Institutional Analysis of Foreign Aid Policy-Making Rules, Incentives, and Commitment Problems in South Korean Government

This electronic thesis or dissertation has been downloaded from the King’s Research Portal at https://kclpure.kcl.ac.uk/portal/

Institutional analysis of foreign aid policy-making rules, incentives, and commitment problems in South Korean government

Yi, Jisun

Awarding institution: King's College London

The copyright of this thesis rests with the author and no quotation from it or information derived from it may be published without proper acknowledgement.

END USER LICENCE AGREEMENT

Unless another licence is stated on the immediately following page this work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ You are free to copy, distribute and transmit the work Under the following conditions:  Attribution: You must attribute the work in the manner specified by the author (but not in any way that suggests that they endorse you or your use of the work).  Non Commercial: You may not use this work for commercial purposes.  No Derivative Works - You may not alter, transform, or build upon this work.

Any of these conditions can be waived if you receive permission from the author. Your fair dealings and other rights are in no way affected by the above.

Take down policy

If you believe that this document breaches copyright please contact [email protected] providing details, and we will remove access to the work immediately and investigate your claim.

Download date: 25. Sep. 2021 0

INSTITUTIONAL ANALYSIS OF FOREIGN AID POLICY-MAKING: RULES, INCENTIVES, AND COMMITMENT PROBLEMS IN SOUTH KOREAN GOVERNMENT

Jisun Yi

Dissertation Submitted to the Department of International Development King’s College London in Fulfilment of the Requirements for the Degree of Doctor of Philosophy in Development Studies

June 2018 1 Abstract

Why do some governments provide less foreign assistance than they originally pledge? This research explores the origins of the ‘big-plan-small-action’ issue through an institutional analysis. The research dwells particularly on how donors’ aid management design (by examining specifically the legal-regulatory rule arrangements for policy formation and implementation) can help explain such problems. The aim is to identify the logic and design of rules for foreign aid policy-making that induce a particular set of incentives and collective actions of state agencies that, in turn, cause a gap between planning and practice in the management of foreign aid. To this end, this research critically reviews Elinor Ostrom’s Institutional Analysis and Development (IAD) framework and comes up with a refined version of it. The fine- tuning of the framework is necessary to take rule institutions more seriously. Given that existing IAD-based empirical analyses in the domain of development cooperation focus on the effects of formal rules on foreign aid policy-making, the earlier application of the framework appears ineffective at acknowledging the broader institutional contexts of policy-making rules and traditions across government units. Considering the unique contexts of emerging/young donor governments, whose policy-making tends to be governed by greater ‘discretion (room for judgment)’ under a relatively insufficient and abstract set of written rules, the IAD research requires some sophisticated analytical procedures to discern the impacts of various types and layers of policy-making rules on the outcomes. The chosen country case is South Korea (2000-16), which has displayed one of the most severe commitment problems. Korea’s public laws and regulations have generally lacked clear logics and definitional details for the management of official development assistance (ODA). In comparison to conventional donor cases, such characteristics have left greater room for undefined policy areas to be decided at the discretion of state executives and bureaucrats. Since the greatest decision- making/corrective powers have been centered on the president, ODA planning at the top policy level has tended to be swung by his/her immediate diplomatic-political needs. Through the two nested cases of knowledge aid and multilateral aid, it is further demonstrated that the presidential ODA initiatives in those sub-policy areas have legally overridden the procedures of reflecting long-term operational considerations. No legal restrictions over the entry and exit of participants have encouraged a high level of opportunism among state agencies to rush into knowledge aid and multilateral 2 cooperation regardless of their organizational talents and traditions. More prominently, KOICA has displayed distinctive strategic behavior in order to reduce its chronic political-financial instability. Korea’s substantial gap between annual commitments and disbursements might be a by-product of its ODA policy system, which is rationalized by a loan-grant division. However, this study argues that the commitment problem has fundamentally resulted from a lack of coordinating mechanisms, which were neither effectively put to work by the laws nor took place by the consensual norm within the system. The coordinating/oversight body remained nominal, since it performed the mandate to a limited extent without proper legal power and resource. Considering the national legacy of having had a very centralized public policy-making structure, successful government- wide coordination relied heavily upon the discretion of the president. All things considered, Korea’s ODA policy system is a case that fails to encourage reciprocal collective actions among the inherently spat-prone bureaucrats. The research is also useful if one wonders what institutional-structural conditions need to be fixed or re-adjusted to make foreign aid policies more credible and sustainable. The empirical analysis of contexts, two action situations, and outcomes suggest that to properly understand the underlying sources for the expansionary rhetoric and associated incentive problems with a country’s development cooperation, due attention ought to be paid to the rule configuration. This will provide an insight into the extent to which hard rules might create room for informal rules to come into play and induce a unique set of incentives, and collective actions, and to end up with undesirable policy outputs. 3 Acknowledgements

My journey to completing this PhD study took me five years. For those years I have been indebted to many people and institutions. I would not be able to count all my blessings given through them. Among those names, Dr. Nahee Kang comes first and foremost. With her intelligent guidance and warm support, I was able to go through life difficulties without losing faith in my research. I also wanted to deeply thank Dr. Ramon Pacheco-Pardo, who guided my early-stage research and generously continued to work with me. Prior to and during the doctoral study, I was extremely lucky to meet great scholars and practitioners in my home country, South Korea, and elsewhere: Dr. Byung Sung Ahn (Korea Institute for Defense Analyses), Dr. Jae Ok Paek (KIDA), Dr. Jiyoung Kim (Soongsil Univ.), Dr. Tony Addison (UNU-WIDER), Dr. So-yeon Kim, Dr. Sang Tae Kim (Kyung Hee Univ.), Dr. Woongjo You (National Assembly Research Service), Dr. Dae In Kim (Ewha Univ.), Dr. Jung Hwa Seo (EDCF), Dr. In Mook Choi, Dr. Jae Een Lee (KOICA), and Dr. Lai-ha Chan (Univ. of Technology Sydney). They spent hours with me, listening to my humble ideas and providing enlightening feedback. Beyond the academic boundaries of my university, this research benefited greatly from my research experiences at global institutes—the United Nations University World Institute for Development Economics Research (2014), the Asian Development Bank (2014), and the World Bank (2017). I would like to thank those who gave me the privilege of working with such highly insightful and committed professionals in the development cooperation field. Colleagues at King’s were great supporters. I am truly grateful for their presence to my life in London: Sun-sook, Danielle Hyeon Ah, Seon Young, Yukyung, Hyungguen, Ji-myoung, Ping, and Feiying. I extend a special note of appreciation to a group of Korean scholars at Princeton University for their kind welcome to me as a guest researcher at the Empirical Studies of Conflict (2016): Dr. Jong Hee Park (Seoul National Univ.), Jeongseok Lee, Jun-hyun Lah, Song Ha Joo, and Yunkyu Sohn. I would also like to thank Dr. Avra Cohn for her careful editing of my drafts. Without unconditional support from my beloved family and old friends, I would have not been able to make it this far. My greatest thanks go to Yoo-sun, Jin-Hag, Ian, Ahn-sun, Jung-ho, Min-jun, and Jae-hyun. My in-laws, Yeojin, Kibum, Ae-rin, Tae-rin, and my parents-in-law have kindly understood my long absence from family duties. My 4 good friends, Hyun-Joo and Yoon-Young, have kept their prayers coming to me at all times. More than anything and anyone, I dedicate this thesis to my father, Bong-hee Lee, and my mother Yun-suk Park. I deeply admire my father’s great legal mind and faith in justice and my mother’s patience and unfailing love for family and others. Finally, I wanted to thank my dearest husband, Jung-jun Yun, for being my true sanctuary.

Bush House King’s College London January 2017

5 List of Abbreviations

AIIB The Asian Infrastructure Investment Bank ADB The Asian Development Bank BOK The Bank of Korea CIDC Committee for International Development Cooperation CSM Case Study Method DAC Development Assistance Committee, the OECD DEEP Development Experience Exchange Partnership DFID Department for International Development, U.K. EDCF Economic Development Cooperation Fund Eximbank Korea Export-Import Bank IAD Institutional Analysis and Development IFIs International Financial Institutions JICA The Japanese International Cooperation KDI Korea Development Institute KOICA Korea International Cooperation Agency KSP Knowledge Sharing Program MDBs Multilateral Development Banks MDGs Millennium Development Goals MOFA Korea Ministry of Foreign Affairs MOSF Korea Ministry of Strategy and Finance ODA Official Development Assistance OECD The Organisation for Economic Co-operation and Development PMO Prime Minister’s Office, ROK ROK Republic of Korea SDGs Sustainable Development Goals SIDA The Swedish International Development Agency WB The World Bank Group

6 Table of Contents

1 Introduction: The Nexus between Rules and Policy Outputs in the Foreign Aid Policy Process

1.1 Research Question 14

1.2 Argument 18

1.3 Theoretical and Methodological Approach 20

1.4 Country Case 26

1.5 Structure 30

PART ONE: THEORY AND RESEARCH DESIGN

2 Theoretical Review: The Institutional Analysis and Development (IAD) Approach to the Development Cooperation Challenge

2.1 Revisiting the Importance of the Research 34

2.2 Positioning and Foundations of IAD: Policy Process Literature and Agency Theory 37

2.3 Introduction to IAD Framework: Analyzing Incentives at Sequential Stages of Policy-making Process 44

2.4 Critique of Previous IAD Study: Envisioning More Practical and Wider Application of the IAD Framework 47 2.5 Conclusion 53

3 Conceptual Framework: Refining the IAD Framework to Take Rule Institutions Seriously

3.1 Theoretical Concepts to be Incorporated 54

3.2 Introducing a Fine-tuned Framework 64

3.3 Research Action Plan 67 7 3.4 Propositions 69

3.5 Conclusion 79

4 Methodology and Data Collection Technique: In-depth Case Study on Intra- Governmental Policy-Making Dynamics in South Korea

4.1 Revisiting the Values of Qualitative Research 80

4.2 Methodological Considerations 82

4.3 In-depth Case Study in Action 85

4.4 Strategy for Data Collection and Presentation 91

4.5 Conclusion 97

PART TWO: EMPIRICAL ANALYSIS

5 Analysis of ‘CONTEXTS’ of South Korean ODA Policy-Making

5.1 Overview of the Chapter 99

5.2 Korea’s Historical Experience of Foreign Aid (1950s-90s) 101

5.3 The Earlier ‘ACTION SITUATIONS’ of Korea’s ODA Management Design (1970s-90s) 106

5.4 The Contemporary ‘CONTEXTS’ (2000-15): Physical/Material Conditions of Aid Resources and Attributes of Aid Bureaucracy 110

5.5 The Contemporary ‘CONTEXTS’ (2000-15): Rules-in-use 113

5.6 Conclusion 126

6 Analysis of ‘ACTION SITUATION’ in Knowledge Aid Management (2000-16)

6.1 Overview of the Chapter 127

6.2 Knowledge Aid Practices in the Korean Government 129 8 6.3 Action Situation at the Individual-Choice Level 134

6.4 Action Situation at the Collective-Choice Level 149

6.5 Conclusion 155

7 Analysis of ‘ACTION SITUATION’ in Multilateral Aid Management (2000-16)

7.1 The Overview of the Chapter 156

7.2 Multilateral Cooperation Activities in the Korean Government 158

7.3 Action Situation at the Individual-Choice Level 163

7.4 Action Situation at the Collective-Choice Level 176

7.5 Conclusion 178

8 Conclusion: Evaluating the Impacts of Rule Institutions on the Commitment Problem

8.1 Significance of the Research 180

8.2 Evaluation of the ‘Rule-Policy Output’ Nexus 182

8.3 Extended Discussion: Implications of Bureaucratic Discretion and Incentives Beyond the Case of Korea 189

8.4 Concluding Remarks 191

Appendix

Bibliography

9 List of Figures

Figure 1.1 Aid Scale and the Commitment Problem 15

Figure 1.2 The Institutional Analysis and Development (IAD) Framework 22

Figure 1.3 Refined IAD Framework for Policy Process Analysis 25

Figure 1.4 Korean ODA Growth (1999-2015) 26

Figure 1.5 DAC Donors' Structures for ODA Management 28

Figure 1.6 Organizational Structure for Knowledge Aid Management in the Korean Government 30

Figure 1.7 Organizational Structure for Multilateral Aid Management in the Korean

Government 30

Figure 2.1 Principal-Agent Model 41

Figure 2.2 Delegation Chain in the Study of Gibson and Others 43

Figure 2.3 The IAD Framework: Formation, Interplay, and Effects of Incentives 46

Figure 2.4 Complexity with IAD Research 50

Figure 3.1 A Hypothetical Delegation Structure within a Donor Government 60

Figure 3.2 Previous and Refined Applications of the IAD Framework 66

Figure 3.3 Research Action Plan under the Refined IAD Framework 68

Figure 3.4 Seven Rule Types Affecting Varied Elements of an Action Situation 70

Figure 3.5 Delegation Structure for Knowledge Aid Management 73

Figure 3.6 Delegation Structure for Multilateral Aid Management 77

Figure 4.1 Targeted Mechanisms of Foreign Aid Policy-Making: The Korean

Case 87

Figure 5.1 Korean ODA Growth (1988-2013) 111

Figure 6.1 MOSF’s Strategy to Promote KSP and EDCF Projects 140

Figure 6.2 KOICA’s Operational Orientation in Transition (1991-2012) 147 10 Figure 6.3 ODA Planning and Budget Processes in the Korean Government 151

Figure 7.2 Annual Disbursements of Multilateral ODA in Korea (1990-2013) 158

11 List of Tables

Table 1.1 Korea's Performance of Making a Credible Aid Commitment (2000-15) 27

Table 3.1 Rule Typology: Discerning Rule Types in the Context of Foreign Aid

Policy Formation and Implementation 56

Table 5.1 Early Years of Korean Development Partnerships (1960s-80s) 104

Table 5.2 Korea’s ODA Scale-up in the 1990s 105

Table 5.3 International Development Cooperation Staff in the Korean

Government 113

Table 5.4 Rule Configuration of ODA Laws in Korea 116

Table 5.5 Partial Amendments of ODA Laws (2010-17) 122

Table 5.6 Review of Annual ODA Strategy Papers: Chronic Gaps between ODA

Plans and Disbursements (2006-16) 124

Table 5.7 Mid-Term Planning of Korean ODA 125

Table 6.1 Knowledge Sharing Program (2004-13) 130

Table 6.2 Regional and Thematic Allocations of KSP (2004-12) 130

Table 6.3 Development Experience Exchange Program at KOICA (2003-12) 131

Table 6.4 Fragmented Practice of Saemaul ODA in Korea 133

Table 6.5 ODA Performance in Comparison: The Roh and Lee Administrations 135

Table 6.6 Korean ODA: Annual Commitments and Disbursements (2006-15) 154

Table 6.7 Korea’s Performance of Making a Credible Commitment:

Knowledge Aid 155

Table 7.1 Accumulated Disbursements to Multilateral Organizations (2000-13) 159

Table 7.2 Top Ten ODA Multilateral ODA Distributors in Korea (2006-13) 160

Table 7.3 Net ODA Disbursements to MDBs: By Bank of Korea (2006-13) 161

Table 7.4 Components of Multilateral ODA in Comparison: The Roh and Lee 12 Administrations 164

Table 7.5 Performance of Korean Businesses in the MDB Procurement Markets 169

Table 7.6 KOICA’s Multilateral Funding in 2011 175

Table 7.7 Multilateral Cooperation Strategy (2011-15): MOSF and MOFA 176

Table 7.8 Korea’s Performance of Making a Credible Commitment:

Multilateral Aid 178

13 List of Boxes

Box 1.1 List of Selected Research on Incentives with Foreign Aid 24

Box 4.1 Korea’s Legal-Regulatory Frameworks that are Reviewed 88

Box 4.2 Analytic Tasks to Understand Actors’ Rationality and Behavior 90

Box 4.3 List of Interviewees 92

14 CHAPTER ONE INTRODUCTION: THE NEXUS BETWEEN RULES AND POLICY OUTPUTS IN THE FOREIGN AID POLICY PROCESS

1.1 The Research Question 1.1.1 The Puzzle: The Problem with Donors’ Aid-giving Manner Why do some governments provide less foreign aid1 than they originally pledge? This research explores the origins of the ‘big-plan-small-action’ issue when donor countries provide foreign aid to developing countries for a mixed set of values and purposes. This research is intended to answer the question by singling out how and to what extent the traits of ‘aid management design and strategy’ within donor governments—which are embedded in formal and informal rules (i.e. constitutional laws, public laws and regulations, traditions and norms, and other rule statements) that are imposed on the processes of public policy formation and implementation—can explain such state behavior. On the theoretical front, Institutional Analysis and Development (IAD), one of the policy process theories, provides the conceptual framework for this research. Adopting the case study approach, this study takes a close look at the country case of South Korea (hereinafter ‘Korea’). Then, why rules? This particular focus of this institutional study is taken since the effect of public laws, regulations, and norms (as the description of organizational design and agency autonomy for policy formation and implementation) on the public policy process has been well assumed but is not that much empirically explored: One assumes that there are a set of rule institutions that delimit the boundaries of policy makers’ rationales, dynamics, and outcomes of interaction, apart from domestic and international conditions that also affect policy makers’ strategy-building and decision-making. However, we do know insufficiently about how and to what extent those rules lead human/organizational actors to what courses of action and what end points of policy output on the policy process; and how they are configured and evolve while interacting one another. In an attempt to track the effect and interplay of rules on state’s policy

1 ‘Foreign aid’ is a colloquial term for official development assistance (ODA), which was coined by the Development Assistance Committee (DAC) of the Organisation for Economic Co-operation and Development (OECD). Such financial instruments of donor governments are often recognized as an essential part of ‘development cooperation’ between the developing and developed worlds. Literature shows a mixed use of those terms. For the sake of reducing any confusion among the terms, foreign aid, foreign assistance, development aid, development assistance, ODA, and development cooperation are used interchangeably throughout the thesis. However, conceptual distinctions among the terms are explicitly explained in the analysis where necessary. 15 behavior, this study examines the logic and design of rules that reshape policy makers’ behavior and their incentive structures that particularly create and even widen the gap between aid planning (commitments) and practice (disbursements). Then, why foreign aid and the commitment problem with donor countries? Making overambitious aid pledges for country/organizational aid givers might seem to have little value for serious theoretical discussion if dismissed merely as rhetoric gestures or whims of state’s political leadership and state bureaucrats. Such behavior is not limitedly observed in public policy fields, rather, is prevalent in almost every public/private domain. What is particularly interesting in the development cooperation field, however, is that each country donor’s performance of delivering its promised aid varies as demonstrated through the following measurement2. While the majority of the OECD’s donors perform fairly well, delivering 85 to 100 percent of what they initially plan and pledge, some other donors seem to have the ‘commitment problem’ to a significant extent. It also appears that the performance of making credible foreign aid delivery appears to have little relationships with country’s total aid scale.

Figure 1.1 Aid Scale and the Commitment Problem for 28 DAC Member Countries (Calculated by author; OECD QWIDS (Accessed Nov. 12, 2017))

GBR 120.0 15)

110.0- SWE DNK

100.0 GRCLUXIRL PRTAUT AUSCAN POLESP 90.0 BEL ITANOR USA NZLFIN CHE NLD 80.0 FRA DEU

Disbursements Disbursements 70.0

60.0 KOR (Percent, (Percent, Averaged from 2006 JPN Annual Annual Ratios of Commitments(100) to 50.0

40.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 Annual ODA Disbursements (USD billions, Averaged 2006-15)

What can explain the exceptional gaps between annual commitments and actual disbursements for the U.K., France, Germany, Korea, and Japan? Is this simply because

2 To indicate the performance of donor governments in measuring up to their annual commitments, the figures are calculated as a percentage. 16 of ODA policy makers’ incompetence/miscalculation; or because of their strategic/cultural/traditional considerations that favor overambitious aid plans (or otherwise, conservative ones); or, regardless of the capability or strategic/preferential choices of policy-makers, are there structural issues with the internal processes and systems of foreign aid policy-making involved? Foreign aid that comes in as planned is critical for achieving its intended objectives. As some studies previously pointed out, unpredictable and unstable aid flows to recipient countries harm aid effectiveness (Celasun and Walliser 2008; Bulir and Hamann 2008; Barder 2009; Handley 2010). That is, the predictability of aid determines the quality of aid. Aid-giving manner per se is now being importantly regarded as a key strategic aspect of making ODA effective and sustainable along with aid scale and destination. But with regard to the origins of such problematic aid-giving manner of country donors, there are not too many studies being done, in fact. To a limited sense, the existing explanation in the literature was that the behavior might result from donor governments’ attempt to achieve incompatible aid goals (self-interested vs. altruistic purposes of foreign aid provision). As an effect, for some countries, humanitarian motives are apt to be expressed in rhetoric at the expense of donors’ pursuit of national interest (Breuning 1995; Hook 1997; Hook and Zhang 1998, p.1061). Aside from donors’ motivational dilemma, other scholars have emerged to empirically discuss the gap between the rhetoric and reality of aid delivery at bilateral and multilateral development agencies. Despite the fact that donors increasingly create and join high-level policy pledges for better foreign aid delivery, they find no clear evidence of improvements with aid agencies in terms of transparency, specialization, selectivity, and cost-effectiveness (Thiele and Nunnenkamp 2006; Easterly and Willamson 2011). Another work (Aldasoro et al. 2010) contends that the gap between the rhetoric of political declarations and donors’ actual aid allocation depends on whether aid agencies concentrate on a few recipients and sectors. Upon this under-explored but critical research area for foreign aid to work better as intended and to be sustainably delivered for the developing world, this thesis revisits the issue of making overambitious aid planning or pledges by examining the rule settings (institutional design) of donor government’s aid management and by learning the outcomes of bureaucratic incentives that are created and played out under a particular aid management design. This attempt is not limited to identifying the institutional causes for the rhetorical aid policy and practice. Adding to this, this research opens up a discussion 17 on what institutional change or readjustment of the aid management design would be meaningful and feasible to cause foreign aid to be delivered in a more sustainable and predictable manner. 1.1.2 Re-framing the Puzzle from the Bigger Debate of ‘Control vs. Judgement’ (as Donors’ Choice of Aid Management Design and Strategy) As some important studies on donors’ aid management strategies and designs are now being introduced, this study anchors to the contemporarily emerging the ‘control- versus-judgement’ debate. As the choice of foreign aid management design and strategy, it is a question of decision-making power allocation within donors’ aid management systems: It is whether donors’ development/aid agencies should be given more autonomy (discretion) to make their development interventions flexibly and efficiently formulated and delivered; or whether their aid operations should be under tighter control from above (either by giving more power to supervisory or monitoring agencies or by further developing regulatory tools) in order to make foreign aid more predictable and accountable. The big debate was driven from the discipline of management studies that models the allocation of formal and matter-of-fact authority within organizations and the resulting impacts of the authority allocation on organizational performance (Aghion and Tirole 1997). After some time, the discourse has gained momentum in the context that the development community refocusing their development policies and operations under the Sustainable Development Goals (SDGs) (2016-2030). A group of scholars and practitioners project a critical voice that for foreign aid to be developmentally favorable to unstable, unpredictable contexts in aid recipient countries, donors’ foreign aid policy systems should be strategically flexible by giving more decision-making powers to the field staff at bilateral agencies and country offices (Gulrajani 2015; Honig and Gulrajani 2016, 2018; Honig 20183). This area of concern is importantly raised in academic and professional circles since this addresses our future options for better aid effectiveness through reform/restructuring of donors’ aid management systems and agencies. Nonetheless, across the donor community, the management designs and strategies are variedly configured and evolving to a different degree—some donor countries prefer top-down management and continuously cherish that, but for others, the bottom-up

3 Honig’s recent book, titled Navigation by Judgement: Why and When Top Down Management of Foreign Aid Doesn't Work (2018), does well capture the ironic challenge that international development organizations face: On the one hand, tight top-down control over field operations appears to generate negative effects; on the other hand, however, more autonomous field agents have greater potential to use operating slack and information at their own interest against organization objectives. 18 approach is embedded in their aid management systems. Due to such variation in aid management design and strategy, donors do development cooperation differently. Back to the core debate, the challenge here is that if donors are willing to enhance the aid effectiveness on the ground, one may wonder what specific change (i.e. reform or restructure of aid agencies or legislations) and to what extent to be re-adjusted might be necessary. Likewise, U.K.’s Department for International Development (DFID) introduces ‘Smart Rules (2014)’ which are expected to empower and allow more room for its staff to use discretion and judgement. The United States Agency for International Development (USAID) implements the Local Systems Framework (2014) in order to embed adaptation and flexibility into its working principles. Despite some great risks of corruption and wrong-doing among aid staff, such movement of ‘moving autonomy to a lower level of aid management’ arose in an effort to readjust the management structures properly if the situations of underdevelopment which would benefit from ‘flexible’ responses are deeply re-considered. The debate gives rise to the starting point of this research as follows: Donors’ planning and delivery manner is determined by donors’ choice of management strategies or institutional preferences. The aid management design/strategy of a donor government can be understood in terms of how the government allocates (centralizes or decentralizes) its ‘control’ (decision-making and corrective powers) within the system and allows ‘judgement’ (discretion) for state policy makers and practitioners in policy-making and implementation processes. This is extremely important since how organizations are organized and regulated also determine the quality of policy products.

1.2 The Argument To further explore the notion of donor’s management design determines its performance of delivering ODA as planned, the case of Korea4 is selected. Nevertheless, not serving as a stand-alone case for the puzzle, the selection of the country case is intended to have wide implications for other donor countries. Representing a group of new-coming donor countries to the OECD-DAC, Korea constitutes a relatively young, ‘eager-to-learn’ member country to the West-dominated donor club. On the other hand, in contemporary discussion of ‘emerging donors’, Korea is considered one of them, which seem to share some important institutional traits with other non-DAC/Asian donors, like

4 Now, Korea, the 11th largest economy (by GDP) (2017), annually provides a 15th-largest sum of official development assistance (ODA) (USD 2.1 billion, 2017) to the developing world among the 29 member countries (as of 2018) of the Development Assistance Committee (DAC) of the Organisation for Economic Co-operation and Development (OECD). 19 India and China, in terms of historical/contemporary aid recipientship, dualistic policy- making structures, and informal legality traditions (i.e. ODA managerial-ship shared between finance and foreign affair ministries, no strict legal guidelines for ODA allocation and delivery imposed) (Jerve 2007). Although Korea’s ODA donorship has been described and perceived under various names: the second Asian DAC member (after Japan), small donor, emerging/new donor, etc., this study re-positions the country in two aspects: first, as one of the cases with the severe commitment problem in the traditional DAC donor group; but, as one of young/new donors whose aid-giving experience is relatively slim and whose institutional arrangements (including rule settings) for aid policy-making are largely either underdeveloped or in the making from Western-DAC standards. When it comes to Korea’s aid management design and strategy, as predicted and confirmed by the theoretical-empirical perspective of this study, Korea is the case of a management design (considering its rule settings for aid policy making) whose control is disproportionally concentrated on political leadership (top-level policy makers) as the legacy of its state-directed development experience (See Chapter Five). When reviewing legal-regulatory frameworks of foreign aid policy-making in the Korean government (2000-2016), Korea’s public laws and regulations have generally lacked clear logics and definitional details for the management of official development assistance (ODA). In comparison to conventional donor cases, such characteristics have left greater room for undefined policy areas to be decided at the great discretion of state executives and bureaucrats (See Section 5.5). Such heavy concentration of power at the top-level of policy-making, as this thesis argues, politicizes the use of foreign aid to a distinctive extent given that aid scale and destination is decided at the discretion of the president by the laws: Hence, his/her immediate diplomatic and political interest in the international arena can be well served through foreign aid provision. As a result, it leaves aid agencies at lower levels more likely to suffer from the lack of ‘flexibility’ in managing resources for foreign development operations. In a nutshell, the chronic gap between annual commitments and disbursement for Korean ODA appears to result mainly from its top-down aid management in which state bureaucrats at a higher level are incentivized to aim high for setting aid targets beyond their capacity and specialization, which is bound to deliver less against aid targets. Nevertheless, for Korea, this management design favorably works for aid expansion and aid diversification within the system. This arrangement helps the country 20 transform into a recognized donor with impressive aid growth records. On the other side of the coin, such top-down management of foreign aid inherently creates bureaucratic incentives that go against the reliable and sustainable delivery of ODA as this thesis unfolds its incentive analyses of state agencies on the policy-making process: under the particular institutional design, state bureaucrats were apt to compete rather than cooperate one another and tended to behave opportunistically—when they build organizational strategy and allocate resources, they are poised to expand their scope of operations regardless of their existing organizational tradition and talent, rather than focusing on fewer objectives or aid operations. The following chapters (See Chapter Six and Chapter Seven) unfold the details of evidence that supports the argument. Overall, this qualitative study contributes to the ‘control-vs-judgement’ debate in the following senses: first, this case study brings the debate into the context of new/emerging donors whose institutional settings are considered heterogeneous from those of Western-traditional donors. In that respect, this study newly pinpoints how the discussion can be useful and applied for our understanding of unconventional donors’ management design/strategy. Second, Korea’s top-down aid management design is re- evaluated in terms of its vulnerability to the commitment problem. Third, the following analyses of bureaucratic incentives provides more detailed evidence and anecdotes regarding the mechanisms of how the top-down approach harms donor’s long-term performance of making a reliable aid delivery at multiple levels of government.

1.3 Theoretical and Methodological Approach 1.3.1 Using Policy Process Theory Strategically: Assumptions and Conceptualization of Policy-Making Process Leaning on the discussion of ‘incentive and collective action problems’ that can arise in the processes of public policy formation and implementation, this thesis conceptualizes the public policy-making process as the collective action situation among a number of state agencies (organizations) within a government. Based on the conceptualization, this is in an attempt to identify the institutional causes of collective action outcomes that particularly undermine the ‘planning-practice’ correspondence5 of policy outputs. Apart from exogenous/international factors that have been vibrantly

5 The aid planning-practice gap is referred to as the ‘commitment problem’ as well. In this thesis, the commitment problem occurs with a country donor which fails to meet its pre-set goals of aid disbursement. It can be measured in either quantitative or qualitative ways in the following fields of aid contribution and practice: ODA scale, geographical/thematic allocation, aid delivery-related policy agendas for reform and change (e.g. reducing a ratio of tied aid), etc. 21 discussed as the determinants of foreign aid policy outputs in the existing international development studies, it is intended to examine how and to what extent the legal-regulatory conditions of the public policy process can help explain donor governments’ performance of carrying out a ‘credible and predictable’ practice of foreign aid so as to be reflective of the proclaimed intentions and planning goals. Speaking of how public policies are formulated, it is widely understood that a range of formal and informal institutions—from constitutions, laws, regulations, executive decisions, rule statements, social-cultural principles/norms, to traditions—are set to guide policy-makers, thereby leading to some expected points of policy output and policy performance (Birkland 2010; Weible 2014). However, what is still in need of further study in this field are the mechanisms of how and to what extent such endogenous-institutional factors are held accountable for unintended or even problematic collective outcomes within a donor government or even across donor government units with diversely configured rule contexts. Therefore, the intentions of this research—as an evaluative study of legal/regulatory frameworks of foreign aid policy-making by a selected criterion of ‘alignment’ between policy and actual practice—include a provision of policy implications and institutional resolutions that could possibly mitigate the incentive problems of interest. 1.3.2 Taking into Account the Wisdom of the Institutional Analysis and Development (IAD) Framework An Introduction to the IAD Framework The IAD framework6 was introduced by Vincent and Elinor Ostrom in 1971 (Ostrom and Ostrom 1971, p.211). The framework is designed to track the mechanisms of how a selected set of institutional factors affect actors’ decision-making process. By utilizing the framework, myriad research programs have examined dilemmas (or collective action problems) with the governance of common-pool resources in diverse settings (e.g. overexploitation of natural resources in groundwater basins, forests, fisheries, and grazing lands) (Cox and Schlager 2014). Not only addressing the common pool resources, the IAD research has extensively touched upon the governance of other

6 The IAD school of thought arises to examine and deal with ‘the tragedy of the commons’ (the term was first used by Garrett Hardin in 1968). Hardin’s article effectively brought our attention to situations where self-interest actors act independently and, as a result, they collectively harm public goods by using up or spoiling shared-resources such as rivers, fish stocks, oceans, forests, etc. Against such collective action problems that frequently take place in our daily lives the scholarship has taken up this issue and theorized the phenomena and discussed possible solutions. 22 types of resources—private, toll, and public resources—and resulting collective action problems.

Figure 1.2 The Institutional Analysis and Development (IAD) Framework: The Formation, Interplay, and Effects of Incentives throughout Hypothetical Stages of CONTEXT, ACTION ARENA, INTERACTION, and OUTCOME

(Source: Ostrom, Gardner, and Walker 1994, p.37)

As a thread of policy process literature, the IAD research has its distinctive merits in accumulating knowledge in terms of (i) context categorization, (ii) a concept of ‘action situation’, and (iii) development of rule typology. Categorization of Institutional Factors In the IAD framework, ‘CONTEXT’ (See Figure 1.2) refers to the initial conditions or the environment that structure actors’ ways of thinking and course of action in conscious or unconscious manners. The analytical framework (Kiser and Ostrom 1982) recognizes three categories of institutions: (a) physical/material conditions of resources in use, (b) formal and informal rules that are imposed on given activities, and (c) collective attributes of involving actors. The categorization of institutional factors is the product of accumulated research findings and theoretical discussion among the IAD scholars. This universal set of institutional variables, in fact, is useful to inclusively review explanations from various theories. The Concept of an Action Situation An ‘ACTION SITUATION’ refers to the situations where actors make their choices through interaction and exchange of information. This is a conceptual unit that allows researchers to reconstruct how a set of institutional factors shapes incentives, interactions, and collective outcomes ‘within particular institutional arrangements’. Action situations can be variedly delimited by the intention of researchers which also 23 generates great flexibility to any levels of analysis. An action situation can be defined by the following elements: what nature and what group of participants are included; what positions and allowable actions are assigned for the participants, what potential outcomes are delimited when specific tasks are assigned and carried out, how much authority (the level of control over choice) is given, what information is available, what the costs and benefits (punishment or rewards) are subject to actions taken and outcomes occurred. In assuming actors’ behavior within a defined action situation, there is room allowed for researchers to synthesize varied theoretical traditions of rationality depending on the nature of actors (i.e. human, organizational, state actors or public/private-sector actors) and involving contexts. Development of A Rule Typology Another important contribution of the IAD research is the development of a rule typology. Among IAD scholars, rules are defined as the enforced prescriptions about what physical and material conditions are required, prohibited, or permitted (Ostrom et al. 2001). As a main category of institutions, rules are assumed to have universality across time and space for institutionalists. Accumulated research findings give rise to a rule category which allows its adopters to produce a more systematic analysis. The seven types of working rules are so far recognized as common elements of rules-in-use at any given settings: (i) boundary rules, (ii) position rules, (iii) authority (choice) rules, (iv) scope rules, (v) aggregation rules, (vi) information rules, and (vii) payoff rules (Ostrom 2005, p.189; Cox and Schlager 2014, p.278). According to the rule category, ‘action arenas’ might show varied rule configurations in terms of who is legitimately involved [boundary rules], what positions they are given [position rules], what actions are allowable or otherwise [authority rules], what potential outcomes are delimited [scope rules], what information is available [information rules], what decision requires mutual agreement [aggregation rules] and what particular actions and outcomes are rewarded or penalized [payoff rules]. Limitations of the IAD Research Reflecting on the scholarly efforts of using the IAD framework and other relevant works (See Box 1.1), this study finds some critical problems with the previous application of the IAD framework (Further see Chapter Two). Given that the IAD-based empirical analyses in the domain of development cooperation focus on the effects of formal rules on foreign aid policy-making, the earlier application of the framework appears ineffective to acknowledge broader institutional contexts of policy-making rules and traditions across government units. 24 Considering the unique contexts of emerging/young donor governments whose policy-making tended or tends to be governed by greater ‘administrative discretion’ under a relatively insufficient and abstract set of written rules, future IAD research might require some sophisticated analytical procedures to discern the impacts of various types and configurations of policy-making rules on the outcomes. The next chapter presents a more detailed review of those studies and pinpoints the specific contribution of this research to the literature.

