The GATT/WTO Secretariat: An analysis of the influence of the GATT/WTO Secretariat on the establishment of the Appellate Body

(1)

The GATT/WTO Secretariat

An analysis of the influence of the GATT/WTO Secretariat on the

establishment of the Appellate Body

Name: Fleurke Exterkate

Student number: 1662066

Thesis Supervisor: Prof.dr. A.K. Yesilkagit Second reader: Dr. M. Beerkens

University: Leiden University

Master Program: Public Administration Track: International and European Governance

(2)

2

From 1986 until 1994 the Contracting Parties of the General Agreement on Tariffs and Trade (GATT) came together to undertake the eighth round of multilateral trade negotiations in the history of the GATT. Not only was this the longest round, but also the most comprehensive and ambitious round. The final agreement, signed on 15 April 1994 in Marrakesh, transformed the GATT into the World Trade Organization (WTO) (Yi-Chong & Weller, 2004, p.102).

The Contracting Parties did not only discuss trade matters during the Uruguay round. The

establishment of a new dispute settlement procedure was also on the table. In 1995, the Contracting Parties decided to establish an Appellate Body, which transformed the former procedure into a two-tier procedure. This decision can considered to be historical, as the world’s most powerful trade nations granted an unrivalled amount of power to a third party. The revised dispute settlement

procedure entails that if two parties are battling a dispute and wish to bring this to the WTO, they have immediate access to a panel consisting of neutral parties that has the mandate to make a legally binding ruling. If one or both of the parties do not agree with this ruling, they have the right to proceed to step two; the Appellate Body. This body reviews the rulings made by the panel and can either decide to overturn the ruling of the panel or uphold the ruling. Moreover, when a party fails to carry out the ruling, the Body has the power to impose penalties (Hudec, 1998, p.101).

The establishment of a new procedure was a long and laborious process. The Secretariat of the GATT played a key role in this process. For almost forty years, the Secretariat has strongly influenced the rulings of the panels. With the decision to establish an Appellate Body, the Secretariat lost a large part of its influence. While it is still involved in the first step of dispute settlement, the Secretariat is no longer involved in the second step as the Appellate Body became a separate institution.

In the past few decades, a large body of literature has been established on the functioning of international organizations. International organizations become increasingly important for global governance which has generated attention from scholars and policy makers alike. However, in general, little attention has been given to the functioning and politics of international organizations internally. This is especially true for the Secretariat of the GATT/WTO and its modus operandi. The case of the GATT/WTO Secretariat can be considered a part of a broader case that analyzes how bureaucrats who are working within an organization can potentially influence policy decision-making. According to the popular opinion and some theories, such as (neo) realism, bureaucracies are simply weak. They only exist to serve the interests of the most powerful states. They have no meaningful existence on their own because states are not willing to hand over substantial power to a third party (Walt, 1998, p.31). However, as Inis Claude claims, ‘administration and policy inevitably merge in defiance of analytical dividing lines, and the real question – in international organization as in national government – is never whether bureaucrats will influence policy’ (Claude, 1971, p.206). The question is how and under which circumstances they influence decision and policy making. Specific features such as amount of expertise can have a huge impact on which policy proposals go up into the organizations and are being considered and potentially end up in the text of a treaty.

The Secretariat of the GATT/WTO has often been portrayed as having a mere clerical role, while in fact the transition from GATT to the WTO could not have happened without the Secretariat (Yi-Chong & Weller, 2004, p.102) Moreover, it has been rarely discussed under which conditions the Secretariat has influenced the establishment of the revised dispute settlement procedure. By focusing on the specific features of the Secretariat of the GATT/WTO, it becomes possible to uncover the potential for bureaucratic influence.

(4)

4 What we see in the example of the GATT/WTO can also be applied to other cases. For example, staff from the International Monetary Fund is able to influence the monetary policy of the most powerful states by demanding certain standards in their fiscal and income policies. Moreover, they rearrange on a domestic level the political and monetary institutions of their members (Barnett & Finnemore, 2004, p.45). Also the environmental department from the Organization for Economic Cooperation and Development (OECD) has substantial influence through its own bureaucracy, their secretariat. Via publishing of reports, development of environmental policy and advisory committees the secretariat was able to put the development of environmental policy in developing countries high on the agenda (Biermann & Siebenhüner, 2009, p.76). These are just two of many examples where international bureaucracies are able to influence policy. Thus, it must be noted that there is a general and broader question at play, namely the influence of international bureaucracies on policy making and/or institutional design. This analytical puzzle leads us to the following general research question: What is the influence of secretariats of international organizations on the development of institutional design?

Societal relevance:

By focusing on the internal politics of the Secretariat, it becomes possible to better understand the underlying mechanisms behind influencing policy making. Moreover, the insights gained from this study can also be used to improve bureaucratic functioning. Knowledge about the internal politics of the Secretariat can also help to better understand the organizational performance of the GATT/WTO. Moreover, the growing importance of international organizations has led to questions about their democratic legitimation. After all, the staff of these organizations has not always been elected democratically. However, they do play an important role in international politics and can influence policy within a particular field. This is particularly true in the case of the European Union. The European Union faces severe criticism as it is being described as a cumbersome institute that is only interested in achieving results they unilaterally wish to achieve. Next to that, the public sees it at the same time as an incredible powerful institute and believes the internal bureaucracy has more power than their own national bureaucracy. This can be seen when you look at the primacy of European law. When there is a conflict between European law and national law, European law will prevail. This reinforces the suspicious feeling European citizens have towards the European Union as an institution (Ellinas & Suleiman, 2012, p.195). This study can attribute knowledge to the more general debate of democratic legitimation of international (governmental) organizations. Additionally, it can shed light on the internal functioning of international bureaucracies and how the staff working within these organizations operate.

Academic relevance:

Even though the influence of public bureaucracies on national policy has been studied extensively, this literature mainly focused on the functioning of national bureaucracies rather than international

bureaucracies (Ellinas & Suleiman, 2012, p.13). This inattention can be explained by the emphasis on state-centric theories describing international politics, that prefer the role of states in international relations and minimizes the potential influence of (international) bureaucracies. Moreover, important international relations scholars such as Keohane argue that international organizations are merely facilitators of international cooperation, and do not possess the autonomy to influence decision making (Keohane, 1984). In line with this view, states create international organizations in order to gain from their knowledge and reduce collective decision-making problems. In recent years, more scholarly attention has been paid to the role of international organizations within international politics (Biermann & Siebenhüner, 2009, p.2). Several scholars have focused on how, and if, international bureaucracies can influence decision-making and institutional design. Whereas some scholars have focused on the relationship between international bureaucracies and member states within an

international organization (Hawkins et al., 2006), others focused on the internal logics of international bureaucracies to understand their functioning (Barnett & Finnemore, 2004). This thesis fits within the

(5)

5 wider range of research concerning the functioning and possible influence of international

bureaucracies. In contrary to state-centric theories, this study will argue that it is indeed possible for an international bureaucracy to influence decision-making.

