICY. THIS SOFT AND INDIRECT MODE OF GOVERNANCE IS BEING DESCRIBED AS ORCHESTRATION. MY QUALITATIVE RESEARCH SETS OUT TO SPECIFY THE RELATIONSHIPS, MECHANISMS AND OUTCOMES OF THE O-I-T MODEL BY OBSERVING CASES ON HARASSMENT AND VIOLENCE AT WORK AND TEMPORARY AGENCY WORK.I FIND THAT THE POWER RELATIONSHIPS CAN DIFFER FROM THE ASSUMPTIONS AND FAVOUR THE INTERMEDIARY.THIS LEADS TO SUBOPTIMAL OUTCOMES IN THE REGULATION PROCESS.
1. Introduction
Traditional regulation has been challenged through the process of globalization.1 Domestic government regulation that is characterized by clear jurisdiction, great capacity, and clear targets2 is losing importance. Instead, many problems of governance such as nature protection, security and migration have diverged onto the transnational level. In the field of labour regulation change has come through the emergence of multinational corporations who access substantive global value chains. As market integration led to intensified global competition in many industries, the power relations between private firms and domestic labour market regulators changed, diminishing domestic governments ability to regulate and monitor. To address these new challenges, new forms of public and private governance have emerged between civil society organisations, businesses and regulators. New strategies of governance have been implemented by national governments and Intergovernmental Organisations (IGO). These are designed to cope with structural changes and attempt to reinstate effective regulatory systems on the inter- and transnational level. While private forms of governance have flourished in a number of policy fields over the last decades34, many scholars are critical of the efficiency of the new regulatory schemes due to the power asymmetry between companies and civil society actors5. Some studies of private labour regulations show that workers’ rights have improved only
1 Trubek et al. 2000, p. 1188 2 Abbott & Snidal 2013, p. 96 3 Trubek et al. 2000
4 Sabel et al. 2000
slightly, unevenly, or not at all and that even strict private auditions are not able to identify violations of process rights such as the freedom of association.6 To overcome the governance gap that stems from weak oversight through states and Intergovernmental Organisations and the lack of efficiency of private regulation schemes, such as transnational codes of conduct, one mode of governance has been theorized to be especially promising: orchestration. In the international system there often exists no possibility for mandatory action, as IGOs are weak in capacity compared to states. This sometimes leaves them with the option to use facilitative orchestration as their last resort.7 Intergovernmental Organizations seem to use this strategy especially in situations where they have a capability deficit on a specific issue but still wish to regulate.8
Facilitative orchestration for IGOs means reaching regulatory goals through the use of different soft mechanisms to influence the actions of intermediaries who then regulate the target. This research is focused on the mechanisms and interactions that are inherent to this mode of governance in the field of labour regulation in the European Union (EU). That a powerful supranational organisation like the European Union would use this mode of governance seems to be puzzling at first sight, but closer inspection of the reality shows that orchestration is a regularly used strategy. One institution that is set up to facilitate transnational regulation of labour and social affairs on the level of the European Union is the European Social Dialogue. This institutional tool has been created by the European Commission in its role as orchestrator to coordinate and support representatives of employers and workers on the European level. Subsequently, these organizations use their capacities to engage in joint transnational negotiations, implementation processes and monitoring. Zooming into this field, my thesis sets out to answer the following research question: How does the European Union use the mode of orchestration, which mechanisms are being used and what are the outcomes of EU orchestration? I argue that answering these questions is valuable for two reasons. First of all, it has social value because it adds to a better understanding of how the governance gap can be addressed and how IGOs can use new modes of soft governance to engage into transnational regulation efforts. This could ultimately lead to the better regulation of social and labour issues in particular. Secondly, my research contributes to the academic debate on new forms of governance in general and improves the understanding of the strategy of orchestration in particular. Orchestration is not at all a new mode of governance, as it has been effectively
6 Egels-Zanden & Lindholm 2015, p. 38-39 7 Abbott & Snidal 2009b, p. 565
practiced by many IGOs since the post-war era9. Nevertheless, it is an emerging analytical concept that has not yet been fully theorized and which still lacks important links to be completely understood.10 Especially the use of intermediaries by IGOs and the inherent power relations as well as the employment of specific mechanisms to orchestrate need further clarification. My research is adding to the understanding of this governance mode in various ways: Understanding the interactions between the European Union and its social partners in the field of labour regulation will help to clarify the relationship between orchestrators and intermediaries and inform future research as well as policy-makers. Analysing the outcomes of this governance mode is valuable for understanding its potential and its shortcomings compared to other modes of governance used by IGOs.
The paper is structured as follows: Chapter two introduces the reader to the theoretical framework of new governance and the strategy of orchestration that my research is based on and from which my research question emerges. In chapter three I elaborate on the qualitative research design of my study which includes a document analysis of policy documents and transcribed expert interviews. Chapter four describes the results of my research and finds answer to the question of how the European Union uses the mode of orchestration, which mechanisms are being used and what the outcomes of EU orchestration are. I find that there are four main mechanisms of support by the orchestrator and three modes of regulation by the intermediary and how the process can work in the favour of the intermediaries preferences in the relationship. In chapter V I discuss how this change in the power relationship can lead to suboptimal outcomes. In chapter six I conclude on my research and give an outlook to further fields of investigation.