Box 1.1 List of Selected Neo-Institutional Research on Incentives with Foreign Aid (by author)

× Aid, Incentives, and Sustainability: An Institutional Analysis of Development (Ostrom et al. 2001) × The Institutional Economics of Foreign Aid (Martens et al. 2002) × The Samaritan’s Dilemma: The Political Economy of Development Aid (Gibson et al. 2005) × Delegation and Agency in International Organizations (Hawkins et al. 2006)

1.3.3 IAD Framework in Calibration: Taking ‘Formal/Informal Rules’ More Seriously In response to the critiques of the existing IAD application (See Chapter Two; Section 2.4), this study proposes further measures to allow the IAD research to draw clearer and more feasible policy implications. Apart from that, the most general elements of the IAD framework—from context, action arena, incentives, interactions, outcomes, and to evaluation—remain central in this refined version of the IAD framework. As a matter of fact, few IAD research activities put in much effort to identify what types of, and under what conditions, respective rule institutions affect the patterns of bureaucratic behavior and the outcomes of the policy-making process (McGinnis 2011). In this regard, this study recognizes two groups of rule institutions: one is formal institutions that include written constitutions, laws, regulations and other policy/rule statements [hard rules]; the other includes informal institutions which are not written down but perform as working rules (or ‘rules-in-use’) such as social/organizational norms, customs or traditions [soft rules]. Thus, this study adds extra measures to the IAD 25 framework in order to explore the relations7 between legal and informal rules. Also, it is to capture the varied impacts of the hard/soft rules on policy outputs. Another intention behind the fine-tuning of the IAD framework is to incorporate a simpler and more heuristic decision-making model of public policy. Instead of adopting the existing three-level decision-making model (i.e. operational, collective-choice, and constitutional stages of decision-making), the proposed framework assumes that public policy-making within government occurs primarily at two levels: Intra-organizationally and inter-organizationally.

Figure 1.3 Refined Framework for Institutional Analysis on Public Policy Process (by author)

In hypothesizing, explaining, and predicting the decision-making behavior of governmental agencies, the notions of bounded rationality and delegation theory are centrally applied. On the basis of those theoretical ideas, it is assumed that policy actors (as humans or organizations) are boundedly rational since they have limited information and capacity in performing their policy-making activities (e.g. reallocation of resources, interaction with other policy actors). 1.3.4 Methodology In order to understand the working mechanisms between rules and policy outputs on the policy process—in order words, what particular rule features induced (or otherwise)

7 The relations between hard rules and soft rules are regarded as complementary. This is because formal-legal rules approximate the most widely held social norms and soft rules fill the vacuum of legal rules in guiding actors’ decision-making and action (Cole 2014). For some exceptional occasions, however, some legal rules are set without having evident relations to working rules or social norms. 26 the gap between planning and practice with foreign aid policy formation and delivery in a chosen country case—, this research finds a qualitative research approach necessary and beneficial to be taken. Through gathering data mainly from state documents and interviews with policy elites and experts, this study is in an attempt to understand the policy-making dynamics under a certain rule and organizational design in terms of how organizational stakeholders or policy actors (i.e. the president, the National Assembly, prime minister office, ministries, implementing agencies, and other state agencies) interact and finalize aid policies. For this study, most evidence is gathered primarily through a comprehensive review of government documents and reports. The second largest body of evidence is drawn from interviews: Nearly two dozen government officials and experts participated.

1.4 Country Case 1.4.1 South Korean ODA Donorship The chosen case for the following IAD analysis is South Korea (hereafter Korea). Earlier, this country graduated from aid recipientship in the late 1990s. Since then, Korea has grown into a recognized donor, joining the Development Assistance Committee of the Organisation for Economic Co-operation and Development (OECD/DAC) in 2010. For the last decade, Korea has impressively scaled up its ODA contributions, reaching an annual aid disbursement of USD 1.9 billion (in net ODA) as of 2015 (OECD 2016).

Figure 1.4 Korean ODA Growth (1999-2015) (OECD 2016)

Despite such a compressed growth in foreign aid, the country has displayed a persistent and significant gap between annual commitments and actual disbursements 27 over the last decade, with a record of failing to reach its annual and multi-year aid targets by 60.7 percent in 2015.

Table 1.1 Korea's Performance of Making a Credible Aid Commitment (2000-15) (Calculated by author; OECD QWIDS (Accessed Nov. 24, 2017))

Net Total

Disbursements Commitments Performance Year (USD millions) (USD millions) (Percent) 2000 212.1 325.0 65.3 2001 264.7 262.7 *100.8 2002 278.8 368.8 75.6 2003 365.9 437.2 83.7 2004 423.3 780.5 54.2 2005 752.3 771.7 **97.5 2006 455.3 892.0 51.0 2007 696.1 1381.9 50.4 2008 802.3 1667.8 48.1 2009 816.0 1980.1 41.2 2010 1173.8 1996.3 58.8 2011 1324.6 1922.7 68.9 2012 1597.5 2295.7 69.6 2013 1755.4 2646.4 66.3 2014 1856.7 2688.4 69.1 2015 1915.4 2931.6 65.3

(Note: Those exceptions (*/**) were made in the occasional contexts of a surge in bilateral grants to Iraq and Afghanistan in 2001 and multilateral contributions upon Korea’s newly acquired membership in IDA and the Inter-American Development Bank in 2005)

Along with Japan, Korea’s performance might look disturbing when compared to other DAC donors. In 2015, for 28 DAC member countries, the average ratio of annual commitment to actual disbursement stands at 85.5. Non-DAC countries, appear over- performing (224.5) and EU-DAC countries turn out to perform better (95.9 percent) than the G7 country group (82.4 percent). On the theoretical front, the Korean case has some added value for institutional research on emerging donorship. As frequently viewed, the aid giving of emerging donors is considered widely heterogeneous compared to traditional donorship. As a possible reason for that heterogeneity, many point out that their aid policies and practice appear to arise in the absence of prolonged aid-giving traditions or a well-established set of formal institutions. Through an in-depth study of a young donor like Korea, a wider spectrum of institutional environments can be captured for existing and potential incentive problems 28 in the sphere of foreign aid policy. Furthermore, this single country study is expected to add contextual knowledge about the logic, design, and performance of rule settings in young donors; and incentive problems that young donors might frequently suffer. 1.4.2 Legal and Regulatory Setting for Foreign Aid Management in Korea As the primary legal foundation for ODA policy and practice, The Framework Act on International Development Cooperation was enacted in 2009 and later frequently revised. Although Korea’s ODA legal framework is evolving rapidly, the widely recognized problems of missing integrative objectives, insufficient co-ordination, and fragmentation remain largely intact (OECD 2012; OECD forthcoming). Despite frequent modifications, the law still displays a lack of important details, champions a state-centric understanding of development cooperation, and allows significant presidential leverage over annual aid allocation and mid-term strategy building. The Korean ODA Act embraces two incompatible objectives of economic cooperation and global responsibility. The loan and grant communities respectively champion each of these mutually exclusive ODA pillars. Also, the ODA Act establishes a coordinating body across the aid system, which is given limited authority and capacity to exercise its mandate (OECD 2012). Other subordinate laws specify tasks for implementing agencies under the supervision of ministries. 1.4.3 Two Nested Cases: State Agencies’ Action Situations with Knowledge Aid and Multilateral Cooperation OECD’s reports (1999, 2009) identify a diverse structures of donors’ aid administration (See Figure 1.5): Korea, falling into the category of model 3, has a dualistic structure in which the finance ministry and the foreign affair ministry equally design and manage ODA activities.

Figure 1.5 DAC Donors' Structures for ODA Management: by Four Models

Model 1: Integrated within the Ministry of Model 2: Development Co-operation Foreign Affairs (Denmark, Norway) Department/Agency within the Ministry of Foreign Affairs (Finland, Ireland, Italy, Africa Asia Latin Netherlands, etc.) Department Department America Department • Foreign • Foreign • Foreign Ministry of Foreign Affairs policy policy policy • Trade • Other • Other • Other • Foreign Affairs • Develop • Develop • Develop • Development (directorate) ment Co- ment Co- ment Co- operatio operatio operatio n n n

Model 3: Policy Ministry with separate Model 4: Ministry/Agency responsible for policy implementing agencies (Belgium, France, and implementation other than Ministry of Germany, United States, etc.) Foreign Affairs (Australia, Canada, U.K., Japan)

• Ministry of Foreign Affairs • Ministry/Agency for Development Co- • Implementing Agencies operation

(Source: OECD 1999, 2009)

To provide more specific cases and evidence, this study sheds light on two sub- policy areas of foreign aid: the knowledge-sharing initiative and multilateral cooperation. The chosen policy fields represent the fastest-growing aid sectors over the last decade. More importantly, the selection of the sub-policy areas is highly relevant to the unique structure of the Korean aid administration, where the Ministry of Foreign Affairs supervises all grant activities and the Ministry of Finance oversees loan operations. Exceptionally, the two managerial ministries (the foreign affairs ministry and the finance ministry) and their subordinate organizations equally manage knowledge and multilateral funding. Thus, the two cases are useful to include the most comprehensive group of state agencies that are involved in the foreign aid policy process. The two policy arenas allow us to comparatively examine the behavior of state agencies. Nested Case (1): Knowledge Aid Knowledge aid8 is an umbrella term that embraces a range of Korean technical assistance projects that aim to disseminate Korea’s know-how for industrial upgrade and economic growth. Since the early 2000, knowledge aid has emerged as one of the fastest- growing aid sectors within the Korean government (OECD 2012). In the sphere of knowledge aid policy formation and operations, the following state organizations are the dominant actors from the legal and budgetary aspects: the president, the Ministry of Foreign Affairs (MOFA), the Ministry of Strategy and Finance (MOSF), the Korea Development Institute (KDI), the Economic Development Cooperation Fund (EDCF), and the Korea International Cooperation Agency (KOICA). The Committee for International Development Cooperation (CIDC), under the Prime Minister’s Office, is assigned to take the role of cross-government coordination and evaluation.

8 This term is used interchangeably with ‘knowledge sharing (KS)’ or ‘development consulting’ in the aid community in Korea. 30 Figure 1.6 Organizational Structure for Knowledge Aid Policy Planning and Implementation in Korea (by author; as of 2016)

Nested Case (2): Multilateral Aid In the Korean aid administration, a similar set of organizational actors are involved in the aid policy process.

Figure 1.7 Organizational Structure for Multilateral Aid Planning and Implementation in Korea (by author; as of 2016)

1.5 Structure This thesis is divided into two parts. The first part consists of theoretical review (Chapter Two), conceptual framework (Chapter Three), and the chosen methodology (Chapter Four). The other part presents empirical evidence from the three stages of the IAD analysis—CONTEXTS (Chapter Five), ACTION SITUATIONS (Chapter Six and Chapter Seven), and EVALUATION (Chapter Eight)—to help explain the Korean government’s contemporary commitment problem (2000-2016). This is to see how the legal and regulatory frameworks of foreign aid policy-making are configured and what 31 type of incentives/collective actions are induced under the rule configuration as the direct cause of the commitment issue. 1.5.1 PART ONE: Theory and Research Design Chapter Two: Theory Review This chapter introduces the Institutional Analysis and Development (IAD) approach, one of the new institutionalist variants. This chapter captures the emergence and development of the IAD tradition in the domain of policy process and development studies. Most importantly, this chapter discusses the benefits and limitations of the IAD in understanding the institutional causes of incentive problems in cases of new/emerging donors. Chapter Three: Conceptual Framework Chapter Three redefines theoretical elements and analytical measures of the refined IAD framework. The refined framework arises to address the limitations of the earlier application of the IAD approach in the field of development cooperation. Also, it is intended to enhance the utility of the framework for broader studies on diverse settings of development cooperation. Chapter Four: Method and Research Strategy Chapter Four unfolds methodological considerations and data collection techniques. For the suggested institutional analysis that examines ‘intra-governmental dynamics’ in terms of how state organizations individually and collectively behave and make policy decisions under a configured set of hard and soft rules, the case study method (CSM) appears beneficial. Major data collection techniques in use are documentary review and elite/expert interview. 1.5.2 PART TWO: Empirical Analysis Chapter Five: Analysis of Historical and Contemporary Contexts The first empirical chapter sheds light on the ‘CONTEXTS’, an initial analytical area of IAD research. This chapter examines the three contextual factors—physical and material conditions, attributes of the community, and rules-in-use—that structure various action situations of Korea’s foreign aid policy formation and implementation in the past (1960s-90s) and present (2000-16). Chapter Six: Analysis of an Action Situation with Knowledge Aid Policy-Making Among a number of ACTION SITUATIONS that take place in the ODA policy process, Chapter Six offers detailed evidence on policy actors’ rationale and behavior through a nested case of knowledge aid management and delivery. It is an attempt to see what kind of respective and collective actions have been induced under the legal- 32 regulatory arrangements and how the patterns of collective actions are directly or indirectly held accountable for a long-term gap between aid planning and practice in Korean ODA. Chapter Seven: Analysis of another Action Situation with Multilateral Aid Policy-Making This second ACTION SITUATION analysis reveals policy actors’ incentives and collective action problems that cause the gap between planning and practice in a sub- policy area of foreign aid, multilateral ODA. In tandem with the pervious chapter, the chapter provides us with some comparative lessons in terms of how variedly the incentive and commitment problems occurred and what explains the variations in outcomes in each policy area. Chapter Eight: Evaluation and Further Discussion The last chapter concludes the thesis by providing the overall findings and am extended discussion on the relations between rules and policy outputs on the public policy process. This thesis explores the question of how rules determine state organizations’ collective actions that are attributable to a planning-practice gap in the national system of foreign aid management and delivery. The thesis is recapped as follows: it is an evaluative study that focuses on the logic, design and performance of policy-making rules as the culprit of country donors’ rhetoric aid policy; and secondly, some additional analytical procedures are suggested to the existing use of the IAD framework to capture the multi- dimensional effects of rule institutions; and finally, this thesis is intended to make a unique contribution by adding knowledge about new donors.

PART ONE: THEORY AND RESEARCH DESIGN

34 CHAPTER TWO THEORETICAL REVIEW: THE INSTITUTIONAL ANALYSIS AND DEVELOPMENT (IAD) APPROACH TO THE DEVELOPMENT COOPERATION CHALLENGE

This chapter introduces the IAD approach, one of the neo-institutionalist traditions, that is theoretically chosen for this research. Distinctively, among many other policy process theories, the IAD research provides a useful framework for our better, holistic understanding of aid management design or strategy of donor governments or aid agencies. More importantly, this approach allows us to pay attention to the ‘incentives’ of state agencies on the policy process and to examine how they are created and induce particular policy outputs within the aid management system with regard to donor’s big- plan-small-action performance. To begin with, this theoretical review opens up the discussion by pointing out the importance of the research on rules, incentives, and interactions among policy actors in the foreign aid/development studies literature (Section 2.1). Next sections (Section 2.2 and 2.3) capture the emergence and so-far developments of the IAD theory and framework. Despite the unique contributions of IAD research to the literature, this review includes a critique (Section 2.4) of the existing IAD-inspired research, in particular, critically reviewing the IAD study of Swedish International Development Agency (SIDA). This chapter (Section 2.5) concludes that there is some significant room for the previous use of the IAD framework to be refined for its wider application to aid-giving government units and organizations as potential cases for IAD analysis.

2.1 Revisiting the Importance of the Research on Rules, Incentives, and the Commitment Problem in the Foreign Aid Context To recap the introduction chapter, this study is intended to examine the mechanisms of how donor governments’ aid management design (by allocation of control and autonomy) affect their performance of delivering aid commitments. Upon the theme, this study sheds light on ‘bureaucratic incentives’, specifically how they are created, interplayed, and reflected into policy outputs that are accountable for the aid planning- practice gap within donor governments’ aid system. Prior to the introduction to the IAD tradition, the following section discusses why incentive study is particularly important for our better understanding of today’s challenge of aid givers in designing and delivering foreign aid effectively and sustainably. 35 2.1.1 Why Donors’ Aid Management Design/Strategy Matters Previously, the notion of how donors’ political economy determines their aid allocation strategies and management performances have been critically explored. For instance, donors’ preferences of aid delivery channels (i.e. between flowing ODA through government-to-government channels and outsourcing ODA delivery to non-state development actors)—as donors’ management strategies—are revisited and argued that donors’ embedded idea of the role of the state (i.e. state-led development model vs. market-led one for economic growth) largely affects their choices of aid delivery. When some country cases, like the U.S., the U.K., and Sweden, are considered, they appear to have the traditions of championing market efficiency so that they prefer to deliver foreign aid with the strong involvement of non-state actors and groups, whereas those countries, such as France, Germany, and Japan, emphasize heavy state intervention on their national development processes so that they tend to prefer a state provision of foreign aid (Dietrich 2013, 2016). Likewise, regarding donors’ strategic choice between channeling more foreign aid through multilateral organizations and giving it more through bilateral channels, Milner and others argue that this is rather a political decision—not an ‘aid effectiveness’ or ‘aid accountability’-driven issue—between the two competing motives, ‘national interest’— through an expanded provision of multilateral funding it is expected to gain more political and economic capitals at international organizations—and ‘burden-sharing’—growing multilateral aid represents increased domestic support for altruistic/humanitarian values against global challenges such as poverty and peace—(Milner 2006; Milner and Tingley 2013). Along with the discussion, Gulrajani (2016) and others (Michaelowa et al. 2015; Eichenauer and Reinsberg 2017; Eichenauer and Hug 2018) point out the growing use of multi-bi aid—which is a hybrid form of bilateral and multilateral ODA and is given in the form of multilateral funding, but the specific use of the fund is not determined by the multilateral organizations but by the country funders. She argues that this strategic choice of aid allocation and delivery channel undermines greatly the performance and credibility of multilateral institutions and operations (Gulrajani 2016). 2.1.2 Why Incentives Matter: Rethinking Donors’ Inherent Challenge Being a staple subject of public policy and public administration studies, bureaucratic incentives, in fact, represent a re-emerging theme in contemporary development studies. In the relatively short history of foreign aid, ever since the Second World War, whenever the role of foreign aid in the developing world is challenged or 36 even seen as counterproductive, many tend to locate the major causes of the ‘failing aid’ primarily on the side of recipient countries. Against such mainstream narratives of ‘corrupt, incompetent recipient governments misusing foreign aid’, some self-reflecting voices were echoed among international aid practitioners and scholars in the 1980s. The idea was that donor aid agencies, many of which are either believed to be or appeared to be competent to perform as ‘solution makers’ for global underdevelopment issues, might not be free from their own challenges in aid management. A group of scholars made the strong point that the aid agencies of donor governments often suffer from structural problems that induce perverse incentives and ineffective practices; therefore, the opportunities to make things better also rely on donor countries’ determination and self-correcting efforts to identify and tackle their own structural constraints (Ufford et al. 1988). However, as such initiatives are highly politically sensitive within the domestic political arena, it was rare that they were proactively debated or openly conducted across the donor community in recent decades. Persistent public requests for thorough scrutiny of their aid agencies remained largely absent, as well. Another wave of this movement arrived in the early 2000s. Around that time, the expectation was high, once again, in the sphere of development cooperation, as global aid donors jointly set eight millennium development goals (MDGs) at the United Nations, and reaffirmed their future stances towards reorienting and scaling up development assistance for the developing world. As part of the effort, the incentive problem was revisited in order to identify the unique incentive structures that development professionals at aid agencies might be put under; the incentives and resulting problems are considered distinctive when compared to those created in any other public or private organization. Noticeably enough, the Swedish International Development Agency (SIDA) commissioned an unprecedented research project in which a band of political scientists at Indiana University was invited to produce a pioneering study—titled Aid, Incentives, and Sustainability: An Institutional Analysis of Development Cooperation—on incentives that work against the sustainability of aid within SIDA (Ostrom et al. 2001). William Easterly (2002; 2011) also played a part in drawing public attention to the ‘structure-performance’ sensitivity at aid agencies: Based on his direct observation, he elaborates on the effects of organizational and procedural design on the quality of aid programs at multilateral aid agencies. He argues that the good intentions of aid professionals or country donors do not necessarily guarantee good outcomes when 37 efficient, adequate institutional arrangements (i.e. organizational structure and procedures) are absent. 2.1.3 Aid Motivation Probably Matters Less Prior to the discussion on the incentive problems associated with foreign aid, foreign aid has been far more discussed regarding why it is given (or why some third world countries attract more aid than others) and what aid motivation dominates others in respective donor societies. On this popular theme, numerous research projects have provided quantitative and qualitative evidence on the multifaceted motives of country donors (representatively, Hook 1995; Schraeder et al. 1998; Alesina and Dollar 2000; Lancaster 2007). Aid scale and allocation are predominantly explained and predicted from a dimension of ‘national interests versus altruistic/humanitarian motives’, which allows us to redefine the motivational orientations of major donors: The U.S. and Japan, for instance, are regarded as the aid givers who prioritize their geopolitical and commercial interests, while Nordic countries traditionally emphasize the humanitarian use of foreign aid. Arguably, the ‘aid motive’ has been the major window through which to understand the policy choices and behaviors of bilateral donors. However, if we are to examine why foreign aid fails to achieve attached objectives, or to what extent aid delivers its goals, the ‘aid intention’ literature gives a partial answer to that (Easterly 2002; Williamson 2010). It is no surprise that aid initiatives with benevolent and clear-cut intentions often turn into something else when they are delivered on the ground and, even prior to that, when they are finalized as the result of internal policy discussions among donors’ aid policy-makers and practitioners. To enhance our knowledge about the intermediate dynamics of how aid motives, at their conceptual stage, correspond to specific management strategies and delivery options during the policy process, the following discussion is designed to provide a gap-filler answer to such puzzles regarding particular policy behavior and performance of bilateral donors.

2.2 Positioning and Foundations of IAD: Policy Process Literature and Agency Theory 2.2.1 IAD in Policy Process Literature Among a variety of theoretical traditions that model the human/organization interactions involving the public policy process, the school of the Institutional Analysis and Development (IAD) provides the theoretical foundations of this study. The tradition, 38 one of the new institutionalist approaches, initially emerged to explain the human phenomenon of governing common-pool resources within communities or even wider geographical/social boundaries. Later, the IAD research extended to explore collective action problems in the settings of public policy formation and implementation. Its debut in public policy and public administration studies is significant since it offers a refreshing insight through its epistemological framework (Cox and Schlager 2014, p.267). Although there are a number of theories that have varied points of utility and limitation, IAD, if the management of public resources comes to the fore, has stronger merits in making a systematic and pragmatic analysis than other conventional and emerging approaches9 in the sphere of policy process research. Common Elements Across Policy Process Theories In general, scholars in this field have shared interests in the following elements to explain and predict the ‘interactions’ and resulting ‘outcomes’ on the policy process (Schlager 2007; Weible 2014; Cairney and Heikkila 2014). Most theories define ‘key actors’ who can be individuals or collectives; this is regarded as a necessary step to singling out a group of influential policy-making actors or organizations, since this task allows us to manage the complexity of policy-making realities where numerous people and organizations are involved. Besides the identification of key actor group, some theories highlight the role of a secondary participant group (‘network or subsystem’10) which functions to feed advice and information to key actors. Policy process theories variedly define the scope, types, and categorization of ‘institutions’. Also, policy process theories touch upon ‘ideologies’: Evolving beliefs and ideas among bureaucrats that constitute some explanations for policy change. ‘Policy context (policy environment)’ — which refers to the conditions that unsystematically influence policy decisions, such as the macroeconomic performance of a country, public opinion, election outcomes, wars/terror—is an inclusive element in policy process research. Various Foci and Concepts in the Policy Process Studies Scholars on different theoretical fronts have developed their own definitions and supporting concepts with regard to the aforementioned elements of policy process research. As active research programs, policy process theories explore a varied selection

9 The following theoretical approaches in the policy process literature are mainly reviewed and considered: the multiple streams approach (MSA), the punctuated equilibrium theory (PET), the Diffusion of Innovations theory (DOI), the narrative policy framework (NPF), the advocacy coalition framework (ACF), and the social construction frame (SCF), and the policy feedback theory (PFT). 10 This element of the policy process research captures the procedural-democratic feature of modern public policy-making, in which state bureaucrats are required to exchange ideas and information with interest/expert groups. 39 of relationships between variables, producing varied scopes and levels of observation and analysis. To reconfirm the merits of the IAD, a brief cross-theory comparison might be useful to begin with. The IAD corresponds with the purpose of this research in that this study specifically examines the contextual relationship between ‘rules-in-use’ and policy outputs of the foreign aid policy process. Fundamental differences among a variety of policy process theories stem from the conceptualization of policy-making systems in the first place. The advocacy coalition framework (ACF) and the punctuated equilibrium theory (PET) find answers to the question of policy change or stability mainly from changed dynamics in the competition between coalitions (ACF) or the influence of subsystem on core policy-making venues (PET). Others might provide more sensible explanations regarding the impacts of established state policy narratives/designs on mass opinion (the narrative policy framework: NPF) or on targeted populations (the social construction frame: SCF; the policy feedback theory: PFT). The diffusion of innovation theory (DOI), which depicts policy processes as interactions of a group of nation-states, can serve to better understand governments’ policy adoption through the frame of how and why one country emulates policy choices of others. When hypothesizing the behavior of actors, some theoretical stances favor the notion of ‘bounded rationality’. The MSA, for instance, stresses time constraints and its impacts on policy outputs. The SCF presumes that policy-makers are prone to make quick and non-rational decisions, thereby being subject to past policy designs and selectively screened information. In the ACF discussion, the ideas and policy learning of policy- makers are considered critical factors in determining the outcomes of coalition formation. The DOI, on the other hand, is the approach that the logic and behavior of actors are assumed to be either boundedly or comprehensively rational when both country and subnational actors are considered. Similar to the DOI approach, the IAD appears to have greater room to incorporate a flexible notion of rationality (from bounded to perfect rationality) than any other approach. The IAD has developed a framework that can be flexibly applicable to actors that exist of various types, levels, and institutional environments. To address the question of how rule institutions shape the collective behavior of state bureaucracy, the IAD is considered the most optimized to conduct generic investigations over the creation and direct impact of institutions on the policy process. On the contrary, the MSA and SCF are the approaches that anchor to the most loosely defined concept/role of institutions. 40 However, regarding the role of subsystems, the IAD turns out to be the one that least develops ideas on that aspect, together with the SCF. Policy Process Research on Foreign Aid Policy In recent decades, relatively few policy process research projects have explored foreign aid policy. One of them is the work of Travis and Zahariadis (2002). They conduct a qualitative study of U.S. foreign aid allocation, utilizing Kingdon’s multiple streams model, in which both external stimuli and domestic-political factors are believed to determine policy outputs. Empirically suggested in this study is that U.S. foreign aid policy outputs during the pre-2000 period tended to be sensitively subject to external factors than domestic ones. In addition, Travis (2010) provides further insights regarding U.S. foreign aid behavior towards Africa, reconfirming his previous argument. Milner and Tingley (2010) take a similar point of view in their understanding of the US’s shifted multilateral aid policy. The accumulated policy process research in this particular policy area of foreign aid allows us to better understand the interactions and impacts of the human and conditional factors behind foreign aid policy decision-making. However, the existing works focus primarily on the U.S. and leave substantial room for researchers to scrutinize aid policy process by utilizing a wider range of policy process theories. 2.2.2 Agency (Delegation) Theory: As the Notional Foundation of IAD As a theoretical foundation of new institutionalist organization/public administration studies, agency theory is outlined in this section. Agency theory emphasizes the idea of interdependence, which is a close reflection of our realities (Eisenhardt 1989; Neilson and Tierney 2003; Shapiro 2005). Considering the functions and existence of respective human or organizational actors in any social and political setting, none of the actors stands alone: They are, rather, intermediate agencies that either delegate or are delegated to perform a defined set of jobs and they collectively work for common goals. Therefore, institutions that structure such interdependent relationships among actors, and that engineer the whole delegation situation in societies lie at the heart of agency analysis. In that sense, agency theory is also labelled as delegation theory. Delegation, a common practice in modern public and private management, takes place between two parties—an agent (the hired) and a principal (the one hiring)—both of which have mutually agreed that the principal delegates defined tasks to a capable agent to perform the tasks. Such principal-agent (PA) relationships are equivalent to a mutual contract, and exist everywhere in our daily lives (e.g. electorates and politicians in a representative democracy). 41 Agency theorists try to explain social and political phenomena as problems that may occur fundamentally from conflicts of interest between a principal and an agent. One of the main concerns rests on how the principal can control the agent through proper institutional tools that can possibly curb agents’ self-interest, therefore achieving the objective of the principal.

Figure 2.1 Principal-Agent Model

(Source: Laffont and Martimort 2009)

Principals and Agents Principals and agents can be relatively defined: One may perform the role of either principal or agent, depending on whom the actor delegates a mandate to or who is delegated a mandate from. In other words, actors are defined by their relationships to one another (Hawkins et al. 2006). For instance, a foreign affair ministry in a donor government performs its mandate of making foreign aid policies. In this case, the ministry serves as the role of the agent who is given direct authority from a higher position (e.g. the state executive). At the same time, when the ministry hires agencies to have them perform aid delivery, the ministry becomes the principal for the subordinate agencies. Delegation Chain A delegation chain refers to an interconnected array of delegation-ships between principals and agents. A delegation chain functions as a collective-action situation where a number of PA relations interplay. Besides observing respective PA relations, an observation of a delegation chain is expected to offer a holistic view for understanding collective incentives among all involving actors. Foreign Aid Management from the Perspective of Delegation Theory When it comes to foreign aid management—from the very initial stage of foreign aid planning to the implementation of development interventions overseas—agency theory serves as a good prism to explain and predict the behavior of those who are involved in the cross-national transfer of capital and resources. As for foreign aid research, 42 agency theory is applied to various levels of actors and delegation: it can be about an inter-governmental relationship between donor and recipient countries; it can be limitedly applied to delegation-ships between sub-state organizations within a donor country; or to the micro-level relationships between managers and employees within a particular aid agency. Those country-, organization-, and human-level actors are equally assumed to have self-interest and preferences as the principal and agent. Across donor government units, it is generally the practice that foreign aid resources are transferred through a delegation chain that starts from the taxpayers (the final principal), and extended through elected parliamentarians, public administrations (president, ministries, and subordinate agencies), NGOs and private-sector bodies (subcontractors), recipient governments, to foreign citizens (the final beneficiary). None of them are stand-alone in the delivery process of foreign aid. Along with the common structure, each donor country might have unique features in its delegation chain. Those unique-nesses depend on what funding sources (e.g. taxpayers (non-voluntary funders) or voluntary funders) are procured for ODA spending; how many intermediate organizations take part in ODA policy-making and implementation; and whether governments prefer either exercising ODA on their own or delegating the task to international organizations with the expectation of benefiting from the alleged high competency of international organizations (Milner 2006), etc. 2.2.3 Delegation Theory Applied to Foreign Aid Policy Analysis Although the new institutionalist studies have embodied established explanations on a wide range of human/mass/country affairs in the modern social sciences, the tradition has yet to be widely embraced by the existing scholarly effort of understanding the dynamics and problems of foreign aid. The typical institutionalist questions—such as how institutions guide the behavior of organizations and persons, or how incentives play out between institutions and agents—have been revisited in settings of aid policy formation and implementation since the early 2000s. On the basis of delegation theory, several analytical attempts11 have been carried out to help understand incentive problems with foreign aid management and delivery. Those models for incentive analysis are varied in mainly two elements: one is the evaluative criteria (e.g. effectiveness, accountability, or sustainability of foreign aid delivery); the other is the scope and focus levels of analysis (e.g. either focusing on a cross-national relationship between donor and recipient governments or delimitedly

11 For further information, a critical review of other delegation theory-oriented research in the context of foreign aid delivery (except for IAD research) is attached in the Appendix (Go to Appendix A). 43 looking at in-donor delegation that spans an aid agency). In other word, the variations primarily come from researchers’ delimitation of a delegation chain for their research. Analysis of Policy-Making Contexts and Collective Actions A publication, titled the Samaritan’s Dilemma: The Political Economy of Development Aid (Gibson et al. 2005) carries great theoretical and policy importance for foreign aid studies and IAD research. Its theoretical discussion was based on empirical evidence collected from an initial study commissioned by the SIDA; as a self-evaluation, the SIDA report (Ostrom et al. 2001) was prepared to identify perverse incentives that worked against sustainable outcomes of development assistance. To the core, Gibson and his collaborators share the idea that the good intentions of donor countries do not necessarily guarantee good results. They argue that even altruistically motivated development interventions might fail to meet good outcomes due to the structure of foreign aid delivery that created perverse incentives against making foreign aid developmentally sustainable in the processes. Regarding the structurally driven incentive problems within various actors (ranging from donors, recipients, and intermediate agencies), agency theory is revisited. The scholars have come up with a network map that includes those who are involved in the resource transfer from donors to recipients at the global level (See the figure below). They categorize all the involved actors into eight different groups. It is assumed that those key actor groups have multiple relationships with one another. Not all relationships are defined as formal delegation (i.e. a contractual relationship) since relationship might exist to very different degrees, and vary in their contents of obligation and rights.

Figure 2.2 Delegation Chain in the Study of Gibson and Others: Interactions among Eight Key Actors in the International Development Arena (Named the International Development Cooperation Octangle)

(Source: Gibson et al. 2005: 64)

This network map is useful in keeping us aware of the scale and complexity of associated interactions in the field of global development cooperation. With the knowledge of the global-level structure of foreign aid delivery, Gibson and his colleagues (2005) applied the IAD framework to identify undesirable incentives at various levels and occasions through specific cases: policy-level incentives for donor countries to allocate aid resources upon recipient conditions; incentives that are differently induced by aid modalities; incentives inside an aid agency (SIDA); incentives for subcontractors in bidding and implementation processes; and on-the-ground incentives for recipient governments (in India and Zambia) in working with SIDA. Although the scholars highlighted incentive problems that were confined neither to particular actors (either donors, recipients, or intermediates) nor aid types, one of their institutional analyses, which touches upon the internal incentive structure of donor’s agency is of great importance and relevance to this thesis. The IAD’s three contexts— physical/material conditions, attributes of community, and rules-in-use, all of which are believed to jointly affect decision-making among actors—are unprecedentedly taken into consideration to explain an observed set of routinized behaviors for SIDA staff’s daily and long-term decision-making. The context-oriented explanation speaks to the outcomes, after all. More details about the operationalization of the framework and findings follow.

