The Next Level of Regionalism. Deep Integration of Regio-Regio PTAs

Content

Tables

Figures

Abbreviations

1 Introduction
1.1 Problem
1.2 Goal of the Thesis
1.3 Methodology

2 Literature
2.1 Asymmetries and Power
2.2 Actorness in Interregionalism
2.2.1 Types of Interregional Relations
2.2.2 Institutionalization

3 Methodology
3.1 Qualitative Comparative Analysis
3.2 Fuzzy-Set QCA
3.2 Operationalization of Depth
3.3 Sample and Case Selection
3.4 Methodology for the case study

4 The Empirical Chapter
4.1 Measurement of the Conditions
4.1.1 Condition 1: Asymmetric Power
4.1.2 Condition 2: Actorness
4.2 Results of the QCA-Model
4.2.1 Fuzzy-set membership and calibration
4.2.2 Tests for Necessity and Sufficiency - transformational calibration
4.2.3 Tests for direct calibration
4.2.4 Assessment of the Results
4.3 Case Study: The EFTA-Central America PTA
4.4 Future PTAs: A short prognosis

5 Conclusion and further remarks

Literature and Sources

Annex I: XY-Plots

Annex II: R code

Tables

Table 1: PTAs under Analysis

Table 2: Merchandise trade statistics of the PTAs

Table 3: Trade Dependencies for all PTAs of the Sample

Table 4: Actorness related to decision processes in trade

Table 5: Recognition Values of PTA Parties

Table 6: Type of interregional relations for the PTAs

Table 7: Single Actorness Value for all regions

Table 8: Aggregated interacting actorness at PTAs

Table 9: Final Empirical Values

Table 10: Fuzzy-set membership scores

Table 11: Key Data about EFTA and Central American States

Table 12: Exports for Upcoming Agreements

Table 13: Raw values and fs-membership scores

Figures

Figure 1: Merchandise export as %-share of total world export

Figure 2: Output for individual test for necessity

Figure 3: Output for individual test for sufficiency

Figure 4: Output for neg. conditions for sufficiency

Figure 5: Output for neg. outcome for sufficiency

Figure 6: Output for necessary condition combinations

Figure 7: Output for testing necessity for ~DPT

Figure 8: Truth Table for transformational calibration

Figure 9: Output for the complex solution - transformational calibration

Figure 10: Output for individual test for necessity - round 2

Figure 11: Output for individual test for sufficiency - round 2

Figure 12: Output for neg. conditions for sufficiency - round 2

Figure 13: Output for neg. outcome for sufficiency - round 2

Figure 14: Output for necessary condition combinations

Figure 15: Output for testing necessity for ~DPT - round 2

Figure 16: Truth Table for direct calibration

Figure 17: Output for the complex solution - direct calibration

Abbreviations

Abbildung in dieer Leseprobe nicht enthalten

1 Introduction

Free trade is under attack. Left-wing and right-wing politicians in Western democra- cies alike criticize the current system harshly. Even the US president, the leader of the country that was responsible for installing the multilateral trade system in the first place, is sceptic of it and his administration frequently criticizes the World Trade Organization (WTO) for its decisions and regulations (Swanson 2017). What further connects all these critics beyond their rejection of free trade, is their answer to it. They dug it up from a very deep pit of long forgotten policies: protectionism. This new wave of protectionism goes even so far as the US president threatening to leave current Preferential Trade Agreements, namely the North Atlan- tic Free Trade Agreement (NAFTA). At least president Donald Trump wants to renegotiate this important FTA with Canada and Mexico (Swanson and Granville 2017). Even trade wars are nothing US President Trump refrains from announcing right now, even at the cost of an- gering transatlantic partners (Swanson 2018). Within the EU, the most institutionalized PTA to date, critics blame free trade and globalization for economic and even societal problems in their domestic countries (Vinocur 2017) or as responsible for the inequality in the world (Schreiber 2017). In the West, free trade never was so much under observation as it is today.

1.1 Problem

Economists on the other side are massively in favor of free trade. From their perspec- tive, global free trade is the most efficient solution for all allocation problems in the world, because the most efficient supplier will meet the demands in the end. But in contrast to this widespread opinion, regional PTAs are getting more and more common. In fact, PTAs are on the rise for decades and grow in space and depth. Nowadays, trade treaties are more and more frequently negotiated by groups of countries instead of single countries. Often these countries already share a treaty or are otherwise deeply connected and act as a common actor in trade negotiations. Some of these negotiations already led to PTAs, even if they pose a small mi- nority in the numbers of trade agreements around the world. However, they are steadily on the rise, especially since the 2000s. These group of PTAs are referred to as Regio-Regio-PTAs. This term shall clarify that no single country negotiates about a treaty, but groups of countries that might be connected by a more or less institutionalized organization.

Some of these treaties are even beyond mere questions of customs. They tackle fields such as intellectual property rights, standards and public procurements. So, regarding many issues they go even further than old fashioned bilateral treaties. But what are the conditions for deep integration in regio-regio PTAs? The thesis assesses this question with asymmetry and actorness. Those conditions are sufficient for deep integration in regio-regio PTAs. These trade deals of a new type can be called the next level of regionalism, because groups of coun- tries negotiate outside of the multilateral framework of the World Trade Organization (WTO).

In general, regionalism is on the rise for quite some time. Since the 1990s more and more regional trade agreements (RTAs), which are regional PTAs, see the light of day and since that time economists realized that a competing trade system is gaining popularity. The roots of regionalism stem from Europe and its long-lasting unification project. But in other regions of the world countries started to adapt these organizational forms and especially the trade liberalization and its benefits. The number of WTO registered RTAs more than quintu- pled from 54 in 1995, when the WTO was founded to 284 currently listed as active RTAs (WTO 2018).