2. Theoretical Framework
In post-war times, regulation was a clear competency of domestic governments. It was predominated by top-down structures, strong regulatory state-capacity and the possible use of sanctions towards easily accessible targets.11 The state-centric and centralized approach to governance relied heavily on mandatory rules in its jurisdiction, coupled with immense bureaucratic expertise.12The domestic model of regulation can’t be successfully translated to the international level, due to substantial systemic
9 Abbott et al. 2015, p. 355 10 Abbott et al. 2015, p. 347 11 Lobel 2004, p. 344
differences, inherent to international politics. In the anarchic international structure no state-like authority exists to set binding rules or impose sanctions as each state is defending its own sovereignty.13 While some efforts have been made to create regulatory bodies on the international level, including the United Nations and the WTO, their power stays far beyond the possibilities of domestic regulation, as they lack the power to enforce mandatory rules and sanction misbehaviour by members.14 Meeting these challenges on the international level has meant introducing new forms of regulation, based on the concepts of Re(New) Governance15 16 or Responsive Regulation.17 These new approaches are based on ‘soft law’, accompanied by ‘small guns’ or ‘benign big guns’, and dispersed expertise and focus on private self-regulation. Private regulation is often coordinated and monitored by public actors.18 While these approaches and corresponding strategies have at first been introduced at the domestic level, they have had important implications for the debate on inter- and transnational governance which I will address later. A second shift in the debate on international regulation has been the emergence of abundant private and public-private regulatory schemes on the transnational level.19 20 21 In the following chapters, I will give an overview of how the understanding of governance has changed over the course of the last two decades22, describe how private governance relates to state and IO regulation and then discuss the strategy of orchestration that has been theorized and enhanced by leading scholars in the field of transnational governance. 23242526
2.1. Transnational New Governance and Private Standard Setting
The limited regulatory power of states and IGOs was matched with new approaches to regulation, leading to the emergence of private regulatory schemes such as voluntary codes of conduct or certification schemes that are implemented by firms and monitored by civil society actors27. Private regulatory schemes have emerged for different reasons. First of all, as free trade enabled multinational companies (MNCs) to
13 Abbott & Snidal 2009b, p. 534 14 Abbott & Snidal 2009b, p. 534 15 Lobel 2004, p. 343
16 Abbott & Snidal 2009b, p. 520 17 Ayres & Braithwaite 1992 18 Abbott & Snidal 2009b, p. 520 19 Trubek et al. 2000
20 Sabel et al. 2000 21 O’Rourke 2003 22 Lobel 2004, p. 344 23 Abbott & Snidal 2009a 24 Abbott & Snidal 2010 25 Abbott et al. 2013 26 Abbott et al. 2015
heavily rely on transnational supply chains, a regulatory gap developed. The reason for that is a lack of states access to targets, a lack of Intergovernmental Organisations regulatory capability and the weakness of many states to regulate - or their reluctance to do so. While MNCs tend to gain profit from this gap in regulatory capacity, many civil society actors who are concerned with environmental, health or labour issues have adapted to this situation. They shifted their approaches and went from contestation of business practices to collaboration with firms28 and tried to influence business practices through the implementation and monitoring of voluntary codes of conduct throughout the global value chains of lead firms. In many industries, these multi-stakeholder codes or certification schemes are now widely spread and have been embedded as a part of companies corporate social responsibility (CSR). Nevertheless, the effectiveness of these private standard setting schemes has been extensively discussed and questioned.293031 Research has shown that private regulation can effectively buttress and support transnational regulation in some industries32. Nevertheless, many scholars are very critical of solely private regulation efforts and recent studies have cast substantial doubt on the effectiveness of private regulatory codes33 or institutionalized certification. Instead they ask for stronger state-based regulation compared to “soft law” practices employed by activists and NGOs, as the power relations between civil society actors and businesses are one-sided.34 Some scholars have argued that MNCs will adopt private regulation because they see competitive advantages as first movers or use CSR as part of their image communication or risk management strategies35. Others find that CSR compliance is not relevant to the economic performance of companies. They conclude that it is therefore up to governments and not NGOs to put more effort into regulating firms, while civil society and social movements have to strengthen market-based incentives.36 One recent study assessing the outcomes of the multi-stakeholder foundation FWF, for example, draws an even darker picture when it comes to the identification of violations of process rights such as the freedom of association.37 Trubek et al. found that national-level regulation can effectively be buttressed by transnational regulatory structures through legal strategies but within their study they
28 Doh & Guay 2004, p. 8 29 O’Rourke 2003, p. 23-25 30 Bartley 2007
31 Vogel 2010 32 Locke 2013, p. 543
33 Egels-Zanden & Lindholm 2015, p. 38 34 Abbott & Snidal 2013, p.102
35 Doh & Guay 2004, p. 12 36 Vogel 2010, p. 81
also address limitations such as the lack of legitimacy of NGOs.38 Locke et al. find that regulatory efforts by private actors, that mostly occur through civil society pressure, can efficiently “substitute” or “complement” domestic labour law, operating in interdependence with public structures.39 Private and civil society actors have set up numerous and innovative governing institutions on the transnational level and created networks that fulfil many of the tasks that used to be exclusively executed by states.
Nevertheless, summarizing these findings, it seems that their scope as well as their capacity to create substantial and sustainable improvements in minimum standards has shown to be insufficient. This may be due to the fact that many of the private actors do not have the capacity to fulfil all of the tasks required for successful regulation on their own. Following Eberlein et al. 2014, there are six main components of regulatory governance: framing the regulatory agenda, formulating rules, implementing rules, gathering information and monitoring, sanctioning non-compliance and lastly evaluating policies and review rules.40 Inspecting the regulatory components in transnational regulatory schemes, Abbott and Snidal describe a strategy deployed by IGOs to solve the problems of territorial jurisdiction and insufficient authority that occur in traditional national and private transnational regulation. In orchestration, transnational networks and domestic civil society actors are being supported to gain access to information and to build up capacity. The intermediary actors then apply pressure either on domestic bureaucrats to regulate companies to abide to international norms or on multinationals to formulate codes of conduct.41 Supporting their case for the use of indirect governance mechanisms, Vogel finds that governments tend to support civil regulation efforts especially in situations where their own power is constrained by neoliberal trade laws which they want to bypass by transnational regulation42. In the following section, I will give a more detailed assessment of the strategy in question and summarize its most important implications.