2.3 Introduction to IAD Framework: Analyzing Incentives at Sequential Stages of Policy-making Process 2.3.1 Incentive Analysis of Foreign Aid Policy Process: An IAD Logic The basic rationale behind this IAD research is that one needs to start her/his learning of the incentive/collective action problems with institutional conditions. The institutional conditions are the structural forces that shape actors’ choices and actions. Incentive structures associated with public policy-making are not just determined by formal institutions, such as, laws, regulations, or rule statements. They are also structured by cultural and traditional factors, which are less invisible forms of institutions than formal ones. Although the policy-making process is admittedly highly complex, with multiple endogenous/institutional factors to be considered, the IAD suggests a framework that is useful to track the logically sequential stages of policy-making: from contexts, action arena, incentives, interactions, outcomes, to evaluation. Those stages follow the formation, workings, and end-outcomes of incentives for collective decision-making for 45 policy-makers. In other words, the best interests of the framework lie in its systematic understanding of the logic, design, and performance of institutional arrangements in various settings and scales (Ostrom, Cox, and Schlager 2014). 2.3.2 Analytical Stages in the IAD Framework: ‘Contexts’, ‘Action Situations’, and ‘Evaluation of Outcomes’ Contexts: First Stage of the Analysis As the opening set of factors to be analyzed, ‘contexts’ refer to the initial conditions or the environment that structure actors’ ways of thinking and course of action in conscious or unconscious manners. The IAD scholars classify a group of contextual factors into three categories: (i) physical/material conditions, (ii) rules-in-use, and (iii) attributes of a community. The three groups of contexts provide an important starting point for the analysis. ‘Physical/material conditions’ are directly linked to the nature of resources-in-use. This is about what type of goods or services are involved. Identifying whether they are private goods, common-pool resources, or public goods is central, since the type of the resources greatly determines actors’ incentive structures for decision-making. ‘Rules-in- use’ are shared understandings among those involved that refer to enforced prescriptions about what actions (states of the world) are required, prohibited, or permitted (Ostrom et al. 2001). Although there are many different sources of the rules that exist for daily life, governments’ activities, including its major task of public policy formation and implementation, are governed by a set of rules-in-use that are drawn from a wide range of externally imposed and self-regulating institutions: ranging from constitutions or national laws to its accumulated experience/learning with past practices or self-crafted norms and principles. The third part of the contexts, ‘attributes of a community’, is about the characteristics the participating actors display as a group (or a community) in terms of size of the group, human/organizational makeup, capacity to self-organize/self- regulate or to recruit participants, cultural-historical backgrounds (e.g. education, ethnicity), etc.

46 Figure 2.3 IAD Framework: Formation, Interplay, and Effects of Incentives (Source: Ostrom, Gardner, and Walker 1994, 37)

Actors and Action Situations: Second Stage of the Analysis Once the contextual factors involved are revealed, the next step is defining a specific ‘action arena’ in tandem with the pre-identified contextual understandings. This task is central to delimiting the scope and conceptual unit of analysis. An ‘action arena’ refers to a social space where actors perform their mandates by interacting and exchanging knowledge with others. At this stage of analysis, the researchers are expected not only to clarify who is participating and what action situations occur in the given action arena, but also to observe the behavior of respective actors. Action arenas can be variously defined by the intention of researchers. Patterns of Interactions and Outcomes (Evaluation): Third Stage of the Analysis Incentive structures for multiple actors can be further understood by identifying patterns of interaction and resulting outcomes. Interaction among actors, as the key activity in the public policy process, is a focal subject at this stage. Interactions take place in various forms within a government, not only by exchanging information between state agencies, but also by collaborating and coordinating their policy choices and actions. This stage of analysis is evaluative since its main analytical objective is to conclude what institutions induced what types of incentives and collective actions, and what direct outcomes resulted from repeated collective actions in the community. There is an additional sequence between the outcomes and institutions, as the IAD framework acknowledges. This is an optional stage of the analysis for the IAD research, depending on the intentions of the study.

47 2.4 Critique of Previous IAD Study: Envisioning More Practical and Wider Application of the IAD Framework 2.4.1 Lessons from the IAD Study on Swedish Development Cooperation Returning to the discussion on the applications of the IAD framework, this section further elaborates on the actual operation and empirical lessons of an aid agency-focused IAD analysis. The study was initially released in the form of an external evaluation report at SIDA (Ostrom et al. 2001). Later, the IAD scholars advanced their theoretical discussions based on the empirical evidence and republished them as a book, titled The Samaritan’s Dilemma: The Political Economy of Development Aid (Gibson et al. 2005), and a journal article (Andersson 2009). Among donor agency problems that arise at various levels and sectors, the study examines SIDA’s internal incentive structures for individual and organizational learning about sustainable outcomes in aid-receiving fields. The findings are based on the information obtained from interviews and web-based surveys with randomly selected SIDA staff. The study locates a focal collective action problem at the information exchange between the Stockholm headquarters and field offices. In the process of identifying resulting collective action patterns, some theoretical hypotheses are confirmed with motivational biases observed in the budgetary-administrative and evaluation processes within the Swedish aid agency. Empirical Findings The fieldwork (2000) for the research came after the rationalization of the Swedish aid system, where an old SIDA and other multiple aid-managing agencies were merged into a new SIDA in 1995. Considering the timing of the study, it served as an opportunity to evaluate the impacts of the performed structural reform (1995) and subsequent legal modifications on budgetary rules (1998) to Swedish development cooperation. As frequently emphasized in the analysis, the researchers duly followed the IAD procedures: by beginning with three contextual variables (constitutional-legal, cultural, and material factors) and later discussing resulting effects of those institutional factors on day-to-day decision-making of managerial and desk officers at SIDA. Constitutionally, the Swedish government is obligated to devote 1 percent of the country’s gross national product to global development cooperation. Upon the constitutional mandate, the government prepares a budget proposal and the Swedish parliament revises/approves annual budgets for aid spending. At the top policy-making level, the Ministry of Foreign Affairs (MFA) sets overall guidelines and plays a 48 monitoring role over the performance of SIDA. Within the ambit of MFA guidelines, SIDA enjoys autonomy in its operational decision-making. Some significant changes were made in budget rules to reduce budgetary flexibility in 1998 and afterwards— through the elimination of reserved funds and the set-up of rigid budget ceilings. Apart from highlighting the legal institutions in place at the constitutional and higher policy-making levels, other cultural and biophysical factors—i.e. solid public support, preference for former socialist countries for development partnership, strong private-sector advocacy, the time/geographically-sensitive nature of aid resources given to recipients—are peripherally incorporated into the analysis. At the subsequent stage of discerning the patterns of interactions, the research acknowledges that SIDA is full of highly committed, open-to-learning professionals. However, the institutional scholars interpret this as a necessary but not a sufficient condition for learning-inductive incentives. Considering the principal-agent relationships that span the agency, SIDA is a relatively flat organization where information loss can be minimized throughout the delegation chain, unlike in other types of public agencies. According to the analysis, nonetheless, negative incentives existed against staff’s continuous learning about the needs of beneficiaries due to the following conditions of the aid agency: (i) its personnel policies that gradually reduced permanent positions—as a consequence, the ratio of short-term positions increased from 11.7 in 1995 to 15.1 in 1999—and encouraged frequent internal turnover (preferring to foster generalists); (ii) internal-formal evaluations whose findings were viewed irrelevant and ineffective by a majority of the staff—partially because the evaluative studies entail few meaningful indicators that properly measure and demonstrate the success of aid projects, and, oftentimes, those reports came too late to be reflected in aid planning in the following fiscal year; and (iii) due to the administrative pressure12 to disburse aid quickly and reliably within the budgetary year, there were stronger incentives towards particular aid projects that were either large-scale undertakings, or aid was continuously given to the same recipients, therefore, discouraging innovative and non-conventional types of development intervention. Another important structural factor that worked against aid sustainability is pointed out, albeit one hardly in the direct control of the aid agency, with regard to the

12 The ‘moving-money’ pressure (Easterly 2002) might prevail across all types and fields of public agencies, not limited to development cooperation. According to SIDA’s internal reports, the organization also suffered from it in a more severe manner after the Swedish budgetary cuts and legal reforms in the late 1990s. Since then, almost 40 percent of aid disbursements took place during the fourth quarter of the fiscal year. 49 interaction between MFA and SIDA. The study reveals that the two organizations tended to mutually promote close and fruitful collaboration. Nevertheless, the foreign ministry, in the pursuit of its diplomatic-geopolitical interests and the commercial interest for Swedish trade, industry, and services, had a strong preference for expansive aid planning and larger investment undertakings, whereas SIDA favored developing aid portfolios that focused on a fewer partner countries and sectors in its operational interest. The authors argue that the conflict of interest between MFA and SIDA in the Sweden’s centralized aid policy-making system generated a risk for harming sustainable outcomes at field settings. As a deep-rooted and hardly modifiable factor, the non-altruistic set of donors’ embedded interests in foreign aid disbursement is constitutionally determined (Gibson et al. 2005: p.131). Limiting the Utility of the SIDA Case Study: Difficult Language and Complex Analytic Procedures This pioneering work proves that the IAD approach can be applied with great flexibility and compatibility as remarked by IAD scholars and others (Kiser and Ostrom 1982; Ostrom 2005; Andersson 2009, p. 343; McGinnis 2010; Weible 2014). Although increasingly recognized as a flexible and inclusive approach in the policy process literature, the IAD application to foreign aid policy settings also reveals some critical weaknesses in the operation and usefulness of the approach. Admittedly, previous IAD analyses involved the language and procedures that made them far less comprehensible. As noted at the beginning of the SIDA report, the institutional approach to development cooperation is itself very new to aid practitioners. The theoretical concepts presented are largely too considered intricate by the readers at SIDA and other development organizations. Apart from the disciplinary exclusiveness, the comprehensive scope of the IAD analysis—which not only covers macro-level incentive problems but also micro-level ones within an aid agency—prevents researchers from drawing specific, actionable lessons from the research. Speaking of the empirical case study on SIDA, multiple horizontal/vertical layers of incentive analysis (See the figure below) fail to generate a clear, integrated sense of implications. In addition, given that the research adopts a decision-making model of public administration that is conceptualized to take place at three inter-linked levels (constitutional, policy, and operational levels), the effects of the three institutional-contextual factors are discussed on each decision- making level. The multi-linked, intentionally systematic incentive analysis, however, is somewhat farfetched and repetitive in capturing the resulting patterns of collective actions.

50 Figure 2.4 Complexity with IAD Research: Multiple Levels of Analysis and Outcomes (Cox and Schlager 2014, p.285)

As far as the above-mentioned aspects are concerned, the applied framework, which aims to ascertain the effects of the Swedish institutional contexts on policy choices in fields, does have some obvious constraints in performing as an effective, intuitive tool to fulfil the attached objective. Considering its unique orientation of assessing the design of institutional arrangements, future research in this area would benefit greatly from recalibrating the framework with a much-narrowed research interest in the impacts of a selected set of contextual factors, rather than reviewing all relevant contextual variables in a descriptive manner, and with a simplified decision-making model of public policy. Other inherent constraint stems from the fact that this type of public agency study comes with insufficient quantitative data sets, thus heavily depending on the individual opinions of the staff, gained through interviews and surveys. There are very weak incentives in place in that few public agencies are willing to keep and open the records of their managerial and day-to-day activities (e.g. the frequency of internal meetings, meeting memos, retention/recruitment guidelines, overseas travel expenses and frequency, etc.). Together with the critiques of the existing IAD-based foreign aid research, the following sections add more specific discussions on the potential benefits 51 and risks of following the operational steps of the earlier IAD application for other development cooperation settings where donor governments’ decision-making structures and contexts are widely varied. 2.4.2 Established vs. Emerging Donors: from the Perspective of Rule Context Similarities and differences that lie between traditional and young/emerging donor governments are an endless story. As foreign aid has been much diversified in the post-Cold War period in terms of who is giving it, there is a growing body of literature on non-traditional country donors’ policies and practices. Many pieces of conducted research dwell largely on what might explain the observed differences in aid practices between traditional and emerging donor groups (or China and other new donors) (representatively, Jerve 2007; Lancaster 2007; Woods 2008; Sörensen 2010; Brautigam 2011; Dreher et al. 2011; Mawdsley 2012; Potter et al. 2012; Sato et al. 2013; Watson 2014). In the literature, the term ‘emerging/new donors’ per se has been interpreted in a variety of ways and ensuing concepts are still in the making (Yi 2015). Despite the ambiguity of defining emerging/new donors, one of the shared features within this group of donors is that they all have a national experience of having received foreign aid, which, in turn, constitutes a sui generis factor to shape the underlying motives and preferred approaches for their aid provision overseas. From the neo-institutionalist perspective, some institutional criteria and categorizations might be useful for a meaningful comparison. In light of the most tangible types of institutions that structure a country’s development cooperation, the legal- regulatory conditions for public policy-making across government units are worthy of attention. The governments in Western, advanced economies/democracies tend to have a relatively well-developed set of written rules and disciplines that govern their policy- making activities. On the contrary, some other governments, particularly in transitional economies/democracies, show a tendency for state executives and bureaucrats to enjoy greater leeway in their specific policy choices (Welch and Wong 1998; Pistor, Wellons, and Sachs 1999; Ginsburg 2000; Jauasuriya 2006; Ohnesorge 2016). The relative prevalence of administrative discretion in those countries with a weak tradition of a regulatory state is highly relevant to our better understanding of emerging/new donors’ aid policy-making contexts. Administrative discretion, a type of autonomy given to state bureaucrats, allows soft rules (e.g. socially, culturally, or traditionally accepted norms, self-crafted/self-regulating disciplines, rules of thumb based on past experience) to work as guidance to their policy-making, filling the vacuum that written rules create. 52 Some empirical evidence supports such a view in the foreign aid policy field. A series of OECD reports (2013; 2015, p.70) examines the extent to which DAC donor governments have developed policy guidelines/strategies in written form for multilateral cooperation. Those reports reveal that most of new DAC member countries (Korea, Poland, the Slovak Republic, Slovenia) fall into a bottom group whose ODA strategies are the least prepared and the most loosely defined among the 29 member countries13, when assessing the contents and scope of government policy documents and official statements. 2.4.3 What Happens if We Follow the Analytical Steps of the SIDA Study for other Country Cases This section dwells on the specific risks of the earlier IAD research on SIDA’s incentive problems if its operational procedures were applied to other country/aid agency cases of foreign aid policy formation and implementation. Centrally argued is that the previous IAD-based empirical analyses in the domain of development cooperation focus limitedly on the effects of formal rules. Furthermore, the empirical evidence on the linkages between contexts, incentives, and outcomes is given in a very descriptive manner. Either synthesis of findings from observations and analysis at each level, or re-taking of incentive/collective action problems in the national aid system (beyond a single aid- managing organization) is absent in the study. For those shortcomings, the earlier operation of the IAD analysis is likely to be ineffective in acknowledging the broader institutional contexts of rules and traditions across government units. More importantly, following the current analytical steps seems problematic when it comes to exploring the unique contexts of emerging/young donor governments whose policy-making tends to be governed by administrative discretion under a relatively insufficient and abstract set of written rules. To more rigorously clarify the associated linkages between rule institutions and outcomes, the SIDA research might require focused measures to identify what types of rules induced what types of incentives, collective actions, and outcomes. Secondly, other contextual elements of the physical world and community, which remained peripheral in the SIDA study, need to come to the fore, and their historical and contemporary interplay with rules-in-use should be incorporated into the analysis (Ostrom 2015).

13 Except for Greece, Korea, Poland, Slovak Republic, and Slovenia, the other 24 DAC member countries have formulated either sector-/ organization-specific strategies or stand-alone multilateral aid strategies besides setting up overarching development strategies. 53 2.5 Conclusion At the core of the critique, it is argued that one needs to recognize some risks of making biased inferences with the previous operation of the IAD framework: this is mainly because the existing IAD study largely focuses on the effects of formal rule arrangements on the policy process, dismissing the importance of informal rules. Empirically speaking, for some young or non-Western donor countries whose aid policy- making is governed far less by written rules and policy statements than in traditional/Western donor governments. Given the varied traditions, the existing use of the IAD framework appears to be inappropriate for understanding the aid management design in those countries. The next chapter introduces some measures to add more sophisticated analytical procedures to discern the impacts of various types and configurations of policy-making rules on the outcomes.

54 CHAPTER THREE CONCEPTUAL FRAMEWORK: REFINING THE IAD FRAMEWORK TO TAKE RULE INSTITUTIONS SERIOUSLY

This chapter lays out theoretical considerations and operationalization of the fine- tuned institutional framework. The refined framework arises to address the limitations of the earlier application of the IAD approach in the field of development cooperation. As mentioned in the previous chapter, the existing IAD research on foreign aid (Snidal 1985; Ostrom et al. 2001; Gibson et al. 2005; Andersson 2009) displays critical weaknesses in differentiating the effects of rule institutions by their types and configurations on resulting collective action problems and outcomes. Besides, there have been few empirical attempts to understand the institutional causes of the collective action problem historically and interactively; in other words, few studies highlight the formation and earlier developments of rules-in-use in interaction with the other two contextual-factor categories (i.e. physical/material conditions and attributes of the aid policy community), which have been a marginalized area of study in the previous IAD application. To make IAD research a more ‘cause-and-effect’-sensitive and in-depth analysis, this chapter re-defines the theoretical elements of the IAD framework and introduces analytical measures to enhance the utility of the framework for broader studies in diverse settings of development cooperation.

3.1 Theoretical Concepts to be Incorporated: For Explaining and Predicting the Behavior of Organizational Actors on Foreign Aid Policy Process To begin with, this section introduces five theoretical concepts that are key to the application and interpretation of the proposed institutional analysis: rules-in-use, rationality, delegation, collective action, and the commitment problem. In an attempt to delve into what set of working rules are at play, motivating or demotivating particular collective choice making, and so causing a planning-practice discrepancy, those concepts are set to work interactively in reconstructing the logical sequences of decisions and actions taken by the respective and collective forms of state bureaucracy in the policy- making process. 3.1.1 Rules-In-Use Definition Among the numerous institutional factors to be involved in public policy settings, rules are the prioritized independent factor in this study. Rules are shared understandings 55 among those involved that refer to enforced prescriptions about what action or physical and material conditions are required, prohibited, and permitted (Crawford and Ostrom 2005). In the public policy domain, rules, in their most tangible forms, are enshrined in constitutions, laws, administrative regulations/statutes, or other types of rule statement. Normally, hard rules are nested at multiple layers, therefore, working interdependently. The formal/written rules legally and procedurally govern the top-level and day-to-day operational decision-making of state bureaucrats. Speaking of rules-in-use (or working rules) in the public policy process, they are not limited to hard rules, but include other types of rules (called soft rules). Working rules, in fact, have been one of the challenging subjects for institutional researchers, since the identification of such fluid-type rules requires deeper institutional analysis. Soft rules do not necessarily exist in written form and are often crafted by participants (not by lawmakers) as a conceptual or social consensus. They also exist as the unique culture or tradition of an organization or a confined group (community). Drawing on the nature of soft rules, this research makes those far less obvious rules explicit and explores their impacts on policy-making. To gain proper evidence, this study depends on a comprehensive review of numerous documents that state organizations have produced and person-to-person interviews with state officials and agency staff. Rule Configuration So far, previous IAD projects have identified an array of rules that affect action situations in various social and political frameworks. To capture the universality along with those rules, the IAD scholars have developed a rule classification. From accumulated empirical evidence, Ostrom and her colleagues come up with a seven-rule-type system (Ostrom, Gardner, and Walker 1994; Ostrom 2005b, 2010, 2011; Weible et al. 2010; Siddiki et al. 2015). Drawing upon more than 1000 research projects that cover diverse settings and scales of collective action problems, the seven types of rules are identified as universal elements (Ostrom 2011). The rule categories are specified as follows: (i) boundary rules, (ii) position rules, (iii) authority (choice) rules, (iv) scope rules, (v) aggregation rules, (vi) information rules, and (vii) payoff rules. The rule types are determined by their respective functions: boundary rules define who is legitimately involved (identifying participants); position rules prescribe what positions the involved/participating actors are given within a certain organization/delegation structure; choice rules delimit actors’ activity by specifying mandated tasks or by listing what the allowable/non-allowable actions are in various 56 circumstances; scope rules delimit the range of possible outcomes by defining what degree of control each participant is allowed over the potential outcomes of performed tasks; aggregation rules delineate what specific occasions (of collective or operational- choice decision-making) require mutual agreement and what requirements should be met to constitute mutual support; information rules control the flows of information and assign particular actors to collect and disseminate the information; and payoff rules assign rewards or penalties to particular actions. The rule categorization can be applied to foreign aid policy-making. A comparative study14 on foreign aid regulations (2012) is found particularly useful to confirm what laws and regulations have been established and practiced in the development cooperation field. With various formal rules in traditional and new donor countries identified, the table below categorizes those rules by their function. Nonetheless, respective donor countries do not have all of the listed functions of formal rules, and therefore display varied rule configurations.

Table 3.1 Rule Typology: Discerning Rule Types in the Context of Foreign Aid Policy Formation and Implementation (by author)

Category Exemplified Functions of Respective Rule Types × Limit or control the number of participating actors/organizations × Define required attributes (qualifications) of state agencies to manage Boundary foreign aid Rule × Define conditions for actors to leave the ODA system × Set minimum or maximum amount of ODA to manage × Clarify the chains of command in the ODA system (i.e. levels of delegation) × Assign positions and define the legal entity of organizations (e.g. quasi- Position Rule state agency) × Set the number of managerial and subordinate organizations × Assign mandates and list designated (or non-allowed) tasks Authority × Prevent the multiplicity and incompatibility of tasks that are delegated (Choice) Rule × Consider organizational talent/capacity/resources with regard to assigned tasks × Delineate the logic and design for division of labor or division of resources among state agencies Scope Rule × Delimit the authority of organizations over potential outcomes × Define required chains of events over potential outcomes

14 A research team at the U.S. Law Library of Congress (2012) published a comparative analysis of foreign aid regulations in 18 donor countries. This study is useful for understanding the legal and policy frameworks of foreign aid management and delivery in the following countries: Australia, Brazil, China, Finland, France, Germany, India, Israel, Japan, Kuwait, New Zealand, Norway, the Russian Federation, Saudi Arabia, South Africa, South Korea, Sweden, and the United Kingdom. 57

× Clarify the policy agendas that require unanimous or majority agreements × Set up a coordinating body and define its authority, tasks, and given Aggregation resources Rule × Define the coordination procedures (e.g. makeup and frequency of coordinating meetings) × Define mandatory or discretionary information sharing (reporting & monitoring, gatekeeping) Information × Establish an integrated information platform and assign a particular agency Rule to control the information × Identify the types and areas of information to be collected, screened, and disseminated × Establish internal systems for monitoring and evaluation (oversight Payoff Rule mechanisms) × Assign rewards or sanctions for particular actions

Those legal regulations and policy guidelines contain great deals of information about one country’s constitutional/legal/administrative structures and appropriation processes pertaining to foreign aid provision. Considering the legal and policy frameworks, different rule types appear closely interlinked. For instance, if a country has a regulative, meticulous set of boundary rules that strictly control the entry and exit of actors, this feature of the boundary rules may affect position rules and are highly likely to establish a flat, concentrated structure of ODA management within a government. Likewise, just looking at countries’ public-sector aid provision (excluding private aid contributions), donor countries display varied rule configurations where, for example, particular rule types are extensively or uniquely developed, or, for some other countries, most of those rule types are distinctively underdeveloped. France is a case with boundary rules that invite/retain an expansive group of ministerial and implementing agencies into the system, whereas Sweden legally promotes an ODA system with fewer actors and fewer aid goals through facilitating a structure of single ministerial supervision and implementation. In this study, it is hypothesized that the way in which rules are configured shapes not only the incentives of bureaucracy but ultimately determines the quality and features of government performance in formulating and delivering public services. In that sense, the seven rule types are useful in helping us figure out what particular rule sets are accountable for widening the problematic gap between aid planning and practice, and into delving into what rule adjustments can be proposed to curtail such effects. Therefore, to improve the performance of a government in terms of aligning its ODA planning and 58 practice, careful attention should be paid to rule configurations (Ostrom and Ostrom 1971, 211). Although the suggested rule category system is not an absolute one, it allows us to gain a systematic understanding of the rule configurations in the concerned action situations, respectively and comparatively. Considering the particular contexts of new donors’ policymaking (i.e. lack of policy experience and lax legal-regulatory frameworks in comparison with Western, traditional counterparts), the behavior and incentives of policy-makers are reconsidered through the refined IAD framework. Along with the formal rules, it is also critical to see what other informal rules (in multiple rule types) might come into play and shape the incentives of aid policy-makers and practitioners. 3.1.2 Rationality Definition Rationality, a mode of reasoning, has served as the dominant measure in human- choice and organization studies, including studies of public administration and public policy. Rationality for IAD research gives rise to the following assumptions (Ostrom and Ostrom 1971): (i) individuals are assumed to be self-interested (not merely selfish, but having their own preferences which affect their decision-making); (ii) individuals are assumed to be rational, which means the ability to rank all possible strategies and alternatives for their choices; (iii) individuals are assumed to adopt maximizing strategies, where maximization means a strategy that results in the highest net benefits; and (iv) individuals are assumed to react on the condition of risk and uncertainty (depending on the level of information, individuals react so as to overcome risk and uncertainty). With regard to those assumptions, some scholars raise the point that the rational-choice idea interprets human motivations and interactions in an oversimplified way so that it cannot be a unitary measure of human and organizational behaviors (Simon 1947, 241; Boden 1994, 188; Jones 2002; Townley 2008). Bounded Rationality Centrally argued is that formal rationality (or comprehensive rationality)—that assumes all decision-makers are equally rational and competent in choosing strategies and taking action for best outcomes—might fail to generate satisfactory explanations and predictions in complex political and social situations. Therefore, this thesis attempts to address the shortcomings of the conventional rationality assumptions and to embrace a flexible type of rationality called bounded rationality. Bounded rationality provides a sensible answer for the sub-optimal choices of policy-makers in that policy actors are boundedly rational since, they have limited 59 information and capability in performing their policy activities. Unlike the general assumptions of perfect rational-choice theory, those constraints are critical to recognize in our understanding of policy-makers' behavior. The proponents of bounded rationality contend that actors (persons or collectives of persons) are intendedly rational but only limitedly so, since they are bound to make decisions based on incomplete knowledge and limited capacity. Also, they often pursue sub-optimal strategies, unlike what axiomatic choice-theorists assume (Simon 1965, 1972; Williamson 1985; Jones 2001; Ostrom, Cox, and Schlager 2014, 273). This notion leaves some room to flexibly expound that the actors can make mistakes and, overtime, they can learn by trial and error. Such learning is a critical element to understand the decision- making in terms of how actions intend to pursue their self-interest in repeated decision- making/action settings over a certain period of time. This approach is expected to be further able to explain gradual changes in policy choices and patterns of interaction. Rationality of an Organization Jones’ bounded rationality (2002), as a critique of rational-choice theory, arose in the discipline of political behaviorism (Simon 1976). Jones argues that the majority of public organization studies care little about the fine details of the specifics of human cognition without linking between the human-level process of decision-making and that of organizational choice. In this regard, Jones suggests that we need to pay attention to organizational culture and tradition (as the complementary measure of rationality) that has been dismissed as the residual category of rationality in the existing literature. Hence, we have those diverse contexts of rationality, beyond the rational-choice context, to be highlighted in this thesis; that is, the relational and historical (i.e. the traditions of organizations) elements of rationality. Those elements of rationality do not depend on the calculability of action or statement only. 3.1.3 Delegation Delegation is a modern logic of practice that public and private organizations adopt to efficiently and effectively achieve the objectives imposed from the top. Delegation, a chain of principal-agent (PA) relations, also constitutes an important condition for participants’ decision-making and rationality in the public policy process. Likewise, for foreign aid policy-making, a series of PA relations among sub-state organizations within a donor government is particularly resonant for understanding the incentives of aid bureaucracy and of a whole ODA system (Gibson et al. 2005, p.70). Delegation theorists explain the decision-making behavior of state bureaucrats through a frame of being either a principal or an agent of state agencies. Across the 60 bilateral aid community, donor governments commonly display a hierarchical structure of intra-government ODA management (See Figure 3.1). The policy choice/reaction of each sub-state organization can be explained and predicted by their relative relationships and conflicts of interest, according to delegation theory.

Figure 3.1 A Hypothetical Delegation Structure within a Donor Government (by author)

Principal Behavior To exercise effective control over agents, principals have an incentive to prevent agents from increasing their autonomy, capacity, or specialization. Suppose, in a working relationship between a foreign ministry and aid agencies, what is generally practiced is that the ministry assigns the head positions at its subordinate organizations. Also, the behavior of principals can be differently hypothesized if the principal is more than one, controlling a single agent, so that the joint managerial-ship imposes multiple, incompatible objectives over the performance of the agent (Seabright 2002). With regard to cases where principals fail to write detailed, clear rules for the delegation, so that it allows agents to carry greater potential for agency slack, the agency theorists argue that it might result from principals’ incompetence (a lack of knowledge or inexperience) or lack of determination (Nielson and Tierney 2003) 15 . From the perspective of principals, particularly in the public sector, practicing strictly controlled delegation can be very expensive and time-consuming. This is partly because public organizations usually work on multiple agendas rather than focusing on a few objectives.

15 Nielson and Tierney (2003) demonstrate this point through the case of the World Bank, which previously enjoyed broad discretion until reforms in the mid-1980s. Since then, the World Bank’s executive board has further elaborated rules and institutions to effectively restrict the organization’s policies and operations that might work against the interests of the board. 61 Therefore, often, public principals with limited information and weak political incentives allow their subordinate agents to have greater policy-making roles. The three key sub-state actors within a government (president, ministry, and aid agency) are assumed to have a distinct set of preferences, traditions, and capacities. The control mechanisms are in kept place through employment/secondment policies, monitoring, reporting requirements, and evaluation. Agent Behavior The current discussion of agency theory is centered largely on the characteristics of principals (Martens 2002). This is because the behavior of principals has been believed to primarily determine the quality and outcomes of delegation. Therefore, many of the relevant studies have aimed to bring about some improvements by adjusting principals’ control systems. However, agents can be more important when they are given a high level of autonomy in particular contexts. In the regulatory delegation settings, agents are more likely to perform routine tasks, given their tightly restricted autonomy. On the other hand, in some policy-making environments with great leeway given, non-passive actions are likely to be observed among agents. Such non-passive actions can be interpreted as the product of agency slacks16 but do not necessarily lead to undesirable outcomes; rather, the resulting impacts on the outcomes can be varied and dynamic. With the attention to the role of agents rising, agents’ strategic behavior/reaction under the control of principals is newly considered a critical factor to the outcomes. Those strategic reaction of possibility can be manifested differently depending on agents’ organizational preferences, capacity, and level of specialization. In the absence of well-defined delegation rules, theorists predict that agents would behave as follows: they are likely to secure/increase their autonomy (i) by reinterpreting the imposed mandates in favor of their organizational interests, or (ii) by inviting a third party (non-principals) into their policy-making processes so as to increase or justify their political indispensability/influence, or (iii) by buffering principals’ monitoring and evaluation on their performance (Hawkins and Jacoby 2006: 202). As one of the possible effects, rule-based delegation in the public sector can foster a policy-making climate in which innovative practices are rarely induced. Greater discretion given to policy-makers, on the contrary, can induce greater opportunities for agents to explore unconventional policy options (Hawkins et al. 2006: 28).

16 It is defined as the cost of having agents who have the possibility of performing against principals’ interest. 62 Delegation Chain Characteristics: Tightly Constrained Delegation vs. Open-ended Delegation The length and scope of a delegation chain matter as well. Hypothetically, the more complex and the longer chain a delegation structure has, the more the system is likely to induce agency slack. The presence of well-shared and well-defined objectives throughout multiple layers of delegation can show how effectively the delegation chain works. In addition, it is important to consider whether proper coordination/task allocation mechanisms are in place in the delegation system. Free entry or exit of agency in and out of the delegation system is counted as a factor to increase agency slack. 3.1.4 Collective Action This research is keen to observe not only the individual behavior of organizational actors, but, more importantly, the patterns of action collectively taken by multiple policy actors. Public Policy Process: Two-Level Dynamics As a key premise, the following institutional analysis conceptualizes the policy process as two-level action situations. Before an action arena is defined and analyzed as a social space where multiple actors interact, exchange information, and compete one another, this analysis posits that there is a prior situation in which individual organizational actors have their own processes of decision-making without interacting with other actors. At this individual action level, respective organizations decide on their intra-agency agendas (on the occasion of making completely independent or autonomous decisions). Those policy activities include redefining organizational mandates, strategy- building, allocation of human and material resources, knowledge management, capacity- building, etc. In performing those activities, organizational actors are examined as to whether they display routinized/passive behavior or opportunistic behavior. On the other level, respective organizational actors are assumed to interact and compete with other agencies in their inter-agency policy activities. At this collective action level, inter-organizational interaction is the key area for observation and inference. Importantly, at this stage of policy-making, it is assumed that policy coordination takes place and policy products are finalized so that the patterns of collective decision-making can be observed. Collective action can be divided into two main types: reciprocity vs. non-cooperation. In some action arenas, multiple actors tend to cooperate and build mutual trust. In contrast, in other action arenas, it is hard to observe inter-agency cooperation; rather, there is intense competition among actors arises. Patterns of Interaction and Policy Outputs 63 Some collective actions are considered problematic and become the subject of the social science studies due to the negative or unintended impact they cause on policy outputs. For instance, non-cooperative actions among multiple public agencies (e.g. hesitance to share information or to collaborate on operations) in a collective-choice arena of foreign aid policy-making is hypothesized to induce not only a higher level of fragmentation but also a planning-practice gap in aid disbursement. 3.1.5 Commitment Problem (as Evaluative Criterion) As a chosen evaluative angle, the performance of donor governments in holding up their commitments of aid scale or aid reform is the dependent factor in this thesis. Previous IAD research projects focus on the incentive problems of aid policy-making that undermine aid effectiveness or aid accountability as the major evaluative criteria. In contrast to that, this thesis examines donor governments’ aid practice of failing to keep their annual or multi-year commitments and explores whether this is either the intentionally chosen strategic behavior of a donor government, or an unintended but structure-driven (by organizational features or insufficient capacity/resources) results, or a product of bureaucratic politics. The commitment problem is interchangeably used with the aid planning-practice gap. In this thesis, the commitment problem specifically refers to the behavior of a country donor that fails to honor its pre-set goals of aid disbursement. It can be measured in either a quantitative or qualitative manner by considering the following fields of aid contribution and practice: ODA annual-total scale, geographical/thematic allocation, and aid delivery-related policy agendas for reform and change (e.g. reducing a ratio of tied aid). Previous Interpretations of the Commitment Problem The issue of donors’ commitment has not been extensively explored yet. Nonetheless, this issue is increasing being taken to the table in various academic and policy discussions. In policy process studies, a low level of ‘planning-practice’ correspondence in policy outputs is discussed as the result of problematic institutional arrangements. Previous foreign aid studies highlight that the largest aid donor countries, such as the US, Germany, France, and Japan, tend to show a behavior of only half-keeping aid promises, in comparison with Nordic countries. As Hook argues (1995), such aid practice can arise when donor’s national interests and humanitarian/altruistic interests are in conflict in the foreign policy-making process. From an implementational perspective, Swedlund (2014, 2017) suggests that the major reasons that donors’ commitments are not 64 upheld might lie in the aid negotiation and delivery processes between donor and recipient governments. For scholars of foreign policy studies in the context of East Asia, such policy behavior was once explained by the power struggle between the Ministry of International Trade and Industry (MITI) and the Ministry of Foreign Affairs (MFA), each of which champions different motives regarding the use of foreign aid in Japan (Hook and Zhang 1998). There have been systematic, multi-year efforts to measure donor governments’ aid planning and disbursement in recent years. One of the relevant activities in the policy circle is the Commitment to Development Index, developed by the Center for Global Development (CGD). Through the Index, 27 donor countries have been evaluated every year since 2003. In seven policy areas (aid, trade, finance, migration, environment, security, and technology), the performance of countries in the aid policy domain is evaluated by a combined set of quantitative and qualitative measures (Roodman 2012). While the quantitative factor is simply counted by an annual ratio of net ODA transfer to GNI, the chosen qualitative measures to evaluate donors’ practice cover four sub-areas17 CGD 2017). According to the Index in 2017, Denmark (which scored at 6.46) tops the list, while South Korea (3.96) ranks at the bottom. Although the Index has been continuously revised, it remains largely dubious whether the selected sub-components of aid quality and quantity might constitute, together, a legitimate overall evaluation of donor government’s contribution to making development-friendly policy. Similar to the intention of the Commitment Index, the OECD aid community has been discussing the idea of Policy Coherence for Development (PCD). Previously, the concept appeared in the two OECD reports that aim to evaluate the aid management systems across the DAC community (1999, 2009). The notion of PCD has been newly introduced and integrated into an overall development goal for many donor governments. Beyond the domain of foreign aid policy, the concept is discussed to measure the coherence between aid and non-aid policies (King 2012).