Scientifically, the topic of regionalism and its effects on the World Trade System is widely debated. Especially in economics, dozens of renowned experts (e.g. Grossman and Helpman 1993, Bhagwati 1995a, Panagariya and Findlay 1996) debate the pros and cons of regionalism. This argument is known as the multilateralism vs. regionalism debate. There, economists argue about the effects of trade creation and diversion due to PTAs and point out the long-lasting effects of regionalism. A central point is the question, which way is better to achieve global free trade. Political scientists (like Janusch 2010) criticize the dominance of economists in the discussions about world trade and regionalism and opt for more interdisci- plinarity and the involvement of more political scientists. For Janusch, the economists lack a deeper understanding of political power structures in the international arena. So, they lack alternative explanations for the formation of PTAs as bargaining strategies (Mansfield 1998) and hegemony (Krasner 1976), as political analysists deliver them.

1.2 Goal of the Thesis

This thesis intends to analyze the current importance of the next level in regionalism. Trade blocs already negotiated trade agreements for themselves (this makes them a bloc in the first place) without direct involvement of the WTO and are currently doing so with other blocs (e.g. negotiations between the EU and MERCOSUR). This particular topic is aforemen- tioned in the literature. Baldwin (Baldwin 2011) views such regio-regio-PTAs as a dangerous development of power-based trade policies. But detailed research on regio-regio-PTAs is very scarce. In order to analyze them further, this paper uses the Design of Trade Agreements (DESTA) database by Andreas Dür, Leonardo Baccini and Manfred Elsig. This database de- livers a yet unseen overview over almost all PTAs existing in the world. Within those data is a 26 cases strong pool of regio-regio-PTAs. The thesis had to exclude some of them for various reasons (e.g. 6 PTAs are listed as “currently under negotiation”) and I focused my analysis to the new wave of regionalism that started in the nineties (Grimmel and Jakobeit 2015: 15, Mansfield and Solingen 2010: 148). Hence, the pool is reduced to 10 cases.

As I looked at the various regional organizations (ROs) that are part of my sample I realized that all of them generally support trade facilitations with other partners and mark those as their goals.¹ But this might only be lip service. Therefore, this thesis shall analyze structural as well as unique reasons for regions that might play a role in the success of eco- nomic cooperation between two regions. Political hypotheses about trade relations are the key to this thesis’ analysis of regio-regio, or interregional² trade relations. From a wide spectrum of literature, I derived two hypotheses that shall help me clarifying the conditions for deep integration between two regions. First, I followed the general arguments about interdepend- ence from Keohane and Nye (1999) and the use of this concept on the field of trade from var- ious other authors (Hirschman 1980, Mansfield 1998). This led to my first hypothesis, that power asymmetry between regions may account for deep PTAs. Since this is a common hy- pothesis or theory about international relations in general, the uniqueness of regions or ROs should be taken into account as well. So, secondly, I support the idea that actorness of regions should play a crucial role for the outcome of interregional negotiations. Actorness is a concept that considers that the role of regions as international actors remains an ambiguous one. It is composed of a variety of factors like cohesiveness, recognition and institutionalization (Bretherton and Vogler 1998, Stone Sweet and Sandholtz 1998, Allen and Smith 1990). Im- proved actorness should enable regions to negotiate more efficiently, leading to better and, according to their own goals, deeper PTAs.

1.3 Methodology

This thesis tests these hypotheses within a Qualitative Comparative Analysis (QCA) model. The model will test the 10 available regio-regio-PTAs taken from the DESTA data- base against the hypotheses derived above. Therefore, the thesis will carve out critical values for the hypotheses (or conditions, as they are called in the QCA context) for each bloc and will test them against the depth values for the PTAs taken from the DESTA database. The creators of the database formed an additive depth index based on seven provisions that are or are not included in the treaties. These are: tariff reductions to zero, intellectual property rights, services trade, public procurements, competition, standards and investments. Consequently, each PTA is measured with a value ranking from 0 (very shallow integration) to 7 (very deep integration). (Dür et. al.: 2014).

To measure the asymmetry, I formulated a trade dependency value based on the con- cepts of Zeng (2003). The necessary trade data were gathered from official databases of the IMF and the comtext Eurostat database. By comparing the exports of each bloc for each PTA, I created values for asymmetric trade dependencies. This value served as the cornerstone for the measurement of asymmetry in the final model. For actorness, I formed an additive index myself by combining concepts of actorness by various authors (Allen and Smith 1990, Breth- erton and Vogler 1998, Stone Sweet and Sandholtz 1998, Wunderlich 2012). The thesis based the index on institutionalization in the area of trade, recognition and type of interregional rela- tion and calculated a score for each bloc and summed it up for a total actorness value during the PTA negotiations. For my QCA model, I chose to do a fuzzy-set analysis. I transformed the values from the hypotheses in set membership scores of conditions and tested these against the given values for depth.

My QCA model delivered the finding that the occurrence of high actorness or high asymmetry is sufficient for a deep interregional PTA. Nevertheless, one case of my sample contradicted that finding. In order to determine the reasons for that, I conducted a qualitative case study for that case where the thesis combined the methods of Document Analysis and Process Tracing to gain better insights. To gather the necessary data for the case study, I tri- angulated information from academic articles, journalistic reports and official statements from the participating blocs. This served not only as a test of my own original hypotheses, but also led to further assumptions or alternative explanations that can help expanding my model in the future.

The thesis proceeds as follows. The second chapter reviews the relevant neoliberal in- stitutionalist literature on asymmetrical trade and the relevant literature on the concept of ac- torness. It further derives and operationalizes the hypotheses. Chapter three elaborates the research design and methodology and explains fuzzy-set QCA and the DESTA depth index in detail. The first section of the fourth chapter provides background information on the devel- opment of trade and interregional trade policy and the detailed presentation of the measure- ment results for the conditions. Section two delivers the QCA model for the case sample and interprets the results and assess the hypotheses. The third section revolves around the qualita- tive case study, its results and retropercussions for the model. In the last section the thesis intends a prognosis on future regio-regio PTAs. Chapter five concludes, reflects on the re- search process and suggests further enhancements of the model for future research.