Methodology and outline of the thesis

This thesis can considered to be a qualitative single-case study. The data collected for this thesis are mostly primary and secondary sources. More specifically, official documents from the GATT/WTO and documents from the Uruguay round have been collected to undertake a thoroughly document analysis through process tracing to establish a high degree of knowledge on the case. Additionally, several books written by experienced authors who wrote extensively on this topic have been consulted to establish a complete body of thought to answer the research question.

The outline of this thesis is as follows. Chapter two discusses the theoretical framework of this thesis and argues why the principal-agent (PA) theory has been chosen to help answer the research question. Chapter three is dedicated to the research design, data collection and conceptualization of this thesis. Chapter four is divided in two sections. The first section discusses the history of the dispute settlement procedure of the GATT. The second section will take a closer look at the role of the GATT/WTO Secretariat in the revised dispute settlement procedure. The actual influence of the GATT/WTO Secretariat will be discussed by using the PA theory. In chapter five, some concluding remarks will be given concerning the research question, value of PA theory, the generalizability of the research and directions for further research.

(6)

6

Chapter 2. Theoretical framework: Review of the literature

There are several approaches that can be used to analyze the influence of international bureaucracies in international organizations. This chapter will discuss these perspectives and argue which one is most suitable for this research.

2.1 PA theory

PA theories have been employed in a wide variety of economic and social settings. Only later entered PA theory the political domain (Miller, 2004, p. 203). PA theory is a general economic theory of behavior. More specifically, the theory discusses the complex and asymmetric relationship between authority on the one hand and expertise on the other. PA theory argues that the agent has an advantage compared to the principal due to its expertise on a particular matter. Whereas the principal is in control over the authority, PA theory mainly focusses on the authority of the principal to appoint certain stimuli upon the agent. The theory then analyzes if the principal can tempt the agent to act accordingly to the preferences of the principal, which decreases the chance that the agent will develop its own interests that are not in line with those of the principal (Miller, 2004, pp.203-204).

Several scholars have used PA theories in the context of international relations literature. Garrett (1995) took a PA approach in analyzing the politics of legal integration in the European Union. He argues that the European Court of Justice (ECJ) functions as an agent to the member states of the European Union. The ECJ can act as a rational agent and is able to develop its own interests. He claims that in the history of the European Union it can be noted that the ECJ’s rulings favored the expansion of European law and at the same time strengthened the ECJ’s reputation in curbing

dominant member states (Garrett, 1995, p.173). Majone contributes to the knowledge of PA theory by analyzing the governance structure of the European Union. He argues that there are two specific reasons why principals delegate to agents: to reduce the costs of decision making and to enhance the trustworthiness of policy commitments (Majone, 2001, p.103). Moreover, Majone argues that the Treaty of Rome, the framework upon which the European Union was build, was merely a set of vague and incomplete specifications and objectives. It is then left to the European institutions to specify these tasks. This incomplete contract led to problems of imperfect commitment between the several

institutions. This has resulted in more implementation and initiating of legislation powers for the European Commission. To understand why the European Commission has received such powers, Majone argues that it is important to go beyond the regular principal-agent relationship and focus on a fiduciary relationship. In the case of the European Union, member states want to show that they are committed to the integration of the European Union, which in practice has translated into the European Commission having a role as trustee and the member states as principal, who deposit confidence and reliance in the trustee (Majone, 2001, pp.113-119).

In more recent years, the creation of large influential international organizations have spurred the interest of PA theory scholars. Contrary to realist assumption claiming international organizations have a mere clerical role, an increasing amount of scholars is convinced that international

organizations do in fact play an important role in the international arena. In turn, scholars lifted the domestic focus of PA theory to the international domain, and applied it to international organizations. The international organization then takes on the role of agent, and member states become the principal. Also in international organizations, the interdependence between the principal and the agent can be noticed. The question of expertise and information asymmetry becomes particularly apparent in international organizations, as they often deal with matters that are very technical in nature. Especially the task the agent is assigned to will influence the relationship the principal has with the agent and consequently the ability for the principal to control the agent. Specifically, when an agent is assigned with a judicial task, the agent often enjoys substantial discretion. This makes it harder for the principal to put control mechanisms in place. Next to that, particular processes can be hard for the principal to monitor. Especially during negotiations, agents have substantial room to maneuver as negotiations are

(7)

7 often a long and unclear process (Hawkins et al., 2008, p.36-37). Member states are often

disadvantaged, as representatives frequently change while the officials of the organizations build a long career within the organization. Moreover, when a large number of principals is needed to authorize certain policy, which is often the case in international organizations, autonomy for the agent also increases because preferences among these principals often differ which makes it harder to reach a common agreement (Hawkins et al., 2006, p.22).

2.2 Institutional perspective

Institutionalist place, in contrast to realist scholars, a great emphasis on the importance of international organizations. They believe that international cooperation between states is possible due to the uniting characteristics of international organizations. They believe international organizations have the power to facilitate cooperation and maintain peace. Moreover, international organizations can function as referees, by punishing those states that do not obey by the rules (Kolodziej, 2005, pp.150-155). Most international relations theory have little interest in the functioning of international organizations, because they regard them as only serving the interest of the most powerful member states. Reinalda and Verbeek on the other hand claim that there is certainly an important place for them. Reinalda and Verbeek take an institutionalist perspective as they underline the idea that international organizations are of great importance. By analyzing three factors, globalization of the economy, increased regional integration and the end of the Cold War, they seek to understand if international organizations have increased their autonomy in recent years (Reinalda & Verbeek, 1998, p.4). They find that some organizations indeed saw an increase of their autonomy, and others not. Interestingly enough,

politically salient organizations such as the International Atomic Energy Agency and Organization for Security and Cooperation in Europe gained autonomy.