2.2. Orchestration: A regulatory strategy
The concept of orchestration has been introduced into the field of International Relations by Abbott & Snidal43 who draw the term from scholars of legal thought who used it as one organization principle of New Governance.44 In their following work, the two scholars have further elaborated on the concept of orchestration. For my further
38 Trubek et al. 2000, p. 1209 39 Locke et al. 2013, p. 520-21 40 Eberlein et al. 2014, p. 6 41 Abbott & Snidal 2013, p. 105 42 Bartley 2007, p. 315
43 Abbott & Snidal 2009a 44 Lobel 2004, p. 400
research I will use the following definition of the concept as proposed by Abbott & Snidal:
We define Orchestration as the facilitation and coordination of intermediary actors on a voluntary basis by providing them with material and ideational
support in order to achieve governance goals with respect to target actors.45
Subsequently, I will explain how the strategy of orchestration can be used by IGOs to engage in new forms of transnational governance using intermediaries to regulate private targets and under which circumstances actors use this strategy. The authors Abbott & Snidal define orchestration as being one mode of governance compared to three other modes. These consist of the classical hierarchy which is direct and hard as the regulator wields power over the targets and has access to them.46 In comparison to this classical mode of regulation, it is referred to the mode of collaboration when the regulator has direct access to the target but has to use soft power to regulate. Another mode of governance is delegation which is used by regulators to govern in instances where they have resources to wield power over targets but delegate this power to other bodies that indirectly act in their place.47
2.2.1. Intergovernmental Organizations as orchestrators
While states possess authority and capacity to regulate on the domestic level, they as well have the possibility to use orchestration as a tool of ‘soft law’ to expand their resources through intermediaries. This is being called directive orchestration. In the international system there often exists no possibility for mandatory action, as IGOs are weak compared to states, leaving them only with the option to use facilitative orchestration.48 They appear to use this strategy especially in situations where they have a capability deficit on a specific issue but wish to regulate.49 Facilitative orchestration for IGOs means reaching regulatory goals through the use of different mechanisms to influence the actions of intermediaries. These intermediaries are mostly civil society actors such as non-governmental organisations (NGO) or trade unions
45 Abbott et al. 2010, p.6 46 Abbott et al. 2015, p. 8 47 Abbott et al. 2015, p. 8 48 Abbott & Snidal 2009b, p. 565 49 Abbott et al. 2012, p. 31
(TU), business federations (BF) and sometimes even other IGOs.50 The targets of IGO orchestration can be states, who are being managed via orchestration, or private actors that can be addressed to bypass the state.51 While orchestration is often welcomed by intermediaries52 states may try to impede this strategy. This is especially the case when the regulatory goals of principals are highly divergent and when states exercise ex ante oversight.53 Multiple case studies show that orchestration is a widely used mode of governance throughout many fields of International Relations. In her study of the G20 Viola finds that the institution manages to orchestrate other actors such as the International Monetary Fund (IMF) and the World Bank to help it with the implementation of policy recommendations as the G20 lacks the bureaucratic apparatus to do so.54 One example is the field of economic surveillance, where the G20 enlists the IMF to monitor the member states economic policies and report on inconsistencies with the aim of balanced and sustainable growth.55 While the G20 use orchestration to overcome capability deficits, other IGOs use this soft mode of governance to increase their focality in a field where they initially lacked access to. In their study of the United Nations Environment Programme (UNEP), Van der Lugt and Dingwerth find that the environmental organisation has endorsed and supported the Principles for Responsible Investment and assisted it to grow in numbers of participants as well as in quality of participation.56 Advancing its influence from the coordination of environmental cooperation into the field of more general sustainable development where focality was low has been only possible through the use of orchestration.57 Assuming that there are multiple cases in which IGOs use orchestration to govern, it has to be specified under which circumstances orchestration occurs and what the mechanisms are.
2.2.2. Properties and mechanisms of orchestration
Abbott & Snidal describe multiple mechanisms that can be used by IGOs to orchestrate. IGOs can for example promulgate standards that can be adapted by regulatory schemes, leading to an coordination of schemes around specific norms. They can act as “honest brokers”58 to help convening stakeholders and overcome bargaining problems. Strong IGOs also are able to provide intermediaries with material 50 Abbott et al. 2012, p. 5 51 Abbott et al. 2012, p. 11 52 Abbott et al. 2012, p. 11 53 Abbott et al. 2012, p. 28 54 Viola 2015, p. 98 55 Viola 2015, p. 108
56 Van der Lugt & Dingwerth 2015, p. 241-42 57 Van der Lugt & Dingwerth 2015, p. 261 58 Abbott & Snidal 2009b, p. 575
support and give technical assistance to build up capacity. By acknowledging specific schemes or codes, IGOs have the possibility to provide legitimacy and empower their bargaining capacity towards private targets. IGOs also have the option to use knowledge-based approaches, such as using their expertise to spread information on best practices.59 In their following work, Abbott et al. summarized these mechanisms to “convening, agenda setting, assistance, endorsement and coordination”60. While the concept of orchestration is new to the realm of international studies, empirical findings show that the strategy is widely spread throughout many IGOs, including the International Labour Organisation (ILO), UNEP and the European Union. After reviewing multiple case studies the following properties can be attributed to the orchestration process.
Orchestration seems to take place in situations where IGOs lack capabilities in a field that they want to regulate.61 The ability of an IGO to use this mode of governance also strongly depends on the availability of intermediaries or the ability of the IGO to create fitting intermediaries.62 Abbott et al. also find that often intermediaries like to be orchestrated as it increases overall efficiency in regime complexes and helps them to achieve their organizational goals while IGOs prefer situations where they can choose from multiple intermediaries as this gives them more power over the outcomes of the regulation process.63 Another property of orchestration is that it is the preferred governance mode in cases where the goals of IGOs member states show a high divergence, as it requires the IGO to use orchestration as a soft governance mode instead of direct hard governance modes that will not be supported by its principal members. Goal divergence among the member states also decreases the chance for blockage of IGO initiated orchestration.64 While these properties describe well under which circumstances orchestration as a mode of governance is being used by IGOs, there is not enough knowledge about how exactly the relationships between the partners of orchestration are shaped. Attending to fill this gap Mattli and Seddon propose a framework to help understand the power relationships between orchestrator and intermediary or intermediaries. As preferences are not automatically harmonious, and aims do not always align, there is a distribution of power which gives the partners specific and changing roles defined by the authors as makers and rule-supporters.65 The rule-maker is more powerful as he is deciding on the course of action
59 Abbott & Snidal 2009b, p. 575-77 60 Abbott et al. 2012, p. 13
61 Abbott et al. 2015, p. 365 62 Abbott et al. 2015, p. 367 63 Abbott et al. 2015, p. 368 64 Abbott et al. 2015, p. 371 65 Mattli & Seddon 2014, p. 318
and can establish an asymmetric information relationship towards the rule-supporter.66 Another way to assess the power relationship is to look at the number of partners involved in the orchestration. In cases where only one intermediary is available, the power distribution can favour the intermediary.67 In the following I will use these properties to describe when and how the European Union as one of the most powerful IGOs is expected to act as an orchestrator.