2 Literature

This thesis analyzes an originally economic topic, interregional trade, from a political- scientific perspective. This chapter derives the hypotheses political the thesis refers to for the analysis. However, the entire topic is naturally researched thoroughly in economics as well. Trade and regional integration were and remain important topics in economics. Yet econo- mists focus mainly on the economic effects of regional integration and neglect political change or institutionalization, hence the term regional economic integration is more precise (as in Baldwin and Venables 2007). A huge number of famous scientists and experts have published on that topic in the still ongoing multilateralism vs. regionalism debate. This is one of the most virulently led academical debates to date. It originated in economics but was later adopted by political scientists and political economists alike. Since this argument has been raging for several decades now, many scientists already tried to collect the arguments and to give an overview or a chronic about the debate. Melo and Panagariya (1992) and Bhagwati (1995a) deliver a good overlook from the economical-theoretical side. Freund and Ornelas (2010) on the other hand give a short overview of empirical and theoretical studies about the topic.

Many authors wrote about the trade creating and diverting effects of Free Trade Agreements (FTAs) and Custom Unions (CUs) (Kemp and Wan 1976; Viner 2014³ ), others about the long-lasting effects of Preferential Trade Agreements (PTAs) in a country’s econo- my (Freund 2000; McLaren 2002). In addition, other economists also developed political- economic models to test the importance of lobbying for PTA policies (Grossman and Help- man 1993; Panagariya and Findlay 1996). Richardson (1993) explicitly refers to the NAFTA and the European unification process to empirically show how lobbying affects the spreading of RTAs.

Beyond classical economic approaches, some economists analyze other aspects of PTAs. The EU and the US have such bargaining power that they frequently include areas or policy fields that go beyond WTO trade talks. This gives these actors enormous standard set- ting power, because they enforce their own standards on other countries or groups of coun- tries that seek market access to Europe or the US (Horn et al. 2010). This development is dangerous for the multilateral trade system with its ‘everybody-is-equal’ approach in the WTO arena. It suggests that the classical economical multilateralism vs. regionalism debate is out of date and that RTAs are apparently better in fulfilling the demands for regulations than the multilateral WTO. But this could lead to a world where powerful trade blocs compete for supremacy, a situation comparable to the pre-World War I scenario in Europe (Baldwin 2011).

For exactly this lack of understanding on power structures in the international arena, scholars demand a stronger involvement of political science (Janusch 2010). Janusch (2010) analyzes the different positions of the debate but criticizes the economic point of view on the topic in general. For him the economists lack a proper explanation why RTAs are formed in the first place. Furthermore, they neglect the role of the WTO, especially its norm-setting power (Ibid. 2010: 371). The missing explanation is a political issue, therefore political scien- tists need to delve into this topic deeper. However, the debate did not wholly avoid political science. Especially in the fields of International Political Economy, Institutionalism and Re- gionalism a lot of scientists dedicated their work to the analysis of international and regional trade relations, how trade agreements come to pass and how these developments influence the world.

This study focuses on interregional integration. Yet, regional integration is of course the pillar of interregional integration. Regional integration is studied by various disciplines, among them law, economics and political science (Grimmel and Jakobeit 2015:10). Regional integration can be understood as a process of unification that is focused on development and permanency, even broadening and deepening and not limited to a certain geographic space (ibid.: 11). As political scientists focus on political cooperation, the institutionalization of the integrational process and maybe even supranational elements, the discipline developed sever- al approaches to research regional integration (Grimmel and Rüland 2015). This thesis focus- es on one policy field, trade. This field is traditionally a mix of politic and economic interests and is a potential source for international conflict (Dieter 2015). Trade often serves as the first step in regional integration and hence, it is an important cornerstone in all regional integration processes.

Not only frequently is economic integration the first integrational step. It was and still is a goal, if not the primary objective, of virtually every regional organization all around the world (e.g. for ASEAN see Feske 2015: 185-186; for the GCC see Holthaus 2015: 197-199; for ECOWAS and WAEMU see Tetzlaff 2015; for CARICOM see Thiery 2015: 362-364; for MERCOSUR see Bodemer 2015: 376-377). Reasons for this are, among others, the high costs of cross-national transactions. To lower such costs, a higher degree of integra- tion is often regarded as a promising solution, as it brings more coordination and dispute solv- ing mechanisms. Therefore, increased economic integration is among the first reasons for regional integration (Stone Sweet and Sandholtz 1998: 14). Promoting better relations and closer links to other regions and to negotiate trade agreements are also goals frequently men- tioned in the treaties of PTAs (Vaduz Convention Treaty Establishing the EFTA: Article 43; Treaty of Asuncion establishing MERCOSUR: Article 13; SACU 2002: Article 31; Treaty on the Functioning of the European Union 2007: Part 5 Title II). This shows that in interregional cooperation economic integration is an important matter, too. A point that is further stressed by negotiators themselves. In an Interview, Sveinung Røren, director of the trade policy de- partment in the Norwegian Ministry of Trade and Industry pointed out:

Moreover, the structure and scope of our FTAs has undergone a transformation - from FTAs that mainly focus on trade in goods to FTAs encompassing virtually all relevant aspects of modern trade (EFTA 2013a: 23).

Since many ROs value trade facilitation, their role in the world economy should be under precise analysis. Still it is interesting to observe that interregional cooperation of any kind does not play a major role in the research on regional integration. Since the literature suggests that recognition and presence (see below) can shape the role of an actor, the lack of hints at interregional PTAs is confusing (Feske 2015, Holthaus 2015). Researchers should give interregional cooperation more credit, as it is the next logical step in non-multilateral international relations and it is also seen as a logic development in a lot of economic research about trade (Freund 2000, Baldwin 2011, Neumair 2006).