This is in sharp contrast with the (neo)realist view, that claims that states want to keep salient matters into their own hands. Moreover, even when international organizations have found a way to increase their autonomy, this is often in the field of ‘low’ politics. Reinalda and Verbeek show that this argument simply does not hold. The International Labour Organization did not manage to increase autonomy. The authors claim that this is due to competition between international organizations, that are all interested in being the most prominent one (Reinalda & Verbeek, 1998, p.5). Reinalda and Verbeek demonstrate that international organizations have various tools available to influence policy making. First, their expertise on technical matters to come up with specific policy. Second, their ability to transform into bridge builders to unite coalitions (Reinalde & Verbeek, 1998, p.6). The former relates to the fact that international organizations are often more acquainted with the topic than their member states, as the staff that serves the organization has worked there for many years and representatives from the member states often rotate. Thus, an information asymmetry often exists which enables international organizations to develop expert knowledge. The latter refers to the ability of international organizations to unite transnational coalitions. These coalitions consist of interest parties within states that are able to influence the policy interests of the member states (Reinalda & Verbeek, 1998, p.22). This result also corresponds with the general idea of institutionalism, as institutionalists claim that international organizations have the ability to facilitate cooperation. Reinalda and Verbeek discuss the role of international organizations within a range of other trends international decision making is facing. Traditional theories of international relations claim that nation states are the most important actors. However, Reinalda and Verbeek argue that international decision making has become a game where government actors and non-government actors play alike. States can no longer play along to their own interest, but have to take domestic preferences into account. This is especially important when discussing international decision making (Reinalda & Verbeek, 1998, p.21).

(8)

8 Another important scholar in the institutionalist field describing the functioning of international organizations is Keohane. He takes an institutionalist approach as he underlines the idea that

international organizations have the ability to align the interests of governments to make arrangements between them possible. More specifically, Keohane identifies two central problems within

international cooperation, transaction costs and informational imperfections, that international organizations can help to address (Keohane, 1989, p.110).

First of all, international organizations can lower the transaction costs of international cooperation by providing a framework that establishes legal liability, and thus gives more confidence to governments that rules are adhered to. Secondly, in international politics it is only seldom the case that governments are fully aware of each other’s intentions and preferences. This causes distrust, because governments are afraid that others might deviate from the agreement. Distrust amongst governments makes it difficult to find a common ground to cooperate on. International organizations can be useful as they can function as a platform where repeated interaction can take place. When governments regularly interact on a familiar platform, governments can become more informed on the intentions of each other, which decreases informational imperfection (Keohane, 1989, p.115).

Archer adds to the understanding of international organizations by taking an institutionalist approach to discuss their functions. He underlines the ideas of Keohane in the sense that he agrees that

international organizations play an important role in offering a platform where cooperation can arise (Archer, 1994, p.11). However, he also emphasizes their role as independent actors. By analyzing several institutions of the European Union, such as the European Commission, Archer argues that international organizations can normalize relations between their member states and in turn create a sense of belonging, a community feeling which also enhances cooperation. Archer claims that this can also be seen in the history of the European Union: since the end of the Second World War a

‘Community Europe’ has developed (Archer, 1994, p.270).

2.3 Constructivist perspective

The constructivist perspective developed as a reaction to the individualist and rational ideas of how international politics work, that are often present in traditional realist and liberalist approaches. Whereas rational approaches believe that international organizations are simply arenas where action takes place, constructivists believe that international organizations are also actors in their own right and are able to shape the behavior and ideas of actors working in these organizations. Moreover, international organizations define the formal and informal rules where actors mirror their behavior upon (Guzzini, 2000, p.147-153).

Until the 1980s, there was little scholarly interest for the creation and functioning of international organizations. Especially during the late 1980s, international politics was focused on the Cold War period which influenced international relations theory through the ideas of Kenneth Waltz’s neorealism (Barnett & Finnemore, 2004, p.7). The first scholars that paid attention to international organizations mainly saw them as arenas in which states acted. They were no actors of their own, but simply a place where states had the possibility to cooperate. When international organizations started to receive more attention, many scholars used theories of organization to uncover the mechanisms through which international organizations function. According to organizational theory, international organizations can reduce transactions costs and help when problems of incomplete information arise. These theories can help explaining why international organizations exist in the first place, but are less suitable to understand how they function in practice (Barnett & Finnemore, 2004, p.8).

As organizational theories lack the ability to fully understand the functioning of international organizations, Barnett and Finnemore turned to a constructivist approach to provide for a better understanding of international organizations. They concluded that the ‘logic of bureaucracy’ is helpful in providing explanations for the autonomy of international organizations and in understanding how

(9)

9 they are created, expand and sometimes fail. The logic of bureaucracy means that Barnett and

Finnemore treat international organizations as bureaucracies. This means that each international organization has its own internal dynamics. Barnett and Finnemore align themselves with the ideas of constructivism as they claim that international organizations have the ability to create their own rules and thus are actors in their own right. With these rules, they regulate and construct the social world. Thus, international organizations become actors of their own and are capable of shaping the people within the bureaucracy (Barnett & Finnemore, 2004, p.3). International bureaucracies are also able to adapt their way of working to a changing environment which makes it likelier they persist over time. However, when they fail to do so this might lead to a collapse of the organization. This also explains why some international organizations fail. Change does not only occur when powerful states demand so, as is assumed in traditional international relations theory. Instead, change is often path dependent due to a robust organizational culture that has persisted for many years (Barnett & Finnemore, 2004, p.8).

Finnemore and Sikkink elaborate on the ideas of constructivism by emphasizing the importance of norms and ideas, that are not only developed by states, but also by international organizations. They take a constructivist perspective as they argue that the power of international organizations lies within the fact that they are able to set standards of appropriate behavior for states to abide to. Moreover, international organizations can also take the role of norm entrepreneurs. Norm entrepreneurs focus on particular issues and call attention to it. At other times, international organizations can become the organizational platforms from where norm entrepreneurs are able to advocate and publicize their norms (Finnemore & Sikkink, 1998, p.899). The authors describe that norm emergence can be divided in three stages; norm emergence, norm cascade and internalization. They claim that international organizations also play a large role during the internalization, as norms only have a chance of surviving when they are formalized into a set of rules within an organization (Finnemore & Sikkink, 1998, p.900).

March and Olson also underline the ideas of constructivism by stating that international organizations have increasingly taken an important role in contemporary life. They elaborate on the constructivist perspective as they state that the way our society is ordered does not only depend on the particular economic and social conditions, but also how we as a society have designed our institutions, such as our parliament (March & Olson, 1984, p. 738). International organizations help to arrange and understand international politics and the influence it has on society. Marc and Olson argue that whereas rational approaches argue that the functioning of international organizations depends on the individual choices actors make, it actually are the interactions between the actors and the organization that influences behavior of not only the organization but also the actors within (March & Olson, 1984, p.736). How actors behave and how society is mapped should be seen as a circle, as ‘political

institutions affect the distribution of resources, which in turn affects the power of political actors, and thereby affects political institutions’ (March & Olson, 1984, p.739).