2.3. The European Union and orchestration
The European Union is a much stronger institution compared to many other IGOs. Its supranational structure and the strongly institutionalized bureaucratic apparatus make the EU a distinct organisation. However, it has been shown to be very active in using orchestration strategies throughout many policy fields to enhance its capacity to govern. One example can be found in its efforts to regulate biofuel in the EU. Schleifer finds in his analysis of governance in the biofuel sector that EU policymakers have used orchestration to coordinate private certification schemes68 using a mix of directive and facilitative measures, such as implementing standards for sustainability schemes as well as recognizing and funding specific schemes.69 While recognizing the approach of the EU as cost-efficient, Schleifer argues that orchestration efforts of the EU failed in the biofuel sector due to the lack of policymakers to establish a level playing field.70 In another instance of orchestration the European Commission has set up European regulatory networks (ERN) that support its efforts to regulate, for example, in the policy areas of competition and telecommunications by acting as intermediaries to help put pressure on national regulatory agencies that the Commission has no direct access to.71 Blauberger & Rittberger found that these ERN’s are highly effective in instances where the European Commission owns strong regulatory competences but does not possess the means to single-handedly implement and monitor regulations that are being transposed on the national level.72 In the case of competition policies, the authors conclude that through giving up some of their competences for regulation and monitoring to national regulation bodies the Commission gained immense operational capabilities in the field it wanted to regulate.73 Similarities to these instances of orchestration by the European Union can be found in
66 Mattli & Seddon 2014, p. 321 67 Mattli & Seddon 2014, p. 324 68 Schleifer 2013, p. 538 69 Schleifer 2013, p. 539-40 70 Schleifer 2013, p. 542
71 Blauberger & Rittberger 2015, p. 40 72 Blauberger & Rittberger 2015, p. 47 73 Blauberger & Rittberger 2015, p. 63-64
the field of industrial relations and labour policy, which is strongly shaped by the European Social Dialogue (ESD), the European equivalent to national collective bargaining structures between business federations and trade unions. I assume that the relationship between the European Commission and its Social Partners in the European Social Dialogue is an instance of orchestration.
The ESD has a legal foundation in the Treaty on the Functioning of the European Union (TFEU), which determines when the social partners have to be consulted. From this point on, the European Commission only work as a supporting partner and has no influence on the process of negotiation or implementation. The social partners also have the right to set their own agenda on other social and labour issues not addressed by the Commission and negotiate these autonomously. This supports my assumption of indirect and soft governance. The European Commission acts in the role of an orchestrator through enlisting and supporting the transnational European Social Dialogue an intermediary to ultimately regulate labour practices on the national business level. The European Commission has no direct access to and only limited power over the national social partners who are in many countries responsible for collective labour agreements. The soft and indirect way of governing that is inherent to orchestration seems to be an efficient solution. As it is clear why and when the European Union uses orchestration, my research sets out to investigate the relationships and mechanisms between the actors in the O-I-T model. Referring to the theoretical framework of Abbott et al., I formulate the following research question: How
does the European Union use orchestration to engage in transnational private governance of labour rights and collective bargaining using the European Social Dialogue? To answer this research question, I formulate two sub questions. What are the mechanisms that are being used by the actors in the mode of orchestration? What are the outcomes of the EU Orchestration efforts? In the following I will outline my
research design as well as the methods used to collect and analyse data.
3. Research Design
To answer my research question, I employ a qualitative comparative historical case study. This design enables me to couple the strengths of an in-depth within-unit case study with the advantages of a small N across-unit comparison. Compared to a cross-sectional study, the within-unit case study has the following advantages74: Firstly, it is more efficient in finding descriptive inferences than proving causality. A case study
also provides the researcher with more in-depth knowledge, which is preferable if the field is not yet fully theorized and crucial interactions have not been explained. While the lack of representativeness and therefore the external validity of a case study is often voiced as an inherent weakness of case studies, I hope to ameliorate the representativeness of my results through analysing more than one case. I do acknowledge that this will only be effective, if the chosen cases actually are comparable.75 Another advantage of the case study is, that its design enables the researcher to analyse causal mechanisms, which will provide further research with the missing links between cause and outcome, such as actors motivations or mechanisms deployed. Based on these findings, quantitative work such as cross-sectional studies can be employed more precisely in this field in the future. As the field that I am employing my research in is not yet fully developed, the exploratory character of a case study design promises to yield more insights than quantitative studies, which are designed to confirm theories. Evaluating the advantages and disadvantages of the case study design, I conclude that my design will yield the most promising results in answering my research questions. In the following, I will further elaborate on the selection of my cases.
3.1. Cases and justification
As explained previously, my research is situated in the field of transnational governance of labour regulation, focussing on the European Union as the main actor, who engages into orchestration of intermediaries as a mode of soft and indirect governance.76 This constitutes my research population as instances of orchestration by the European Union in the field of labour regulation. Following from my research design, my studied sample must consist of more than one unit, to allow me to make comparative statements. As stated, my research question asks for a qualitative approach, which implicates in-depth within-unit studies. This means that I pick two predominantly comparable units while studying their within-unit spatial and temporal covariance, comprising my cases.77 The two units observed are two different policies in the social and labour related sector initiated by the European Commission and negotiated by the European Social Partners as part of the European Social Dialogue, enabled and facilitated by the European Commission and its agencies, predominantly the Eurofoundation. Within these units, I will make observations about different variables that the introduced cases consist of. The variables will arguably be such as
75 Gerring 2004, p. 348 76 Abbott et al. 2015, p. 9 77 Gerring 2004, p. 342
the inclusion of different stakeholders, the employment of varying mechanisms as well as the change in power relations between orchestrator and intermediary depending on availability and focality of the actors.78 To answer my research questions, I need to employ scientific methods, that are fit to produce data that enables me to make valid statements about the variance of these variables within and across units. As explained above, the European Social Partners are being consulted on social policy under Article 154 TFEU. The two cases that I chose differ widely in their outcome, which gives me the possibility to observe differences in the process that lead to this divergence. My first case looks at the cross-industry negotiations on the topic of harassment and violence at the workplace. These negotiations concluded in the implementation of an autonomous framework agreement in 2007 through the European Social Partners and it’s national member organisations. The second case observes the issue of temporary agency work on the European level. As cross-industry negotiations on this topic have failed, a sectoral social Dialogue between the sectoral social partners emerged who have been able to give input to the drafting process of the Declaration 2008/104/EC on Temporary Agency Work and have since then developed a strong Social Dialogue on the sectoral level. In the next chapter I will explain the methods deployed for my research.