The question about the necessities for ROs to engage in Preferential Trade Agreements arises, especially in deep PTAs, that might lead to sustaining trade facilitation. By reviewing the literature, I concluded that two factors matter most. First, the interdependent power relations regions or ROs have to each other. Second, how great their ability is to effectively negotiate trade agreements. These two factors that will serve as the variables of my thesis are further derived in the sections below.

2.1 Asymmetries and Power

Political scientists see the sources for liberal trade less in the welfare effects of trade itself, but more in the power structures of the international arena, especially the power asym- metry between states. Global economic decline, but also the fall of hegemonies cause the rise of regional trade agreements (Mansfield 1998: 537). International trade relations, dependen- cies and power are deeply interlinked. The best situation for country A is to be an (almost) irreplaceable trading partner. If it is extremely costly for countries to trade with other coun- tries than A or to replace A as a supplier or demander, scholars speak of asymmetric depend- ence in the field of trade (Hirschman 1980: 13-18). Such dependencies give country A power over its trading partners, making “commerce […] an alternative to war” (Ibid.: 15), because they can affect the economic wellbeing of their partners quite directly. This makes trade an instrument for coercion.

The power to coerce other actors at an acceptable cost to do something they would not do otherwise is one of the most basic definitions of power (e.g. in Keohane and Nye, 1999: 11). For Keohane and Nye (ibid.: 8-11) interdependence is a key structural element of interna- tional relations to understand power asymmetries. Interdependence is the state of mutual de- pendence, but not limited to beneficial cases of mutual dependence only. The standard cases are asymmetric dependent relations between states, meaning that country A needs country B more than vice versa for some reason. In these asymmetric relations lie the roots of power in the international arena. International interdependence generates policy externalities that can convince national executives to pool their sources and hence, such externalities can trigger processes of regional integration (Stone Sweet and Sandholtz 1998:7). Keohane and Nye pro- ceed to differentiate between sensitivity and vulnerability interdependence . Sensitivity refers to the connection between changes in two countries and looks at how quickly and how deeply changes in country A affect country B and if country B has to undertake costly actions to adapt to these changes. Vulnerability on the other hand looks at how costly the adaptions are and if alternatives are available in the first place (Keohane and Nye 1999: 12-13).

Although asymmetric interdependence is identified as a potential source of power, it cannot fully explain an outcome alone. It merely offers a view on the advantages of either side if countries come together to negotiate. The process of political bargaining determines out- comes in the political arena nowadays and is influenced by many more factors, such as politi- cal unity in a state or the probability of military engagement (ibid. 18-19). Still, asymmetry may play at least an equally important role in bargaining than other factors like actors’ unity. For Conceição-Heldt, this proves to be true for the EU in trade negotiations under different bargaining configurations. Under asymmetric bilateral conditions the outcome for the EU is better than under symmetrical multilateral conditions (Conceição-Heldt 2014: 991-992).

Building on those insights about asymmetry in trade relations, various authors (Heron 2011; Meunier and Nicolaïdis 2006; Shadlen 2008) conducted studies to support the theory. To measure power asymmetries in international trade relations, researchers use concepts like asymmetrical trade dependence (Zeng 2003: 67-69). This concept compares country A’s per- centage of export to country B in A’s GDP with vice versa.⁴ However, Zeng concludes that the structure of trade is at least equally important. Countries with a more competitive trade structure are more likely to yield to trade threats, countries with a more complementary trade structures can evade threats easier. The importance of trade structures is equally shown by Shadlen (Shadlen 2008: 4-8) with his more complex concept of political trade dependence. In this concept he compares the share of a country A’s total exports that go to country B which is based on GSP or GSP-like concessions and is easily revocable by country B. The revocability of such rules gives country B a believable threat in trade negotiations since it can revoke those concessions without fearing consequences from the WTO. Normally, such concessions under GSP refer to certain goods under the SITC scheme, but naturally the countries that are de- pendent on these concessions favor more enduring agreements.

Economists also clearly pointed out the asymmetric relations of trade negotiations in the so called “spaghetti bowl effect” (Bhagwati 1995b, Baldwin 2006). According to this ap- proach to regionalism, the rise of RTAs leads to “hub-and-spoke” systems with competing rules and regulations that might end up tangled like a bowl of noodles (Bhagwati 1995b: 4-5). Furthermore, the hub-and-spoke system is massively asymmetric, with the hub countries or blocs like the USA or the EU dominating negotiations with spoke-countries, which are usual- ly developing countries. These asymmetric export dependencies of the spoke-countries force them to make concessions to the hub-countries they would not support on the multilateral arena of the WTO framework (Baldwin 2006: 1501,1511). As shown above, many studies confirm these findings and the conclusions of economists and political scientists go hand in hand.

The key question is, what is necessary for free trade? According to Krasner (1976:323), open trade relations are most likely to occur in a system with a powerful hege- monic actor. Since openness in trade is directly connected to political power, it is the reason of state of hegemons to use their power to guarantee free trade (ibid.: 320). But this rationale changed to guarantee free trade that benefits the strongest the most. Especially in north-south PTAs, the developed countries use the asymmetric bargaining power to their advantage (Gal- lagher 2013: 148). The difficulties of maintaining the multilateral trade order in a multipolar world without hegemons also hints at that. Geopolitics about power and influence are back in international relations and trade policy has been re-politicized and is now used as a tool of coercion (Dieter 2015: 95-96, 103).

As Tussie and Saguier (Tussie and Saguier 2016: 2-3) describe it, asymmetric relations are the status quo in trade negotiations and unavoidable. On a bilateral and on the multilateral level, participating countries have to deal with structural asymmetries. Countries or blocs do use such asymmetries to impose or include their own rules, which are accepted by the minor partner. Because of its dependency disadvantage, the minor partner agrees to concessions it would not agree to otherwise (Drezner 2008: 32-35). Such concessions may lead to advantages for the stronger partners some may regard as “unfair”, but they also lead to deeper PTAs, that go beyond mere questions of trade in goods and tariffs. Hence one can derive the following hypothesis from the literature discussed in this chapter:

H1: Power asymmetry in trade between partners leads to a deeper integration in the final PTA.