(10)

10

2.4 The decision for PA theory

After I had looked into three perspectives that analyze how international bureaucracies can influence international organizations I came to the conclusion that PA theory would be most suitable to answer the research question. More specifically, I will use PA theory as described by Hawkins et al. to provide a testable theoretical framework. PA theory as described by Hawkins et al. gives me the necessary tools to understand to what extent an international bureaucracy has influence. PA theory focuses on conditions why a principal would delegate autonomy to an agent. The institutional and constructivist perspective simply do not look at this. Thus, PA theory specifically focuses on the delegation aspect which is very useful in this research as it is about the member states delegating tasks to the Secretariat in order to have an effectively functioning dispute settlement. The benefit of PA theory in comparison to the other two perspectives is that it emphasizes and explains under which conditions after delegation the agent is able to initiate and influence decision-making. Consequently, PA theory is able to go beyond the mere focus on a contract between the principal and agent. Whereas the institutional and constructivist perspective do add valuable knowledge in understanding the influence international bureaucracies within international organizations can have, they do not give me the tools needed to develop a testable theoretical framework to answer the research question.

First of all, the argument given by constructivists that international organizations are able to develop norms and values that actors can mirror themselves to and thus influences the actors is in itself interesting, but very hard to test. Second of all, the argument that international organizations can both function as a platform that enables states to cooperate and act as a neutral actor to punish and guide cooperation, as institutionalists argue, is also an interesting perspective but again, very hard to test. Moreover, PA theory dives deeper in the underlying mechanisms of influence than the other two perspectives do. First, PA theory considers the discretionary room to maneuver which possibly can be a source of influence for the agent. Second, PA theory as described by Hawkins et al. dives deeper into two underlying mechanisms of the theory. It shows that PA theory can be divided into two

perspectives; the autonomy perspective and the political control perspective. The autonomy

perspective relates to the attempts of the agent to gain autonomy. The political control perspective on the other hand highlights the side of the principal and whether the principal is willing or not to restrict that autonomy. These two perspectives do not contradict themselves but are part of the whole that is PA theory. Applying these two perspectives to the case of the GATT/WTO Secretariat allows me to understand under which conditions either the member states as principal were able to restrict the autonomy of the GATT/WTO Secretariat or the Secretariat was able to gain autonomy, and thus which perspective in this case gained dominance.

The PA framework as described by Hawkins et al. provides several building blocks upon which PA theory is based. I will now highlight three aspects that I think will be valuable in answering my research question.

1. Discretionary room within the contract

Interactions between the principal and agent are always driven by a contract. This contract does not need to be explicit, but can also be based on an informal agreement. According to PA theory, agents are not merely the extension of the principal. Agents in fact can develop their own interests and may act accordingly. This is also known as agency slack, the undesired maneuvers of the agent that are not in line with the preferences of the principal. It can appear in two ways; through ‘shirking, when an agent minimizes the effort it exerts on its principal’s behalf, and slippage, when an agent shifts policy away from its principal’s preferred outcome and toward its own preferences (Hawkins et al., 2006, p.8). When the agent has established some control, autonomy can describe the range of possible autonomous action. When the contract is based on an informal agreement, more discretionary room is available to the agent. Discretion differs from autonomy in the sense that ‘discretion entails a grant of authority that specifies the principal’s goals but not the specific action the agent must take to

(11)

11 accomplish those objectives’ (Hawkins et al., 2006, p.8). Discretion increases the autonomy available to the agent, but decreases autonomy when control mechanisms are set in place.

2. Delegation due to expertise

Hawkins et al. identify several reasons why principals would delegate authority. Amongst others, principals delegate tasks when substantial policy externalities, such as problems with cooperation, arise. Agents may then serve as a coordinating mechanism, thereby limiting the costs of externalities (Hawkins et al., 2006, p.15). Moreover, the principal might delegate to an agent when, due to fixed preferences, decision making is troubling. Thus, when it is difficult for states to agree on a certain agreement, they might appoint the principal to set the agenda (Hawkins et al., 2006, p.16). Other reasons to delegate can be identified, however, one particular reason stands out. This reason will help me answer my research question, as the Secretariat is an actor with considerable expertise. Principals delegate because the agent is a specialized actor that has the knowledge, ability, competence and/or time to execute the delegated task. Often, especially with highly technical topics, states lack the expertise to handle these themselves (Hawkins et al., 2006, pp.13-14).

3. Control mechanisms

Hawkins et al., identify five mechanisms principals can use to control their agent. In general, principals aim to guide the incentives of agents beforehand to ensure the preferences of agents are in line with those of the principals. Besides this effort, the principal can try to specify the contract as clearly as possible, to limit discretion. Second, the principal can opt to install monitoring demands, meaning the agent needs to regularly report on their actions. Third, principals can control who becomes the agent by imposing strict screening programs. This way, they can choose which agent looks most reliable. Fourth, the agent can design an institution with built-in checks and balances that create an opportunity for the principal to control the agent and reduce agency slack. Fifth, principals can impose sanctions against those agents that do not act in line with the preferences of the principal. Most often this is done by cutting the available budget.

From the above mentioned information I can conclude that it is indeed possible for the agent to influence the principal. The three aspects I highlighted above led me to the following conclusion. The agent is able to influence the principal when no formal contract is established which increases the discretion available to the agent and when there is a lack of control mechanisms set in place by the principal to control the agent. Moreover, delegation is likelier when the agent is a specialized actor that has expertise on the topic and is able to create an information asymmetry which is in favor of the agent. Delegation is also likelier when the agent has expertise in a particular subject.

From the decision to use PA theory as main argument to develop my theoretical framework I expect that the Secretariat was able to influence the revised dispute settlement at least to some degree. This expectation comes from the fact that no formal agreement was established between the GATT/WTO member states and the Secretariat can considered to be a specialized actor with considerable

knowledge on the GATT/WTO history and law. Therefore, the general research question is as follows: To what extent can the influence of secretariats of international organizations be attributed to their expertise and the design of the contract between the secretariat and its member states?

(12)

12

Chapter 3: Research design and data collection 3.1 Design and Methods

Qualitative single-case study

My thesis is a qualitative single-case study design. Instead of looking across multiple units, I will analyze several pieces of literature which are discussing a single unit. Whereas a comparative design looks at a small number of variables for many cases, my thesis will analyze many variables that all concern one case. By conducting a single-case study, I am able to analyze alternative explanations and focus on the underlying mechanisms which allows me to develop a deep understanding of the case. Moreover, I am able to compare competing explanations of a particular outcome. My case mostly resembles the conditions for being a crucial case, based on a most-likely design. This type of case is appropriate when the goal of the study is to test a particular theory, in this case PA theory (Levi, 2008, p.12). With a single-case study design I aim to explain in detail the outcome of a single case, in my case the establishment of the Appellate Body.