3.2. Methods
In order to collect data, I have employed two different methods. Firstly, I have collected all the significant policy documents concerning the process of the drafting and transposition of the policies and analysed these in regard to the relationships between orchestrator and intermediary, the mechanisms employed, the reasoning behind the choices of specific intermediaries and the changes in the relationship between actors as well as outcomes of orchestration outcomes. In a second step, I have carried out stakeholder interviews with experts that have been involved in the aforementioned programs and framework negotiations or the process of their implementation. One of the reasons to include policy documents as well as interviews into my study is the possibility to triangulate between the two different sources and therefore reduce potential bias in the data overall.79 The collected documents have been analysed using the Computer-Aided Qualitative Data Analysis Software (CAQDAS) Atlas.ti 8.0.
78 Abbott et al. 2015, p. 366 79 Bowen 2009, p. 28
3.2.1. Policy documents
The documents acquired are all openly available on the websites of the European Commission, its agencies or the partner organisations. To figure out which documents are most valuable to my research I firstly acquired all policy documents that relate to one of the two aforementioned cases. Reading these documents one finds hints to other related documents that are being mentioned in the present paper. Following this strategy, I was able to get an overview of the existing policy papers, create a timeline to follow the process and decide which documents are most valuable. A second way to validate the selection process of the documents to be analysed is to ask stakeholders involved in the process for further material. Most of the documents that were acquired on this way have already been included in the first round, leading to the assumption that the collection of policy documents is adequate and sufficient.
3.2.2. Stakeholder interviews
In the second step I identify the stakeholders involved in the two negotiation processes to lead interviews with experts that have been part of the processes and can give me additional insights. These stakeholders include the European Commission experts in the field of collective bargaining at the DG Employment, Social Affairs and Inclusion, policy officers of the involved business organizations (Business Europe, CEEP, UEAPME, WEC), and the trade unions (ETUC, Uni-Europa). These interviews have been carried out in person and aimed to enhance my understanding of processes and mechanisms inherent to orchestration as well as reveal possible shortcomings and unexpected repercussions of this specific mode of governance, such as selection biases and capture.80 For each of the cases one expert from the three stakeholders were set out to be interviewed to get a balanced overview of the mechanisms and relationships, leading to six stakeholder interviews in sum. The interviews were in-depth and semi-structured. I have used an interview guide with four separate blocks that consisted of three questions each. The questions asked vary slightly in between Commission and social partner interviewees. The semi-structured interviews enabled me to skip questions if necessary or go deeper into one topic or question if the answers seemed to be promising. The interviews have been manually transcribed to obtain text documents of the conversations.
3.2.3. Document analysis in Atlas.ti 8.0
The policy documents and the transcripts have been uploaded into the document analysis program Atlas.ti. 8.0. The decision to use a computer-aided qualitative data analysis software (CAQDAS) was taken as it holds many advantages compared to manual document analysis. With large amounts of documents and raw data, the program enabled me to easily structure and analyse the collected data. Although the use of a prefabricated program can possibly restrict the method of analysis compared to manual coding, I found that the design of the program Atlas.ti is well suited for my theory building research design.81 Therefore the advantages clearly outweigh the disadvantages of computer-aided analysis. The method of document analysis is being used in qualitative research to evaluate and interpret meaning in written data to gain a better understanding of concepts and give them meaning in relation to the issues researched.82 Following Bowen, the document analysis is designed to “find, select, appraise, and synthesize data contained in documents”83. This data has been produced in multiple steps. First of all the documents have been scanned for interesting or important parts and marked with quotations. This process is highly subjective and depends on the prior knowledge of the researcher about the topic in question. The second step was a strategy based on grounded theory and involved linking codes to the quotations, giving them meaning apart from the obvious statements. This process of coding in the program Atlas.ti resembled the strategy of open, axial and selective coding developed by Strauss and Corbin.84 In open coding the researcher choses parts of the data and gives them distinct meaning on a more abstract level. Going through the documents many codes emerged. The codes have later been merged, dropped or categorised.85 After the first round of coding, the existing codes have been grouped into categories. This process had to be repeated until patterns emerged from the data that gave the researcher new answers to the questions that lead the study. In the following chapter I describe the data that has been collected using the mentioned methods.
3.3. Data
Following my research design, I have collected numerous documents and conducted multiple interviews. In the next sections I give an overview of the collected
81 Smit 2002, p. 68 82 Bowen 2009, p. 27 83 Bowen 2009, p. 28
84 Strauss & Corbin 1998, p. 55-143 85 Smit 2002, p. 70
data to increase the transparency of my research. For the case of policy on harassment and violence at work eight documents have been collected and analysed. The documents are ordered on a timeline ranging from 2005 till 2015 and include various different types, sources, authors and publishers. These include the “Consultation of the social partners concerning violence at the workplace and its effects on health and safety at work” (Document H&V #1); “Framework Agreement on Harassment and Violence at Work” (Document H&V #2); “Communication from the Commission to the Council and the European Parliament” (Document H&V #3); the “Press release by the Social Partners” (Document H&V #4); “Yearly joint table summarising ongoing social partners activities” (Document H&V #5); “Autonomous Framework Agreement on Harassment and Violence at Work: An ETUC interpretation Guide” (Document H&V #6); “Implementation of the European Autonomous Framework Agreement on Harassment and Violence at Work” (Document H&V #7); “Study on the Implementation of the European Autonomous Framework Agreement on Harassment and Violence at Work” (Document H&V #8). For the policy process on temporary agency work I have collected and analysed six documents including the “Eurociett/UNI-Europa Joint Declaration on the Directive on working conditions for temporary agency workers” (Document TAW #1); “Directive 2008/104/EC of the European parliament and of the Council of 19 November 2008 on temporary agency work” (Document TAW #2); “Eurofound report: Temporary agency work and collective bargaining in the EU” (Document TAW #3); “Eurociett/UNI-Europa Joint Declaration: Training for Temporary Agency Workers, Joint actions developed by sectoral social partners play a key role in facilitating skills upgrading” (Document TAW #4); “European sectoral Social Dialogue Committee on ‘Temporary Agency Work’: Work program 2009-2010” (Document TAW #5); “European Sectoral Social Dialogue factsheets on temporary agency work” (Document TAW #6).