To operationalize this hypothesis, this thesis collects the trade data of the mentioned ROs and groups. It will show their relations according to Zeng’s (2003) concept of asymmet- ric trade dependence. The exports of region A to region B as a share of A’s total exports and vice versa determine their respective trade dependencies. Changes of these values change the dependency of the actors on each other, so that these values reflect the sensitivity dimension of interdependence. The higher the value, the higher the sensitivity of a region to losses in market share to another region. As a result, one gains the values of the trade dependencies of the varying parties to each other. The values for the parties of a PTA are then subtracted from each other and used as the base for the condition in the QCA analysis. The higher the resulting value from the subtraction, the higher the asymmetry between the two regions.

The analysis refrains from conducting research beyond goods trade but does not limit itself to certain areas of goods trade. Data on trade in services is difficult to measure and therefore excluded. Even big organizations lament the lack of precision in service data and warn explicitly about these flaws in their database.⁵ Accordingly, base of the research are the numbers in total good exports in a three-year’s average. These years refer to the three years prior to the installment of the respective PTA. This way, the dependencies reflect the situation of the parties in the years next to the foundation. Since many trade talks take a long period of time, it is rather difficult to pinpoint the exact years of the most important trade talks, espe- cially with the high numbers of PTAs analyzed and high number of countries involved. But the last three years prior to the installment should include the trade data the parties were fa- miliar with before signing the treaty. Additionally, with a three-year average one can mini- mize the influence of extremes and include the asymmetry during a negotiation period. The thesis assumes that the parties are aware of the difference in trade power, so to speak, at the moment of signing the contracts.

For the goods classification, the thesis will make use of the Standard International Trade Classification (SITC). Of the many different systems of goods classification like the Combined Nomenclature of the EU, the SITC has the advantage of a higher aggregation of different groups of goods and it is very statistics-friendly (Statistisches Bundesamt, 2017) Data on these goods and other trade related data were gathered from Eurostat and the Direc- tion of Trade Statistics (DOTS) of the International Monetary Fund (IMF). Due to their preci- sion, these databases share the same values up to at least the second decimal place. Since this thesis will not make use of more precise numbers, these databases are compatible and hence sufficient for this thesis. Eurostat only allows to look at the exports from the European part- ner, either EFTA or the EU, in various groupings. To collect the data for the exports of other regions to Europe, the thesis had to use the DOTS database. For non-European regions or ROs, the DOTS database was used as a standard.

For the sake of data assembly, the trade values of the member states of each RO were grouped together, and their trade statistics were aggregated⁶. In the DOTS database one had to rule out intraregional trade by subtracting the exports from the member states to each other from the total exports to the world. This was done for all participating groups and ROs, to establish their common ground as blocs in the negotiations. In case of the European Union, the bloc size varied. For the European Economic Area (EEA) agreement the EU15 group was analyzed, for the other PTAs the EU27, so without Croatia. In the DOTs database, the exports values of Croatia were subtracted from the EU bloc value. For the EEA agreement, we chose to look at the exports from and to every country single-handedly. This thesis used all the val- ues up to the second decimal place. Generally, for the DOTS database, it aggregated the ex- ports of one country of region X to all member states of region Y. This procedure was used for all countries under research. The resulting values are aggregated with all other member states of a RO and used as the data to determine the trade dependence of a certain region. In Eurostat, grouping countries by bloc was possible and done. So, the thesis looked at EU and EFTA exports as a single bloc. The exception, again, is the EEA agreement: the intra- European trade data date only back to 1995.

Unfortunately, the data situation is not satisfactory. Especially in the cases of SACU and ASEAN there was a remarkable lack of data at the DOTS database and at the WTO data- bases and trade profiles. In part, you could avoid this problem by referring to datasets provid- ed by the organizations itself. In case of SACU we had the most profound lack of data but could partly substitute it thanks to the SACU database. Unfortunately, only for the years of 2007 and 2008, since the SACU database does not go back further. But for the mentioned years you could retrieve the export data and the intra-trade data of the African organization. The SACU database refers to the SITC system as well and hence is also compatible to the other datasets.

2.2 Actorness in Interregionalism

As this thesis discusses interregional trade relations, it shall look at the literature on interregionalism. Interregionalism can be defined as a “process of widening and deepening political, economic and societal interactions between international regions” (Roloff 2006: 18). This sub-discipline of International Relations takes, as the name suggests, a closer look at the relations between regions and especially regional organizations. It saw the light of day during the late 90s and early 2000s due to the growing number of ROs. Moreover, these organiza- tions increasingly started to interact with each other and the organizations started increasingly to act on their own accord as well. (Grimmel and Rüland 2015: 41). Since then, interregional- ism identified three main types of interregional relations, bi-regional, transregional and hybrid forms of interregionalism. The “classic” case are bi-regional relations that appear when two regional organization interact with each other (e.g. EU and MERCOSUR) (ibid.: 42). Being a relatively young approach to understand this special type of international relations, “a con- vincing theory of interregionalism is still outstanding” (Hänggi, Roloff and Rüland 2006: 10). Despite the growing interregional relations, Hänggi, Roloff and Rüland deny the institutional- ization of such relations. But at least in the area of trade, the growing numbers of PTAs since the 1990s tell otherwise. According to the Design of Trade Agreements (DESTA) database, eight interregional⁷ PTAs were signed between 1945 and 1990, 12 after 1990 and six are cur- rently under negotiation (Dür, Baccini and Elsig 2014: Meta-datasheet⁸ ). This hints at a light trend towards growing interregional trade relations and a growing institutionalization of those relations in trade agreements, hence an institutionalization of some sort takes place. The De- sign of Trade Agreements database accounts for 26 interregional PTAs as to today⁹.