3.2 Introducing a Fine-tuned Framework Together with the aforementioned concepts that help to explain the rationales and policy-making behavior of state agencies, the following section introduces a refined IAD framework. 3.2.1 Intention

17 The subcategories are maximizing efficiency, fostering institutions, reducing burdens, and transparency and learning. 65 In response to the critiques of the existing IAD application (See the previous chapter; Section 2.4), this study proposes analytical procedures in applying the IAD framework. Apart from that, the sequential stages and subjects of the IAD framework— from contexts, action arena, incentives, interactions, outcomes, and to evaluation— remain intact in the framework. 3.2.2 Comparing the Existing and Refined Uses of the Framework The major difference between the existing and revised operations of the framework rests on to what extent the researcher attempts to capture the impacts of various institutional conditions under the framework. With the existing operationalization of the IAD framework, the three institutional factors are considered in an isolated manner and, in particular, the two contextual conditions (i.e. the nature of resources in use and the attributes of participants)—that are assumed as secondary factors while rule institutions are the direct one in IAD research (Ostrom, Cox, and Schlager 2014, 280)— are rarely incorporated into the analysis in the earlier study of foreign aid policy. Also, regarding the linkage between CONTEXTS and ACTION SITUATIONS, this area of analysis—on how those three institutional factors directly affect action situations—is limited to the effects of formal-legal rules only, and remains largely descriptive in the previous studies (See Figure 3.2). Thus, first, the supplementary importance of other two contextual factors is brought back into the analysis and the three contextual conditions are attempted to be examined in an integrative manner under the suggested framework. While the importance of rules-in-use is primarily emphasized, the following analysis includes historical and contemporary pictures of the creation and interplay between those three contextual factors (The corresponding analysis is introduced in Chapter Five). If applied to the national system of foreign aid policy-making, it is non-negligible conditions to be considered that how the government budget for ODA is financed and secured or what qualifications and cultural or educational backgrounds of policy-makers are concerned are significant to policy-makers’ chosen behavior.

66 Figure 3.2 Previous and Refined Applications of the IAD Framework (by author)

Second, to grasp the ‘institution-action’ linkage in a more systematic, explicit manner, requires some analytical steps to be taken further, as centrally argued in this research. As illustrated in the figure above, for instance, when the researcher examines the CONTEXTS under the refined framework, he or she (i) identifies an array of working rules (i.e. formal-legal regulations (hard rules) and social norms (soft rules)) and (ii) figures out what specific areas18 of policy-making action are affected by the hard and soft rules. What is further attempted here is to (iii) understand the creation and developments of those rules by taking the nature of aid resources and the attributes of an aid policy- making community into account19 (The corresponding analytical parts are included in Chapter Six and Chapter Seven). Another consideration behind the refinement is about incorporating a simpler and more heuristic decision-making model. Instead of adopting the existing three-level decision-making model (i.e. operational, collective-choice, and constitutional stages of decision-making), this framework understands that decision-making among state policy- makers and practitioners occurs primarily at two levels: intra-organizationally and inter- organizationally. Besides the additional processes of discerning the varied impacts of rule institutions, the rest of the differences between the two operational models are explained

18 As highlighted in the previous chapter, the IAD’s rule typology identifies the seven distinctive areas of action. 19 This study adds historical reflections in Chapter Five accordingly. This is to provide an in-depth understanding of the rules that have evolved and are continuously shaping the behavior of policy- makers. 67 by the varied interests of the researchers. At the stage of analyzing ACTION SITUATIONS and OUTCOMES, the previous model is intended to see the micro-level impacts of institutions on the day-to-day, operational decisions of the development staff. On the other hand, the revised model serves its interest in the government-wide, collective impacts of institutions on the patterns of policy choice and performance in delivering planned ODA, not limiting to those to a single organization or individuals (The corresponding analysis is present in Chapter Eight).

3.3 Research Action Plan According to the refined IAD framework, a research action plan can be drawn to systematically and sequentially understand the logic, design, and performance of legal/regulatory frameworks for aid policy-making as the core cause for the commitment problem (See Figure 3.3). By performing the tasks of [A] to [F], we ultimately discern the impacts of different rule types on collective action outcomes in various action situations of foreign aid policy-making. Drawing on the mutually supplementary relationship between formal-legal institutions and social-cultural norms in guiding the policy formation and the implementation of state bureaucracy, the agendas of [A] and [B] are centrally discussed in the analysis of CONTEXTS—in the following empirical part, Chapter Five includes the analysis. In the next round of the ACTION SITUATION analysis, each state agency is observed in terms of their decision-making and interaction with other policy actors through [C] and [D]. As assumed, the two areas of decision-making are closely linked since intra-agency decisions can constrain the inter-agency affairs and vice versa. Chapter Six and Chapter Seven present the findings of [C] and [D] with the two selected sub-policy agendas of foreign aid. The conclusion chapter (Chapter Eight) covers [E] and [F] to provide the holistic findings on how and which specific legal-regulatory policy rules induced some problematic collective actions that caused the commitment problem. After all, those analytical procedures and research plans suggested are intended to draw clearer and more feasible policy implications out of the IAD study.

68 Figure 3.3 Research Action Plan under the Refined IAD Framework: From [A] to [F] (by author)

[A] Rule Configuration: Hard Rules To examine the effects of rules that travel throughout the analytical stages of the framework, the suggested analysis starts with a review of existing constitutional/legal frameworks that are enshrined in various forms of formal institutions; such as constitutions, statutes, laws, and administrative regulations that are imposed on various levels of public governance. At the initial stage of analysis, rules-in-use in written and hard form are reviewed, and the direct impacts of those rules are examined. What is empirically observed, for instance, is that some donor countries have developed a detailed and restrictive set of boundary rules (e.g. legislation to cap the number of ODA-managing organizations with strong boundary rules). Other donors, in contrast, have established a very lax form of boundary rules in their ODA laws and administrative regulations. Such a tendency appears highly associated with young donors. [B] Rule Configuration: Soft Rules and Administrative Discretion Once hard rules delineate the boundaries of actionable options for participants, this thesis assumes that when written rules are absent in some areas for detailed policy choices there is still remaining effects of soft rules. Those soft rules (i.e. regulatory frameworks) are created and redefined through executive decisions, official government statements, and internationally-acknowledged pacts or agreements (e.g. white papers, the U.N. Charter, etc.). In many cases of public policy formulation and implementation, administrative discretion is necessarily given to policy-makers in order to fill the legal vacuum. The key task at this stage of analysis is to figure out which areas of policy choice or to what extent administrative discretion is given, and what justification state bureaucrats attach to their actions (or non-actions) in the policy process. [C] Action Situation (1): Intra-Agency Decision-making for State Agencies At the individual choice/action level—in which respective organizational actors redefine their mandates, build strategies, and take action on intra-organizational areas of policy choice—this level of analysis examines the following routine policy activities of 69 state agencies: reinterpretation of rules, reallocation of human/material resources, knowledge management (i.e. R&D, data collection, interactions with domestic/international interest or expert groups), and reporting to/monitoring other state organizations. [D] Action Situation (2): Inter-Agency Decision-making With individual organizational actors’ preferred strategies and behavioral patterns identified, the next level of analysis (called collective choice/action level) is conducted on the interactions among a group of organizational actors in the aid policy systems. Inter- organizational interactions are examined at either mandatory or autonomous procedures of whole-of-government policy coordination, inter-agency collaboration, and information collection/dissemination/exchange. [E] Patterns of Collective Action & [F] Resulting Outcomes Based on the findings from the two-level analysis, this study is ultimately geared towards capturing the system-level impacts of the collective action dynamics on the long- term aid practice of a donor government. [G] Overall Evaluation: Logic, Design, and Performance of Hard Rules Against Commitment Problem Previous IAD research projects have explored varied evaluative angles of aid sustainability, efficiency, accountability, or transparency (representatively, Ostrom et al. 2000, Martens et al. 2002, Gibson et al. 2005, Hawkins et al. 2006, Milner 2006). This thesis advances the literature by examining the relationship between the design of legal/regulatory frameworks and donor government’s performance to measure up to their pre-set aid plans and intentions.

3.4 Propositions 3.4.1 When Different Types of Rules-in-Use Affect the Decision-making of Aid Bureaucracy The following section discusses specific hypotheses that are applicable to each analytical stage of the framework. The effects of hard/soft rules on the incentives of organizational actors are variously hypothesized depending on which areas of action/decision-making are concerned, mainly focusing on the seven areas that affect the disparate elements of action/decision-making (with reference to Section 3.3 and Figure 3.4), regarding (i) qualifications and entry/exit conditions for state agencies to manage ODA [boundary rule category]; (ii) the structure of delegation that spans a public administration [position rule category]; (iii) the specification of designated and non- 70 allowable tasks [authority rule category]; (iv) the control of potential outcomes [scope rule category]; (v) inter-agency coordination procedures and requirements [aggregation rule category]; (vi) knowledge management [information rule category]; and (vii) rewards or sanctions on specific actions [payoffs rule category].

Figure 3.4 Seven Rule Types that Affect Varied Elements (Participants, Positions, Actions, Information, Control, Net Costs & Benefits, and Potential Outcomes) of An Action Situation (Ostrom 2005, p.189)

It is also important to capture the changed impacts of rules over the incentives as rules, particularly hard rules, are modified or gradually evolve overtime. On the occasions of rule change (e.g. legal-regulatory reforms, public administration restructuring), this study assumes that incentives, interaction, and policy outputs are subject to the change accordingly. Boundary Rules and Hypothesized Outcomes The boundary rule20 category functions to control the number of participants (or actors) and to delineate qualifications or conditions to legitimatize the decision makers. From the delegation theory perspective, the boundary rules are highly relevant to actors’ incentives by developing capability and specialization. For instance, easy entry and exit to the aid policy arena (with no strict requirements for joining the aid policy formation and implementation community) can attract prospective participants with no proper

20 Boundary rules are also called ‘entry and exit’ rules because this category of rules often serves as the barrier for actors to join/leave action situations. 71 organizational capacity and resources. To prevent the proliferation of aid-managing organizations, some established donor countries either cap the number of implementing agencies or strengthen the conditions attached to the management of foreign aid projects. Position Rules and Hypothesized Outcomes Position rules define respective actors’ assigned positions. Also, the rule category delineates the relations between positions, thereby determining an organizational structure (i.e. level of delegation) among the participants. Also, varied natures of position (i.e. legal entity) can be given to actors. The rule category includes functions to determine the number of managerial and subordinate organizations. In some donor countries, such as Korea, two ministries (the foreign affairs ministry and finance ministry) share a managerial-supervisory position. In this case, the position rules of the Korean government structure a dualistic management of foreign aid. Such divided supervision of foreign aid management in donor governments is presumably subject to inter-ministerial rivalry and competition. On the contrary, donor governments with a single aid managership of foreign affair ministry are less likely to suffer, either positively or negatively, from bureaucratic competition but more likely to have a weak checks-and-balances system. Authority Rules and Hypothesized Outcomes Authority (choice) rules specify designated tasks for each participant. The choice rules can negatively affect the performance of actors when an incompatible or complicated set of tasks is assigned. In some cases, authority rules fail to prevent problems that are caused by a mismatch between given tasks and organizational talent. Authority rules with a low level of clarity can motivate participants to reinterpret their mandates according to their own interests. As a possible result, reinterpretation of designed tasks among participants can cause conflicts of interest within the aid policy system. Scope and Aggregation Rules and Hypothesized Outcomes The scope rule category delimits the potential outcomes of actors’ action. To control the outcomes, the scope rules assign a particular division of labor or resources among state agencies to delimit the scope of their authority. Aggregation rules define specific policy decisions that require agreement or coordination among actors. Those rule categories are critical to determine the degree of policy coordination and interaction among actors. If aggregation rules worked effectively as proper (either informal or formal) mechanisms of policy coordination and collaboration among state agencies, the final policy outcomes would be less fragmented and better aligned with aid planning. Information Rules and Hypothesized Outcomes 72 This category of rules affects actors’ behavior of collecting, exchanging, and disseminating information and knowledge in the system. The rules clarify the types and areas of information to be mandatorily managed by a designated actor. Establishment of an integrated information platform within the aid policy system is likely to promote better policy and operational cooperation among participants. On the other hand, a fragmented management of knowledge and ODA information is assumed to foster a hostile environment for inter-agency cooperation. Payoff Rules and Hypothesized Outcomes Payoff rules prescribe the benefits and costs that will be assigned to combinations of actions and outcomes. The rule category determines the incentives and deterrents for bureaucratic action in that it establishes an internal monitoring and evaluation systems over actors’ policy-making and implementation activities. If a strongly enforceable set of payoff rules were in place, there might be less room for agency slack. 3.4.2 Rule Contexts and Hypothesized Bureaucratic Behavior in Knowledge Aid Management Legal/Regulatory Frameworks: Delegation, Task Allocation, and Coordination In the sphere of knowledge aid policy formation and operations, the following state organizations are identified as dominant actors from a legal and budgetary viewpoint: the president, MOFA, MOSF, KDI, EDCF, and KOICA. According to the Framework Act, the Committee for International Development Cooperation (CIDC) under the Prime Minister’s Office (PMO) is assigned to take the role of coordination and evaluation across the government. As emphasised earlier, KS is the exceptional sector where two managerial ministries, MOFA and MOSF, which leads respective policy communities of grants and loans, are equally involved in knowledge aid projects. At the implementing level, in total, more than 30 state and local agencies carry out KS projects (ODAWatch 2012).

Figure 3.5 Delegation Structure for Knowledge Aid Planning and Management (by author)

In performing public services in general, state agencies simultaneously delegate and are delegated a task of policy formation and implementation—which means that none of the state agencies work stand-alone in the public policy process. The working relationships among those actors are hierarchically structured in a four-tier delegation system in the case of knowledge aid management. However, there are no further detailed sense of delegation set by law: For instance, the CIDC does not perform as a strong type of principal for implementing bodies because the coordinating body has no budget allocating power, in general (as of 2016). Neither does MOSF work as the exclusive supervisor of KDI. Rather, KDI, as a governmental research agency for economic policy, has a cooperative and occasional working relationship with MOSF in the promotion of overseas knowledge projects. In the knowledge aid sector, a heavy and systematic control of supervising agencies over implementing agencies is difficult to find. Occasionally, some episodes of conflicted interests between high-level policy-making bodies and implementing agencies have occurred. According to the interviews, some experts point out that this might be relevant to ministries’ indifference (lack of interest) or incompetence in controlling and monitoring subordinate organizations. This observation might confirm the theoretical notion that a lack of technical knowledge and manpower on the side of principal organizations leads to a higher level of agency slack among subordinate aid agencies and increases potential conflicts of interest between the principal and agent organizations. As for the relationships between peer organizations (in other word, the relationships between state organizations that are at the same tier of delegation), these might be chronically vulnerable to the pressure to compete. In the context of an impressive, decade-long budgetary expansion 21 for KS operations, bureaucratic

21 The budgetary growth in KS operations has been hardly interrupted by any regime change since the early 2000s. 74 competition in this aid sector has grown substantially. This has effectively brought about the reorientation of thematic focus for implementing agencies. Information-sharing and operational collaboration between state agencies, albeit neither mandatorily required nor culturally promoted, were close to none in this KS field. Such collective behavior of non- cooperation in informational and operational senses, without an effective coordination effort, speaks to the highly fragmented, incoherent practices of KS over the last decade. Bureaucratic Autonomy, Rationality, and Incentives with Knowledge Aid With an opportunity window that is wide opened in the field of KS, there is a decade-long tendency that aid agencies have increasingly competed with one another for more resources and have generally enjoyed great leeway in managing and promoting knowledge aid. Such bureaucratic behavior of pro-activeness is not universally observable across donor government units, in fact. Interestingly, the opportunistic behavior of state organizations in the Korean ODA policy sphere is self-portrayed variously: during the interviews with incumbent state officials, some say it might represent praiseworthy ‘entrepreneurship’, other describe it as a series of episodes of ‘moral-hazard’ to some extent, reflecting on the undisciplined management and uncontrollable proliferation of knowledge projects across the country. This type of development intervention has been discussed as inductive to a particular set of perverse incentives among aid policy-makers and practitioners in the literature of bureaucratic incentives (Gibson 2005). For instance, SIDA’s Contract Financed Technical Cooperation (KTS)22 represents Swedish commercial motivations (Schedvin 2001). Since KTS was carried out in close collaboration with Swedish private companies—those who are keen to expand their markets overseas—SIDA’s policy- makers have strong interests and preferences in deciding on the duration and scale of the knowledge intervention in favor of the participatory businesses. What is centrally demonstrated in this analysis is that the formal rationality approach23 might not provide us with a satisfactory explanation either on the above- mentioned opportunistic/sub-optimal behavior or the resulting policy outputs of discrepancy between the planning and practice of ODA. Neither does information (or technical knowledge) constitute major explanatory factors to such organizational behavior, in contrast to what is commonly hypothesized in existing delegation studies.

22 This project facilitates knowledge transfer between Swedish consultants and local partners in recipient countries. 23 In this approach, it is hypothesized that state organizations make a rational choice and set up optimal strategies based on ‘information’ and ‘talents’. 75 Likewise, the budget-oriented interpretation24 of bureaucratic politics is considered too simplistic to make proper inferences about bureaucratic rationality25 as well. The orthodox rational-choice explanation is unsuccessful in that it neglects the reality in which decision-makers are put under multiple limitations in terms of resources, time, information, and capability. Those constraints greatly reshape bureaucratic rationality accordingly (Kato 1987, Jones 2002). The prism of bounded rationality is facilitated regarding (i) how bureaucratic rational choices are formed when discretionary room for policy-making remains with overseas knowledge-transferring programs; and (ii) how a variety of interests and rationalities of bureaucratic organizations interact (under political negotiations that might lead to either competition or (non-)cooperation). In a nutshell, in the Korean case, the bureaucratic proliferation of knowledge aid is newly explained by the collective action problems of those who are incentivized to behave opportunistically and are discouraged from cooperating and coordinating with one another. Such opportunism appears to have negative impacts on policy outputs in that the surviving strategies of respective agencies do not match with their capacity or specialization. As a result, the collective behavior of opportunistic bureaucrats might result in uncoordinated, unintended policy outcomes. Expected Patterns of Interactions, Collective Actions, and Policy Outputs A key proposition is that bureaucratic organizations often make political (non- economic or culture/power-sensitive) decisions so as to increase their policy influence in relation to others (Kato 1994; Martens 2002; Chibber 2002)26. In this regard, what is further examined is how politicized decision-making becomes intensified or reduced— when the scope and details of bureaucratic autonomy are arbitrarily interpreted—under the conditions of lax legal-regulatory frameworks. At the collective action level, it is assumed that a sum of individual organizations’ intentionally rational choices does not itself determine the quality of the overall policy product; rather, the level of inter-ministerial/agency coordination determines the quality of aid policy in terms of in-donor fragmentation (whether policy produces duplicated spending); policy coherence (whether policy is set and implemented in a coherent manner

24 Suppose that the size of budget either determines or indicates the relative power of bureaucratic organizations. 25 From the perspective of ‘bounded rationality’, the scope and details of bureaucratic autonomy can be another important piece of political capital and exert influence over other policy actors in the policy- making arena. 26 This proposition generally runs against the long-established view of state bureaucracy whose decision-making is believed to be determined from ‘apolitical-technical’ considerations based on high quality information and superior capacity (Weber 1978). 76 overtime and across the national system); and policy rhetoric (how well actual practices are aligned with the pre-set or officially proclaimed policy objectives and directions). Hence, it is vital to understand the mechanisms of how inter-agency coordination or cooperation became discouraged or encouraged as the direct effects of the legal- regulatory frameworks that are expected to assign a proper set of resources and autonomy for a coordinating body to perform its mandate. Therefore, the autonomy, rationality, and incentives of aid bureaucracy are re-evaluated as the effects of multi-layered, configured rules. 3.4.3 Rule Contexts and Hypothesized Bureaucratic Behavior in Multilateral Aid Management Legal/Regulatory Frameworks: Delegation, Task Allocation, and Coordination While both the Framework Act and the Presidential Decrees lay out the overall foundations for ODA management and delivery, the Law on the Measures for the Admission to International Financial Institutions (enacted in 1963; hereafter, IFI Act) provides the detailed guidelines for the management of multilateral ODA. As stated in the Framework Act (Article 2-7), multilateral development cooperation27 is defined as the practice by the nation-state of making contributions to selected international organizations, and is differentiated from bilateral development cooperation. Korea’s multilateral development cooperation aims to serve the following objectives: (i) to indirectly help developing countries promote poverty eradication and economic growth; and (ii) to increase Korea’s influence over the decision-making at funded IOs. As of 2016, the IFI Act lists 15 international financial institutes (IFIs), limiting Korea’s multilateral partnerships to the group. Considering the administrative and budgetary practices with regard to multilateral activities within the Korean government, a hypothetical delegation structure can be identified as follows (See Figure 7.1); major decision-makers and practitioners (president, CIDC, ministries, and implementing agencies) collectively work in a four-tier delegation system.

Figure 3.6 Delegation Structure for Multilateral Aid Planning and Management (by author)

27 The definition of multilateral development cooperation in the Framework Act is aligned with that of the OECD-DAC (OECD 2015, p.24). 77

Across the global development community, a variety of structures for managing multilateral aid exist. Compared to other DAC member countries (OECD 2008, 2012, 2015), the contemporary Korean multilateral aid system is distinctive in the following aspects. First, neither the Framework Act nor the IFI Act articulate the scope and areas of multilateral activities for respective state agencies; rather, a simple division of labor is at work in that MOSF enjoys autonomy over the partnerships with the 15 IFIs, whereas MOFA supervises non-IFI partnerships. Secondly, the scale of multilateral funding is decided primarily by the president (by issuing presidential decrees). No hard criteria are written down in the IFI Act regarding Korea’s entry to IOs, budgetary limitations for capital subscriptions to IOs, or withdrawal of IO membership. As an OECD report28 (2015) reveals, Korea is a case that is classified as the least prepared among DAC donors when considering the country’s manner and extent of goal-setting and strategy-building in the domain of multilateral aid policy (OECD 2015, p.70-71). In light of the concentration of multilateral practices within the government, Korea is found highly fragmented from the DAC standards. In total, 17 state agencies are involved in multilateral aid policy formation and its implementation processes, whereas an average of 9 state organizations manage multilateral cooperation in other DAC donors (OECD 2015, p.76). Annual funding for multilateral development banks (MDBs) is greater than funding to the United Nations (UNs) and other IOs. Like the cases of other young DAC member states, Korean aid agencies tend to disburse a large portion of

28 This report reviews the policy documents and statements of DAC donor governments and examines how clearly and coherently donor governments set up their goals and strategies in enhancing their cooperation with IOs. The number and contexts of relevant policy papers are reviewed. Based on the findings, the report includes a list of donors that is divided into four groups depending on their performance. 78 funding in the form of core contributions. Earmarked contributions remain relatively inconsiderable. Bureaucratic Autonomy, Rationality, and Incentives of Multilateral Aid Korea has a dualistic ODA management structure29, in which MOSF and MOFA equally supervise and perform multilateral activities. As articulated in Framework Act (Article 9-2), MOSF supervises contributions to selected MDBs as specified by the IFI Act, and MOFA is mandated to cooperate with non-MDB IOs, representatively, the United Nations’ agencies. Given the expansion of multilateral cooperation in Korea, rationality and incentives for aid policy-makers and practitioners are reconsidered from the perspectives of bounded rationality and principal-agent relationships (delegation). Although the regulatory-legal frameworks for ODA policy formation and implementation have been institutionalized in Korea since 2010, multilateral cooperation constitutes an example of how ODA is mobilized by bureaucratic autonomy without substantial legal foundations and policy precedents. The aid policy-makers and practitioners are alleged to enjoy great autonomy, and their policy decisions and interactions are likely to be subject to the interest of their own organizations. Expected Patterns of Interactions, Collective Actions, and Policy Outputs In the public policy process, written rules30 (in forms of laws and regulations) are expected not only to determine the areas and extent of policy-making autonomy for respective actors, but also to delineate the power relationships among them. Under hard rules, state agencies interpret their mandates, build strategies, allocate internal resources, and delegate tasks accordingly. Whether the behavior of state agencies can be understood as proactive-opportunistic or passive-routine, the organizational contexts—such as traditions, talents, information and resources—should be taken into consideration to understand their organizational rationality and incentives towards the chosen policy decisions and interactions. When state agencies interact with other actors in the collective process of coordinating and finalizing policy decisions, the following analysis is focused on seeing to what extent the behavior of the respective organizations is affected under peer pressure, and to what extent the observed collective actions might have contributed to widening the gap between planning and actual disbursements in the sphere of multilateral cooperation.

29 The following DAC countries—Japan, the U.S., France, Germany, Spain, and Belgium—are examples of having both finance and foreign ministries involved with the disbursement of multilateral ODA (OECD 2015). 30 As previously highlighted (See Section 5.5), Korea’s ODA laws leave detailed objectives and task allocation to be decided exclusively by the president. 79 Also, regarding the role of a coordinating body (CIDC)—a legally-imposed mechanism to coordinate and supervise the performance of state agencies—, the following analysis scrutinizes to what extent the CIDC affects the collective behavior of policy actors, i.e. whether it promotes cooperation or not. Hypothetically, the performance of a policy coordinating body is highly linked to its organizational attributes; the limitations and improvement in the performance of the CIDC are included. Apart from the legally imposed coordination mechanism, the study also sheds light on what alternative types of coordinating mechanisms might be fostered or discussed in the aid community in Korea.

3.5 Conclusion To properly and systematically find out the impacts of rules on policy outcomes, this research suggests a refined version of IAD framework. This attempt is new, since the existing institutional analysis pays little attention to discerning rule institutions by specific functions and effects. To recap, in hypothesizing, explaining, and predicting the decision-making behavior of governmental policy actors, the notions of bounded rationality and delegation theory are centrally applied. The fine-tuning of the framework is necessary to take rule institutions more seriously. Given that existing IAD-based empirical analyses in the domain of development cooperation focus on the effects of formal rules on foreign aid policy- making, those earlier operations of the framework appear ineffective to acknowledge the broader institutional contexts of rules and traditions across government units. Therefore, the previous attempts of utilizing the IAD framework seem problematic if we wish to explore the unique contexts of emerging/young donor governments whose policy-making tended or tends to be governed by greater administrative discretion under a relatively insufficient and abstract set of written rules. To properly tackle the concerns, this study adds more sophisticated analytical procedures to discern the impacts of various types and configurations of policy-making rules on the outcomes.

80 CHAPTER FOUR METHODOLOGY AND DATA COLLECTION TECHNIQUE: IN-DEPTH CASE STUDY ON INTRA-GOVERNMENTAL POLICY-MAKING DYNAMICS IN SOUTH KOREA

The purpose of this chapter is to explore the merits of qualitative research though a single country case. In particular, this chapter unfolds methodological considerations and data collection strategies that are chosen to highlight the benefits of IAD research. For the suggested institutional analysis that delves into the intra-governmental dynamics—in terms of how state organizations (as policy making actors) respectively and collectively behave under an array of hard and soft rules—the case study method (CSM) appears beneficial and even irreplaceable in providing in-depth knowledge about one county’s unique contexts.

4.1 Revisiting the Values of Qualitative Research 4.1.1 Qualitative Research for Public Policy and Public Administration Studies In comparison with the proven benefits of quantitative research in policy and organization studies, this proposed qualitative approach promises to exceed what quantitative research can offer under the following two conditions. First, considering the nature of public administration with regard to its internal process of formulating public policies, the decision-making process tends to be led by a small, closed group of policy elites. This fact might prevent relevant data from being available in a quantifiable and open manner. An identical issue has been discussed among those who have investigated ODA policy processes within donor governments or multilateral organizations (Ostrom et al. 2001; Martens 2002). In this regard, CSM can be a strategic way to solicit and incorporate evidence that is scattered around multiple levels of government and that can be approached by multiple techniques (ranging from elite interview, expert interview, direct/indirect observation, documentary analysis, and archival study). Second, CSM appears strategically beneficial if the approach is intended to shed light on the logical/sequential mechanisms behind a cause-and-effect relationship. Running against the passive interpretation and usage of quantitative research in Designing Social Inquiry (King, Keohane and Verba 1994; KKV)31, a group of methodologists

31 In the publication of KKV it is mainly argued that the rigor in qualitative research can be fortified through a better alignment to the logic of statistical and mathematical approaches. Although the work of KKV promotes qualitative research methods, the merits of qualitative research are revisited in a narrow way. 81 contends that qualitative research can excel in its own distinctive logic of inference that emphasizes causal relationships (Mahoney 2010; Collier and Brady 2010). The relevant attempts are largely made through an advanced research technique, process tracing, which is being discussed vibrantly with regard to its potential contributions to theory testing and theory development. Therefore, case study research, when a case study is designed to track causal sequences between a chosen set of independent and dependent factors, is favorable for generating knowledge that is not only country-specific but also has implications for a wider group of countries. 4.1.2 Avoiding the Risks of Making Incorrect Inferences about Bureaucratic Rationality and Incentives When it comes to operational matters, this research is required to tackle the following issues of how to collect, screen, and interpret data in order to reconstruct how a set of hard and soft rules affects bureaucratic rationality, incentives, and collective actions at the organizational and system levels. Suppose that an organization has a rationality of its own. It is important to figure out under what principles of discriminating evidence one may properly understand such organizational rationality and behavior. In addition, it is also critical to give due attention to the extent to which the researchers take the opinions or insights of policy makers or practitioners—if evidence was gathered through interviews with aid agency-affiliated officials or staff—into account as hard evidence of those of targeted bureaucratic organizations. This is because it is difficult to draw a line between personal opinions and those that might be well shared or supported throughout an organization. As Jones (2003) points out, there are no firm foundations that can either conceptually or methodologically link between individual-level processes of choice making and that of an organization in the existing organization studies. The existing institutional analysis is vulnerable to such methodological dilemmas as well. Although the incentive structures for respective aid agencies are explored32, there has been insufficient attention to this issue so far. Against the backdrop, this chapter add some discussion regarding how to minimize the risks of inappropriately explaining and predicting bureaucratic policy choices based on the

32 In existing IAD analyses on the foreign aid policy process—representatively, Ostrom et al. (2001), Martins et al. (2002), Gibsons et al. (2006), Hawkins et al. (2006), Park (2012)—individual aid agencies in those existing studies (the European Commission, SIDA, KOICA) are examined in an isolated manner without highlighting their working relationships with other key state agencies. Although they are the leading organizations within their national governments, their incentive systems do not necessarily represent those of a donor government. Therefore, the previous studies provide limited evidence and discussions when it comes to formulating an explanation of a country-level policy behavior. 82 evidence that speaks from one level (of human decision making) to another (that of organizational choice making). What is suggested in this study is to make the processes of data collection and data analysis as cross-checked and transparent as possible. Primarily, two tasks are proposed. First, with reference to the process tracing method, the probative power of qualitative evidence is evaluated depending on theoretical relevance and data collecting channels or methods. Second, to collect data on the collective action of multiple organizational actors at the government-wide level, the distributional influence of involving bureaucratic organizations on aid policy process is measured not just by size of allocated budget or the law-assigning role but also by the actual scope and detailed areas of policy-making autonomy. Nonetheless, this research also recognizes the limitations of the qualitative approaches in providing a superior, full explanation of bureaucratic behavior. And even the subjectivity of data in non-numerical forms cannot be fully eliminated. Rather, a major intention behind the suggested strategies is to leave room for the reader to re- evaluate presented data and ensuing inferences themselves in addition to what institutional theorists might interpret.

4.2 Methodological Considerations 4.2.1 The In-depth Case Study Approach to Delegation/Collective Action Dynamics The following discussion elaborates on the in-depth case study method, as this study prefers to call it. This study draws on an eclectic use of CSR methods and techniques that are optimized to shed light on the logical mechanisms of how the legal- regulatory frameworks of ODA policy making affect the collective actions among sub- state organizations that result in the commitment problems. The chosen case study method is labelled ‘in-depth’ for the following reasons. First, the logic of process-tracing is employed in order to distinguish existing explanations by varied powers of evidence—where evidence is collected at various methods and levels, therefore, presumably, having varied values for the research. Second, in a single country context, two nested cases (or within-cases) are provided. This enables the researcher to validate findings in a comparative sense. Third, this method intends to serve as an exploratory case study so as to discover and incorporate new evidence and variables into analysis. Beyond the traditional use of CSM, this open-ended research design is expected to contribute to theory refinement in terms of hypothesizing bureaucratic rationality and collective action. 83 Incorporation of the Logic of Process Tracing For the last decade, process tracing (PT) has been a buzzword for qualitative research methodologists. This non-statistical approach is expected to contribute to making rigorous inferences. By testing the probative power of evidence, it aims to single out a surviving explanation from dismissing weaker hypotheses. PT functions as a process of testing multiple hypotheses which explain a causal mechanism and narrowing down to a single, confirmed hypothesis that is competitively supported by the quality of evidence, not the quantity of it. By this approach, evidence can be evaluated by the purpose of research so the data are not assumed to be equally created or equally taken into consideration, unlike the statistical approaches. Also, by definition, the PT approach can be considered with a wide range of data sources—histories, archival documents, interview transcripts—that can confirm or reject given hypotheses (George and Bennett 2005; Bennett and Checkel 2015: 6). Previously, Allison and Zelikow (1999), in their work of Essence of Decision: Explaining the Cuban Missile Crisis, carried out PT-embedded research, which enabled the researchers to test hypotheses derived from three major theories of foreign policy making. The authors screened out poorly-warranted explanations based on multiple sources of evidence. The PT technique is now being practically but variably operationalized. In most explicit forms of PT research, four empirical tests33 are invented and often selectively conducted to identify the necessity and sufficiency of each hypothesis for a given causation. For this research, the benefits of the PT method are embraced when evaluating the hypotheses of sub-state organizations’ policy behavior from the theories of delegation and bounded rationality. Such an evaluation entails the task of finding supportive evidence to competing explanations by facilitating multiple data collecting techniques. On the other hand, there are some risks with the PT technique to be recognized. The major criticism of PT is its incapability to discover and incorporate variables that were not previously identified in theories. In addition, the value of data from multiple sources still can be subjectively compared. Also, when tracking back to causal steps, it can be endless to count intermediary sequences, and is difficult to identify a direct starting point for the given causation—hence called the problem of infinite regress (Bennett 2010). Providing Nested Cases

33 (i) Smoking gun test [to find hypotheses that are sufficient but not necessary to establish causation], (ii) hoop test [not sufficient but necessary causal hypotheses], (iii) straw-in-the-wind test [neither sufficient nor necessary], and (iv) doubly decisive test [both sufficient and necessary] (Collier 2011).