Although my research design provided for the conduction of six stakeholder interviews, only three personal interviews with experts from stakeholder organisations have been conducted. The first one was conducted with a Senior Adviser from the European employer federation BUSINESSEUROPE for the case study on harassment and violence at work (Interview H&V #1). The second interview has been conducted with a Public Affairs Manager from the European sectoral employers federation World Employment Forum on the topic of temporary agency work (Interview TAW #1). The third interview has been conducted with a Senior Researcher from the ETUC. (Interview H&V #2). No interviews have been conducted with experts of the European Commission as I have not been able to make an appointment for an interview in time. Resulting from these shortcomings in the data collection process is a undeniable
imbalance concerning the interview data as only one side of the tripartite governance process has been interviewed. This limits the validity of my research on the mode of orchestration. The data gathered from stakeholder interviews is supposedly unbalanced. Another problematic spot of my research is the under specification of the intermediary – target relationship that is inherent to my research design. Studies at the company level and of national member organisations could provide further research with important links to the implementation process. Having described my research design, the methods used to collect and analyse data as well as the nature of the data collected, I proceed with presenting the results of my research in the next chapter.
4. Analysis and Results
The aim of this research is to shed light onto the mechanisms, relationships and processes as well as outcomes of the newly theorized mode of governance that is orchestration. The findings that emerge from the qualitative study of the processes behind the two policies on Harassment and Violence at Work and Temporary Agency Work are numerous and illuminative of the way in which the European Union as an Intergovernmental Organisation is using orchestration to govern the field of labour and social affairs in the Eurozone and beyond with the help of institutionalized European Social Dialogue. The way the European Commission and its agencies facilitate the regulatory efforts of the European Social Partners and their members by convening and supporting them clearly meets the definition of orchestration as proposed by Abbott & Snidal.86 In the following chapters, I am presenting and discussing the results of my qualitative analysis in order to answer the questions that are leading this inquiry. These questions were: How does the European Union use the mode of orchestration? Which mechanisms are being used and what are the outcomes of EU orchestration?
First of all, I will give an overview of the institutions and actors involved in the process. Secondly, I will introduce the reader to the cases that have been observed and give an comparison. I will then examine the relationship between the European Commission and the European Social Partners with respect to the mechanisms that are being used to facilitate regulation. The fourth chapter focuses on the relationship and mechanisms between the European Social Partners and the companies which are the targets of regulation. Afterwards I analyse further findings of my research that are related to organizational issues. I will conclude with explaining how the results of my research add to the theoretical knowledge that has so far been available on orchestration. While much of my research is confirmative of the assumptions made by
the scholars that have written on orchestration, my results reveal the complexity of the process and show how the construction of the process leads to feedback loops that can hold unexpected results which may be concerning for the effectiveness of this mode of governance. In cases where the power relationship between orchestrator and intermediary is shifting towards a more powerful rule-making intermediary, the described O-I-T relationship is disturbed and the outcomes do not resemble the aims of the orchestrator.
4.1. European Social Dialogue
European Social Dialogue is deeply rooted in the DNA of the European Union. From the early years on joint committees have been formed that represented employers and workers in different sectors throughout the European Union to discuss issues of collective bargaining and industrial relations.87 With further integration of the European Union, the number of sectoral committees increased. Parallelly a cross-industry social dialogue between transnational business organizations and trade unions emerged. The European Commission decided to further institutionalize the bodies and strengthen Social Dialogue on the European level to enable the European sectoral and cross-industry Social Partners to negotiate bilateral agreements.88 This was supposed to facilitate transnational governance. The legal basis for European Social Dialogue can be found in the articles 154 and 155 of the Treaty on the Functioning of the European Union. Article 154/1 requires the Commission to consult the European Social Partners on social policy issues before submitting proposals. Article 154/4 gives the ESPs the possibility to halt community action and initiate Article 155 which gives them the right to negotiate binding agreements that can be implemented by national collective agreements or by a Council decision.89 While the Social Dialogue has been implemented and is being facilitated by the European Commission and its agencies such as the Eurofound, the European Social Partners as well as their national members are largely autonomous in their negotiations90 which supports the assumptions of soft and indirect modes of governance associated with orchestration.
The European Foundation for the Improvement of Living and Working Conditions (Eurofound) was established in 1975 and is the European agency that is responsible for facilitating the European Social Dialogue on the sectoral and the cross-industry
87 Perin & Leonard 2011, p. 160 88 Perin & Leonard 2011, p. 161
89 Treaty on the Functioning of the European Union 2007, Article 154/155 90 Perin & Leonard 2011, p. 161
level. It’s main task is providing the European Social Partners with information and advice as well as conducting cross-national research on topics such as collective bargaining and industrial relations. Since 2006 the agency is also responsible for carrying out the representativeness studies that determine which organizations are eligible to be a Social Partner on the European level Social Dialogue and therefore be consulted under the article 154 TFEU. To be representative, a European Social Partner has to be cross-industry or sectoral-specific, needs to have the capacities to negotiate agreements on the European level and has to consist of social partners of all member states as far as possible to ensure even transposition of agreements throughout the Community.91
The social partners on the European level are the umbrella organisations of the social partner organisations on the level of national member states as well as other EEA and future accession states such as Turkey. The European trade unions and business federations comprise the main users of the Eurofound’s services and have seats in the governing board of the agency. At the moment the official representativeness study of the Eurofound has identified the ETUC on workers side and BUSINESSEUROPE, CEEP and UEAMPE on the employers side as the primary European Social Partners on the cross-industry level.92 In 2015 a relaunch of the cross-industry social dialogue was issued with strong political support by the Commission president Juncker, with the European Trade Union Confederation, a transnational organisation representing workers from 89 trade union organisations in 39 European countries, and three employer associations – Business Europe, CEEP, UEAPME. These organisations signed a statement in June 2016, assuring that they will further engage into Social Dialogue and capacity building for collective bargaining. This involves their regional and national members to ensure implementation of their autonomous framework agreements in member states. The Commission wants to promote and support the efforts of the social partners, including technical and legal information and encourage knowledge- and capacity-building as well as exchanges of good practice.93 Meanwhile the social partners on the sectoral level are evolving in quantity and quality.94 In the following chapter I will give a short overview over the course of the two policies and give explanations for the differing outcomes.