A fundamental question for any research on international relations is the actorness. Who is an actor? How do we recognize actors? The classical answer of Realists and Neoreal- ists is the state (Bretherton and Vogler 1999: 18-19). But this state-centric concept of actors changed over time and was tackled by institutionalists, behavioralists and constructivists alike (Sibeon 1997, Keohane and Nye 1999, Wendt 1994). The new approaches to actorness wid- ened the perception of it and included many more players on the international arena besides states. Among them are regional organizations like the European Union. The EU stands in the center of many modern actorness concepts and the research on this specific topic is extremely euro-centric. This creates problems, as it is often difficult to adapt research based on the EU to other international organizations in other regions of the world (Wunderlich 2012: 655-657).

Before delving deeper into interregionalism, the term “region” needs some clarifica- tion. Scholars have debated about the definition of region for a long time (Mansfield and So- lingen 2010: 146,147). Constructivist approaches describe regions usually as “social and cog- nitive constructs” (Hemmer & Katzenstein 2002: 578). Hence regions are not geographically limited, they have an ideational facet as well. That explains why the NATO could speak about the common values of the NATO members, even when the different members are separated by thousands of miles of water (The North Atlantic Treaty 1949, preamble). Regions are also defined by interaction. By interaction with a regional “peer”, a region may shape its own identity and gain acceptance as a region in the first place (Gilson 2005: 310). Thompson (1973) defines a region as a group of states that interact with each other in a, to some extent institutionalized way, they are geographically proximate and share certain social and cultural features, among other points.

Nevertheless, to analyze interregional relations, defining the actorness of regional or- ganizations is essential. Especially since, as the literature suggests, one can expect varying degrees of actorness capability for ROs. But in case of negotiations, internal cohesiveness that leads to a single voice is an important factor for the success of such negotiations (Conceição- Heldt 2014: 980-981). Also, the interaction of regions shapes the role of regions or regional organizations as actors (Gilson 2005: 322). Hence the process of mutual recognition and the habit of interacting with each other increases the international standing and sharpens the pro- file of regions. This directly refers to two key concepts of actorness: presence and recognition (Allen and Smith 1990; Wunderlich 2012). Presence comes in two forms, firstly, the physical presence in institutions, meetings or fora. And Secondly, but not less powerful, the intangible way. This refers to how policy makers or institutions perceive another actor or actor-to-be and its notions and how this influences their own decisions (Allen and Smith 1990: 21,22). For Western Europe, the presence aspect of actorness played a crucial role for the perception of the EC as an actor in the international arena (ibid.: 36-37). Recognition on the other side re- fers to how other actors perceive an actor-to-be. For Thompson (1973: 97), external recogni- tion is a significant attribute of a region. According to Jupille and Caporaso (1998) recogni- tion includes acceptance and interaction by and with other actors, meaning multilateral organ- izations or states.

A constructivist approach to actorness for the European Community is delivered by Bretherton and Vogler (1999: 32-44). Building on the pioneer work of Sjöstedt (1977), who was the first to conceive actorness as a concept, they developed five basic prerequisites for actorness. These are:

1. A shared commitment to a set of overarching values and principles.
2. The ability to identify policy priorities and to formulate coherent poli- cies.
3. The ability to negotiate effectively with other actors in the international system.
4. The availability of, and capacity to utilize, policy instruments.
5. Domestic legitimation of decision processes, and priorities, relating to external policy (ibid.: 38).

According to them the EU¹⁰ fulfils prerequisites one, three and four. For the second prerequisite they describe the internal consistency and coherence problems the EC had to face (ibid.: 38-40). Their analysis is based on the EC/EU of the Nineties, so the institutional and organizational reforms of the Treaty of Lisbon are not included. Thus, an update is necessary to test their prerequisites for the current state of the EU. This is even more important if you consider that a more formal institutionalization may enhance regional actorness (Wunderlich 2012: 664). The EU’s institutionalization is already advanced in comparison with other re- gional organizations, yet it is important to find if even further progress in this area enhances the EU’s actor capability even more. For the fifth point, they remind the reader of the ongoing “democracy deficit” debate of the EU and put the legitimation of the European decision pro- cesses at question (ibid.:42). But they conclude to describe the EU as an “actor sui generis” that “remains in course of construction” (ibid.:44).

Coherence or cohesiveness is often mentioned as important in international negotia- tions (Conceição-Heldt 2014: 980). Thomas (2012) describes coherence as a major factor for the actorness of a regional organization. In his case study he analyzes the EU’s performance in the International Criminal Court. For him “EU foreign policy coherence is best defined simply as the adoption of determinate common policies and the pursuit of those policies by EU member states and institutions” (ibid.: 458). He measures coherence twofold. Firstly, by policy determinacy for which he looks at documents published by the Union and analyzes how specific the Europeans word their commitments and goals. Secondly, by political cohe- sion, which is based on research on the compatibility of the position of the member states and the Union itself. This research is highly interpretative and requires a nuanced understanding of the European Union (ibid.: 458,459).

For the EU trade is the main area of independent acting and the EU has regarded itself as an actor in this area for quite some time already (Bretherton and Vogler 1999: 64). Despite the problems of coherence, the EU faces in trade due to competing interests of its member states, trade is the topic where the EU acts consistently the most (ibid.: 78-79). Even after the Lisbon Treaty reforms trade and economic foreign policy remained the main areas of EU ac- torness, even though other areas such as security gained a more important role as well (Smith 2017: 15-16). In other fields than trade the EUs willingness to exert only soft power and not use economic threats or incentives limits its effectiveness (Thomas 2012: 471,472). Despite that, the EU is regarded as an economic superpower and seen on par with the other heavy- weights USA and Japan. In this triadic context, the trade superpowers have very special, yet mainly informal relations (Bretherton and Vogler 199: 66-74). Also, in multilateral trade ne- gotiations at the WTO the EU could prove its capabilities as an actor. In parts, the entire Uru- guay round was dominated by bilateral US-EC trade talks. Yet the Europeans have weakness- es due to the complex decision-making system and the dependence on the confirmation of the council, where a veto on trade deals struck by the commission is possible (ibid.: 74-78). To be precise, if the agreement covers topics of mixed responsibility, the Council can conclude it only after ratification by all member states.