84 Although a single country case study (or N=1 study) is perennially criticized for its weak explanatory power for other cases in a context-free manner, it is counter-argued in this study that a large number of cases might not automatically guarantee a better sense of explanation or theoretical understanding. Considering individual contextuality, which is never negligible in reality, this can get bigger enough to adversely affect the conclusions of the research (George and Bennett 2005: 6). To respond to the dismissal of single case research from the statistical viewpoint, this study employs the idea of providing multiple observations, or called embedded cases, according to Yin (2009). Two sub-policy fields are selected to present specific observations on how the lack or insufficient clarification of rules, for instance, affected aid policy making at individual and collective levels in the Korean government from 2000 to 2016. Along with the theoretical consideration 34 , long-term observations are to be presented on aid bureaucrats’ decision-making behavior in the fields of knowledge and multilateral aid policies. This selection is made against other potential selections, such as regional ODA policies (towards geographical regions) or quantitatively significant ODA projects (i.e. the largest aid modalities or initiatives). Considering the commitment problems, the two cases have commonalities in that in recent decades, the two aid initiatives have been dramatically and exceptionally grown in scale but have turned out to be highly fragmented practices. 4.2.2 Strategies to Make Theory, Method, and Region Interactively Work Apart from the strategies to enhance the internal and external validity of qualitative research, there is another methodological caveat to be mindful of for case studies that pay attention to regional or country-specific uniqueness. A great volume of area studies relies on the interpretivists’ approach. This prevailing orientation tends to leave regional studies largely descriptive and context- limited. Primarily based on South Korean evidence, this empirical research needs some strategies to avoid being descriptive and theory-free. On this point, Kuhonta and his colleagues (2009)35 provide useful advice to area study researchers, introducing three major strategies to embrace (i) a causal argument, (ii)

34 As previously mentioned, in Korea, knowledge and multilateral aid cases are chosen on the grounds that their policy networks within the aid administration are comprehensive enough to include most of the key state agencies. 35 This publication explains the academic phenomenon in which area studies, in particular, in non- Western or developing region contexts, are relatively neglected because of the stark absence of good case studies that display any great consideration of how theory, methods, and region interact. 85 conceptual and typological analysis, and (iii) interpretivism 36 . A causality-focused research design and repeated proposition testing might be favorable to provide a better rigor for region/country-specific investigation. Given that the positivist approach also runs risks of generating wrong conclusion based on inappropriate hypotheses, this research also takes the interpretivist ideas into consideration. In this respect, the following context chapter serves to bring a dimension of tradition into analysis. This is to add a deeper understanding of given phenomena within a historical context.

4.3 In-depth Case Study in Action 4.3.1 Targeted Logical Mechanism for the Case Study: ‘Rule(X)-Incentive- Interaction-Collective Action-Policy Output (Y)’ A key proposition to be tested in the context of young donors is that a lack of legal-regulatory guidance (which results in ample room for bureaucratic leeway) for policy makers leads to policy practices that are poorly aligned with planning. Delegation theorists posit that administrative laws and regulations imposed on state bureaucracies are the critical rule institution. The rule institutions are assumed to work as the direct mechanism of preventing state agencies from potential agency slack in the performance of delegated tasks (Aranson et al. 1982; McCubbins et al. 1989; Epstein and O’Halloran 1994, 1999; Bendor et al. 2001). Suppose that a rule configuration is set as a dependent variable (X). X creates a particular set of bureaucratic incentives (M) for interaction (and collective action), and, in turn, determines collective policy choices (Y) 37 . Given that donor governments’ challenges are understood as structurally driven, a possible resolution against the problem of Y is that X, with appropriate adjustment (X’), would change the dynamics of bureaucratic incentives (M’) and resulting outcomes (Y’).

36 In a similar vein, Vennesson (2009, p.229) also makes suggestions rooted in the work of the French philosopher, Bachelard. Driven by applied rationalism, there are three epistemological acts that facilitate a dialogue between theory and empirics: epistemological (i) rupture (in other word, to address the rupture with conventional wisdom, proactively carving out a new aspect or correcting an incorrect image of reality); (ii) conceptualization (by questioning what it is a case of reflecting on previous theoretical concepts); and (iii) observation (to make multiple observations on the targeted event or object). Those suggested elements are widely used and are incorporated in this research design. 37 Particularly regarding the quality of aid policy, one of the most popular dependent variables in relevant research, there are a great many global attempts to quantify and compare it. For instance, the OECD-DAC initiates research projects that measure policy coherence and fragmentation among its member countries (OECD 2004; 2011). The Center for Global Development (CGD) develops a platform that has annually evaluates 27 donor countries’ commitment to development (Commitment to Development Index; CDI) based on their annual aid volume and aid component (CGD 2017). 86 Furthermore, this study points out potential problems with the existing understanding of the relationship between rules (X) and policy products (Y). Exploring the nexus in the context of policy making in the Chinese government, Rothstein (2012) argues that with non-Weberian type38 state bureaucracy in the central government of China, Y is far less likely to be explained by a hard/written form of X only. This is because a great many foreign policy choices are determined at their discretion and coordinated by authoritarian leadership under a loosely established set of national laws and administrative regulations. Another example speaks to this, as well, in a more cross-national sense. Among 29 DAC member countries, their degrees of articulating multilateral aid objectives and building strategies in written form turn out to be widely varied according to an OECD (2015) report39. Interestingly, five young small-scale DAC donors—South Korea, Poland, the Slovak Republic, Slovenia, and Greece—are rated low and placed in the bottom group category (‘having overarching development strategies’) (OECD 2015, p.71). Other new DAC donors (e.g. the Czech Republic) are rated slightly better in the context that they exceptionally develop engagement strategies with the EU. In fact, the number and contents of the policy documents per se do not automatically indicate the quality and strategic value of ODA that these countries provide. However, this indicator is useful in that it reveals the tendency that shows new donors have relatively insufficient elaboration of their ODA objectives and strategies compared to experienced donors. In such unique contexts where firm, meticulous legal and policy guidance is missing—in the parlance of delegation theory, discretion-based delegation settings—, the intermediate sequence of how state organizations autonomously reinterpret their mandates and perform their policy-making roles at their discretion is instrumental in explaining Y. Therefore, for this non-Weberian-type bureaucracy case, the following analysis inclusively examines how administrative discretion works.

Figure 4.1 Targeted Mechanisms of Foreign Aid Policy Making Dynamics within a Donor Government: The Korean Case (by author)

38 Weber (1978) defines state bureaucracy mainly by its competence and rule-following nature. According to him, the administrative functioning of bureaucracy is legitimized by legal authority. This type of understanding of state bureaucracy is called Weberian(-type) bureaucracy. 39 DAC donor governments are evaluated by three criteria: (i) whether they have stand-alone multilateral strategies, (ii) whether they have strategies specific to individual MOs or to a group of MOs, and (iii) whether they build sectoral or thematic strategies. 87

Legal and Regulatory Frameworks (X): To What Extent Hard Rules Leave Room for Bureaucratic Discretion to come into play Public laws and regulations are identified as the primary, direct tool of structuring the incentives and constraints of bureaucratic policy elites. As the independent variable, regulatory-legal frameworks that govern state agencies’ aid policy-making and implementation are examined in the first place. The reason that this research uses two terms interchangeably—legal-regulatory frameworks and delegation rules—is that the administrative laws and regulations do perform the function of not only clarifying the given authority for each actor, but also of defining relationships (of delegation) among actors. The legal-regulatory frameworks to be reviewed include documented policy statements (rule statements) pertaining to foreign aid provision, such as mid- and long- term strategy/white papers.

88 Box 4.1 Korea’s ODA Legal-Regulatory Frameworks (in Written Forms that are Reviewed in the Analysis (Source: www.odakorea.go.kr)

Legislations of International Development Cooperation and ODA

• Framework Act on International Development Cooperation (IDC) • Presidential Decrees on IDC • Law on the Measures for the Admission to International Financial Institutions • EDCF Law • KOICA Law

Aid Policy Strategy Papers

• The Strategic Plan for International Development Cooperation (2010) • Mid-term Strategy for Development Cooperation (2011-2015, 2016-2020) • Sectoral Strategies for International Development Cooperation (2011-2015, 2016-2020) • Annual Comprehensive Implementation Plan for International Development Cooperation 2006–15 (Annual Implementation Plan Paper) • Country Partnership Strategies • The Guidelines on the Integrated Evaluation (2009)

Memorandum and Ministerial Announcements

• CIDC (Committee for International Development Cooperation), meeting agendas at the 1st– 21st meeting (2006–15). Available at http://www.odakorea.go.kr/) • Foreign Affairs and Unification Committee, National Assembly, minutes from committee meetings at the 16th–19th National Assembly (2000–15). Available at http://likms.assembly.go.kr/record/mhs-40- 010.do?classCode=2&daeNum=19&commCode=CI&outConn=Y • MOSF, ODA or development cooperation related announcements. Available at http://www.mosf.go.kr/policy/policy06.jsp • MOFA, ODA related announcements. Available at http://www.mofa.go.kr/ENG/policy/oda/index.jsp?menu=m_20_110

Configuration and Descriptive Style of Written Rules Delegation rules are primarily reviewed in term of who has what scope of authority in the domain of foreign aid policy formation and implementation. Locations for inter-ministerial (or inter-agency) policy coordination, budgetary allocation, monitoring and evaluation of aid programs and projects, and information management are identified. The location, scope, and details of mandated or not-described autonomy of 89 state actors at multiple levels of delegation are critical to clarify rule configurations. Also, the manner of articulating objectives and tasks in hard rules is critical to figuring out how and to what extent the descriptive style of rules affects bureaucrats’ reinterpretation of the existing policy guidance. Incentive Dynamics (M): How Discretion Shapes Interaction As hypothesized, the given discretion in some areas of policy choice has effects on intra-governmental relationships that exist vertically (e.g. relationships between ministries and subordinate aid agencies) and horizontally (e.g. inter-ministerial or inter- aid agency relationships). As the core arena of ODA policy-making and implementation, the intra-governmental delegation chain in Korea entails a four-tiered structure in which the task of foreign aid delivery is delegated from the executive (president) to ministries and further to subordinate aid agencies. Once again, delegation theory appears sensible when analyzing the political dynamics among policy-making bureaucrats, and explaining resulting policy products. Delegation theory, an expanded version of agency or principal-agent theory, provides us with an analytical framework for identifying legally powered, multi-tiered relationships of delegation within an aid administration and for tracking individual and collective effects of bureaucratic relationships on foreign aid policy formation. Principal-Agent (Delegation) Situations among State Agencies A primary methodological task here is to find evidence of bureaucratic rationality that matches the actual practice of delegation. It is assumed that the lax nature of delegation rules gives great room for each organizational actor to re-interpret its own mandates and to redefine tasks to increase their organizational interest and autonomy. At the same time, the activities of a state agency can be incentivized or restricted by its principal organization(s). For respective state organizations that are involved in aid policy making, official statements and documents released are reviewed to understand their preferences, chosen strategies, and behavioral patterns in working with principal and agent organizations. Reflecting on traditional specialty and capacity of each organization, it is intended to see whether chosen strategies and activities for an organization are seen as routine/passive or opportunistic. Peer Situations among State Agencies With regard to inter-ministerial and inter-agency interaction, it is hypothesized that no clear division of labor (or resources) would lead to intensifying bureaucratic competition. At this level of interaction, inter-agency negotiations take place and 90 collective policy decisions are coordinated and finalized. Therefore, a key analytic task is to delve into the collective rationality behind cooperation or hostility in the system. Likewise, it is important to learn about the gap between the actual and nominal functions of the policy coordinating body (the Committee of International Development Cooperation; CIDC). Collective Action Outcomes (Y): What Policy Patterns are Induced Overtime This research is intended to capture long-term trends in aid policy products as the outcomes of collective action among policy makers. The features of policy products are analyzed by sub-criteria, which are suggested by the CGD (2017) and QuODA (2014). Those indexes are developed to examine donor governments’ performance in the realm of aid effectiveness principles (See Appendix). 4.3.2 Correspondence between Concepts and Data A proper match between given concepts and available measures is one of the key tasks for this qualitative research (Goertz 2005). Rationality has many degrees and multiple dimensions (Boden 1994: 188). Bureaucratic rationality is a rather relational concept given that the behavior of a state organization, for instance, beyond its means- end thinking, can be significantly reshaped by its relationships with other agencies (Townley 2008). Therefore, each organizational actor is examined by the following proposed scheme.

Box 4.2 Analytical Tasks to Understand the Rationality and Behavior of Organizational Actors

A. Organization’s reinterpretation of goals and tasks • Behavior: Match between goals and means • Context: Budget, capacity, and tradition

B. Organizational behavior and rationality towards principal and agent organizations • Behavior: Observance or control • Context: By principles of intended rationality, adaptation, uncertainty, and trade-off40

C. Organizational behavior and rationality towards peers • Behavior: Cooperation or rivalry • Context: By principles of intended rationality, adaptation, uncertainty, and trade-off

Incorporating Human-Level Evidence for Organization-level Analysis

40 The four principles are suggested by Jones (2003) as elements of bounded rationality. 91 In conducting the first analytic task (Task A), major sources are official documents and announcements that are produced by respective organizations. Official websites and national archives are visited in order to find organizational goals (including overall, sectoral, time-specific objectives), strategic areas, performance records, size and allocation of budget, human resources (employment policy), and internal decision- making structure). Although Task B requires data on inter-organizational interaction, there are insufficient quantitative data available in this area. Thus, this analysis depends on personal accounts. This study found proper candidates for interview on the basis of their direct relevance to ODA policy making. Data that are required for Task C are available in a better sense compared to those of Task B. The relevant records of inter- ministerial and inter-agency committee meetings were collected at the official website of the Prime Minister’s Office (www.odakorea.go.kr).

4.4 Strategy for Data Collection and Interpretation 4.4.1 Obtaining the First Layer of Evidence: Documentary Sources This following analysis drew upon a wide range of documentary information that is available at governmental websites and archives. Quantitative and qualitative data were primarily collected through the National Statistics Office and documents from governmental sources. Evidence on overall policy-making (e.g. policy making legal and administrative structures, inter-agency meeting agendas and involved personnel/organizations, budget distribution standards, stipulated ODA goals and mandates, strategy papers, and policy coordination procedures) and respective organizations’ operational preferences and strategies (e.g. historical ODA performance records, executives of organization, strategy papers) were taken into consideration. Data on inter-organizational interactions within government (e.g. informal policy negotiations, exchange of information or personnel, and operational cooperation) were the least accessible and were insufficiently available. In this area of information, interviews were major sources. 4.4.2 Obtaining the Second Layer of Evidence: Interviewing Aid Policy Elites and Experts To address missing data and data inconsistency that the documentary review left, in-depth interviews with state officials, policy advisers, and experts were conducted in 2014 and 2015. As a result, a total twenty-two personnel participated in semi-structured interviews (for the questionnaire, See Appendix B). Interviews were used as a secondary technique to fill in the gaps for missing quantitative/qualitative data. Also, this technique 92 is suitable when studying the complex policy formation process where probability sampling is insignificant and, instead, multiple, sporadic data points exist (Tansey 2007). Expert interviews took place with those who worked closely with aid-managing ministries and agencies, such as R&D researchers, scholars, international organization staff, and domestic NGOs. From the very early stages of this research, expert interviews were conducted to collect a variety of explanations and insights. Their opinions on the potential reasons for Korea’s commitment problem were not necessarily theory-based but found to be empirically interesting. I found this inductive research approach very useful to find new variables and rival explanations that were contextually driven and empirically understood among a wide range of aid policy stakeholders. Elite interviews were with those who were directly involved in the foreign aid policy formation and implementation processes.

Box 4.3 List of Interviewees: Selected ODA Policy Makers, Practitioners, and Experts in Korea

Elite Interview (13 interviewees)

MOFA 1 officer (Division of Int’l Development Cooperation) MOSF 1 officer (Division of Int’l Development Cooperation) EDCF 2 head of department (Planning & Coordination Dept.) deputy dept. head (Planning & Coordination Dept.) KOICA 3 section chief (Office of Assessment and Evaluation) researcher (Office of Assessment and Evaluation) researcher (Public-Private Partnership Department) NARS41 1 officer (Foreign Affairs and National Security Team) PMO 2 staff CIDC 2 non-governmental committee member KDI 1 researcher (Center for International Development)

Expert Interview (9 interviewees)

Former MOFA staff * 1 researcher (Dr. Lee, Jae Eun) Former KOICA officer * 1 Dr. Kim, Sang-Tae (Kyung Hee University) Academic * 4 Dr. Kim, Soyeun (Sogang University) Dr. Kim, Jiyoung (Soongsil University) Dr. Park, Hae-yun (Ewha W. University) Dr. Kalinowski, Thomas (Ewha W. University) KIEP 1 researcher (Center of Int’l Dev’t Cooperation) NGO 2 researcher (ODA Watch)

41 The National Assembly Research Service, Republic of Korea. 93 researcher (KoFID42)

(*Some of the names and holding positions of the interviewees were permitted to be disclosed for the research)

To control the quality of such narrative data—in terms of relevance, reliability, validity, and representativeness—several measures were applied. Although this research included interviews with a policy group, which is a very small, closed elite group, most of the suggested evidence from interview was cross-checked by interview participants. Contradictory data or acquired from interviews were presented as they are in the analysis. 4.4.3 Note for Effectively Managing Evidence from Elite/Expert Interviews The last but not the least methodological concern for this qualitative research is rested on how to effectively present and use the evidence from interviews for clarifying the ‘rule-policy output’ nexus on the aid policy-making process. First of all, structuring the whole questionnaire (See Appendix B) by the logic of IAD—by the sequential- analytical stages (from CONTEXT, ACTION ARENA, INTERACTION, and to OUTCOMES) of how bureaucratic incentives are formed, interplayed, and determined into policy outputs under a particular rule setting—appears ineffectively difficult and time-consuming for prospective interviewees. Due to that reason, the questionnaire was structured from big/easy questions (for asking about the overall impression on Korea’s aid system and performance) to smaller/ more specific questions (about their specific operations and insights within their affiliated organizations) to effectively retain the attention of the interviewees throughout the interview.43 Second, although a number of interesting ideas and perspectives were gathered in each area of the questionnaire, I evaluated them depending on whether they are concurred with others when it comes to assessment of the overall aid system and inter-agency interaction. This is because each agency has different natures and degrees of organizational interest considering particular questions. Third, double-checked evidence was re-evaluated by contrasting with the suggested hypotheses and propositions. In this process, this research left some room for the evidence to be supportive of other explanations as well. Oftentimes, this study found that a better explanation for some

42 The Korea Civil Society Forum on International Development Cooperation 43 For the mid-level area question, a series of questions were asked in order to understand the perceptions and actual interaction between state agencies. This area of research (on inter-agency relationships within Korea’s aid management system) is particularly short of quantitative evidence and regarded as politically sensitive in the context of an intensive level of bureaucratic competition and rivalry. 94 specific bureaucratic incentives and behavior at the individual and collective action levels might not come exclusively from rational choice theory and delegation theory. However, overall, interviews confirmed that the institutional perspective this study holds does help explain the somewhat irrational aid planning behavior of the Korean president and aid bureaucrats largely through the idea of bounded rationality. Fourth, regarding the core argument of this research—Korea’s top-down aid management design induces the country’s poor performance of delivering its aid goals and pledges—, it is somewhat an after-thought after analyzing all the documentary and interview evidence, thus, it was not clearly presented at the time of interview. Luckily or not, it allowed me to encourage interviewees to free talk about how they define their aid management design in the first place, to discuss the benefits and costs of having such a structure (most defined as top-down) with less prejudice, and to share their relevant policy-making and operational experiences. I also added a hypothetical question like if Korea had a bottom-up ODA management design (or something different from the existing aid management design), asking what the consequences or effects on their operations and overall ODA performance would be. Fifth, in the interest of better understanding of the reader, I tried to present the interview evidence as close and concise as it was in terms of who (i.e. by category of interviewee and by organizational affiliation) answers what question and how the answers mean to the theoretical explanation of this research or others (The contents are followed). Although there might be various ways of making the implication of the interview evidence more straightforward and valid (e.g. counting the number of positive or negative answers among interviewees), for this particular case study that conducted elite/expert interviews with less than thirty personnel during the research period, quantifying the evidence is strictly limited and does not make good sense of the key puzzle and possible explanations. As earlier mentioned, the sample questionnaire (See Appendix B) was used for the general guideline. Most of the time, the interviewees and I ran through a different series of questions. Sticking to three area-focused questions, the interview addressed a wide range of inquiry and answers about bureaucratic incentives as follows. Questionnaire Area (1): Overall Perspective of Korea’s Aid Management System As implied through the questionnaire, the interviews were usually started not with the pinpointing of Korea’s commitment problem but with Korea’s experience in aid expansion. This is because aid expansion is the most familiar, widely accepted perception of Korean ODA among policy makers and practitioners. As far as the state documents 95 and academic literature are concerned, under the shadow of the ‘aid expansion’ story, the commitment problem has been rarely recognized or dismissed in policy and academic circles. The major area of question was the interlinked bureaucratic perception about government’s aid management design, aid expansion (performance), and aid motivation. At the expert interviews (addressing Question 1 and 2), most of them agreed that Korea has a distinctive top-down structure of foreign aid policy-making compared to other countries given its legal-regulatory setting and state-led development tradition. The interviewees from state agencies, on the other hand, tended to partially or hesitantly agree on Korea’s top-down aid management design but emphasized the merits of the top-down management (i.e. ‘state entrepreneurship’ in ODA expansion and diversification; See empirical chapters (Section 5.1 and Section 6.4)). For Question 3, in particular, almost all the aid practitioner interviewees recognized that Korea’s aid system appears to have insufficient manpower and resources to fulfill the initially-set aid goals (See Section 5.4.4)—despite that the aid system has grown in terms of personnel and expertise, albeit relatively slowly. Under the structure in which no restrictions or monitoring over the discretionary aid planning practice are in place, they found little problem with setting overambitious aid targets at respective organizations–rather finding it strategically and routinely driven. In the discussion of Question 4, the president was pointed out as the most influential actor for and contributor to Korea’s rapid aid expansion. However, some of the interviewees expressed that president’s expansionary aid policy direction has been rarely shifted (unlike predicted by theories that this can be possibly affected by the political orientation of political leadership) at times of regime change. But some aid practitioners at the implementation level showed the fears of regime change which does increase political-budgetary uncertainty for aid agencies. This heuristic idea, in fact, is confirmed in that Korea’s aid geographical and sectoral allocation, besides the quantitative expansion in ODA, has not been significantly changed during the transition from the left-wing to the right-wing government (See Section 6.3.2 and Section 7.3.1). For Question 5, regarding the presence of NGOs on the aid policy-making process, only a few of state official respondents commented that a selected few (from the non- government sector) can intervene the process. Aid policy makers and practitioners seem to be reluctant to work with NGO actors, mentioning their limited capacity and expertise. Questionnaire Area (2): Inter-Organizational Interaction within the Aid System 96 Why inter-agency interaction is important? I asked those relevant questions because it is important to know whether the Korean aid system has some certain policy coordination or monitoring mechanisms to manage the overall aid performance (or to offset the side-effects of aid management design) among state agencies. Evidence from Question 6 and 7 confirmed that there are very strict boundaries between areas of interest/operation among state agencies (i.e. loans vs. grants) by which the Korea’s aid policy community is sharply divided. For the role of a coordinating body (Question 8), some commented that the CIDC has gradually created some positive-coordinative outcomes in the aid system, others argue that the coordination role of the CIDC remains largely nominal with limited resources and expertise in the policy field. Questionnaire Area (3): In-house Discourses within Respective Aid Agencies According to the interview, state agencies tend to reinterpret and redefine their organizational objectives and assigned tasks at their wider discretion which goes beyond the strict interpretation of ODA-pertaining laws and regulations (Question 9 and 10). For some previous aid officials, they justified that Korea’s rhetorical aid practice is somewhat necessary for small donors like Korea in order to exaggerate its limited influence (Question 11 and 12). Although this issue (of allocating control and autonomy which determines aid management design) is quite important for their daily operations, aid practitioners at the operational levels tend to be indifferent in this matter or did not find the connection between them during the interviews (Question 12 and 13). Among others, KOICA is relatively interested in this issue, referencing other country cases with reforms and restructuring that allowed more autonomy to the lower levels of aid management and their better performance in aid effectiveness (Question 14). As a matter of fact, many of the interviewees do not know the details of ODA laws and regulations. Also, not all of them have equal interest in understanding the holistic aid management design and how its configuration would affect their daily jobs and overseas operations. A few of them actually recognized the commitment problem with Korean ODA and that within their own organizations. Nonetheless, those interviews largely confirmed that the concentration of control at the top-level of policy-making does explain the dual sides of Korea’s ODA performance for the last decades, which has been impressively expanding ODA in a short period of time and turning Korea to a recognized donor, but producing the most fragmented, the least reliable aid delivery in comparison with other global donors (OECD 2008, 2012). At the policy coordination and implementation levels, lack of autonomy does increase uncertainty for aid agencies to flexibly manage their resources, therefore, opportunistically behaving by allocating 97 resources for short-term, high-profile aid operations rather than their specialized areas of operations or retention of aid staff.

4.5 Conclusion The proposed research design is in an attempt to embrace on-going discussions on evolving qualitative research methods. With the mixed methodological design, this proposed in-depth case study is intended to form a contextual, selective response to the recurrent criticisms and limitations of CSM. In line with George and Bennett (2005) who highlight the potential role of CSM in contributing to theoretical development, this study is methodologically designed to be both exploratory and explanatory. Also, this chapter includes strategies for collecting data and effectively presenting interview evidence.

PART TWO: EMPIRICAL ANALYSIS 99 CHAPTER FIVE ANALYSIS OF ‘CONTEXTS’ OF SOUTH KOREAN ODA POLICY-MAKING

5.1 Overview of the Chapter 5.1.1 Analysis of Aid Policy-Making Contexts: Past and Present As the first empirical one, the following chapter lays out ‘CONTEXTS’, an initial area of the institutional analysis. As conceptualized earlier, all the relevant contextual factors are categorized into three groups in the IAD research: (i) physical and material conditions; (ii) attributes of the community; and (iii) rules-in-use. Although this study focuses on the direct linkage between rules-in-use, incentive, and the commitment problem, the effects of the other two categories of contextual factors are taken into account as supplementary knowledge. Another layer of CONTEXT analysis is additionally suggested. In fact, historical contexts have rarely been discussed in the existing IAD analysis. Historical developments are instrumental in providing a deeper understanding of past ‘contexts’ that are relevant to contemporary contexts. An analysis of contexts in recent history is also useful since the commitment problem prevailed widely and substantially in foreign policy domains during the Cold-War period in Korea. Therefore, the following analysis is intended to explain how contextual factors have evolved, and how the changed or persistent contextual factors might have contributed to the problem of the planning-practice gap in the provision of ODA. 5.1.2 Foreign Aid Policy Traditions in East Asia and Korea The policy behavior of expansionary aid rhetoric and ambitious goal-setting is not something new to the Korean government. During the Cold War era, the country took a unique path of industrial upgrade and democratization before being recognized as an advanced economy and democracy by the global community from the 2000s onwards. Numerous academic works have explored the puzzles behind such rapid economic and social development in Korea and other East Asian industrializing countries (WB 1993; Johnson 1995; Evans 1995; Woo-Cumings 1999; Kohli 2004). A mainstream explanation speaks to a notion of developmental state, which represents a third type of capitalist economy, one with a strong emphasis on the penetrating role of the state in domestic and international markets. Across public policy areas, the developmental role of the Korean government was possible, and perceived to be successful, allegedly due to having charismatic leadership, highly competent bureaucrats, and a close coalition between the state and businesses 100 (Johnson 1995). Under unconsolidated rule settings, the following two governmental factors appeared to be working complementarily in such a way that an inherently fragmented bureaucracy was effectively orchestrated by authoritarian leadership—during the Rhee (1945-60), Park (1963-79), and Chun (1980-88) administrations—(Kohli 1999, 2004); and the policy initiatives of the state executive tended to face little parliamentary opposition as they were effectively justified by the presence of technocracy44 (Woo- Cumings 1999). In addition, the external assistance that was given to and disbursed by the Korean government was channeled to promote close coordination with domestic industries (Evans 1995; Herring 1999). On the contrary, in the area of international affairs, Korea’s foreign policy stance appeared far less proactive. Such behavior was commonly found across East Asian countries, which displayed a tendency to lack indigenous rationales and to prefer reaction to strategy. In the context of competing with the north (the Democratic People’s Republic of Korea) for international recognition, the Republic of Korea appears to have preferred a reactive and US-dependent course of action. Such a preference appeared well-reflected among state bureaucrats in their practice of aid management, as a recipient and a development partner. In the very early years of initiating development cooperation with other capitalist developing countries, it was the material conditions under which the Korean government implemented very small-scale international development packages, upon the request of and financial support from the U.S.. The involving policy-makers and practitioners involved in the development of global partnerships, and the economic and scientific bureaucrats (at the industrial and education ministries) led the process. During the Cold War period, overambitious policy goal setting—in cases of disbursing aid to the rest of the Third World or making commitments to international schemes—was an accepted strategy within the government to manage the gap between domestic constraints and external pressures. 5.1.3 Contemporary Contexts Embedded in the National System of Foreign Aid Policy- Making ODA now represents one of the fast-evolving, expansionary policy areas in Korea. Since Korea officially graduated from its aid recipient in the late 1990s, the country’s

44 The presidential initiatives during its industrialization period (1960s-90s) entailed a high level of technicality, which prevented politicians from effective monitoring and evaluation. They were put forward by a highly educated, well trained group of bureaucrats. In the post-war Korean society, those government initiatives of controversy, radicality, or even oppression were largely justified by a national cause of economic betterment of a national economy. 101 scale of ODA provision has grown rapidly. The increased scale of ODA provision itself has led to some inevitable institutional changes, such as an increase in human resources in the ODA management system, a series of enactments and amendments of ODA laws and regulations, and the establishment of a policy-coordinating body. Although some scholars explain the impressive growth in Korean ODA by citing its unique state entrepreneurship, revisiting the developmental-strategic state narrative (Sörensen 2010; Kim and Potter 2012; Kalinowski and Cho 2012; Sato and Simomura 2013), this study suggests that for the contemporary institutional settings of ODA policy- making—in terms of physical and material conditions, attributes of aid policy-makers and practitioners, and hard/soft rules pertaining to ODA management and delivery—it might be reasonable to reconsider by analogy with the historical contexts under the reactive state legacy (Yi 2016). There are some sea changes observed in the current material, legal, and human contexts, nonetheless. As of 2015, an annual government budget of a trillion won (equivalent to USD 0.91 billions) has been spent on ODA. The management of ODA policy-making is shared between the foreign affair and economic ministries. The ODA policy system comprises a diversified group of state agencies and non-governmental personnel, which means that the norms and attributes of the ODA community have evolved in recent years as a result. However, the changes that have accompanied the drastic institutionalization of the last decade have largely been made by way of acknowledging the existing informal structure of policy-making without attaching proper political legitimacy or economic logic to the institutional choices. Although the human and organizational participants in the ODA policy system have expanded in number, the expansion of the management system appears insufficient to control the drastically- increased ODA scale (OECD 2012). With all the institutional contexts to be considered, the commitment problem might not be the product of simple miscalculation or mere wishful thinking on the part of ambitious policy-makers. The underlying institutional settings, both present and past, are some important starting points for further analysis of the collective action problem.