91 Perin & Leonard 2011, p. 161
92 European Foundation for the Improvement of Living and Working Conditions (2013),
Representativeness of the social partners in the European cross-industry social dialogue,
93 A new start for social Dialogue. Statement of the ESP, EC, NL PRES on behalf of CNSL 27062016. (27.
June. 2016. Retrieved from
https://www.eurofound.europa.eu/observatories/eurwork/articles/national-capacity-building-initiatives-for-social-partners-experiences-in-five-eu-member-states (23.06.2016)
4.2. The European Social Dialogue in action
The Commission has been concerned with health and safety issues at the workplace for a long time already. In 1997 the Commission requested the Advisory Committee on Safety, Hygiene and Health protection at Work to gather information on violence at the workplace. The Advisory Committee issued its opinion in 2001 and proposed that the Council should set up guidelines to tackle the issue. In 2003 the European Foundation for the Improvement of Living and Working Conditions conducted a research on harassment and violence at the workplace and issued a best practice report. The cross-industry Social Partners included the topic in their joint work program 2003-2005 and started discussing possible policy solutions while the Commission communicated it’s aims to regulate psychosocial risks at the workplace and the European Parliament issued an resolution on harassment at the workplace in 2001.95 In 2005 the European Commission officially consulted the social partners on the topic of violence at the workplace and its effects on health and safety at work96 asking for an analysis of possible preventive measures and an assessment of the scope of an agreement between the European Social Partners. The European Social Partners decided to invoke Article 154 TFEU and regulate the issue via an autonomous framework agreement between the ESPs. On the 26th of April 2007 the Framework Agreement on harassment and violence at work was signed by the responsible partner organisations ETUC, BUSINESSEUROPE, UEAPME and CEEP.97 The framework agreement defined harassment and violence at the workplace, aimed at increasing awareness and put in place an action-oriented framework, determined procedures to cope with harassment and violence and stated how the agreement was to be implemented and monitored. Subsequently the Commission communicated the framework agreement to the Council and informed the European Parliament. The social partners agreed that the framework was to be implemented within three years with an implementation report being issued by the European Social Partners within the fourth year. The final report by the European Social Partners was issued on 27th of October 2011. It relayed the outcomes of the autonomous agreement and the challenges of its implementation. In 2015 the European Commission - DG Employment, Social Affairs and Inclusion - carried out a study on the implementation of the autonomous framework
95 Document H&V #1, p. 3 96 Document H&V #1, p. 3 97 Document H&V #2, p. 1
agreement. This study found that although small improvements have been made Community action is still advisable.98
My second case observes the negotiations around the topic of temporary agency work. While the Commissions main focus was workers in standard work relationships, it has also addressed the rights of workers with temporary contracts in the agency work sector through the Directives on Health and Safety 91/383/EEC and on the Posting of Workers 96/71/EC.99 Through the strong increase in non-standard work relationships, characterized by the triangular relationship between worker, agency and user company throughout many sectors and countries in the European Union, the Commission wished to regulate the temporary agency work sector in particular. Workers and employers were consulted on this topic in 1995 by the Commission to give their opinion on the way of action and officially consulted the social partners in 1996 to find a solution to the ‘flexicurity’ debate.100 After concluding an framework agreement on fixed-term work in 1999 the cross-industry social partners then informed the Commission in May 2000 of their wish to initiate autonomous negotiations on the topic of temporary agency work. After asking the Commission for an extension of the deadline, the cross-industry social partners had to inform the European Commission of the fact that their negotiations had failed to produce any form of autonomous agreement on the topic of temporary agency work.101 Meanwhile a sectoral Social Dialogue Committee on temporary agency work was institutionalized on the European level. The European Social Partner organizations were the World Employment Confederation (formerly Eurociett) on the employers side and Uni-Europa on the workers side. In 2001 the new European sectoral social partners issued their first joint declaration addressing the flexicurity debate. While the Council decided to set up a declaration on the issue, the sectoral social partners accompanied the drafting process by issuing two more joint declarations with their common positions on the topic in 2007 and 2008. In November 2008 the European Parliament and the Council adopted Directive 2008/104/EC on temporary agency work.102 The main aspects of the new Directive were the commitment to equal treatment and equal pay for agency workers as well as the review and lifting of unjustified restrictions on the sector. The Sectoral Social Dialogue Committee has since then adopted multiple joint declarations on temporary agency work and helped monitor the transposition process of Directive 2008/104/EC. 98 Document H&V #8 ,p. 105 99 Document TAW #3, p. 1 100 Document TAW #2, p. 9 101 Document TAW #2, p. 9 102 Document TAW #2
Having described the two observed cases, I will go on to argue that the different power distribution between the actors has led to the different outcomes that were described above. In the case of Harassment and Violence at Work, a clear instance of orchestration has occurred. The European Commission has supported the cross-industry social partners through various mechanisms, while the European Social Dialogue succeeded in negotiating an autonomous agreement and the implementation of it on the national level through its national member organisations. Orchestration successfully occurred because the intermediaries had the capacity to negotiate and implement the agreement on their own, with support by the Commission. The capability of the cross-industry social partners to engage in regulation is key to the process in this case. Although orchestration effectively occurred, the outcomes of the process were not aligned with what the Commission was aiming for, as the implementation report shows.103 This is due to the distribution of power in the relationship between orchestrator and intermediary, as well as the strong influence of the targets on the implementation process. In the case of the negotiations on Temporary Agency Work, the orchestration process failed on the level of the cross-industry as well as the sectoral level. This was due to the incapacity of the social partners to autonomously agree on the framework.104 While the cross-industry partners had problems with finding a common position on the flexicurity debate105, the sectoral social partners on agency work lack the political will and the capacity to effectively negotiate and implement autonomous framework agreements.106 The incapability of the European Social Partners to act on their own led the European Commission to fall back to the governance mode of hierarchy and issue the Directive 2008/104/EC. The sectoral social partners had only partial impact on the text of that Directive through issuing joint declarations107. Pointing out the differences of these two cases shows that the capacity of the actors as well as the power relationships play an important role in understanding when orchestration occurs and how the outcomes are shaped. In the following I further explain these relationships.