To generalize the results for the EU, one concludes a simple rationale. The higher the actorness of a regional organization, the more capable is it to act and negotiate with a single voice and to reliably enforce the results of negotiations. Some scholars attest actor quality to the EU (Grimmel and Jakobeit 2015: 435,436). It is just a small step from intraregional to interregional trade. Countries realize that reducing tariffs in a more and more complex world is just not enough. But because important topics for international trade such as public pro- curements, subsidies or investment are too hard to negotiate in the multilateral WTO-arena or are not even on its table yet, countries started to talk about these topics in a bilateral PTA ap- proach (Hoekman 2014: 258).

As the literature shows, economic integration is a favorable goal of regional organiza- tions. However, this thesis takes a look at economic integration in interregional relations, hence between regional organizations. This is economic integration of the next level. If a bet- ter economic integration is the goal of ROs, it is easier for supranational bodies to achieve this goal when they have the necessary competencies available. Better actorness should enable ROs to achieve better, more concise and deeper agreements due to the avoiding of internal struggles. But what is necessary for a high actorness? In my research of the literature regard- ing interregionalism I came across two important subsections of the debate that affect actor- ness. The type of interregional relation a certain RO upholds with other regions and vice versa is one point. Another is the level of institutionalization. The thesis looks at the different kinds of interregional relations that exist and how exactly institutionalization matters in measuring actorness.

2.2.1 Types of Interregional Relations

Interregionalism comes in all forms and colors and developed over time, similar as re- gional integration processes did. Here too, one can differ between “old regionalism”, which refers to relations between at least two regional organizations (such as the EU and ASEAN), and “new regionalism”, which covers broader and less specific types of interregional relations (Hänggi 2006: 54-56). Commonly researchers understand interregionalism as a generic term that expresses three different kinds of interregional relations, biregional, transregional and hybrid forms (Grimmel and Rüland 2015:42). Hänggi (2006: 34-39) goes a little further and derives five types of interregional relations by the types of regional actors involved. These are:

1. The relations between a regional organization or a regional group¹¹ and a third state
2. The relations between a regional organization and a regional organization
3. The relations between a regional organization and a regional group
4. The relation between a regional group and a regional group
5. The relations among a group of states from more than two core regions

While the types two to four are classified as interregionalism in the narrow sense, all of them show interregional relations, at least in the wider sense. Hänggi labels type one as quasi-interregional relations and type five as megaregional relations. While type one original- ly does not show classic interregional relations, this type skyrocketed in the last decades. Fur- thermore, this type includes relations between regional organizations and important states such as the USA, Japan and China, which accept these relations more and more. Hence it is a trademark of new interregionalism (ibid.: 43). Such a typical development of new interregion- alism are megaregional relations depicted by type five as well. Such relations are usually dominated by a hegemon and include countries and regional organizations from at least two regions. The Asia-Pacific Economic Cooperation (APEC) serves as the prime example here (ibid.: 46-48).

Another taxology divides interregionalism into bi-regionalism or bilateral interregion- alism and transregionalism. While bilateral regionalism refers to the relations of two regional organizations, transregionalism covers the forms of interregionalism that go beyond that. Explicitly included from the concept of interregionalism are the relations many regional or- ganizations uphold with very influential countries, such as the USA, Japan or China. This type of relations is described as external relations of regional organizations (Rüland 2006: 298). The last kind especially is also important regarding the triadic context of interregional- ism. The Triad of trade includes the most prosperous regions in the world, which are North America, (Western-) Europe and Southeast Asia. Each of these regions has its key power, these being the EU, the USA and Japan (Bretherton and Vogler 1999: 67). Many interregional projects and fora can be seen as process of balancing between these three major regions. So, did the USA view APEC as a prevention of the forming of a strong trade bloc in East Asia, while Asian countries on the other side regarded it as a counterweight to the potential harmful effects of the European Single Market and NAFTA (Hänggi 2006: 52). Also, interregionalism with a triadic connotation are usually broader in scope and depth than their south-south coun- terparts. These tend to be rather vague and much less institutionalized. A special kind of rela- tions also in this context remains the transatlantic relations. The countries of the transatlantic community remain the ones with the deepest connections and the best institutional framework (Rüland 2006: 301).

2.2.2 Institutionalization

“Rules define roles (who is an actor) and establish the social context in which actor’s interest’s and strategies take shape. [...] Institutions are systems of rules, and institutionaliza- tion is the process by which rules are created, applied, and interpreted by those who live under them,” (Stone Sweet and Sandholtz 1998: 16). This sums up the ideas of some authors men- tioned above pretty well. If one intends to understand the actorness of a region, taking a closer look at the institutions that shape the region is recommendable. Frequently, regions are en- compassed by regional organization. They vary greatly in power and autonomy, which is the result of each integration process.

The very base of this in context to interregionalism is how ROs or groups of countries come to make decisions. This may vary among ROs, depending on the policy field. While in one field intergovernmental bodies may have the upper hand, in other it may supranational organs that have the last say (Fligstein and McNichol 1998: 62). The treaties or founding documents have an important role in shaping those bodies and decision-making procedures in certain fields have. They lay the basis for the power distribution in international organizations (Stone Sweet and Sandholtz 1998: 17)

In matters of institutionalization and autonomy of supranational authorities, many scholars make use of the principal-agent framework (Stone Sweet and Caporaso 1998) and link this approach to international trade talks (Conceição-Heldt 2011, Pollack 1998). I am aware of this framework’s advantages to understand autonomy of agencies and therefore to better understand certain aspects of institutionalization. In my comparative work it is unable to help, since many of the ROs under research are of intergovernmental character, less institu- tionalized and lack agents. Yet in case of the EU, contributions that use the P-A approach testify the Commission considerable autonomy and influence in the area of external trade (Pollack 1998: 248). Pollack (1998: 238) names external trade one of the commissions most important powers, since is the exclusive negotiator for the Union for all international trade negotiations.