A large problem with qualitative single-case studies is often that they are prone to selection bias. Cases are chosen non-randomly which harms the external validity of the case. Therefore it becomes quite difficult to generalize beyond the particular case. However, it is often not the purpose of case studies to generalize. Instead, it tries to understand the underlying mechanisms of the particular case in depth to provide for a full explanation (Toshkov, 2016, p.285). As my case is a single-case study, it also harms the above mentioned principles. I tried to tackle the issue of selection bias by making use of triangulation. I identified and collected several different pieces of data from different authors to make the possibility of bias as low as possible. I ensured that these pieces of data were generated independently, to avoid possible bias. More specifically, I identified three authors that have extensively written on the Secretariat of the GATT/WTO in relation to the Uruguay round. I

thoroughly read each piece of literature and analyzed which events they agreed and disagreed on. By only using information they agreed upon, I increased the likelihood of it being true and it also decreased the chance of author bias.

3.2 Conceptualisation 3.2.1 Influence

As the theoretical framework has established, two perspectives can be distinguished in PA theory; the autonomy perspective and the political control perspective. In conceptualizing the influence of the autonomy perspective I have chosen to use a definition from another author than the one my

theoretical framework is based upon, Hawkins et al. The reason for this decision is that Hawkins et al., do not provide an understanding of the term influence, and what possibly could be influenced. The authors simply use the term without providing a clear definition for it. The lack of definition would make it impossible to objectively measure the potential influence an international bureaucracy has. More importantly, I could not conclude in my case whether the GATT/WTO Secretariat had any influence. This would leave me with only a vague guess of what Hawkins et al., mean with the term influence. This would substantially damage the quality of conceptualization and eventually the thesis as a whole. Thus, I searched for other authors that used PA theory to establish their case. There is a risk in using different authors for the same conceptualization. Potentially, you could be measuring two different concepts. I tried to tackle this problem by ensuring that the conceptualization of influence is still based on the theoretical framework of PA theory. In the work of Biermann and Siebenhüner I found a good alternative. In their book, the authors too describe the influence international bureaucracies can have in policy making. They do not only look at the degree of influence

international bureaucracies can possible have, but also at the types of influence. In other words, they look at which processes could possibly influence an international bureaucracy. Moreover, their research draws upon several key features of PA theory. Amongst others, Biermann and Siebenhüner

(13)

13 claim, as does Hawkins et al., that the environment created by the principal defines the room to

maneuver for the agent, the saliency and technicality of a problem determines whether or not the principal is willing to delegate the problem to the agent and last, the competences of the agent also determines delegation from the principal to the agent (Biermann & Siebenhüner, 2009, pp.49-52). These are all features that form the basic structure of PA theory. By ensuring that the theoretical framework of Biermann and Siebenhüner is comparable to that of Hawkins et al., I could in fact still measure the possible influence of the Secretariat due to a clear conceptualization which is still based on the PA theoretical framework.

Biermann and Siebenhüner describe that the influence of international bureaucracies can be measured by looking at three processes; ‘the output, which is the actual activity of the bureaucracy; the outcome, which is the observable changes in the behavior of actors targeted by the bureaucracies’ output

(including unintended consequences) and the impact, which is the change in economic, social or ecological parameters that result from the change in actors’ behavior’ (Biermann & Siebenhüner, 2009, p.41). The distinction between these different processes is particularly useful as it allows for the categorization of the different variations of influence within international bureaucracies. It also helps in combining the research on influence of international bureaucracies with the broader literature on the functioning of international organizations. However, focusing on all three processes would be beyond the scope of this research. Thus, this thesis will focus on the second process, the outcome process. The outcome in this case is particularly interesting to research as it provides for more information than solely the output process. In practice this means that I will research if the establishment of the

Appellate Body was in line with the interests of the principal or the agent. Thus, I will make a comparison between the interests of the agent before the establishment of the Appellate Body and after. I will look into the possibility that the agent was at least for a certain degree able to influence the relative change in behavior and interest of the principal. I have chosen to focus on the outcome process rather than the impact process as studying impact is a very demanding process, which unfortunately is beyond the scope of the research. For this thesis it is too time consuming to study impact on policies in the economic, social or ecological domain as impact denotes a very broad scope.

To measure influence of the political control perspective I have chosen to use the conceptualization as described by Hawkins et al. The reason for this is that, contrary to the autonomy perspective, they provide a clear and measurable concept for this. Whereas in the autonomy perspective the reader must guess what is meant with influence on the side of the agent, Hawkins et al. clearly specifies how the principal is able to exert influence for the political control perspective.

The principal is able to exert influence due to the control mechanisms it imposes on the agent. By imposing control mechanisms, the principal hopes the agent will act in line with their interests. By imposing control mechanisms, the principal sets out a specific path of actions and behavior that the agent is supposed to follow. This way, the principal hopes to avoid agent shirking and slippage as much as possible (Hawkins et al., 2006, p.8). The principal is able to exert influence by designing a formal contract with fixed rules that describe how the agent is expected to act. When the principal designs the contract that allows for discretionary room but at the same time imposes control mechanism, there will be less possibility for the agent to influence the principal than when a fixed contract with preset guidelines is in place. Because, as Hawkins et al. argue, imposing control mechanisms is the best option for the principal to exert influence (Hawkins et al., 2006, p.8).As a result, more control mechanisms and a formal contract will enable the principal to exert influence.

(14)

14

3.2.2 Autonomy

Bureaucratic autonomy is the scope of instruments available to the agent to potentially undertake independents actions after the principal has installed a set of control mechanisms. More specifically, autonomy entails the room to maneuver for agents after the principal installed ‘screening, monitoring, and sanctioning mechanisms intended to constrain their behavior’ (Hawkins et al., 2006, p.8).

Autonomy can be practiced by the agent to either act in line with the preferences of the principal or undermine these preferences. Moreover, autonomy is the inevitable consequence of the incomplete control the principal has over the agent. Autonomy also increases when a large number of states is needed to approve a certain decision (Hawkins et al., 2006, p.21).

3.2.3 Delegation

Delegation entails the conditional allocation of power from a principal to an agent. This in order to enable the agent to operate on behalf of the principal. The allocation of power is often restricted in time or applies to specific tasks. Thus, the principal is still able to potentially take back power. Moreover, for delegation to occur the principal and agent must have a mutually constitutive

relationship. Thus, the principal depends on the agent for being a principal and the other way around (Hawkins et al., 2006, p.7).

In my conceptualization I have described three core concepts: influence, autonomy and delegation. I have chosen these concepts because they describe the core values of PA theory. By using influence, I am able to both measure possible influence on the side of the principal and on the side of the agent. I have specified which process the agent can influence and how the principal on the other side can impose mechanisms and design the contract to exert influence itself. This way, I can measure which side was able to exert more influence in my case. By specifying what autonomy entails I am able to understand which instruments the agent is left with after the principal has installed control

mechanisms to in turn exert influence. By describing delegation I am able to understand that there exists a mutually constitutive relationship between the principal and agent, and more specifically, it is a conditional allocation of power meaning the principal can revoke power.