5.2 Korea’s Historical Experiences of Foreign Aid (1950s-90s) To begin with, Korea’s mixed experience as an aid recipient and development partner should be unfolded. This section highlights the earlier contexts that were involved with its aid recipientship and development partnerships. As the following observations reveal, the government managed the external assistance in a highly autonomous and 102 centralized manner. There were strong bureaucratic rationality and incentives towards being externally reactive in accepting and disbursing development assistance. 5.2.1 Aid Receipt (1948-99): Channeling Foreign Aid to Post-War Reconstruction and Industrial Transformation Arguably, post-war Korea (from 1953 onwards) depended heavily on foreign resources for its long-term transformation to an aid-independent economy (Lim 1985; Woo 1991). From 1945 to 1999, an estimated USD 12.7 billion was funneled into Korea in mixed forms of bilateral aid, multilateral lending, and humanitarian assistance (Kim 2011). During the period of aid receipts, 92.7 percent of aggregate aid inflow was given bilaterally—mostly from the U.S. (USD 5.5 billion) and Japan (USD 5.0 billion). The ratio of loans to grants stands at almost one to one throughout the entire period, whereas the major type of foreign aid given gradually shifted from grants to loans. In the aftermath of the Korean War (1950-53), the country benefited from a huge aid influx, exclusively from the U.S., which considered Korea a strategic breakwater to contain communist influence. Under the pro-American Rhee Syngman administration (1948-60), foreign assistance served as the primary income source and political stabilizer (Woo 1991). In the pre-war period, Korea-destined U.S. assistance was delivered by the Economic Cooperation Agency (ECA)—which opened a branch office in Seoul and exerted leverage over the process of national development, strategy-building, and policy implementation. Shortly after the Korean War, the Combined Economic Board (CEB)—a U.S.-led committee that was composed of bilateral donors and U.N. organizations—soon replaced the ECA. Although this change would have led to a useful chance for Korea to be less dependent on the American resources, the Rhee administration remained in close consultation with US’s aid officials. While Park Chung-hee held power (1961-79), the major component of the aid inflows had been gradually shifted from humanitarian support to concessional loans. In anticipation of potential decreases in U.S. aid, which were foretold in tandem with the revised U.S. foreign policy stance under the Kennedy administration, the Park regime struggled to secure other external capital resources to minimize domestic shocks and to continuously stimulate the national economy. After Korea reached war reparation deals45 with Japan in 1962, Japan became the largest

45 The 1962 Ohira-Kim Memo was the outcomes of closed-door negotiations between the Ikeda cabinet and the Park government regarding compensation for the Republic of Korea under Japanese colonial rule (1910-45). This reparations deal, despite that nationalist sentiments against Japan common in Korean society, allowed Seoul to channel foreign capital and technology to local businesses. Japanese aid came in grants (USD 0.3 billion), concessional loans (USD 0.2 billion), and commercial loans (USD 0.1 103 bilateral aid provider, channeling concessional loans and technological transfers to the Korean economy. In the Cold War milieu, in which the two Koreas furiously competed for international legitimacy, the Park regime necessarily and heavily relied on the U.S. in terms of national security and markets. Throughout the Park regime period, American assistance played a crucial role in the national economic development of Korea. From 1946 to 1980, the country was the third largest recipient of U.S. aid, after Vietnam and Israel, according to the United States Agency for International Development (USAID). Although Park’s first five-year development plan (1962-66) was directed at making its economy self-sufficient, the government did not underestimate the fact that Korea was still highly aid-dependent, and external assistance was managed under the strong monitoring of the USAID (Woo 1999). With regard to the governmental structures and mechanisms of aid management under the Park administration, the Economic Planning Board (EPB), after the disbanding of the CEB, performed as the central agency for attracting and managing foreign loans (Kim 2011). Internally, the EPB was a top agency that enjoyed a high level of autonomy, with exclusive support from the president, who seemed to be aware of the great importance of securing and channeling foreign loans for national development. Despite weakened U.S. influence, the EPB maintained its close working relationships with the U.S. and other international aid organizations. 5.2.2 Development Partnership (1960s-80s): Occasional Provision of Technical Assistance The very first development cooperation organized in Korea dates back to 1963. Since then, relevant activities were occasionally organized in the form of technical cooperation. Most funding came from USAID and the Korean government. They dispatched Korean experts to other developing countries and invited foreign government officials to Korea for professional training. Such technical cooperation remained at a very small scale until the late 1980s. That is rather understandable in light of the fact that Korea itself received a staggering amount of development assistance from the rest of the world. Notwithstanding the meager scale of development partnerships, the early Korean aid projects were determined by its relationship with North Korea (Kim and Seddon 2005). Major partner countries were the newly independent countries of Africa and Latin America. It was part of Park’s foreign policy doctrine to promote development

billion). Major beneficiaries were the Pohang Iron and Steel Company and other enterprises in the industrial sectors of railway, expressways, telephones, dams, shipping, and electricity. 104 partnerships with those remote countries against the expanding influence of North Korea through the Non-Aligned Movement 46 . Even under subsequent regimes after Park, technical assistance was initiated mainly for the same diplomatic purposes, as is clearly stated in the official statements47 of the Ministry of Foreign Affairs48 (MOFA).

Table 5.1 Early Years of Korean Development Partnerships with Other Developing Countries (1960s- 80s)

International Development Cooperation Activities First training project for foreign trainees, funded by USAID (1963) 1960s Government-funded training project initiated (1965) Medical and Korean martial arts experts dispatch (1968) In-donor training projects for foreign technicians (1975) 1970s First bilateral grant program initiated by the Ministry of Foreign Affairs (1977) International Development Exchange Program launched by the Korea Development Institute (1982) 1980s Technical cooperation activities initiated by the Ministry of Construction, the Ministry of Labor, and the Ministry of Science and Technology (1983, 1984, 1987) UNESCO Korean Young Volunteer project launched (1989)

(Source: www.koica.go.kr)

5.2.3 New Momentum for Scaling Up ODA in the 1990s From the late 1980s, the country’s economic growth resulted in substantial international pressure to address the ethical obligation of better-off economies towards the global problems of extreme poverty and conflict. At about that time, a burden-sharing initiative between advanced economies and newly industrializing economies (NIEs) officially emerged at the 1987 Louvre Accord in Paris. Set against the growing international demand for contribution, the Korean government started to scale up its annual aid commitments and established two ODA funds—the Economic Development Cooperation Fund (EDCF) in 1987 and the Korea International Cooperation Agency (KOICA) in 1991. Throughout the preparatory processes to launch full-scale aid projects, Japan’s ODA model was the major reference point for the Korean aid system architects, as is

46 The Non-Aligned Movement (NAM) was a multilateral project among Third World countries, which pledged to hold a neutral position in between the communist and capitalist blocs. North Korea was a signatory country to the multilateral treaty, whereas South Korea was not. 47 MOFA’s Standard Operating Procedures on Grant Aid was published in 1987. 48 In Korea, the names and jurisdictions of ministries have been constantly shifted at times of administrative reform or regime change. As of 2015, the foreign affairs ministry was renamed the Ministry of Foreign Affairs (MOFA), previously the Ministry of Foreign Affairs and Trade (MOFAT). 105 confirmed in interviews with retired government officials and relevant sources (Kim and Seddon 2005). Among some of state bureaucrats, the Japanese model was considered a source of positive lessons for the Korean government. This is mainly because the two countries have similar interests and strategic stances in trade and national defense in terms of heavy dependence on international markets and U.S. security umbrella. During the period of institutional buildup, Korean ODA was explicit about fostering Korea’s economic and commercial interests in Asia. On the other hand, due to extremely limited domestic resources and insufficient administrative capacity, the altruistic/humanitarian purposes of ODA were limitedly pursued throughout the 1990s. Overall, in the previous aid regime, the sources of aid increases were largely supplied from the outside rather than domestically.

Table 5.2 Korea’s ODA Scale-up in the Context of Economic Growth in the 1990s

1991 1992 1993 1994 1995 1996 1997 1998 1999 ODA 57.5 76.8 111.6 140.2 116.0 159.2 185.6 182.7 317.5 (USD Million) ODA/GNP ratio 0.020 0.025 0.034 0.037 0.026 0.033 0.042 0.058 0.079 (Percent) GNP 292.0 305.7 330.8 378.0 452.6 480.4 437.4 316.8 402.1 (USD Billion) Bilateral ratio 54.7 58.8 53.8 42.8 61.8 77.4 59.9 68.2 41.3 (Percent) Grant ratio 53.6 50.8 38.5 36.7 55.9 52.4 45.2 38.8 33.6 (Percent)

(Source: Author’s compilation based on data from OECD QWIDS)

Both the abrupt dissolution of the Soviet Union and the U.N.’s recognition of the two Koreas in the early 1990s provided further momentum for Korean ODA policy to keep its expansionary stance. Neither Korea’s first emergency relief to North Korea49 in 1996, or the Asian financial crisis of 1997, appears to have interrupted the long-term upward trend in ODA.

49 In fact, North Korea-destined emergency aid does not count as ODA due to constitutional interpretations, and these humanitarian activities were organized exclusively by the Ministry of Unification. 106 5.3 The Earlier ‘ACTION SITUATIONS’ of Korea’s ODA Management Design (1970s-90s): Focusing on Actors, Structure, External Influence, and Ideas Previously, the external contexts that the Korean government encountered from the 1950s throughout the 1990s assisted our understanding of the country’s mixed experience of aid receipt and development partnership. U.S. alliance, diplomatic competition with North Korea, and the Japanese ODA model were identified as solid factors that shaped Korea’s aid-giving motives and structure, where the planning-practice discrepancy in the provision of ODA was part of foreign policy rhetoric during the Cold War. In the following section, Korea’s inceptive aid policy system (1970s-90s) is further examined. This section highlights the resources in use, bureaucratic culture and norms, legal-administrative structure, and state-society relationships that were involved in the aid policy process. 5.3.1 Failures to Make Credible Aid Commitments: A Product of Preferred Strategy, Political Immobility, or Inexperience? From the early years of aid provision, ODA policies and strategies might have been made in a reactive manner. The reactive behavior of a donor country can be defined by its policy choices that are neither strategically driven nor socially embedded. Therefore, this study posits that expansionary aid policies, for instance, cannot be automatically considered an indicator for the performance of a donor government for its better developmental commitment. During this earlier period, Korea indeed tended to make unrealistic ODA pledges and displayed incoherency between aid planning and practice. An important point for thought in this analysis rests on whether Korea behaved reactively and ostensibly by choice or otherwise. In other words, the question proposed here is whether a donor government strategically chooses a reactive policy behavior despite its full capacity to do otherwise; or whether it is the domestic constraints (e.g. the absence of consensual aid objectives or integrated aid systems, or lack of expertise or human resources) that make it difficult to produce proactive and independent aid policies; or whether the government culturally or ideologically prefers to set its aid targets high in the policy-making process (Miyashita 1999, 727). 5.3.2 Actors: Receptive State Bureaucracy Traditionally, and to date, the president of Korea exerts penetrating leverage over both domestic and international policy areas (Hahm and Plein 1995). During the Cold War, Seoul’s diplomacy was centered on national security, due to the inter-Korean military tension. However, post-Cold War foreign policy was transformed to address the 107 issues of globalization by expanding economic cooperation with the rest of the world. The more widely foreign affairs issues are broadened, the more technical issues, as opposed to high-profile/traditional foreign affair agendas, come under the auspices of state bureaucrats. Korea’s state bureaucracy during the Cold War era was considered distinctive in terms of educational qualification50, recruitment and promotion processes, and positional turnover in comparison with those cultivated in other East Asian industrialized economies (Woo-Cumings 1999; Kang 2002; Wong 2011)51. They worked in a highly competitive and hierarchical environment, under strong presidential leadership. Foreign aid policy hardly attracted the keen interest of either the president or the political parties in Korea. Given that Korea’s national development model was profoundly growth-oriented, redistributive policy discourses were marginalized among political leaderships and policy-makers during the industrialization period. Among state bureaucrats, foreign aid giving was limitedly legitimized only because Korea itself benefited from foreign help. Uniquely, the fear of losing face among state bureaucrats was a key motive for providing ODA, and this notion was well-shared across ministries from the 1960s to the 80s. However, while the government went through the process of dismantling the traditional structures in the face of globalization, state bureaucrats became significantly divided and competitive about taking the lead in the foreign aid discourse. State bureaucrats increasingly accepted the neo-liberal version of capitalism, favoring price stability over growth, and championing reduced state intervention. A U.S.-educated economist, Jae-ik Kim, the former Director-General of the EPB (1976-80) and the Senior Secretary to the President for Economic Policy (1980-83), was an example of this change among economic bureaucrats (Evans 1995). Indeed, such ideas affected all corners of the international economic policy fields, including aid policy.

50 The higher civil servants in the Korean economic ministries (e.g. the finance, the EPB, and the construction ministries) were all highly educated personnel: Out of 314, 53.8 percent of the state officials graduated from Seoul National University and the graduates from top-tier universities (SNU, Yonsei, and Korea University) account for 73.9 percent in 1989 (Kang 2002). 51 Bureaucracy is one of the institutions that help explain the success of Korea’s state-directed developmental model. Kang (2002) compares two Asian countries, South Korea and the Philippines, in terms of how state bureaucracy in varied configurations (e.g. educational qualifications and relationships with rulers, etc.) might explain the trajectories of national development and corruption vulnerability. Kang highlights the high educational qualifications of higher civil servants in Korean economic ministries (p.57) and cabinet turnover under the Rhee and Park administration (p.67). The patterns of recruitment and promotion of higher civil servants (p.68-71) were important in understanding the state bureaucracy in Korea during the Cold War period. Wong (2011) also argues that state bureaucracy’s discretionary policy-making to promote biotechnology in Singapore, Taiwan, and Korea was unique in explaining the innovative action that was taken by East Asian governments. 108 5.3.3 Structure: Fragmented, Discretion-based Aid Policy-Making Environment Previously, the now-defunct Economic Planning Board (EPB)—independently empowered to design five-year national economic development plans and to monitor policy implementation (1963-94)—was the single planner, coordinator, and budgeting body for all ODA affairs. Before the 1990s, other ministries—the Ministry of Science and Technology, the Ministry of Labor, and the Ministry of Construction and Transport— performed as occasional ODA distributors. They provided small-scale technical assistance on an ad hoc basis. Until the late 1980s, none of those ministries developed their leadership or specialization in the ODA policy field. The establishment of the ODA-specialized agencies, EDCF and KOICA, eventually strengthened the legitimacy of the Ministry of Foreign Affairs and Trade (MOFAT)52 and the Ministry of Finance and Economy (MFE)53 to take the lead in the ODA policy sphere. Thereafter, once the single aid manager, the EPB, was phased out by a structural reform that aimed at making a smaller, more efficient government, it gave rise to a dualistic power structure of aid administration. Since then, the Korean aid administration has evolved into a distinctive structure that is conducive to achieving high efficiency in aid expansion without fostering institutions that help to produce coordinated, coherent aid policy strategies and consensuses within the government. The historical relationship between MOFAT and the MFE—the two leading ministries in the ODA policy formation in the 1990s—was one of rivalry. As Seoul’s determination to win the diplomatic competition with North Korea weakened in the post- Cold War period, the role of MOFAT became reduced. While MOFAT attempted to strengthen its presence, the MFE emerged as a balancing power in the national ODA policy-making arena (Kim 2011). Whereas MOFAT stressed the political and diplomatic objectives of aid giving over humanitarian and economic interests, the MFE placed emphasis on measurable economic returns on ODA spending. In particular, economic bureaucrats articulated the positive functions of loans, referring to Korea’s own experience of effectively utilizing concessional loans for its national economic prosperity. 5.3.4 External Influences over Aid Policy Regarding the degree and nature of state-society embeddedness in the ODA policy process, the Korean case displayed no clear evidence that the state-business coalition provided the strongest legitimacy and momentum for aid expansion. Rather, the Korean government rarely experienced any significant levels of domestic interests or pressures

52 The foreign affairs ministry was renamed the Ministry of Foreign Affairs (MOFA). 53 As of 2015, the finance ministry was renamed the Ministry of Strategy and Finance (MOSF). 109 leveraging over its aid policy arena. This contrasts with the Japanese case where ODA provision was promoted in support of the private sector’s commercial opportunities overseas (Orr 1990; Katada 2002). The corporatist nature of Korea’s state-directed developmental model appeared very weak in the foreign development cooperation policy field. For the foreign affairs ministry, which normally drew no substantial domestic interest towards its policy stance, foreign aid could be a double-edged sword: the ministry enjoyed a great deal of autonomy over the management of foreign aid, avoiding potential public criticism or public scrutiny. At the same time, however, the foreign affairs bureaucrats had no firm political support for their operations overseas. On many occasions, the ministry rather strategically used the American reference to expand its clout against economic bureaucrats. For instance, in the 1990s, MOFAT expanded grants to newly diplomatized Asian countries, justifying the increased aid disbursement in line with the Washington doctrine. 5.3.5 Non-consensual Policy Discourse for Aid Giving and Aid Expansion Security concerns associated with North Korea predominated in this subfield of foreign policy. This view is supported by the fact that Korea’s development partnerships were concentrated in Africa and Latin America in its very early days. Although the government’s posture in systematizing ODA outflows hardly appeared persistent throughout the 1990s, aid-managing ministries and agencies started to articulate the purposes and functions of ODA in more diverse ways. With the neo-liberal wave sweeping through, the internal aid policy discourse embraced more pragmatic, commercial notions; in this regard, foreign aid was increasingly seen as a vehicle for stimulating overseas investment. In the aftermath of the Asian financial crisis of 1997, concessional loans won legitimacy over grants. It was also affected that Korea was a student of the commercially-oriented Japanese ODA model (Kim and Seddon 2005). Regarding the patterns of policy goals and agenda-setting, aid-managing ministries turn out to be traditionally overambitious in their aid planning. In 1990, EPB officials released their mid-term plan54 to increase ODA to a level that was equivalent to 0.2 percent of GNP, simply referring to the then-average level of OECD/DAC countries. MOFAT, in 1999, set a target of 0.1 percent ODA/GNP, with no specific timetable attached. A primary rationale behind the aid target was MOFAT’s rule-of-thumb in its aspiration to gain international recognition of Korea’s economic and political status (Kim

54 The ODA/GNP ratio stood at 0.02 in 1990. 110 and Seddon 2005). None of these targets was realistically driven at that time—more than a decade later, Korea’s highest ODA/GNI reached at 0.14 percent in 2012. To find a proper level of aid commitment, few serious discussions took place across the ministries. Unlike the Japanese ODA policy regime, where the MFA takes the lead in providing the overall aid strategies and coordinating tasks, the Korean case lacked policy coordination mechanisms.

5.4 The Contemporary ‘CONTEXTS’ (2000-15): Physical/Material Conditions of Aid Resources and Attributes of Aid Bureaucracy In this part of analysis, the public finance, human, and legal-regulatory contexts structured around today’s foreign aid policy-making are revealed. 5.4.1 Overall ODA Performance Apart from the failure to make credible commitments, Korea’s aid growth per se is far from mediocre. From 2000 to 2014, Korea’s ODA grew by a factor of 8.7, whereas the total ODA volume of 24 DAC member countries increased by a factor of 2.5 in the same period. In the post-global financial crisis period (2009-14), Korea stayed top-ranked in terms of aid growth rates (228 percent). Other crisis-affected DAC countries, by and large, showed sluggish aid increases, (114 percent) in general (OECD 2015). The quantitative expansion of Korean ODA, however, appears to accompany few noteworthy shifts in aid components and regional/sectoral orientations. Throughout the years from 2000 to 2014, annually, an average 75 percent of ODA was given bilaterally. Although Asia always received the largest portion of Korean ODA (52.9 percent in 2014), Africa has gradually emerged as the second largest aid-receiving region (23.8 percent in 2014) since 2006. Korean ODA has constantly been disbursed mostly in the areas of social infrastructure and services (40.7 percent in 2014)—in relation to the other sectors of economic infrastructure (34.2 percent), production (10.8 percent), and the rest (14.4 percent). Neither did a new DAC membership for Korea (2010) seem to result in a significant change to its aid policy directions, principles, or structure: Rather, it might have worked as a factor for the continued aid increase. Bilateral donors’ half-met policy rhetoric does cause concern. Such behavior is argued to work against making ODA predictable, so that aid recipients are highly likely to experience difficulties in planning their aid management (Lensink & Morrissey 2000; Bulir & Hamann 2008; Kangoye 2013). According to a recent OECD report (2015), Korea’s mid-term aid predictability accounts for 46 percent, which is considered far lower than the average of the DAC. As part of responding to the problem of unmet aid 111 commitments, the Korean government recently moderated its aid target (0.2 ODA/GNI55 by 2020, initially set at 0.25 by 2015) when releasing the second Basic Plan for International Development Cooperation (2016-2020) in 2015. 5.4.2 Characteristics of Korean ODA From 2008 to 2012, Korea showed an impressive record of having the most rapid growth rate in annual ODA disbursements (OECD 2012b). Although Korea’s overall volume of ODA remains relatively small56, its performance in annual aid growth (17.4 per cent from 2009 to 2012) is a unique one in the DAC history (e.g. Demark achieved 24.8 per cent from 1960 to 1963; New Zealand, 19.1 per cent from 1973 to 1976).

Figure 5.1 Korean ODA Growth (1988–2013)

2000 0.3 1800 0.25 1600 1400 0.2 1200 1000 0.15 800 0.1 600 400 0.05 200 0 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

ODA volume (left axis*) ODA/GNI ratio (right axis**)

(Source: Author’s computation based on the Query Wizard for International Development Statistics (QWIDS); Accessed Sep. 4, 2014; OECD 2014a)

(Note: *Figures in USD million; **values in percent)

In 2011, Korea’s grant ratio (58.1 percent) was relatively low considering the DAC average (88.6 percent). Annual untied aid ratios are slowly improving, but stay at a higher level (51.1 percent) than any other DAC member country. More than half of the overall Korean ODA (68.4 percent) is Asia-destined. A major sectoral orientation of Korean ODA is economic infrastructure-building (36.8 percent) and this is in sharp contrast with other DAC countries, whose sectoral focuses are largely social infrastructure-building and humanitarian support.

55 ODA/GNI indicates a ratio of gross national income to the volume of official development assistance. 56 It accounts for USD 1.3 billion. The average ODA contribution for DAC member countries stands at USD 45.6 billion in 2011. 112 5.4.3 Physical/Material Conditions: Public Finance, Scale of ODA Funding, and Public Support Financing Resources for ODA Provision As of 2013, a Korean citizen annually spent more than USD 30 for Korea’s development cooperation operations overseas. In fact, Korea largely lacks legal- regulatory foundations for ODA budgeting (OECD 2012, 67). In Korea, ODA spending is secured from the overall government budget (non-voluntary tax revenues), while some other donor countries have separate budgets exclusively earmarked for ODA, or have diversified funding sources for ODA funding not solely dependent on public funds. Public Opinion Although Korean citizens are increasingly aware of the government’s ODA activities overseas, the level of public awareness and public involvement is still far lower than the OECD standards (OECD 2008, 2012). According to a 2011 survey, 52 percent of respondents say they are informed about Korean ODA provision (KIEP 2012). ‘Payback for what Korea received57’ was the strongest rationale for the early years of Korean aid giving in the 1990s and 2000s (Kim 2011). In recent years, the mentality of repaying historical debts has become a relatively weaker motivation in public perception. Rather, global peace and prosperity (28.2 percent), global citizenship (25.2 percent), and national interests (20.1 percent) are top-listed above the repaying rationales (13.4 per cent) according to recent public surveys (KIEP 2012). To a significant extent, the concept of national interest and mutual benefit in facilitating foreign aid is significantly embedded in the Korean public mindset. There are also worrying voices among the public about the rapid ODA expansion. 5.4.4 Attributes of Aid Bureaucracy Since 2000 As pointed out in the OECD peer review, Korea has constantly suffered staff shortage. (OECD 2012, 66). Staff shortage has been a prominent issue, but no proper action has ensued. For high level policy-makers at ministries, the civil servants experience frequent position rotation (remaining for less than one or two years) across ministries and state agencies. The frequency of staff rotation turns out to be relatively higher than those in other DAC member states. As participants in ODA policy-making process are frequently changed, it is less likely that the policy-makers and practitioners have a reasonable chance to gain knowledge and experience in the ODA policy domain. As for the staffing problem at implementation levels, KOICA provides an example. From 2005

57 In rough terms, Korea received an accumulated USD 12.8 billion over five decades in the aftermath of the Korean War. 113 to 2012, the ODA budget for KOICA doubled. However, the number of the staff at KOICA grew very slowly and unresponsively during this period. KOICA tends to depend heavily on contract-based specialists who would be less likely to benefit from occupational stability and continuity. Salaries for them are around 70 percent of the wage average for 250 Korean government agencies.

Table 5.3 International Development Cooperation Staff in the Korean Government (2007 and 2012) (Source: OECD 2008, 2012)

Number of Staff Number of Staff Organization In 2007 In 2012 PMO X* 12 MOFA 35 33 MOSF 22 23 KOICA, Total 213 247 (KOICA, Headquarters) (171) (165) (KOICA, Country Offices) (42) (82) EDCF, Total 65 84 (EDCF, Headquarters) (60) (73) (ECDF, Country Offices) (5) (11) TOTAL 335 399

(Note: *Data available after 2008)

5.5 The Contemporary ‘CONTEXTS’ (2000-15): Rule-in-use The Framework Act and Presidential Decree on International Development Cooperation, which defines the key activities of ODA management in a relatively succinct manner 58 , provides the primary and comprehensive legal basis for ODA management and delivery. Besides that, the following series of strategy policy papers complement the legal institution in guiding Korea’s ODA policy formation and implementation: The Strategic plan for International Development Co-operation (revised in 2010), Mid-term Strategy for Development Cooperation (2016-2020); Annual Implementation Plans (since 2011); and Country Partnership Strategy 59 (www.odakorea.go.kr). The following analysis is aimed at seeing what extent the legal

58 What is constantly pointed out in academic and policy papers is that Korea’s ODA policy and practice are far less articulated in written form than in other established donor cases (OECD 2012). 59 In 2015, 24 Country Partnership Strategy was first introduced. Among 134 partner countries, the strategy paper focuses on 11 Asian, 7 African, 2 Middle Eastern/CIS, and 4 Latin American countries. 114 and regulatory frameworks set out clear aims, priorities, objectives, and intended outcomes for the policy process. 5.5.1 Rule Configuration: By Seven Rule Type Category The ODA Framework Act and Presidential Decrees on the Frameworks are reviewed in this section to identify rule types. The rules are further examined in terms of specific functions, location of decision/corrective powers, and descriptive style. This is to see in which specific areas of policy decision the written laws leave room to be decided at the discretion of designated actors. Hence, it is to figure out to what degree of autonomy is given according to the laws. As delineated in the table below (See Table 5.4), there are several features to be observed with regard to the configuration of Korea’s ODA laws. First, Korean ODA laws have insufficient details in almost every area of rules— from boundary to information rule category—, therefore leaving greater policy areas to be decided at administrative discretion. More importantly, the greatest decision and corrective power rests with the president across the rule categories. Presidential powers can intervene and override those of supervisory ministries and Prime Minister’s Office over annual aid allocation and mid-term strategy building; this is one of the exceptional Korean features which is hard to find in other DAC donors. Third, regarding overall ODA objectives and task allocation, the laws embrace two barely compatible objectives of economic co-operation and global responsibility with no unified ODA objectives attached. Fourth, the ODA laws champion a state-centric notion of development cooperation. As for the boundary rule category, ODA laws develop a very lax form of boundary rules in that there are no legal restrictions or required qualifications imposed on prospective ODA implementing agencies. The Framework Act defines the actors involved in development cooperation in a narrow way, confining them to public agencies only, not effectively emphasizing the role of non-governmental, private actors in the policy formation and implementation processes. For the position rules, the laws function better than other rule categories in the sense that the delegation structure is clearly defined, so that supervisory and implementing agencies are assigned and their working relationships are prescribed. When it comes to authority rules, Korea ODA laws do not set strict standards for annual volume or geographical allocation of ODA, leaving the areas of policy decision mainly to the president. The scope rules are driven by a simple logic to divide the labor and resources between loans and grants. Considering the nature and content of Korean ODA objectives stipulated in the Act, the legal mechanism of how aid tasks are allocated to multiple aid agencies appears too vague to function. This leaves the authority for 115 managing unconventional modalities of ODA undefined and vulnerable to bureaucratic competition. 60 Although the aggregation rules in Korean ODA laws delineate the mandate and specific tasks of the coordinating body (CIDC), its oversight and coordination roles does not appear fully functional since proper legal power and resources for effective performance do not enforce it. In the categories of payoff rules and information rules, the Framework Act does not contain the necessary details. Other subordinate laws (the EDCF Act, the KOICA Act, and the IFI Act) specify tasks for implementing agencies, such as the EDCF, KDI, and KOICA, under the supervision of their managerial ministries. However, they are not restricted in expanding the thematic coverage at their wide discretion. What makes the system expanding itself is that, unlike other DAC member cases, the Korean laws on ODA do not put any limit on the number and qualification of prospective aid-managing agencies. Considering the timing of the ODA law enactment, the introduction of the ODA Act appears merely symbolic to the domestic and international public. Also, one of the strongest rationales behind the sudden enactment of ODA laws might be for Korea’s entry to the DAC—to gain international recognition as a successful aid model and an influential economy. Although the ODA legal framework is still evolving slowly, the widely recognized problems of missing integrative objectives, insufficient co-ordination, and fragmentation remain problematic.

60 For instance, technical co-operation, including knowledge aid, this type of ODA activity is wide open to existing/prospective aid agencies for competition. This loose division-of-labor clause has been a source of controversies among ministries and aid agencies. Although technical assistance is largely categorized as grants in foreign cases, some inspired organizations have attempted to launch technical assistance projects based on their interpretation that technical co-operation is a third type of ODA that is neither grant nor loan, and thereby, no legal restrictions or requirements are applied—according to interviews, news articles, and official releases from MOSF and the National Assembly. 116

Table 5.4 Rule Configuration of ODA Laws in Korea (by author)

Location of Legal-Regulatory Category Functions Decision/Corre Content Reference ctive Power × Actors of ‘international development cooperation’ are limitedly defined to the State, local Define participants for ODA F/A* 2-1/2-8 governments or public institutions President policy-making [Definition] × Implementing agencies are central administrative agencies, local governments, and public institutions Boundary None for overall Rules Limit/control number of ODA management participants and define and practice × KOICA’s Executives shall be persons with Korean nationality, without declared bankruptcy,

required conditions for KOICA Act 9 court judgement, and are not under adult/limited guardianships. participants to enter or leave [Disqualification]

EDCF Act 4 [Financial × Listed financial resources for EDCF: contributions by the Government and public Resources] institutions, contributions from other government funds; deposits from the Public Capital EDCF Act 9 Management Fund, and profits from operations [Operation] × MOSF operates and manages EDCF; the ministry may entrust the Export-Import Bank of Create types of position (e.g. KOICA Act 3/4 Korea (as prescribed by Presidential Decree) (2011) quasi-state agency) and President [Legal × KOICA shall be a juristic person (2010); registered with the registry having jurisdiction over allocated budget Personality/Establis its business (2010) Position hment] × Listed financial resources for KOICA: contributions by the Government and persons; Rules KOICA Act 15/16 borrowings from foreign organizations, and other revenues (2016); The Government shall [Operation] grant optional contributions within budgetary limits by Presidential Decree (2010).

F/A 10 [Supervising × Supervising agencies (MOSF, MOFA) prepare basic plans; preforms what the Committee Identify delegation and Agencies] assign (2011) assign managerial and F/A 20 [Delegation] × KOICA has 15 directors (prescribed by Presidential Decree, appointed by MOFA on the subordinate positions KOICA Act 8 recommendation of the President), including one Chairman; Chairman appointed by the [Executives] President on the recommendation of MOFA; auditor appointed by MOFA (2013)

117

Table 5.4 Rule Configuration of ODA Policy Formation and Implementation in Korea [Continued]

× Listed objectives: poverty reduction in developing countries; human rights betterment for women, children, and the disables; gender equality; sustainable development; F/A 3 [Objectives] humanitarianism; economic cooperation between Korea and partner countries; peace and Define overall ODA objectives F/A 4 [Basic Principles] prosperity × Listed principles: respect for U.N. Charter; support self-help; respect for development necessity; increase knowledge sharing; promotion of multilateral cooperation Define standards for ODA allocations in terms of volume, Authority None destinations, modalities, Rules conditionality, etc. × Expended purposes of EDCF: loans and cooperation projects, extending forms of loans Preside (prerequisite investigation, importation, loan-supporting projects allowed by Presidential Specify organization-specific EDCF Act 7 [Use of nt Decree, compensation), etc. (amended in 2011/16) objectives and designated tasks Fund] (>MOS × Listed project types allowed: invitation of trainees, dispatch of professionals, dispatch of (+Prescribe actions that must KOICA Act 7 [Projects] F, Korea Overseas Volunteers, research in development, disaster relief, assistance with not be taken) MOFA) goods, funds & facilities, international cooperation, government-entrusted projects, education, training, and publicity (2010) × Supervising agencies prepare and submit drafts to the Committee; the Committee Preside integrates (amended in 2014); the contents may be determined by the President (2011); Scope Delimit the range of possible F/A 8 [Basic Plan] nt, the Committee may revise plans within five years (2014); other necessary matters and Rules outcomes (Planning) CIDC procedures shall be prescribed by Presidential Decree (2011) × Relevant procedures shall be prescribed by Presidential Decree (2011) 118

× Each implementing agency submits implantation plan to supervising agency; supervisory F/A 11 [Implementation MOSF, agencies may change the plan before submitting to the Committee (amended in 2011); Plan] MOFA based on the submission, the Committee formulates and reports annual comprehensive plans to the National Assembly (2014) Lay division of labor among F/A 9 [Supervising × Loans supervised by MOSF; Grants supervised by MOFA (amended in 2013); ministries and implementing Agencies] Cooperation with IFIs in charge of MOSF; other international organizations in charge of agencies MOFA (2013) Define required chains of events or actions over potential None outcomes

119

Table 5.4 Rule Configuration of ODA Policy Formation and Implementation in Korea [Continued]

No further than Define occasions that planning and require unanimous or evaluation under majority agreements CIDC Aggregati × Established under the control of the Prime Minister (amended in 2014); the Committee on Rules Assign a coordinating body members do not exceed 25; ministerial members determined by Presidential Decree; the Describe procedures, F/A 7 [Coordinating President Minister of the Office for Government Policy Coordination shall be the secretary (2013); makeup, due process of body] the Committee decides on basic plans and annual comprehensive implementation plans; coordination a working committee established (2013) × The Committee prepares guidelines; implementing agencies submit self-evaluation plans Set oversight/evaluation President to the Committee (amended in 2013); External experts at implementing agencies (2014); F/A 13 [Evaluation] systems (>CIDC) The Committee reporting to the National Assembly (2014); Contents and release of evaluation results shall be prescribed by Presidential Decree (2014) None for overall aid Payoff Define rewards and management and × KOICA’s employees shall be dismissed by Chairman (2010/12); Korea’s employees who Rules sanctions with regard to practice divulge confidential information shall be punished by imprisonment with labor for not more particular actions or KOICA Act 13/30 than two years or a fine not exceeding 20 million won (2015) outcomes [Dismissal/ Penalty] Define executors and President enforcement tools × Implementing agency to report annual statistics to its supervisory agency; supervisory Control information agencies to the Committee; the Committee reporting upon the request of the National F/A 18 [Statistics] management (reporting & CIDC Assembly F/A 16 [Training] Informatio personnel training) × The State’s endeavor to train specialized human resources; but no allocated budget, no n Rules organization in charge specified (amended in 2014) Assign a collector and manager for government- None wide information 120

Specify type and areas of information collected and None shared *F/A refers to the Framework Act on International Development Cooperation; **P/D refers to the Presidential Decrees on the Framework Act) (Note: Legal references are accessible at www.law.go.kr; regulatory references are accessible at www.odakorea.go.kr) 121 5.5.2 The Structure of ODA Administration and Coordination Mechanisms Drawing on the review of the legal foundations, Korea has a highly centralized aid policy system, but, ironically, coordination remains a challenge (OECD 2012, p.47). In principle, Korea’s global development cooperation is overseen by the Prime Minister’s Office (PMO), MOFA, and MOSF. However, MOFA and MOSF are the acting coordinators in charge of inter-ministerial co-ordination. The two ministerial managers of aid-giving delivery have their own line of implementing agencies. This dualistic structure is likely accountable for a high level of fragmentation within the aid administration. As a result, Korean aid is delivered in a most fragmented way among DAC members in that aid resources are thinly spread out, largely in very small projects across 79 countries (OECD 2012). The average size of Korea’s projects is USD 1 million for grants and USD 30 million for loans, according to the OECD peer review (2012). As of 2011, over 32 ministries, agencies, and local governments were involved in ODA activities, carrying out 1073 projects (OECD 2012; see also ODAWatch 2012). As previously mentioned, Korea’s bilateral ODA activities are co-ordinated by and heavily concentrated on two ministries (a combined 88 per cent of total ODA were administered by MOFA and MOSF in 2011; EDCF implemented 41.8 percent and KOICA 40.8 percent). While MOFA–KOICA and MOSF–EDCF–KDI operate grants and loans, respectively, other ministries and aid agencies also manage grant and technical assistance projects. A designated budget distributor in the Korean ODA system is the Budget Office of MOSF. MOSF is also responsible for loan-based aid policy-making and overseeing its subordinating agencies. EDCF is responsible for collecting and managing ODA statistics. However, MOFA largely influences grant aid policy affairs, but has a limited leverage over aid budgeting and planning (OECD 2012). MOFA, on the other hand, performs as a focal point between ODA-requesting foreign governments and the Korean government, through its embassies. Against the strong separatism between the loan and grant-aid policy communities within the government, the role of CIDC was examined previously. For the period from 2006 to 2010—before legal recognition of CIDC in the Framework Act (Article 7.2, 2010)—, the committee appeared to exercise no agenda-setting or monitoring power. The limited autonomy of the committee can be demonstrated through its main performance of setting up cross-government annual implementation plans. From 2006 to 2009, the Annual Implementation Plans were prepared in the rather passive way of reviewing and 122 publishing a collection of individual implementation plans that were submitted by ministries and implementing agencies. 5.5.3 Partial Amendments of Framework Act and Presidential Decrees The subsequent amendments61 to Presidential (Enforcement) Decrees and ODA Framework Act have been made almost every year since 2011 (See Table 5.5). The purposes and scopes of amendments are varied. One of the significant changes is that the CIDC has been gradually empowered, which allows the agency to better perform its coordination and monitoring functions. An array of documents62 from the CIDC (i.e. Annual Implementation Plan and Committee Meeting Agendas) provides the evidence. The modifications made in 2012, 2013, and 2017 were to change the names of state organizations after regime change. Apart from those trivial changes, some meaning changes were made in 2014 regarding the strengthened power of the CIDC. Further details for ODA objectives, task identification, self-evaluation, and penalty provisions were added in 2013, 2014, 2015, and 2016. One significant change would be the removal of a special fund (called Contributions for International Antipoverty) for KOICA in 2016.