The thesis concludes that scholars believe that a better institutionalization towards more supranational elements, meaning a deeper integration increases actorness. To measure actorness the thesis enquires what the decision-making procedures for trade-related issues with third parties are all about. The more important supranational organs are in trade negotia- tions, they should increase the actorness of their ROs. They should bargain more cohesively and as result be able to enable deeper integration with other regions. Despite the multiple ap- proaches to conceptualize actorness¹², scholars seem to agree that high actorness is crucial to succeed in international negotiations. High actorness means acting as a single actor, something that is usually an assumption for states, but is not considered as self-evident for regional organizations. Of course, it is only one of many causes for success and effectiveness in negotiations. Still, the literature on actorness leads to the second hypothesis:

H2: The higher the actorness of the negotiating partners, the higher the integration of the PTA.

The second hypothesis in this thesis regards the actorness level of ROs and regions. Nowadays in the literature many crucial points for the actorness of an organization are mentioned. Cohesiveness or coherence¹³ are named frequently (Conceição-Heldt 2014, Thomas 2012, Wunderlich 2012), as well as recognition (Allen and Smith 1990, Jupille and Caporaso 1998, Thompson 1973) and interaction (Gilson 2005, Thompson 1973). Plus, there are the five criteria devised by Bretherton and Vogler (1999). Unfortunately, all these points or criteria were researched and derived in a qualitative manner, so that grading the level of actorness in a solid way proves somewhat difficult. Still, this thesis attempts it and therefore yet again expands the research tools for interregionalism.

Base of the grading of actorness are the mentioned points, but to put them into a QCA model you need to scale them. This is what I aim to do by giving the founding documents of each of the researched ROs notes if they contain the following points:

1. Institutional Decision Making for external trade
2. Recognition as an actor at other international institutions
3. The general type of the interregional relation between the parties

Since this work researches economic integration based on trade, the thesis narrowed the very general five criteria of Bretherton and Vogler down and included the arguments for recognition and cohesiveness. I acknowledge the argument of Thomas, (2012: 458,459) that authors require extensive knowledge about the details to evaluate the internal processes of regions. Yet, to facilitate analysis and to enable a more quantitative approach, a simplified scale system will have to do. The data will be less precise, but this cost comes at the ad- vantage of easier comparison of many cases. As some authors mention that institutionalization may help to enhance actorness (Bretherton and Vogler 1999, Wunderlich 2012), the first source of grading will be the foundation thesis of the involved regional organizations. There, since this thesis sticks to trade as the primary policy field of interest, we will look especially at the decision-making processes for trade related issues, if available. Then we will look at the recognition of regions by their role or appearance at important global institutions. Lastly, we will grade the type of interregional relations.

This research will revolve around economic and trade related issues; this focus is the only fundamental part to understand the role of regions in trade negotiations. I do not claim to gather precise data about the general actorness of regions and ROs, in fields other than econ- omy and trade. Regions may have varying levels of actorness in different policy fields, de- pendent on their institutions and decision-making rights. Stone Sweet and Sandholtz (1998: 9) follow the same approach when they measure the institutionalization of the EC by policy sec- tor. The exclusion of interaction as its own point is due to the difficulty of measuring it. Sure- ly there are learning processes that enhance the actor capability of regions if they engage in many actions and negotiations but how many will suffice to create a measurable effect? Fur- thermore, since many ROs are rather young in comparison with the old horses like the EC, those would be put at a clear disadvantage. The thesis generalizes approaches derived from the EU and uses it on other ROs. Therefore, it not only delivers an important contribution to the actorness debate, it also helps to test the entire theory. It overcomes the often-criticized euro-centrism by quantifying it.

3 Methodology

Interregionalism is a not yet sufficiently enough researched phenomenon. There is a substantial lack in comparative studies, case studies dominate the field (Rüland 2006: 312). But comparative studies are needed to strengthen empirical basis. This thesis aims to satisfy this demand by delivering a Qualitative Comparative Analysis (QCA) approach to compare the results in negotiations of PTAs between different regional organizations. As QCA is not a traditional statistical based comparative approach but rather serves as an end to bridge the gap between quantitative and qualitative approaches, this section starts with a short overview and explanation of QCA. It follows a description of how the data were assembled, which databases where used and about the use of the statistical program R.

[...]

---

¹ This point is elaborated later with showcasing the treaties of the ROs in question.

² Since there is no accord on the term yet, I use - for the topic of trade - regio-regio and interregional as interchangeable, synonymous terms.

³ Viners Original article dates back to 1950.

⁴ In his study, Zeng compares the values for the trade relations of several countries with the USA.

⁵ For instance, the Trade in Services Database of the World Bank was last updated in February 2015 (World Bank 2017), whereas the DOTS database of the IMF already delivers the data for merchandise trade for the first half of 2017

⁶ In case of the Principality of Lichtenstein, the trade data are measured as part of the Customs Union with Switzerland and therefore added to the values of Switzerland in the databases.

⁷ the authors of DESTA call them regio-regio.

⁸ The authors published several datasets, this refers to the list of treaties available at: https://www.designoftradeagreements.org/downloads/.

⁹ Last checked: 02/08/2018.

¹⁰ Is noted as EC in Bretheron and Vogler (1999). For clarifications, I use the term EU equivalently to EC, whenever the EC of the past is mentioned.

¹¹ Hänggi defines regional groups as “a more or less coordinated group […] of states” (ibid.:39).

¹² For an even deeper overview on this issue, see Niemann and Bretherton (2013).

¹³ For simplification, in the thesis the term cohesiveness will be used from now on.