I will take these core concepts and apply them to the case of the Secretariat of the GATT/WTO, during the establishment of the revised dispute settlement procedure which ultimately led to the Appellate Body. This had led me to the following research question, applied to the case of the GATT/WTO: To what extent is the revised dispute settlement procedure a result of the preferences, interests or dominant ideas of the Secretariat of the GATT/WTO?

3.3 Data collection

In this thesis I mainly collected my data from primary and secondary sources. More specifically, I collected documents that were gathered in the negotiation process of the Uruguay round, official documents from the GATT/WTO and publications from researchers that have studied this topic. This part will go further into detail and discuss which documents I have used in this thesis.

Documents that were produced during the negotiation process of the Uruguay round can considered to be primary sources. These documents will be important for my thesis as I expect that they will provide me with a timeline of what happened during the Uruguay round negotiations. I expect these documents will help me understand what the preferences of each actor were, and how these preferences evolved during this round.

(15)

15 Throughout this thesis, three main authors will be used to analyze the Secretariat starting in the 1950s until the beginning of the 1990s. The authors Xu and Weller, Hudec and Marceau & Azevedo are chosen for their expertise on this topic. Each author has written extensively on the functioning of the GATT/WTO in their relationship with the Secretariat. By identifying multiple authors and see on which events and underlying reasons they agree on, I hope to avoid author bias as much as possible. Moreover, the GATT and WTO have kept a rich online archive in which every document is published regarding their history of dispute settlement, both during the times of GATT and WTO. These

documents will be useful in establishing the evolution of dispute settlement. Moreover, it will help me understand which preferences, that of the principal or agent, prevailed in the end.

3.4 Document analysis and process tracing 3.4.1 Process tracing

In this thesis, I will make use of process tracing. Process tracing is defined as ‘the use of evidence from within a case to make inferences about causal explanations of that case’ (Bennett & Checkel, 2015, p.4). Process tracing focuses upon establishing the causal mechanisms that led to a particular outcome. This is done by linking together various events, that are linked together as a sequence (Toskov, 2016, p.298). The evidence gathered from process tracing will help me to understand if, and how the Secretariat was able to exert influence in the establishment of the Appellate Body. By

applying process tracing, I hope to uncover the underlying information and motivation for the revision of the dispute settlement procedure, and which preferences ultimately prevailed. Process tracing research in this thesis consist of uncovering the chronology of events leading up to the Uruguay round and thereby the establishment of the revised dispute settlement procedure. Furthermore, I will look for details in the institutional context that will help me understand which actors and events were relevant for the final outcome.

3.4.2 Document analysis

Document analysis is in qualitative research particular useful to give meaning and asses documents regarding a certain topic. It is an efficient and effective method to establish and assemble information. In this thesis, three different types of document will be assessed to answer the research question. Official documents coming from a public source; the GATT and the WTO, publications from researchers that studied this topic and documents gathered in the negotiation process of the Uruguay round. Whereas most of the documents from the GATT, WTO and negotiation process are primary sources, the documents from authors should be considered secondary sources. I will assess these documents to find convergence and eventually verification that particular events indeed happened. Each document contains bias. It is important to acknowledge this and address it. I addressed bias in my document analysis by consulting multiple pieces of data and literature to establish which piece could count as evidence and which not. Whereas official documents from an international organization can be regarded as relatively trustful, I still needed to establish the authenticity of these documents. The same applies to documents from authors, which can contain an author bias. By comparing multiple pieces of literature each describing the same event I avoided bias as much as possible. With every piece of data I assessed the original purpose of it, if it was a primary or secondary source, the targeted audience and whether or not the data was edited at a later point in time.

(16)

16

3.5 Summary of research strategy

This is a qualitative single-case study design, where based on the PA theory, it will be researched to what extent the revised dispute settlement procedure is a result of the preferences, interests or dominant ideas of the Secretariat of the GATT/WTO. By carrying out a document analysis in combination with process tracing, I hope to develop a historical overview of the history of dispute settlement within the GATT. Moreover, I aim to provide a conclusion if with the establishment of the Appellate Body, the autonomy perspective or the political control perspective ultimately prevailed. By focusing on the outcome, thus the Appellate Body, I will make a comparison between the behavior and interests of the principal and agent before the establishment of the Appellate Body and after. By analyzing which interests were in line with the establishment of the Appellate Body, I am able to conclude which perspective dominated. As PA theory claims that there is a possibility for the agent to influence the principal, I will look into the possibility that the agent was at least for a certain degree able to influence the revision of the dispute settlement procedure.

(17)

17

Chapter 4: Analysis of the GATT Secretariat

This chapter is divided into two parts. The first will discuss in chronological order the history of the GATT dispute settlement procedure. More specifically, it will start with a background of the history of the GATT/WTO. In the second part I will focus on the analysis. In this part I will apply my two theoretical perspectives, the autonomy perspective and the political control perspective, to the case of the GATT/WTO and argue which perspective and interests are most in line with the establishment of the Appellate Body.

4.1 Historical overview

This part will start with a historical background regarding the dispute settlement procedure. By doing this, I am able to make a comparison between the decades before and after the establishment of the Appellate Body. This comparison will help me understand if, how and in which particular stage the Secretariat has been involved in dispute settlement throughout the years, and how this has changed with the revised dispute settlement. Moreover, this historical background will give me the foundation needed for the analysis in the second part of this chapter.

After World War II, several countries called for the creation of an agreement that would stimulate international free trade. This was supposed to lead to the establishment of the International Trade Organization (ITO). However due to the inability of the US Congress to ratify the treaty, the ITO never came into force. While member states were negotiating the establishment of the ITO, they were simultaneously discussing the GATT, which purpose was to immediately reduce tariffs between participating countries. When ITO negotiations failed, the GATT became the only viable alternative. The GATT then was signed in 1947, in which many principles of the ITO returned, but without the formal institutional character of an organization. Because the GATT was originally designed as a trade agreement, the agreement did not have a full grown legal design (Hudec, 1998, p.102). In practice this meant that there was not a clearly set out dispute settlement mechanism. Disputes were handled ad hoc by plenary meetings of the member states. This lack of a well-developed legal design was meant to satisfy the United States, which had very little room to negotiate and was uncomfortable with the creation of an international organization. The United States preferred the GATT as solely being a trade agreement. Besides having no clear organizational structure, the GATT did not have its own

secretariat. Thus, when the GATT opened its doors in 1947, it continued to function administratively by borrowing the Secretariat of the United Nations that was set in place to conduct the ITO

negotiations (Hudec, 1998, p.103).