Table 5.5 Partial Amendments of ODA Laws (2010-17) (by author)

Year of Laws Amended Clauses Nature of Amendment Amendment Article 3 [Basic ideas and 2013 Added wording (human rights for the disabled) objectives] Added wording (Promotion of transparency in Art. 5 [Obligations of State] 2014 ODA management) Reflecting changed names of participatory 2013 ministries Art. 7 [CIDC] Additional resolution power given to CIDC 2014 along with its existing authority in review and Framework coordination Act 2011 Add Sectoral Basic Plans Art. 8 [Formulation of Basic Added wording (Transparency) Plans] 2014 Additional resolution power given to CIDC Art. 9/10 [Supervisory Reflecting changed names of participatory 2013 ministries] ministries Art. 11 [Formulation of Deleting a phrase that implementing agencies Annual Implementation 2011 shall change plans if planned projects Plans] overlapped with those of other agencies.

61 One of the main reasons for a series of amendments is the strengthening of the CIDC’s decision- making power that is required for its effective and proper policy coordination and supervision (MOLEG 2016) [in English]. 62 Content analysis was conducted on the CIDC’s Annual Implementation Plans (2010, 2011, 2012, 2013, 2014, 2015). Those documents, collections of individual agencies’ annual implementation plans, are prepared for the committee meeting and reviewed for the approval of the CIDC. 123

2014 Additional resolution power given to CIDC

Additional self-evaluation for implementing 2013 agencies and (Reporting to the Committee) Art. 13 [Evaluation] Additional self-evaluation for implementing 2014 agencies (Employment of external evaluation experts) Remove ‘evaluation outcomes’ from public Art. 15 [Public awareness] 2014 releases Art. 16 [Training human 2014 Add elaborations resources] Reflecting changed names of participatory 2013 ministries (of Framework Act) Art. 2 Reflecting changed names of participatory 2014 [Composition of Committee] ministries Reflecting changed names of participatory 2017 ministries (of Framework Act) Art. 4 Reflecting changed names of participatory 2013 [Working Committee] ministries Changes to the ODA scale, selection of (of Framework Act) Art. 8/9 strategic partner countries, other important [Formulation of Basic Plan] 2011 matters shall be decided by Presidential

Decree (of EDCF Act) Comprehensive 2012 Slight changes in wording and elaborations amendments made Enforcement Reflecting changed names of participatory 2013 Decrees (of EDCF Act) Art. 5 ministries [Composition of Committee] Reflecting changed names of participatory 2017 ministries Reflecting changed names of participatory (of EDCF Act) Art. 11 2013 ministries [Procedures for fund Reflecting changed names of participatory request] 2017 ministries (of EDCF Act) Art. 20 Additional elaboration of tasks of MOSF [Operative/Managerial 2016 minister in preparation of Fund Management regulations] Guidelines Reflecting changed names of participatory (of KOICA Act) Art. 2/3 2014 ministries [Executives, Dispatch of Reflecting changed names of participatory Officials] 2017 ministries (of KOICA Act) Art. 10 2016 Delete whole clauses (2007-16) [Antipoverty Fund] 2011 Slight changes in wording Added a phrase, ‘Compensating for any loss Art. 7 [Use of Fund] 2016 incurred from loans extended by the Export- EDCF Act Import Bank of Korea to developing countries.’ Art. 8 [Requirements for 2011 Slight changes in wording Assistance from Fund] 2016 Slight changes in wording Reflecting changed names of participatory KOICA Act Art.2 [Definition] 2013 ministries 124

Art.5 [Establishment of Reflecting changed names of participatory 2013 offices] ministries Art. 6 [Articles of Reflecting changed names of participatory 2013 association] ministries Add a phrase that at least one of civilian members shall have a wealth of experience or Art. 8/9/10/11 [Executive 2013 knowledge regarding development officers] cooperation 2014 Slight changes in wording Reflecting changed names of participatory 2013 Art. 13/15/18 [Operating ministries funds; borrowing of funds] Remove a special funding resource 2016 (‘Contributions for International Antipoverty’) Art. 20/21/23/24 [Approval of Reflecting changed names of participatory 2013 business plan; supervision] ministries Reflecting changed names of participatory Art. 30/31 [Penalty 2013 ministries provisions; administrative Raise amount of fine (from KRW 5 to 20 fines] 2015 million) Add three more IFIs (Inter-American Development Bank, Inter-American 2005 Art. 2 [Investment-related Investment Corporation, Multilateral institutions] Investment Fund) IFI Act Add one more IFI (Asian Infrastructure 2015 Investment Bank) Delete a phrase that ‘Necessary matters shall Art. 7 Deleted 2005 be decided by Presidential Decree’

5.5.4 ODA Planning As far as the annual planning and implementation records of Korean ODA are concerned, as shown in Table 5.6, from 2006 to 2016, there is clear and strong tendency of most of state organizations aiming high for their next year ODA targets.

Table 5.6 Review of Annual ODA Strategy Papers: Chronic Gaps between ODA Plans/Budgets [A] and Actual Performance/Executed ODA [B] (2006-16)

Bilateral ODA Multilateral ODA Bilateral + ODA/GNI, Total, Year Multilateral ODA, MOSF MOFA MOSF MOFA Percent KRW 0.1 KRW 0.1 billion (EDCF) (KOICA) BOK KOICA billion 3315 2302 667 2253 [A] 6290 0.083 1124 (2600) (2041) N/A 0 2006 1427* 2377 530* 324* [B] 4823 0.05 1085 (1361) (1847) N/A 0 3023 2708 1482 340 [A] 7971 0.08 2263 (3000) (2138) N/A 94 2007 1561 2718 1454 340 [B] 6460* 0.072 2189 (1553) (2138) N/A 94 125

3590 3428 2288 329 [A] 10085 0.107 3092 (3565) (2736) N/A 174 2008 3877 [B] 9329 0.086 2371** 3080 N/A N/A (2847) 4418 367 320 [A] 10860 0.11 3500** 2915 (3572) 1955 N/A 2009 4507 N/A N/A [B] 10941 0.11* 3200** N/A (3661) N/A N/A 5205 565 203 [A] 13411 0.13 4700** 3506 (4318) 1955 N/A 2010 5826 [B] 12481 0.12 4107** 1787 N/A N/A (4955) 6157 4843 3600 1025 [A] 17181 N/A 5305 (6047) (4723) N/A 231 2011 N/A N/A [B] 16000* 0.12* 5292** N/A N/A N/A N/A 6561 5391 3946 [A] 18638 0.15 5408 785** 2012 (6367) (5158) N/A [B] 18600 0.14 5600** 5200** 5827 N/A N/A 7076 5690 3982 [A] 20411 0.16 6156 1489** 2013 (6926) (5423) N/A [B] 20411* 0.13* N/A N/A N/A N/A N/A [A] 22666 0.16 7887** 5993** 7243 980 2394 2014 [B] 19552 0.13 5393** N/A 4854 N/A N/A 8774 6339 956 1576 [A] 23782 0.15 6956 2015 (8647) (5916) N/A 414 [B] 21619 N/A N/A N/A 5122 N/A N/A 9358 7787 937 776 [A] 24394*** N/A 2425*** 2016 (9239) (6070) N/A N/A [B] N/A N/A N/A N/A N/A N/A N/A

(Note: This table is restructured based on information provided from a series of CIDC’s annually

published ODA strategy papers (2005-17); 국제개발협력종합계획 [in Korean]); Accessed at

http://www.odakorea.go.kr) (*Tentative or estimated figures; **Combined; ***The sum excluding KRW 1700 billion (new subscription to AIIB))

Table 5.7 Mid-Term Planning of Korean ODA (Source: http://www.odakorea.go.kr)

Title of Agendas Target Performance Document First Basic Plan ODA/GNI × Integrating ODA system No. of coordinated (2011-15) By 2012 0.15 × Bottom-up, Ex post projects across grants (Published in By 2015 0.25 Coordination and loans (2013-16) 2008; Approved Bi: × Constant ODA scale-ups 47, 93, 171, 134 at the 4th CIDC Multi=75:25 × Evaluation system set-up Untied ratios (2011-14) meeting) Untied ratio 126

× Institutionalization of civil (46/68; 47/81; 47/89; group participation 48/92) × ODA awareness

Second Basic × Integrating ODA system ODA/GNI Plan × Top-down Strategy making By 2020 0.20 (2016-20) × ODA quality enhancement By 2030 0.30 × Improving evaluation Bi: N.A. (Published in system Multi=75:25 2015; Approved × Encouraging civil Untied ratio nd at the 22 CIDC participation By 2020 meeting) × Global citizenship (55/95)

5.6 Conclusion Why Korea has been so eager to increase foreign aid and how the country has justified its aid increases? This chapter helps explain the commitment problem from the historical and contemporary contexts of aid policy-making. Notwithstanding its impressive aid growth over the last decade, the country has constantly failed to meet its annual commitment by a significant margin. It is argued that such policy behavior might be explained from its historical contexts, in which the country’s nascent aid policy administration produced expansionary but non-strategic rhetoric, due to its fragmented structure and lack of indigenous policy rationales during the Cold War period. Such traits of the policy system linger today, thereby continuously favoring overestimated aid targets and outward-looking aid initiatives.

127 CHAPTER SIX ANALYSIS OF ‘ACTION SITUATION’ IN KNOWLEDGE AID MANAGEMENT (2000-16)

6.1 Overview of the Chapter 6.1.1 Observing Bureaucratic Behavior in an ‘Action Situation’ of Knowledge Aid Planning and Management The following analysis singles out an intra-governmental policy-making arena for knowledge aid management and delivery. One of the rationales for the nested case selection is that the process of knowledge aid policy-making and operation has the most comprehensive scope of sub-state organizations within the Korean government. Given that the Korean ODA system is rationalized by a loan-grant division 63 (Further see Section 5.5), the selection of undefined sub-policy area (technical assistance) allows us to better understand the incentives and behaviors of dominant policy actors across the system and is optimized to figure out the patterns of government-wide (or system-wise) collective actions. In the hypothesized sequences of policy-making logic and actions—from context, incentive, interaction, and collective action, to policy output—key locations for the observation and interpretation of bureaucratic behavior is four-fold: First, the following analysis focuses on how hard and soft rules interactively come into play—presumably, hard rules delimit the working extent of soft rules, and then determine the level and areas of bureaucratic autonomy—; after that, the analysis highlights to what extent the ideas of bounded rationality and delegation theory explain observed bureaucratic behaviors that are induced under a configured set of hard and soft rules; the third analytical task is to discern the types of collective action (cooperation vs. non-cooperation or hostility) and to evaluate the effects of the collective actions on long-term aid policy outputs; and finally, the analysis offers some lessons on the impacts and limitations of the previously adopted institutional adjustment (i.e. modifications of formal rules and organizational structure) and on what institutional change might alleviate the degree of the commitment problem in future (See Chapter Eight). 6.1.2 Proliferation of Knowledge Aid in the Korean Government

63 As identified in several sources that compare aid systems across DAC countries (LLC 2012; OECD 1999, 2009), Korea is found to have a distinctive organizational-administrative structure that dualizes its ODA policy formation and delivery by loans and grants. 128 Knowledge ODA64, one of the aid modalities, is an umbrella term that embraces a range of technical assistance projects that aim to disseminate Korean industrialization experience to the developing world. Since the early 2000s, knowledge aid has emerged as one of the fast-growing aid sectors in Korea (OECD 2012). The knowledge-sharing (KS) initiatives constitute a controversial policy area, primarily because KS is an open-ended sub-policy field in which no exclusive mandates or tasks of policy formation and implementation are legally assigned to particular organizations. Neither does the aid policy community have any consensual norms or traditions of division of labor (or division of resource) regarding the management of KS. Knowledge aid is often re-interpreted as the third type of foreign aid6566 among ODA commentators and state officials. MOFA and MOSF have competitively put forward efforts to justify their exclusive autonomy over the ODA type on the basis of their respective legal interpretations and operational relevance. As a matter of fact, knowledge-sharing becomes a wide opportunity window for governmental organizations; not only for existing aid-managing ministries and implementing agencies, but also for local authorities and other public agencies that were previously not familiar with development cooperation. For policy-makers at a higher level, they have a strong interest in expanding such a unique type of development cooperation, due to its high visibility to domestic and international publics. From the budgetary and technical angles, for most state agencies that struggle to secure more resources for their survival and expansion, knowledge projects appear a relatively accessible budgetary option; since there are no strict legal restrictions or qualifications— in terms of specialization or required initial conditions for human and material resources—imposed on prospective implementing bodies. Speaking of what specific knowledge is intended to be shared, and how to deliver it, interpretations and delivery methods of Korean expertise are varied, with little consensus across the Korean aid community. Knowledge aid epitomizes a unique gesture of emerging donors that attempt to export their previous experience as former aid

64 This term is used interchangeably with ‘knowledge-sharing (KS)’, knowledge aid, or ‘development consulting’ in the Korean aid community. 65 Relevant comments frequently appear in news articles, interviews with MOSF officials, and official announcements from several ODA-involving ministries; with reference to a news article of A-ju Economy News (Published 17 January 2013; Available at http://www.ajunews.com/common/redirect.jsp?newsId=20130117000532 [in Korean]) 66 It is still debatable whether the knowledge-sharing activities should be considered a uncategorized type of ODA or a sub-type of grants. Article 9 [Agencies Supervising International Development Cooperation] in the Framework Act articulates that all government ODA activities are divided according to the two types of ODA (loans and grants) only, and the supervision of each type of ODA is carried out by MOSF and MOFA respectively. The rest of aid modalities are left largely undefined. 129 recipients to other developing partner countries. On the other hand, what is observed within the Korean government is that such initiatives appear to be proposed and managed by opportunistic state actors in a most expansionary but fragmented and inconsistent manner.

6.2 Knowledge Aid Practices in the Korean Government 6.2.1 Typology of Knowledge Intervention Prior to revealing the observations on the action situations associated with knowledge aid, this section lays out the scope and types of knowledge aid practices in Korea. The idea of sharing the national experience of overcoming extreme poverty and accelerating economic growth gained importance in the very early period of Korean ODA disbursement. Knowledge-sharing has been widely seen as a unique field of contribution among Korean bureaucrats. The two ministries, MOSF and MOFA, have promoted Knowledge-sharing Program and Development Experience Exchange Partnership, respectively. In recent years, other ministries and local authorities—e.g. PMO, the Ministry of Public Administration and Security, and Rural Development Administration- —have also launched similar types of knowledge aid by revisiting the value of Korea’s historical social movement that arose during the industrialization period in the 1960s and 70s. A variety of development consulting activities may be categorized into three distinctive modules67: knowledge-sharing programs, development experience exchange programs, and Saemaul ODA. 6.2.2 Knowledge Sharing Program (KSP) KSP has shown a noteworthy expansion in the volume and scope of partner countries since its debut in 2004. As a public sector-oriented technical co-operation program, its budget has increased over 20 times over a decade (2004-15). As of 2013, 33 partner countries have participated, and 440 development subjects have been covered (MOSF 2012). This program is mainly intended to consult with partner governments on strategically selected areas of development policy and to provide capacity-building training seminars and workshops to high-ranking governmental officials of the partner countries. KSP is composed of three main activities, including in-donor and overseas operations: (i) systemization and modularization of state-led experience of industrialization; (ii) bilateral policy consultation between the Korean government and

67 The categorization is suggested for the purpose of this study. It is neither officially recognized nor confirmed by the Korean government at the time of writing and publication. 130 counterpart governments; and (iii) cooperation with international organizations (KDI and WBI 2011; Lim 2014).

Table 6.1 Knowledge-sharing Program: A Rapid Growth (2004-13)

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Annual budget 0.94 0.94 0.71 1.06 1.62 3.64 7.13 10.38 18.64 24.95 (USD million) Partner country 2 2 4 5 8 11 17 26 33 — (Number)

(Note: All budget figures converted to USD at the exchange rate of KRW 1000 = USD 0.97)

(Source: MOSF 2014)

With the intention to function as a platform to promote peer-to-peer learning across governments, KDI occasionally dispatches its delegates consisting of former policy-makers, business leaders 68 , and academics. Judging from the geographical destinations of KSP, the projects have not taken place only in low-income countries, but have also included high-income countries, such as the UAE, Saudi Arabia, and Kuwait. The projects are mobilized mostly on an annual basis. Thematically, major themes for KSP are export promotion, finance-sector institutionalization, and capacity building for governmental officials in foreign governments.

Table 6.2 Regional and Thematic Allocations of KSP (2004-12)

South Asia Middle East Africa Europe America No. of Partner 12 5 10 9 3 Countries Economic development and growth (50), Industrial organization (48), Marco-economics and financial economics (44), Agriculture, environment Themes* and resources (28), International economics (27), Public, health, and (No. of Project) welfare economics (23), Labour and education (20), Urban, rural, and regional economics (18), etc.

(Source: Kim and Tcha 2012)

(Note: *Each KSP project was intended to touch upon multiple themes of national development. Therefore, the number of project in each thematic category is counted in a repetitive manner.)

68 According to the interview with the KDI staff, the presence of private-sector figures in the inter- governmental dialog is crucial. KSP also aims to encourage vertical interactions within partner countries so that local interest groups, including concerned businesses, scholars, and experts, are invited. 131

Due to the involvement of high-ranking government officials and high-profile business figures, KSP has enjoyed more publicity in the media than any other ODA operation. The knowledge project has often coincided with diplomatic events. According to the interviews with the KDI staff, KSP is expected to trigger opportunities for larger- scale development loan projects of EDCF. 6.2.3 Development Experience Exchange Program (DEEP) DEEP was invented as a counterbalancing initiative of KOICA against the fast- growing KSP. DEEP was introduced in 2012. The likely intention behind the creation of DEEP was for KOICA to re-recognize its previous technical assistance projects that are relevant to the theme of sharing Korean knowledge. It was statistically reorganizing and re-labelling KOICA’s previous technical cooperation projects, but not so as to initiate a new type of knowledge intervention, as far as this study concludes from available sources. DEEP includes a number of KOICA’s complete projects that support economic and social development of partner countries in the forms of policy consultation and development research. KOICA identifies mainly four types of knowledge projects: (i) policy and technical consultation; (ii) institution-building; (iii) infrastructure-related technical co-operation; and (iv) human capacity building (KOICA 2013). However, the definitions and categorization of knowledge intervention that KOICA adopts are fully acknowledged neither within the Korean aid administration and nor by MOFA.

Table 6.3 Development Experience Exchange Programs at KOICA (2003-12)

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

KOICA Overall 147 203 215 184 251 305 355 524 451 497 Budget

DEEP Budget 2.7 6.1 6.2 5.9 7.6 19.2 12.3 21.5 14.8 27.2 DEEP Share 1.8 3.0 2.8 3.2 3.0 6.3 3.4 4.1 3.2 5.4 (Percent)

No. of Projects 2 11 5 7 17 12 8 14 17 12

(Note: All budget figures in KRW billion)

(Source: KOICA 2013)

From 2008 to 2012, 63 projects were re-labeled DEEP, accounting for 4.5 per cent of KOICA’s overall budget. Geographically, DEEP projects are concentrated in the Asia Pacific region (62.7 per cent) and strategic partner countries are Indonesia, Vietnam, 132 Cambodia, Azerbaijan, and the Philippines. Areas of focus are industrial energy (57.2 per cent), public administration (14.8 per cent), and education (10 per cent). DEEP projects are carried out largely in the form of preliminary research activities (i.e. feasibility studies) in alignment with on-going industrial infrastructure projects. DEEP was introduced when KOICA internally revisited the importance of technical co-operation for institutional/human capacity development, according to KOICA’s internal study reports (KOICA 2012, 2013). However, in contrast with its strategic emphasis on sharing Korean experience, KOICA has cultivated little discourse of Korean-style social development. This missing discourse on the essence of the Korean development experience in DEEP policy-making and operations might explain its opportunistic behavior in managing knowledge aid. 6.2.4 Sae-ma-ul ODA Sae-ma-ul ODA has been implemented in a sporadic way, by a wide range of central and local government organizations. The Korean term, ‘Sae-ma-ul’, comes from the name of the historical government-led rural development campaign, Saemaul Undong (translated new village movement). When the authoritarian Park Chung Hee government (1963-79) implemented the development campaign throughout the 1970s, this movement was credited for rapidly increased rural income and agricultural productivity. In fact, the historical Saemaul Undong movement has been evaluated in terms of its direct and indirect impact on national development (Moore 1984). Despite the international and domestic criticisms of Park’s iron-handed way of mobilizing national economic development, the World Bank (1993) highlights the state’s distinctive role in supplying late-industrialization in Korea in comparison with other failing country cases. Since then, there have been national efforts to revisit and capitalize on the positive lessons of the Saemaul Undong as a component of Korean ODA. Globalization of Saemaul Undong values has become one of the top agenda items in the ODA Strategic Plan (PMO 2012) and the Saemaul ODA task force team was established in 2013. Saemaul ODA is agricultural/rural development-focused and designed to share the ingredients of Korea’s successful rural development model with local communities in partner countries. The following organizations manage Saemaul aid projects: the Korea Sae-ma-ul Movement Center (KSC) in collaboration with the Ministry of Security and Public Administration (MOSPA), the Rural Development Administration (RDA), the Ministry of Agriculture, Food and Rural Affairs (MAFRA), and Gyeongsangbuk-Do Provincial Government. One of the most valued project types appears in-donor training (i.e. invitation of foreign trainees to visit Korea). KOICA also implements Saemaul ODA, 133 inviting community leaders from 131 countries for short-term agricultural training. KOICA has allocated more than US 50 million dollars from 2011 to 2015.

Table 6.4 Fragmented Practice of Saemaul ODA* in Korea

Korea Sae-ma-ul Ministry of Gyeongsangbuk- Rural Development Movement Agriculture, Food Do Provincial Administration Center** and Rural Affairs Government Year 2009–10 2007–10 2009–10 2007–10 Volume (KRW 1,890 1,110 7,806 2,570 Million) Invitation of Village Agricultural Village leaders and development development development technician for Activities In-donor training consultation projects short-term training for local leadership Multilateral co- Trainee invitation Agricultural Volunteer dispatch operation projects Volunteer dispatch technical advice Laos, Mongolia, Vietnam, Nepal, Cambodia, Myanmar, Vietnam, People’s Major Myanmar, Vietnam, Uzbekistan, Republic of China, Partner Democratic Cambodia, Laos Kenya, Paraguay, Indonesia, Countries Republic of Congo, Cambodia, Tanzania Uganda Bangladesh

(Source: Committee for International Development Cooperation (CIDC 2010), cross-checked through multiple websites) (Note: *Confined to technical co-operation on agricultural development; Excluding infrastructure- building projects; ** Mandated by the Ministry of Security and Public Administration)

Consisting of roughly 5 per cent69 of overall Korean ODA in 2012, knowledge ODA has been expanding dramatically. Recently, the Park Geun Hae administration (2013-18) released an announcement that it was doubling the Saemaul ODA budget (KRW 26.2 billion) by 2014 (201570). This announcement signifies that the political and financial opportunities for Saemaul ODA implementers are likely to remain high. However, the current expansionary trend of KS seems somewhat disturbing for the following reasons. First, there are no comprehensive policy co-ordination and monitoring mechanisms in place. Second, the insufficient capacity of the implementing agencies is

69 There is no official categorization for knowledge ODA by national and international statistics platforms. The figures have been re-calculated using data available from the websites of KDI, KOICA, MOSF, and OECD/DAC (2012) peer review report. 70 The relevant news article can be found at http://www.pressian.com/news/article.html?no=125255. 134 referred to as one of the perennial issues71 in producing low quality aid delivery (Lim and Lim 2013). Overall, KS practices have been managed and promoted in the absence of consensual discourses of the Korean development model at the policy-making and implementation levels. Rather, each module heavily depends on personal accounts (of those who are involved in knowledge projects as consultants or advisors, utilizing their past experiences in the public and private sectors) and there is little accompanying effort to systemize and digitalize the contents that Korean expertise to share. Considering Korea’s aid-receiving experience, relevant discourse and research activities are largely absent.

6.3 Action Situations at the Individual-Choice Level 6.3.1 When Respective Organizations Decide on Intra-Agency Matters: By Actors The following analysis provides observations on the respective organizational actors’ behaviors in making the following intra-organizational decisions: Redefining mandates, short-term/long-term strategy building, reallocation of human and material resources, and knowledge management. As expected, the state organizations reinterpret their mandates and set up strategies not purely by a tangible set of rules (legal-regulatory frameworks) but also by the perception and specialization each one organization has internally developed and sustained over time. The perceptions that each organization individually conceptualizes about its social mandates towards taxpayers (as the distant but fundamental principal of the government’s ODA delivery) are varied, along with its legally defined role in the first place. To increase their autonomy, the state organizations build up and execute strategies for interaction towards their direct principal and subordinate organizations. The following analysis of respective actors is presented in a top-down manner, starting from the head of government to the bottom of the intra- governmental delegation chain. This order follows the logic of laws and regulations. 6.3.2 President Presidency and Foreign Policy in Korea To a significant degree and traditionally, foreign policy decisions reflect the political orientation of the president in Korea (Lee 2009). Seoul’s diplomacy to North Korea, for instance, has been one of the most inconsistent public policy areas at the times

71 At the operational level, inadequate English communication between the staffs from implementing agencies and their counterparts frequently becomes a critical problem that might work against the overall performance of the project. 135 of regime change. There is a clear tendency in Korea that the more one country’s public policy issues are diversified, the greater is the policy-making role in low-profile, technical policy fields given to state bureaucracy (Calder 1988; Kato 1989; Kim 1993; Lee 1999). Legally speaking72, the president is given exceptional discretionary power to intervene the process of ODA mid-term/long-term strategy setting, in which overall ODA scale and selection of strategic partner countries are decided. With regard to the extent to which the president’s political preferences affect the ODA policy stance—in particular, the introduction and promotion of knowledge aid initiatives—the following analysis reveals that it appears to be not significant. To exemplify, when the two previous regimes (from 2003 to 2012) are compared, it is clear that although their domestic and foreign policy stances were in sharp contrast (Doucette 2009; Shin 2016), ODA strategic directions remained relatively constant (Kim and Oh 2012). Such a status quo aid policy stance over the consecutive periods can be explained in that the two governments were almost identical in terms of allocating ODA in close coordination with trade/investment policy, promoting a slogan of enhancing national prestige, and taking a conformist posture towards international norms and consensuses in the domain of development cooperation. During that period, the government equally supported the dual pillars of Korean aid giving: global responsibility and economic cooperation (Jerve and Selbervik 2009; Kim and Oh 2012). As shown in Table 6.5, no critical changes are observed in grant ratios, geographical allocation patterns, or annual aid growth rates.

Table 6.5 ODA Performance in Comparison: The Roh and Lee Administrations

Roh Administration Lee Administration

ODA/GNI 0.068 0.114 (Percent) (Averaged, 2003-07) (Averaged, 2008-12)

Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Share of Grant* 59.3 64.1 68.6 68.8 73.0 68.3 63.1 63.7 58.1 60.4 Aid (Percent) Share of CPA** 93.9 90.1 88.3 82.0 87.2 80.5 85.5 90.8 88.0. 86.6 Aid (Percent)

72 With reference to the amended Framework Act (Article 8, amended in 2011 and 2014); a comprehensive review of legal and regulatory frameworks for overall aid policy formation and implementation in Korea is previously presented in the Chapter 5. 136

Share of Tied Aid 80.5 80.8 81.3 79.4 63.5 56.7 51.5 64.3 48.9 44.5 (Percent) Share of Aid for NGO Support 1.82 0.41 0.50 1.12 0.91 1.04 0.90 0.05 0.02 0.05 (Percent)73 Share of HIPC***- Destined Aid 11.7 11.8 5.1 5.6 7.9 10.0 9.6 16.3 10.4 18.3 (Percent)

(Note: *Bilateral ODA only; ** Country Programmable Aid; ***A total of 39 highly indebted poor countries)

(Source: Author’s computation based on QWIDS; Accessed Sep. 4, 2014; OECD 2014a)

Development of the Knowledge-sharing Initiative at the Presidential Level Nevertheless, further evidence suggests that the knowledge-sharing initiative might have been facilitated with slightly varied motivations and approaches under the three administrations (2003-15). Despite the fact that there were no fundamental changes observed in terms of ODA makeup (by aid modality) or preferences towards Asian or high-/middle-income countries74 throughout the period, the interests in the sharing of Korean knowledge have grown continuously at the presidential level. Such opportunism with KS was differently labeled and incorporated into their overall diplomatic strategies to varying degrees. The idea of utilizing Korea’s industrial upgrade know-how officially emerged and was picked up as one of the top three ODA policy agenda items under the Roh administration, as articulated in the Improvement Plan for International Development Cooperation75 (2005). As part of Roh’s ODA expansion plan (with an aim to reach a GNI/ODA of 0.1 percent by 2009), the document delineates that Korea’s developmental experience is a strategic resource that allows Korea a comparative advantage in its overseas aid activities.

73 During this period, the degree of government’s openness towards domestic NGOs at the policy- making and operational levels is considered slightly divergent as well. This is an area in which Korea took a regressive step even after joining the DAC. That being said, the gap in this area is not significant enough, since the average level of NGO support ratio throughout the two regimes stayed far lower than that of the DAC. 74 The allocation pattern might reveal Korea’s economic interest-driven principles embedded in its ODA policy in practice. 75 The document was prepared by the Office for Government Policy Coordination (2005) (국제개발협력 개선방안 [in Korean]); Retrieved from Presidential Archives (archives.knowhow.or.kr); Accessed Aug. 31, 2016. 137 The knowledge initiatives gained momentum after Roh’s76 visit to Africa, which broke a 24-year absence of Korea-Africa diplomatic and economic cooperation. After announcing the Korea Initiative for African Development (2006)77 in Nigeria, Roh made a modification of Korea’s mid-term ODA strategy, prioritising a partnership with African countries over previous partner countries. Since 2007, on a regular basis, government activities for knowledge-sharing were organized and carried out through a channel of the Korea-Africa Economic Forum. In line with the earlier aid policy stance, the subsequent Lee government fine- tuned ODA policies. The government fulfilled its planned achievements, such as the entry to DAC (2010), the hosting of a G20 summit meeting (2010), and the High-Level Forum on Aid Effectiveness. From the beginning of the Lee administration, ODA was strategically facilitated to enhance Korea’s global standing. There was a notion that Korea might serve as a knowledge bridge between the global North and South. In doing so, Korea’s dual experience as aid recipient and aid donor can be effectively utilized as a unique ODA resource. According to the policy document ODA Advancement Strategy78 (2010, p.16), the promotion of Korean experience-based ODA projects remained one of top-listed ODA policy agenda items79. The establishment of special presidential committees—such as the Presidential Committee on Green Growth (2009) and the Presidential Council on National Branding (2009)—is a key example that demonstrates the growing interest of the president in knowledge-sharing practice. KSP, peacekeeping operations, and World Friend Korea (a government-funded overseas volunteer program) are the examples of the president’s growing interest in ODA (Choi and Kim 2014, IANA80 2012). 6.3.3 National Assembly The Role of Politicians in the Foreign Affair Domain in Korea

76 In general, ODA policy during the Roh administration appears relatively neutral to his forefront political slogan: It includes emphasizing Korea’s self-reliance against its traditional alliance with the United States and restructuring the Chaebol-favored Korean economy. During this period, ODA increased in scale mainly by non-systematic factors, such as the support for reconstruction in Iraq and Afghanistan (2005), new subscriptions to the Inter-American Development Bank and the International Development Association, and the United Nations (2005), and Africa–Korea economic co-operation (2007) (OECD/DAC 2008). 77 President Roh pledged to triple development aid to Africa by 2008. 78 A joint report which complies reports submitted by PMO, 13 ministries, KOICA, Eximbank, and Korea Communications Commission (개발협력 선진화 방안 [in Korean]) 79 The three strategies to advance ODA policy are (i) reinventing development knowledge to share with partner countries; (ii) reorganizing the ODA system for making it more effective; and (iii) strengthening co-operation with international organizations. 80 The report (2012) was submitted to the Parliamentary Inspection of the Administration (이명박 정부 ODA 정책 평가와 차기 정부에 대한 제언 [in Korean]). 138 In general, party politics is considered a minor factor in forming ODA policy in Korea (Chang 2014). Now that the scale of ODA81 is no longer negligible in the 2010s, parliamentarians are growingly interested in ODA budget issues. As in other donor country cases, senior politicians who have had a long career in politics and have expertise in foreign policy affairs tend to participate in debates over ODA spending. That is, professional experience and the knowledge of politicians matter significantly (Weber 1978; Yasutomo 1987). In a recent study, Chang (2014) argues that the passive role of the National Assembly—given that parties largely failed to issue a strong, consistent evaluative voice against government’s expansionary aid policies—might stem from having politicians with insufficient knowledge and experience. As a possible result, this might be strongly contributing to a long-term budgetary expansion of overall ODA activities and the high-profile ODA practice of knowledge-sharing. The parliamentary committee—the Foreign Affairs, Trade, and Unification Committee—is mandated to function as an ex post mechanism to monitor the bilateral ODA performance of the government and to evaluate the appropriateness of ODA scale. To re-confirm the argument of Chang (2014), this study conducted a comprehensive review82 of committee meeting memos (2004-16), and it reveals that proactive steps have rarely been taken during the period. Rather, the National Assembly routinely performed its approving role but not to the extent to which it works as a watchdog over the performance of ODA. Once again, this is closely linked to the fact that there are few committed/well-informed politicians and interest groups on the issues of development cooperation. The Bureaucratizing of Aid Policy-Making: When the President and the Parliament are Given Great Discretion by the Laws In light of presidential/parliamentary interests in knowledge-sharing ODA, both the roles of the president and the parliament in ODA affairs appear insignificant in Korea. As an effect, the substantial room left for their open-ended intervention, a great deal of aid policy-making role appears to be delegated to non-elected state officials and government agencies, except for the policy decisions on overall ODA volume and geographical allocation. Gradually becoming known to the Korean political leadership, ODA is primarily utilized as a tool for promoting economic cooperation—since it is a perennial, national imperative for the inherently export-dependent Korean economy. In

81 As of 2013, a ratio of ODA to GNI stands at 0.13 per cent. The 2017 government budget for ODA amounts KRW 2.72 trillion (MOSF 2016). 82 The review covers meeting minutes from the 17th National Assembly (2004-8) to the 19th (2012-16); Available at http://uft.na.go.kr/ (an official website of the Committee, National Assembly) 139 this respect, KS initiatives are instrumental in publicizing and justifying the overseas expenditure on domestic electorates and industrial interest groups. 6.3.4 The Finance Ministry (Supervisory Agency) Korea is not a unique case of bureaucratized aid policy-making. In other DAC donor states, ODA policy is often dubbed ‘the policy of public indifference and government elites’ (Otter 2003). The Korean case is rather unique in that, in the absence of strong public support for ODA or interest/civil group engagement, the country’s ODA activities have been significantly expanded over the last decade. In the contexts of low public interest and awareness, on the other hand, the aid bureaucracy tends to enjoy a relatively great level of autonomy by avoiding public scrutiny. According to the Framework Act83, MOFA is mandated to perform its supervision of all grant aid affairs, while MOSF supervises overseas loan projects. When it comes to knowledge-sharing aid activities, exceptionally, both the lead ministries have been planning, managing, and monitoring relevant activities at their discretion. MOSF’s Rationality and Opportunistic Behavior Towards Knowledge Aid As mentioned earlier, the involvement of the finance ministry in ODA affairs underlines the distinctive organizational and legal contexts for aid policy formation and practice in Korea. MOSF is not only a supervising body over the Eximbank and KDI, but also the budget setter for ODA spending. MOSF is found to have a minor or non- systematic interest in ODA, based on its budgetary scale and the allocation of manpower for the tasks of ODA policy. As a part of the International Economic Affairs Bureau, the Development Cooperation Division manages ODA affairs at MOSF. KSP is supervised by the International Development Policy Team. Reinterpretation of Mandate: Chosen Policy Discourse and Strategy The review of relevant policy documents and press releases (available at the official website of MOSF) (2008-15) reveals that MOSF has developed a distinctive policy discourse regarding development cooperation and knowledge aid. Under its exclusively mandated objective—promotion of economic cooperation with foreign economies—according to the Framework Act and presidential decrees, MOSF has fostered a specific model of international development cooperation over the years (See Figure 6.2). What is agreed among economic bureaucrats at MOSF is that knowledge aid

83 The following articles of the Framework Act articulate the division of labor and the role of MOSF and MOFA: Article 9 [Agencies Supervising International Development]; Article 10 [Roles and Functions of Agencies Supervising International Development Cooperation]; and Article 11 [Formulation of Drafts of Annual Implementation Plans for International Development Cooperation]. 140 is recognized as an apparatus for MOSF’s reoriented ODA policy where the ministry intends to expand EDCF operations overseas. In pursuit of KS initiatives, MOSF and its line agencies (EDCF-Eximbank and KDI) aim to interact with domestic businesses with emphasis on the mutual benefit of the public and private sectors (MOSF 2014). Under the international development cooperation model, KSP is expected to provide country-/sector-specific knowledge to domestic industries from the sectors of construction, information technology, communications, distribution, and business consulting. With the three-year economic development plan (2014–17) (See MOSF and KDI 2014) introduced by the Park administration, the scope of EDCF operations expands, including the services of concessional loans, government guarantees, and export financing for exporting companies: This is in order to promote public–private partnerships and to re-align ODA operations in the interest of domestic export industries. On this point, the interviews also confirm that KSP is largely expected to induce future operations of EDCF loan projects.