4.3. The European Commission and the European Social Dialogue
The results of my research on the relationship between the European Commission and its social partners are twofold. Theory assumes that there is a soft
103 Document H&V #8, p. 105 104 Document TAW #2, p. 9 105 Interview TAW #1, p. 8 106 Interview TAW #1, p. 6 107 Interview TAW #1, p. 8
power top-down relationship between the orchestrator and its intermediaries. This means that compared to the mode of delegation, the orchestrator has no direct influence on the work of its intermediaries. The reasoning behind relying on autonomous intermediaries for regulation is the lack of capacity or access on side of the orchestrator combined with an necessary alignment of aims between orchestrator and intermediary. In my case study of the European Social Dialogue, this relationship is confirmed in general but slightly altered. In cases where the power distribution between the orchestrator and the intermediary favours the later, the outcomes of orchestration can be expected to occasionally not align with the aims of the orchestrator. The O-I-T relationship implies that the orchestrator leads the regulation process through indirect and soft power mechanisms. My findings show that in some cases the orchestrator can end up not governing at all, and instead finds himself in a rule-supporting role instead of the one of a regulator. In the following I will more precisely define the relationships between the actors as found in my research and then place this in the context of the bigger picture.
4.3.1. Support
First of all, the European Commission does substantially support its European Social Partners - trade unions and business federations in my case - in multiple different ways. The most important mechanism is financial support. This includes the funding of translations, brochures and meetings108 as well as grants for joint research and projects109 and funds for capacity building on the national member level through the financing of round tables and best practice seminars.110 The social partner organisations admit that without these funds many of their actions would not be viable.111 Especially for the sectoral social partners, the funding gives them the possibility to participate in advocacy on the European level.112113The capacity building is an important tool on both the European as well as on the national level. While the access to funding support the different European Social Partner organisations in being more effective in working together as intermediaries, the funding at the national level also build up capacities that ensure a wider scope of agreements as well as better implementation and monitoring capacities. A second important way of support is being granted on the political level. Through promoting Social Dialogue on the European 108 Interview TAW #1, p. 4 109 Interview TAW #1, p. 4-5 110 Document TAW #4, p. 2 111 Interview TAW #1, p. 4 112 Interview H&V #1, p. 3 113 Interview H&V #2, p. 2
level, the Commission as well as the Council114 help strengthening the position of the social partners on the national level. This form of support is crucial to the process of capacity building for the transnational federations and unions and widens the scope of their influence115 and subsequently the impact of the orchestrated regulation. Two more forms of support have emerged from the data. Through the Eurofound agency, the European Social Partners have access to many forms of informational support such as legal advice and research.116 Lastly, there are forms of logistical support that are being provided by the European Commission and its agencies. This can be seen in the institutionalization of quarterly meetings of the Social Partners where negotiations are being led117 as well as the liaison forum of all social partners.118 While this support is crucial to the process, it is not automatically bringing the intermediaries into a dependent position. Instead, a distinction has to be made between the less powerful sectoral social partners and the stronger cross-industry partners. The sectoral social partners often rely heavily on the support by the European Union and as a result might be more receptive to orchestrator influence. The cross-industry partners on the other hand are less financially dependent and therefore more autonomous in their negotiations. The support mechanisms do help in capacity building but do not automatically influence the intermediaries decision making.
4.3.2. Agenda setting
The European Social Partners jointly set the agenda for the topics they want to negotiate and have all autonomy in their negotiations119. Nevertheless, one feature crucial to the process on the European level has a strong influence on the agenda setting process as well as the preferences and aims of the social partners. In comparison to other Intergovernmental Organisations, the European Union as a supranational actor does have the capacity to regulate directly through directives. While the European Union is often not using this hierarchic mechanism of regulation and instead has given the European Social partners the right to halt Community action and negotiate autonomous agreements instead, the regulative power of the European Union looms large over the process of regulation through Social Dialogue. While the work programs of the social partners are set up independently, the topics discussed
114 Document H&V #3, p. 3 115 Interview H&V #1, p. 2 116 Interview H&V #1, p. 3 117 Interview TAW #1, p. 4 118 Interview TAW #1, p. 5 119 Interview TAW #1, p. 7-8
naturally cover issues that are prevalent in the European Commission.120 Juncker’s ten points plan has for example strongly set the agenda for future Commission work and therefore can’t be ignored by the autonomous partners.121 Even in cases where the social partners are being consulted and decide to regulate without direct influence of the European Commission, the outcomes are being influenced by the shadow that a potential Directive throws. This is due to the possibility that in the case of failed negotiations on transnational regulation or a rejection by the Council, the European Union is able to regulate hierarchically through a Directive. That has happened in the case of temporary agency work. The existence of a ‘benign big gun’122 is the key to understanding how the aims to regulate of orchestrator and intermediary align often enough for orchestration to occur. There exist also some other forms of influence by the European Commission on the negotiation process of the ESPs. One Senior Adviser of the ETUC mentions, for example, that the consultation papers by the Commission already set the direction of the negotiations. Even while the autonomous partners negotiate, the Commission seems to try to indirectly influence the wording and contents of the negotiated agreements through giving informal legal advice.123 Apart from these limitations my research has also shown that the European Commission does in some ways try to influence the process via restricting the budgets of smaller social partners124 as well as monitoring the outcomes of autonomous agreements.125 These interferences show the thin line between the modes of orchestration and delegation.
4.3.3. Intermediary influence
Although the relationship between orchestrator and intermediary appears to be a top-down process, the mode of soft governance gives the orchestrated organisations on the European level multiple possibilities to influence their relationship to the orchestrator. In their study of orchestration instances by the European Commission in the fields of accountancy and environmental regulation Mattli and Seddon describe how the number of available intermediaries significantly shapes the outcomes of orchestration.126 In the case of the European Social Dialogue this has far-reaching repercussions, as the social partners are being selected through representativeness studies which are conducted by the Eurofound agency and are set-up to be transparent
120 Interview TAW #1, p. 8 121 Interview TAW #1, p. 8 122 Ayres & Braithwaite 1992 123 Interview H&V 2, p. 1 124 Document H&V #3, p. 3 125 Document H&V #3, p. 3