Start of dispute settlement in 1947

This lack of legal design did not stop member states to submit cases to the GATT. As mentioned, the GATT did not have a firm legal basis to handle dispute settlement. This is not to say that the General Agreement does not mention which procedures should be followed when member states have a dispute. Article XXⅡ, named ‘Consultation’, described that each member should make a substantial effort to consult the other party when in dispute. If this was not sufficient to solve the conflict, the article states: ‘The CONTRACTING PARTIES may, at the request of a contracting party, consult with any contracting party or parties in respect of any matter for which it has not been possible to find a satisfactory solution through consultation’ (GATT, 1947, p.31). This in practice meant that in the first few years, the Executive Board consisting of representatives from the member states handled dispute settlement.

There is no official document describing the role of the Secretariat in the meetings of the Executive Board. However, as several authors have established already, it is very likely that the Secretariat had a considerable influence (Xu & Weller, 2004, Hudec, 1998). In these days, the Secretariat together with the Chairman, Dana Wilgress, put together the agenda for meetings amongst the member states. The Chairman had an important role. The Chairman was elected from the representatives of the member

(18)

18 states. This role had a predominantly organizational and procedural nature. The Chairman would preside each periodic meeting of the member states, lead the discussion and would have a general control over the course of the meeting (WTO, 1994, p.3). The Secretariat assisted the Chairman with each task it was assigned with. In practice, this meant that the Secretariat could also influence which point and in which order it was put on the agenda. This makes is likely that the Secretariat also did this for meetings of the Executive Board, where in the first few years dispute settlement was handled (Hudec, 1998, p.106). Soon, however, it became apparent that the Executive Board was ill-equipped to analyze and rule on these often complex disputes. Thus, the Chairman worked together with the Secretariat to deal with these cases (Xu & Weller, 2004, p.204). As the member states accepted the rulings of the Chairman, the capacity and expertise of the Secretariat was also confirmed. Member states started submitting an increasing amount of cases. Neither the Chairman nor the Secretariat was able to handle all these cases within a reasonable amount of time. Thus, a new system was needed to handle the large workload.

The 1940s and 1950s

While there was a great need for the establishment of either a legal panel or a tribunal to discuss the cases that were put forward, the General Agreement did not provide for this. Thus, in 1949, the Executive Board decided that working parties should be set up to rule on the cases. Originally, these working parties included the two disputing parties, and three member states that could considered to be neutral. It has been established that the Secretariat had a large influence during meetings of the working parties. After analyzing the case that had been put forward, ‘the Secretariat prepared a draft report in the name of the Working Party, which explained the conclusion of the majority’ (Hudec, 1975, p.79). Moreover, the Secretariat was also responsible for arranging the location and other details of the meetings and provided background information of the particular case that the working party needed to make a ruling on (Williams, 2015, p.101, in Marceau & Azevedo, 2015).

In 1952, the Director-General of the GATT, Eric Wyndham-White, together with the Chairman of the member states decided that a single working party should be formed that would handle all cases instead of multiple working parties (Williams, 2015, p.98, in Marceau & Azevedo, 2015). The name of this single working party also changed to ‘panel of complaints’. The functioning of this panel was fundamentally different compared to that of a working party. The Chairman appointed six member states to take a seat in the panel. Then, the member states chose their representatives to provide an opinion on the particular case. To ensure objectivity and impartiality, the Director-General added several other guidelines. Amongst others, the representatives who had a seat in the panel would not only be selected on their nationality, but also on their expertise of the General Agreement (Xu & Weller, 2004, p.205).

You might expect that the establishment of the panel decreased the influence of the Secretariat because the panel now consisted of skilled individuals with detailed legal knowledge of the General

Agreement. However, this was not the case. The Secretariat continued to work very closely with the panel. In fact, Jean Royer, the Deputy Executive Secretary in these days, who was known for his expertise on GATT law, drafted the final agreement in every case. During the closed doors meetings Royer played a particularly important role, as he participated in discussing the case and its final outcome. Through the years, the continuous involvement of the Secretariat created an atmosphere of respect and appreciation for the work they delivered. Therefore, in almost every case, the disputing parties accepted the final draft of the Secretariat. The influence of the Secretariat was widely

recognized not only by the disputing parties but also by the panel members (Hudec, 1975, p.88). In the more than ten years that followed, the panel accepted and ruled upon 53 cases that were brought forward. It must be emphasized that, as mentioned, dispute settlement was based upon diplomatic practices. Thus, when states had conflicting ideas on a certain topic, they were encouraged to first communicate extensively to try and solve their issues. If this did not achieve the desired result, the Director-General could step in as arbitrator and hold informal meetings in his office, providing each

(19)

19 party with a chance to argue his position (Xu & Weller, 2004, p.207). In this way too, the Director-General as head of the Secretariat had considerable influence in dispute settlement.

The 1960s

The 1960s marked a changing point in the trust the member states had in the ability of the GATT to deliver high quality dispute settlement (WTO, 2018, p.20). Only a handful of states brought cases to the panel. 1964-1970 proved to be an all-time low for dispute settlement, as no cases were brought forward. According to Hudec, two reasons explain this dramatic turn. First, the development of the European Economic Community (EEC). Even though the EEC to a large degree accommodated to the legal requirements of the GATT, two major policy areas did not; the Common Agricultural Policy and their preferential policies towards their former colonies. The EEC argued that member states should retain the right to handle according to their own policies when matters are economically and politically sensitive. Second, the membership of the GATT expanded in the 1960s. Mostly developing countries joined the GATT, which meant that the developing countries outnumbered the developed countries. This created conflict within the GATT as the developing countries demanded rules that were more in line with their interests (Hudec, 1998, p.109).

The 1970s

Then, in the beginning of the 1970s, large trading countries such as the United States adopted non-tariff restrictions that were in sharp contrast with the free trade spirit of the GATT. In turn, this decreased confidence in the ability of the GATT to effectively enforce dispute settlement. The Secretariat of the GATT and member states realized that if the GATT wanted to continue to play an important role in international trade, something had to change. Thus, when member states agreed to launch a new round of trade negotiations in 1979, the Tokyo Round, the dispute settlement procedure was also brought to the table (Hudec, 1998, p.110). Member states decided that in order to have an effectively functioning dispute settlement procedure, this first of all needed to be codified in the General Agreement. On 28 November 1979, the member states agreed to an ‘Understanding Regarding Notification, Consultation, Dispute Settlement and Surveillance’ (GATT, 1979). This ‘understanding’ basically meant the codification of existing practices into formal rules and reestablished the trust parties had in the ability of the GATT to settle disputes. During the Tokyo round negotiations, the Secretariat was also heavily involved. For each negotiating group, a member of the Secretariat was appointed to serve as a chairperson. More importantly, the group that was

responsible for negotiations regarding a revision of the dispute settlement was joined by a secretary from the Secretariat that had the responsibility to formulate the final draft (Williams, 2015, p.100, in Marceau & Azevedo, 2015). The Tokyo round turned out to be very successful. From 1980 till 1993, 42 cases were brought to the panel (Xu & Weller, 2004, p.21).

Whereas the general rules throughout the 1950s, 1960s and 1970s did not change substantially, the environment in which the dispute settlement procedure operated did. In the 1950s the GATT

Secretariat consisted of a small group of people, often career diplomats, specialized in the working of the GATT. Dispute settlement was aimed at consulting and legal rulings, where possible, were mostly avoided. Specialized lawyers were often not involved, because they could harm the atmosphere of consultation (Hudec, 1998, p.111). For the Secretariat, the 1970s proved to be a completely different environment in which they had to operate. Whereas in 1947 only 23 states were member of the GATT, by the time the Tokyo round began, this number increased to 102 members (WTO, 2018). The

changed environment was also reflected in the type of cases that were brought forward to the GATT. The United States was one of the first states that filed their legal cases through professional and specialized lawyers. The Secretariat on the other hand, together with most other delegations, was still used to the old-fashioned way of handlings claims (Xu & Weller, 2004, p.214). Soon, however, they realized that this way of working simply could no longer hold.

(20)

20

Procedural and organizational challenges for the Secretariat in the 1970s

The use of professional and specialized lawyers showed that the Secretariat and the panel procedure did no longer suit the demands of that time. Panels still consisted of six representatives from neutral member states (GATT, 1987, p.7). Whereas these representatives were competent in handling cases due to their extensive knowledge on the GATT, they were not equipped with substantial legal knowledge to provide the disputing countries with a sound legal ruling. Members of the Secretariat that assisted the panels often had more extensive knowledge on technical issues, but also lacked sufficient legal knowledge to provide a sound ruling. The second obstacle the Secretariat had in the 1970s compared to the 1950s was that the organization and their agenda grew heavily. This meant that where the Director-General in the 1950s invited disputing parties to come to his room to

diplomatically solve issues, this no longer was possible in the 1970s (Xu & Weller, 2004, p.216). In practice, this meant that on an ad hoc basis it was decided which department of the Secretariat would be responsible for handling it. Thus, there was no possibility for the individuals of the Secretariat to build up general knowledge and expertise as each case was assigned to the department that resembled it the closest.

This inability to appropriately handle cases soon became visible. One of the most prominent examples is the DISC case. The DISC case discussed the tax systems of the United States and EEC. The panel ruled that the tax system of the United States breached GATT law, specifically article Article XVI:4. Most member states of the GATT agreed with this ruling. However, at the same time the panel ruled that the tax systems of the Netherlands, France and Belgium also violated this article (Jackson, 1978, p.761). This caused great disapproval amongst member states as many believed this ruling was simply a way to diplomatically balance the two sides of the disputing parties. The EEC members simply did not accept the ruling, and continued to fight it. In 1981, the case was settled in favour of the

Netherlands, France and Belgium (Hudec, 1998, p.113).

The most obvious solution for the Secretariat would be to hire some expert lawyers that would guide and assist the panels in making legal decisions. However, this at first sight easy looking move was not very easy. In the 1980s, the United States and the EEC still strongly opposed any formal involvement of lawyers hired by the Secretariat. However, Arthur Dunkel, the former Director-General of that time knew that he needed to find a way to move the dispute settlement procedure forward. His challenge was to come up with a solution that would improve the current procedure but at the same time also satisfied the wishes of the United States and the EEC. Dunkel came up with a solution that can easily be called an impressive balancing act, and was in line with GATT’s long history of favoring

diplomatic solutions. In 1981, Dunkel established a new post within the Secretariat; the Director of Legal Affairs (Hudec, 1998, p.114). This new post was meant to be a temporary body, which was a first step in strengthening the capacity to handle disputes. By appointing an official from the Secretariat that had worked at the GATT for many years instead of a professional lawyer, Dunkel avoided criticism from the United States and the EEC but still managed to build the legal capacity of the Secretariat (Xu & Weller, 2004, p.215). The following two years did not cause any uproar amongst the member states. This led to the decision to make the position a permanent one. The permanent position increased the trust member states had in the ability of the GATT Secretariat to handle dispute settlement. This made it possible for the Secretariat to extend the position with two more lawyers and a secretary, thereby creating its own division.

Because member states brought more cases to the panel, the influence of the Secretariat also grew. In the 1950s the Secretariat was only responsible for providing adequate knowledge of the particular case and making a draft report of the final rule. However, during the ministerial meeting in 1982, it was decided that the responsibility of the Secretariat would be extended. From now on, ‘The Secretariat of GATT has the responsibility of assisting the panel, especially on the legal, historical and procedural

(21)

21 aspects of the matters dealt with’ (GATT, 1987, p.14). The division of Legal Affairs as part of the Secretariat continued to function accordingly in the meantime. Also the Director of Legal Affairs saw its role expanding. He was also asked to join the Chairman of the Negotiating Group on Dispute Settlement to advance compromises and solutions when two parties disagreed on a particular issue. This also extended the influence the Secretariat had on dispute settlement.

Road to a new trade round in the 1980s

The member states decided halfway the 1980s that a new trade round needed to be launched to discuss important matter such as services and property rights. Dispute settlement was also placed high on the agenda because in contrary to other agenda points, the member states were on the same page regarding this topic. In the past decades, the GATT had evolved from a diplomacy oriented organization towards a rule-based organization. This transformation needed to be reflected in the settlement procedure. Years of negotiating the dispute settlement procedure let to the following procedure. Member states agreed in 1994 that in order to have an effective dispute settlement system, a two-tier procedure must be established.

From that moment on, the Council, consisting of all the member states, would be responsible for the appointment of a Dispute Settlement Body (DSB). The DSB is the first step in the two-tier procedure. The DSB consists of an independent Chairman who is guided by several staff members. This body then becomes responsible for the operational procedure. Amongst others, this means the establishment of panels and reviewing the panel reports (Hillman, 2015, p.69, in Marceau & Azevedo, 2015). If one, or both of the disputing parties disagree with the final rule of a particular panel, they have the

possibility to proceed to the next step of the procedure; the Appellate Body. The Appellate Body consists of seven members that represents the membership of the WTO. Each member is appointed by the DSB for a four year, renewable term. When a case is brought to the Appellate Body, three

members are appointed to serve on the particular case (WTO, 1994, p.364). The fundamental change for the Secretariat is noted in the second step of the procedure. Neither the Secretariat nor the Office of Legal Affairs is allowed to be involved in the second step. The Appellate Body has become a separate body, nor its staff or budget is considered to be part of the Secretariat (Hillman, 2015, p.70, in

The GATT/WTO Secretariat: An analysis of the influence of the GATT/WTO Secretariat on the establishment of the Appellate Body