Figure 6.1 MOSF’s Strategy to Promote KSP and EDCF Projects: Enhancing Foreign Market Accessibility of Domestic Businesses (by author)

Evidence-Based Policy-Making?: Technical Considerations behind the Chosen Policy Discourse and Strategy To justify the promotion of industrial interest through EDCF and KSP operations, MOSF finds its references mainly from other bilateral and multilateral development institutions: For instance, U.K. Department for International Development’s Business- Friendly Development Strategy 84 (DFID 2014) was referenced to support MOSF’s

84 This document highlights the U.K.’s recently shifted the policy stance that emphasizes more private sector-driven development. This approach is suggested to make ODA sustainable in the face of the 141 commercialized policy stance (MOSF 201185). However, little evidence was found to further explain or support the theoretical and technical validity of MOSF’s development cooperation model. No concrete data are attached in terms of how many domestic companies have benefited from it or what economic impacts have been actually generated through MOSF’s dual promotion of EDCF and KSP (CIDC 2011). As revealed in the first and second ODA basic and sectoral plans (CIDC 2010, 2015), the finance ministry has constantly supported an expansionary ODA policy. At the same time, MOSF’s ODA policy has been traditionally rhetorical by setting unrealistic aid targets. This problematic pattern of ODA goal-setting was pointed out in the OECD special report and peer review (OECD 2008; OECD 2012). As previously discussed (also see Yi 2016), MOSF, as the ODA agenda and budget setter, tends to release ambitious aid targets, thereby creating its chronic and distinctive gap between ODA commitments and actual disbursement (2000-2015). Such behavior appears, rather, to be culturally and traditionally driven in the financial ministry. The latecomer mentality seems to be widely shared among bureaucrats at MOSF, according to interviews with former state officials. To elaborate, this is a bureaucratic perception that Korea, as a latecomer to the global aid giving markets, needs to find a comparative advantage of its own to compete with other advanced, established aid givers. In this respect, economic bureaucrats believe that KSP is instrumental in promulgating Korea’s own development experience. In addition, to address the low level of public awareness (MOSF 2013)86, MOSF also highlighted KSP as a strategic, cost- effective tool to publicize the country’s overseas activities towards its citizens and the international public. Capacity, Information, and Resource Within MOSF, the staff allocated to the Development Cooperation Division is identified as around 20 personnel, and this number remained almost the same even when MOSF dramatically expanded its ODA spending87 between 2008 and 2010 (OECD 2008, 2012). As reported in the DAC review, a relatively small number of personnel work in the aid policy sector in the Korean government.

U.K.’s economic hardship. Relevant agendas are included as follows: Direct financing to public infrastructure, close partnerships with national businesses and local ones in partner countries. 85 MOSF’s report (2011) submitted to CIDC. 86The document is available at http://www.mosf.go.kr/nw/nes/detailNesDtaView.do?searchBbsId1=MOSFBBS_000000000028&search NttId1=OLD_4016590&menuNo=4010100 [in Korean]. 87 MOSF experienced a drastic expansion in its loan operations: From KRW 145.4 billion (2007) to 546.7 (2009) (Source: Annual Implementation Plan reports (CIDC 2008, 2010)). 142 Apart from the number of the staff, the level of expertise with high-/mid-level officials at MOSF matters. Theoretically and empirically, it is a factor in the quality of MOSF aid policy formation and overall performance. More than anything, the frequent rotation of desks among bureaucrats is considered to be working unfavorably to the effective and coherent performance of ODA management; under the circumstance, it appears rather difficult to promote or keep a good level of ODA expertise within the ministry. Delegating Tasks to and Monitoring Subordinate Aid Agencies MOSF appears to have had a cooperative relationship with KDI in wider economic policy sectors. KDI, an independent governmental think tank in economic and industrial policy domains, was established to provide policy recommendations to economic and financial ministries in the 1970s. Although the KSP management falls under the scope of the International Development Policy Team at MOSF, the actual initiation, implementation, and evaluation of KSP projects are mostly done at the KDI. MOSF appears to exercise lax control over the KDI’s KSP operations, not significantly influencing its employment and recruitment of staff at KDI. KDI implements KSP through one of its divisions, the Center for International Development Cooperation. 6.3.5 The Foreign Affairs Ministry (Supervisory Agency) MOFA’s Rationality and Opportunistic Behavior Towards Knowledge Aid MOFA is the agency that supervises the largest portion of ODA—in the forms of grant aid and multilateral cooperation with U.N. organizations—within the government. MOFA leads the Inter-agency Grant Committee, which coordinates government-wide grant aid activities. As of 2016, under the Development Cooperation Bureau, the Development Policy Division and the Development Cooperation Division mainly play a role in aid policy-making at MOFA (MOFA 2016; PMO 2016). Reinterpretation of Mandate: Chosen Policy Discourse and Strategy In contrast to MOSF’s promotion of economic cooperation, the foreign affair ministry champions the idea that ODA should be a responsible gesture of advanced economies towards the developing world with great emphasis on global responsibility. According to Article 3 of the Framework Act, multiple objectives—from poverty reduction, improvement of human rights, gender equality, sustainable development, humanitarianism, and economic cooperation, to peace and prosperity in the international community—are listed. Also, MOFA’s objectives include those of internationally acknowledged ODA norms and principles (e.g. UN’s MDGs (2001-15)). In MOFA’s 143 mid-term ODA policy documents (MOFA 201088, 36), the ministry reconfirms that the achievement of the MDGs (consisting of 8 goals and 18 targets) is most prioritized. In effectively helping partner countries, five strategic areas are selected for MOFA’s future operations: education, public health, public administration, agricultural-fishery industries, and industrial energy. Although MOFA acknowledges the benefits of Korea’s national experience of social and cultural development from its earlier years of ODA management, the strategic positioning of knowledge-sharing in overall business plans is rarely specified before 2010. Besides that, as part of an effort to promote marginalized types of ODA intervention, MOFA previously planned double KOICA’s research in development projects89. According to the ODA Advancement Plan (201090, p.18), it was suggested that MOSF might be at the best agent to coordinate knowledge-sharing activities. For the operations of Sae-ma-ul ODA, on the other hand, joint leadership by MOSF and MOFA was proposed. In recent years, MOFA’s interest in knowledge-sharing has been significantly strengthened. A review of MOFA’s annual implementation plans (2006-2015) suggests that there were no specific MOFA initiatives from 2006 to 2010 that can be associated with knowledge-sharing. As articulated in MOFA’s 2012 Annual Implementation Plan91, the category of research in development was replaced with ‘development consulting’ and a wider range of technical cooperation types (e.g. public policy consultation, feasibility study, development master plan setup) was incorporated into the policy (CIDC 2011, p.15). In the 2014 Implementation Plan, the four sub-categories of development consulting were newly adopted92. Evidence-Based Policy-Making?: Technical Considerations behind the Chosen Policy Discourse and Strategy The policy stance of MOFA towards knowledge-sharing is largely affected by the fact that the ministry works on multiple objectives and supervises subordinate agencies

88 The document is titled Mid-term ODA Plan for 2011-2015 (분야별 국제개발협력 기본계획 ’11-‘15 [in Korean]); The report was submitted to CIDC for approval. 89 Given that KOICA allocated 2.7 per cent of its overall budget to Development Study projects in 2009, MOFA planned to increase the budget by 5 per cent by 2015. 90The document is titled International Development Cooperation Advancement Plans (국제개발협력 선진화 방안 [in Korean]). 91MOFA publishes overall strategies in terms of selecting strategic regions (or countries), sectors, and project-types. 92 Technical consultation for public policy, institutional build-up, knowledge training associated with infrastructure building, and capacity building (정책기술자문, 제도구축, 인프라구축연계, 역량강화 [in Korean]). 144 at the same time. Although the ministry traditionally enjoys exclusive autonomy in the management of all kinds of technical assistance projects, MOFA belatedly recognized the importance of knowledge-sharing for its overall ODA policy. To support such view, the introduction of development consulting projects was combined with KOICA’s introduction of DEEP (2012). Shortly after, this area became one of the strategic areas for grant activities in the 2010s. MOFA contends that the promotion of knowledge ODA is beneficial to increase the effectiveness of other grant projects and even to find potential opportunities for grant operations overseas. Although MOFA has emphasized avoiding duplication and fragmentation in ODA operations in general, the area of development consulting is an exception, since the diversification of development consulting projects is encouraged, according to MOFA’s operation plans. No specific data or empirical evidence attached to MOFA’s plan to expand knowledge-sharing ODA. Capacity, Information, and Resource Similar to the situation at MOSF, insufficient staffing and expertise at MOFA has been identified as a potential cause of reduced aid effectiveness (OECD 2012). From 2008 to 2012, the number of ODA-related positions at MOFA slightly decreased from 35 to 33. MOFA’s Delegating to and Monitoring of Subordinate Aid Agencies In legal terms, MOFA is the principal organization of KOICA: MOFA is capable of influencing KOICA’s business planning and budgeting, and its appointment to executive posts. According to the Enforcement Decree of the Korea International Cooperation Agency Act, the approval of the Foreign Minister is required in many important operational activities of KOICA; for instance, the dispatch of civil servants (Article 3), the borrowing of funds (Article 10), and annual business plans and budget (Article 11). 6.3.6 Korea Development Institute (Implementing Agency) Political and Financial Uncertainty for Implementing Agencies In Korea, the implementing agencies in the Development Cooperation are quasi- governmental organizations. Although a large portion of ODA operations are directly funded by the government, those aid implementing bodies often need to secure extra funding from non-governmental sources. Given that, Korean aid agencies are likely to be vulnerable to budgetary pressures and regime change. Such a high level of political/budgetary vulnerability may lead implementing agencies to constrain their ability to recruit and retain high quality development experts. According to the 145 OECD/DAC peer review (2012), a majority of staff at Korean aid agencies are paid roughly two-thirds or three-quarters of standard international/domestic pay rates.93 Together with a growing ministerial interest in knowledge-sharing activities, a growing number of state agencies now implement KS projects. As previously mentioned, since there are no strict laws that limit the number of aid agencies or that specify operational requirements for prospective aid agencies, the entry and exit of agencies in this field are rather convenient. KDI: Organizational Mandate and Behavior towards Knowledge Projects KDI is not an ODA-specialized agency. As a government-funded research institute for economic policy, the organization has long been mandated to provide the government with economic policy recommendations since its establishment in 1972. The mandates of the organization are listed in the Act on the Establishment, Operation and Fostering of Government-Funded Science and Technology Research Institutions (1999). KDI initiated the International Development Exchange Program in 1982. Later, KDI set up a specialized unit (named KSP) at the Center of International Development94 to run knowledge-sharing projects upon the request of MOSF-EDCF in 2009. Despite its pioneering role in promoting knowledge aid, the organization’s main mandate remains national economic policy research, therefore making its ODA operations peripheral. KDI’s relationships with MOSF, EDCF-Eximbank, and other ministries appear cooperative. Since the inception of KSP in 2004, KDI has experienced drastic budget increases, and its KSP operations appear to respond to the policy direction that is supported by MOSF. 6.3.7 Korea International Co-operation Agency (Implementing Agency) KOICA was established in 1991, fully committed to carrying out non-lending operations in foreign countries. Among grant-managing agencies under the supervision of MOFA, KOICA manages the largest volume of ODA: In 2012, KOICA received 71.6 percent of total grant aid budget, which account for 26.5 per cent of gross Korean ODA (CIDC 2012). As the legal foundation of KOICA operations, the KOICA Act and the Enforcement Decree of KOICA Act were enacted in 1991, and have been extensively

93 Among 250 Korean government agencies, KOICA’s employees are paid 70 per cent of the average pay rate. 94 Allocated personnel accounts for roughly 60 (including economists, temporary-hired consultants, and interns) as of 2014 (KDI 2010–). 146 modified over the last decades95. Those modifications brought some significant changes in the autonomy and budgetary security of the organization. Also, the working relationships with its supervisory agency (MOFA), sub-contracting bodies, and other peer organizations have been affected under the revised regulatory-legal frameworks. Reinterpretation of Mandate: Chosen Policy and Strategy For KOICA, DEEP arose in the process of revisiting KOICA’s role in the technical cooperation field. KOICA envisions that, in the context of promoting aid effectiveness, DEEP is expected to serve as a flagship project for capacity building in developing countries (KOICA 2014). From Article 7 of the KOICA Act, which specifies the mandated project types, DEEP falls into unspecified areas of ODA apart from KOICA’s specifically mandated tasks as follows: invitation of trainees, dispatch of professionals and overseas volunteers, research in development, disaster relief, and assistance with goods, funds and facilities. Figure 6.3 demonstrates that KOICA used to focus on a few types of technical assistance that included; dispatch of overseas volunteers and in-donor training programs. During the period from 1991 to 2012, development awareness and development study (also called research in development according to the KOICA Act) were marginalized in KOICA’s overall activities.

95 The amended ODA laws expand the scope of KOICA operations (2001), listing an additional funding sources (named Contributions for International Antipoverty (국제빈공퇴치기여금 [in Korean], which was established in 2007 but later abolished in 2016), reporting to the National Assembly (2012), and merit-based appointment of directors (2014). 147 Figure 6.2 KOICA’s Operational Orientation in Transition (1991-2012): By Aid Modality

100% Multilateral

90% Development awareness 80% Administrative costs

70% NGOs

Volunteers 60%

Experts* 50% Invitation of trainees 40% Development study 30% Emergency relief 20% Aid in kind/in cash

10% Infrastructure

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

(Note: *Including medical and Korean martial arts experts)

(Source: Author’s computation based on KOICA Statistics System (KOICA 2014))

In alignment with MOFA’s 2013 policy strategy, the introduction of DEEP serves as a turning point; it allowed KOICA to allocate more resources for knowledge aid. In 2012, KOICA allocated increased resources for DEEP operation (amounting KRW 37.5 billion). KOICA commissioned a research project to help build strategies for the promotion of DEEP. The commissioned study reviews technical cooperation practices in advanced donor countries. The report of the study points out that there are insufficient efforts being made to systematize the lessons from Korea’s development experience (KOICA 2013, 201496). According to KOICA’s DEEP policy, KOICA expected that DEEP would be managed in a manner to reduce the number of standing-alone projects and to increase the synergy effect with other KOICA’s grant projects. Technical Considerations behind the Chosen Policy Discourse and Strategy Despite KOICA’s mobilization of DEEP, it appears rather reactive in regard to the rising influence of MOSF-KDI. As long as KOICA’s Advancement Plan 2010-2015

96 The report was published by KOICA’s ODA Research Team, titled Study on the Strategies to Improve DEEP Program (DEEP 프로그램 발전방안 연구 [in Korean]); Another report is titled Evaluation Report on the Performance of DEEP (개발컨설팅 프로젝트 (DEEP) 성과관리 방안 수립 연구 및 시범사업 평가보고서 [in Korean]). 148 and other policy documents (KOICA 2010, 201397) are concerned, its knowledge-sharing initiative remain rhetorical without displaying substantial efforts to improve knowledge ODA. KOICA has continuously introduced a variety of projects (KOICA 201598) to align itself to the Post-2105 development agendas. KOICA identifies 12 strategic assignments, including the continued operation of development consulting projects. Under the Post- 2015 Brand Project initiative, DEEP and other subtypes of development consulting are continuously prioritized in KOICA’s strategy. Capacity, Information, and Resource Traditionally, KOICA used to prefer person-to-person modes of development cooperation: expert dispatch, overseas volunteer programmes, and invitation of trainees to visit Korea. Since 2000, KOICA has increased its commitments in the sector of social infrastructure building. The organization secures its operating funds both from governmental and non- governmental resources; not only direct contributions from the government, borrowed funds from international development organizations or foreign governments, but also involuntary contributions collected from citizens. The Contributions for International Anti-poverty (2007-17), which annually amounts around KRW 20 billion, contributed to reducing KOICA’s financial uncertainty. In the operation of DEEP, several issues arise with KOICA’s internal allocation and management of human resources. As mentioned in the internal research report of KOICA, the frequent changes in the makeup of the DEEP task force team are seen as negative factors to the performance of DEEP. Working Relationship with Supervising Agency What is suggested in expert interviews is that KOICA might have shown non- conformist behavior towards MOFA. Such behavior of KOICA might be motivated to increase autonomy and to become an independent development agency. Referring to the case of the Japanese International Cooperation Agency (JICA) after the 2008 reform, KOICA emphasized the positive effects of establishing an independent development organization on aid effectiveness and cost efficiency. However, under the delegation

97 The document (2010) is titled KOICA Advancement Strategies (2010-15) (선진화 방안 2010-2015 (2010) [in Korean]); according to the document, KOICA’s top agendas can be boiled down to (i) streamlining organizational structures at headquarters and (ii) strengthening field operations set against KOICA’s recent experience of having reduced staff (from 247 to 222) and teams (from 26 to 20); another document (2013) is titled Strategies to Utilize the Values of the Sae-ma-ul Movement for ODA (KOICA 새마을운동 ODA 추진 안내서 [in Korean]). 98 The document is titled Mid-term Management Strategy 2015-2017 (KOICA 경영중기전략 (2015- 2017) [in Korean]). 149 structure where MOFA legally intervenes KOICA’s important business decisions—e.g. the appointment of KOICA executives, the approval of business plans and operations budget, and the dispatch of staff overseas—KOICA is less likely to resist the interest of MOFA. Working Relationship with Sub-contractor Korea’s high country programmable aid99(CPA) ratio score—Korea is 35 percent and the DAC average is 54 percent (OECD 2012)—indicates that Korean aid agencies tend to depend less on sub-contracts for their overseas operations. As revealed in interviews, a negative perception of Korean NGOs among Korean bureaucrats might largely discourage public–private partnerships in the development cooperation field, in general. Hence, the low level of policy/operational involvement of NGOs and private consulting companies might be also explained by the preferences of the Korean aid bureaucracy. Likewise, according to an internal report of KOICA (2013), the weak capacity and inexperience of domestic NGOs are mentioned as the critical obstacle to KOICA’s intended expansion of public–private partnerships in the promotion of development consulting. As a partial consequence, KOICA works with a small, selected pool of NGOs and academic institutions.

6.4 Action Situations at the Collective-Choice Level The following section highlights the action situation when state organizations coordinate or cooperate. More specifically, this analysis delineates under what mechanisms state organizations interact, and to what extent policy coordination is achieved in the knowledge aid policy field. To understand the policy outputs that display the gap between aid planning and practice, it is important to study the role of coordination (Evans 1989; Kato 1994; Chibber 2002): Good coordination promotes inter-agency cooperation and generates consensus among bureaucratic organizations. Policy coordination and information sharing are focal observation points for the analysis. 6.4.1 Legal Institutions for Coordination CIDC100 was established as a designated policy coordinating body in 2006. The legal mandate and authority of CIDC is recognized by the later-enacted Framework Act. Article 7 of the Framework Act was amended in 2014 in terms of the makeup and working procedures of CIDC. The amendment was designed to further empower the coordinating

99 The higher CPA ratio means that capitals and commodities are directly transferred to recipient countries with less administrative costs and in-donor expenditure. 100 CIDC produces two major policy guideline documents: the ODA Basic Plan (prepared every five years) and the Annual Comprehensive Implementation Plan. 150 body in drawing up effective, coherent, and transparent resolutions (NLIC 2016101). As of August 2016, CIDC held 26 inter-agency meetings102. Drawing upon the relevant documentary evidence and interviews with CIDC members, the following analysis presents how its main tasks of reviewing the business plans of implementing agencies and coordinating operations have been performed in the case of knowledge aid. Also, a focus is to observe CIDC’s control over the flows of information; whether they had proper enforcement power; and to what extent CIDC is involved in the processes of plan formulation, implementation, and evaluation. 6.4.2 Inter-Ministerial Interaction Rivalry between MOSF and MOFA Whereas MOFA stresses the political and diplomatic objectives of aid-giving over humanitarian and economic goals, MOSF puts emphasis on the positive impacts of loans. The bureaucratic rivalry between economic and foreign affair officials has been a somewhat a chronic issue (Kim and Seddon 2005; Lim 2010). The bureaucratic dynamics is largely affected by domestic and international environments. In the post-Cold War period when Korea’s determination to win the diplomatic competition with North Korea significantly weakened, MOFA lost its superior position in the administration. Newly emerging in the ODA policy-making arena, MOSF started to function as a balancing power to the foreign ministry. As long as ODA laws and regulations leave substantial room for reinterpretation, the location for co-ordination, monitoring, and evaluation for knowledge aid practices is contentious within the administration. In the grant community, it is considered unreasonable that a loan-managing MOSF should manage knowledge-sharing initiatives in contrast to its traditional turf. Despite such controversies, KSP is an example of the decade-long, systematic effort of MOSF to consolidate and justify its authority and policy influence.

101 The National Law Information Center 102 All the CIDC meeting memoranda are available at www.odakorea.go.kr. 151 Figure 6.3 ODA Planning and Budget Processes in the Korean Government

(Source: OECD 2012, p.48)

A factor in the bureaucratic politics in the development cooperation domain was the recent modification to the administrative procedures of ODA planning and budgeting. This shift allows MOSF to exercise a veto power against the decisions of CIDC and the Inter-Agency Grants Committee. In some cases, the Budget Office at MOSF has approved EDCF operations in the absence of cross-ministerial consultation. As a result, the augmented intervention of the Budget Office makes the ODA approval process less predictable (OECD 2012). The Role of the Coordinating Body in Supervising and Coordinating Knowledge Aid At the third meeting103 of the CIDC, the knowledge-sharing initiative gained attention upon a presidential request (CIDC 2008104). The committee acknowledged that knowledge-sharing practices had been planned and implemented in the most fragmented and cost-ineffective manner, without proper coordination. In responding to the substantial need for cross-government coordination, CIDC created a list of tasks105 as follows: first, a special subcommittee needs to be set up in order to supervise and coordinate from planning to evaluation processes; also, all knowledge projects require annual review.

103 The meeting agendas and approved resolutions at CIDC are available at http://www.odakorea.go.kr/eng.structure.CIDC.do [In English]. 104 A CIDC report; titled Plan to Promote Development Experience Sharing (2008) (개발경험 공유 활성화 기본계획안 [in Korean]) 105 Ibid., p.11-13 152 During the period from 2010 to 2015, CIDC commissioned a preliminary feasibility study for large-scale ODA projects through KIEP106; and the issue of project duplication was overhauled for most grant projects. Through the document of 2011 ODA Implementation Plan, CIDC recognized once again, the problem of project duplication (fragmentation of ODA) with development consulting projects. As indicated in the document, a total eight implementing agencies107 have run 40 projects in 16 foreign countries. Although CIDC stressed that inter-agency cooperation might be needed, no specific strategies or measures entailed. Repetitively, the 2013 Annual Implementation Plan (CIDC 2012) re-visited the issue in the context of the dramatic ODA growth in 2013. At that time, CIDC evaluated that Korea’s KS operations might be improved in terms of fragmentation on the grounds that the number of implementing agencies reduced. However, it is a by-product of KOICA’s introduction of DEEP. Apart from KDI and KOICA, other ministries and agencies were encouraged to initiate knowledge-sharing projects. In the 2015 implementation plan document, the same issue was raised without no proper resolution plans attached (ODAwatch 2012108). To recap, CIDC’s ODA strategies and plans remain largely rhetorical—no actual activities of the special sub-committee followed afterwards. As pointed out in the OECD special report (2008), CIDC was found to have insufficient capacity and proper authority109 to perform its mandate of supervision and coordination over the leading ministries of MOSF and MOFA. As an effort to facilitate the comparative advantage of Korean ODA110, knowledge aid has been uninterruptedly promoted in the 2000s to 2010s, but no evaluative statement on the problem of fragmentation has been made at CIDC. CIDC tends to produce cross-government policy discussions that include too many agendas and present no clear indicators of progress upon repetitive agendas. CIDC’s

106 In general, ODA projects are self-evaluated by implementing agencies. Therefore, few cross- government evaluations of ODA projects were conducted. 107 In 2011, the Korean government disbursed knowledge ODA mainly through the following organizations: MOSF (KSP) (KRW 19.5 billion); KOICA (DEEP) (KRW 16.8 billion); Ministry of Environment (KRW 6.7 billion); PMO (KRW 1.3 billion); KCC 1.0 (KRW 1.0 billion); Ministry of Labor (KRW 0.4 billion). 108 The document is titled ODA Watch 44th Workshop Report (새마을 운동 ODA 누구를 위하여 새벽종은 울리나? [in Korean]) 109 This includes a veto power against the decisions of the Inter-Agency Committees or MOSF’s budgetary approvals. 110 This agenda has been frequently raised in government ODA policy reports and documents; with reference to a document titled Strategies to institutionalize a Korean ODA Model (한국형 ODA 모델 수립 및 내재화 [in Korean]). 153 documents are produced in inconsistent formats, so that end users and public readers have difficulties in understanding the performance of agencies.111 6.4.3 Inter-Implementing Agency Interaction No Common Informational and Operational Platform EDCF is mandated to develop a national ODA data platform. Data available at the EDCF website are often discrepant from what respective implementing aid agencies offer through their websites112 or reports. The duplication and fragmentation of ODA information is also problematic, contributing to the increasing cost inefficiency, lowering transparency, and creating confusion for policy-makers and other data users. According to Article 18 [Statistics on International Development Cooperation] of the Framework Act, all implementing agencies are required to submit statistics to supervising agencies and supervising agencies, in turn, submit them to CIDC. Although CIDC publishes annual figures on government-wide ODA activities, the presentation of the year-by-year data sets is inconsistent over years. The Korean aid system has yet to introduce an effective, integrative aid data platform. Considering the policy permeability of aid agencies—in theory, public agents are likely to show policy conformity to non-principals so as to increase its own indispensability (Hawkins and Jacoby 2006: 204)—some of the Korean implementing organizations significantly show such behavior as the theory predicts. In sharing ODA data and engaging in policy dialogues with international organizations, both KOICA and EDCF are assessed to have strong incentives to collaborate with them (OECD/DAC 2012). In the specific field of development consulting, however, policy or operational collaboration among aid agencies is hardly observed. Although CIDC has recognized associated problems, no counter-measures have been taken so far. There is a stark absence of policy dialogue and knowledge-sharing between the two leading implementers, KDI and KOICA. There are strict silos between state organizations that effectively prevent any inter-agency co-operation113. 6.5.4 The Outcomes of the Collective Actions: The Commitment Problem

111 With reference to the following documents: (i) Comprehensive Plans for Globalization of Sae-ma-ul Movement (2014) (지구촌 새마을운동 종합추진 계획 18 차 [in Korean]); Plans for Making an Integrative ODA Implementing System for Local Authorities (2015) (지방자치단체 ODA 추진체계 개선방안 20 차 [in Korean]); Strategies for Dissemination of the Values of Sae-ma-ul Movement (2016) (새마을운동의 국제적 확산방안 25 차 [in Korean]). 112 For instance, KOICA runs its own data platform which is open to the public; Available at http://www.oda.go.kr/opo/masc/mainPage.do). 113 With reference to the following documents: Strategies to Improve the Inter-Agency Cooperation (2011) (ODA 관계기관간 협력체계 구축방안 [in Korean]); and Strategies to Enhance Aid Effectiveness through Inter-Agency Cooperation (2013) (협업 활성화를 통한 ODA 효과성 제고방안 [in Korean]). 154 The commitment problem seems severe across the ODA sectors in Korea. Table 6.6 indicates that Korea’s performance of making a credible aid policy is persistently poor for the last decade (2006-15).

Table 6.6 Korean ODA: Annual Commitments and Disbursements (2006-15)

Year Commitments Gross Disbursements Performance (USD million) (USD million) (Percent) 2006 675.5 401.4 59.4 2007 1,053.3 525.5 49.9 2008 1,455.0 578.7 39.8 2009 1,449.4 615.3 42.5 2010 1,809.6 933.2 51.6 2011 1,623.6 1,034.3 63.7 2012 1,753.0 1,232.2 70.3 2013 2,238.2 1,375.0 61.4 2014 2,378.3 1,476.9 62.1 2015 2,311.7 1,548.9 67.0

(Source: OECD Creditor Reporting System) (Note: The above figures capture ODA allocated to all sectors and all channels; Accessed Nov. 25, 2017 from OECD.Stats)

The previous sections reveal that as the collective effects of the opportunistic behavior of the president, the supervisory ministries, and implementing agencies, knowledge aid has been rapidly proliferating. The CIDC appears to have insufficient capacities and legal powers to promote government-wide coordination for this sub-policy area of knowledge-sharing. With the ODA budget for knowledge sharing expanding with great attention of policy makers at higher level, missing coordination and cooperation among state organizations has caused a serious fragmentation problem in this field. Unfortunately, there are no datasets that properly capture the sum of various knowledge-sharing activities within the Korean government. Alternatively, this study obtains OECD data that are close to Korea’s actual performance of making a credible aid commitment in the sector of knowledge aid. Surprisingly, unlike what predicted in the beginning of the study, the figures (See Table 6.7) from 2010 to 2015 show that the commitment problem appears insignificant in this particular area of ODA operation. Such trend can be partly explained by the fact that the scale of knowledge ODA is relatively small (about 10 percent of total ODA) in Korea, the growth rates of knowledge aid are exceptionally high for the last decade. 155

Table 6.7 Korea's Performance of Making a Credible Aid Commitment: Knowledge Aid

Year Commitments Gross Disbursements Performance (USD Million) (USD Million) (Percent) 2010 126.6 127.9 101.0 2011 143.7 151.0 105.1 2012 168.0 167.0 99.4 2013 200.1 196.1 98.0 2014 228.1 227.5 99.7 2015 216.2 211.2 97.7

(Note: Data available from 2010; ODA to all developing countries; Current Prices)

6.5 Conclusion The knowledge-sharing initiative is the critical nested case that shows how bureaucratic incentives comes into play, interact, and shape the quality of policy products in terms of the planning-practice correspondence. As the knowledge-sharing sector is a grey area in ODA laws and regulations, state agencies are incentivized to competitively and opportunistically manage knowledge aid projects. Given that ODA laws that allow exclusive leverage to the president and MOSF over aid volume and geographical allocation, those state agencies at the high policy-making level served as the critical force for the dramatic growth of knowledge aid. That little monitoring/watchdog role taken by the parliamentarians also contributes to an uninterrupted increase in the government ODA budget. The weak social embeddedness of ODA in Korean society leads ODA policy to be an exclusive area for bureaucratic decisions. The rivalry between MOFA and MOSF also triggers intense competition for the policy leadership in the knowledge-sharing field. At the implementing level, KOICA—perennially put under severe budgetary pressures and political uncertainty—tends to show opportunistic behavior by re- prioritising knowledge aid in contrast to their existing operational orientations. Despite CIDC’s efforts to integrate the inherently dualistic aid policy system, the intensified bureaucratic competition continuously works favorably for the diversification of knowledge-sharing projects but against establishing coordinated and coherent ODA policy discourses and practices. As opportunistic implementing agencies have increased in number and in operational scope, without proper coordination, the problem of policy fragmentation and policy coherence has been only exacerbated.

156 CHAPTER SEVEN ANALYSIS OF ‘ACTION SITUATION’ WITH MULTILATERAL AID POLICY (2000-16)

7.1 Overview of the Chapter 7.1.1 Observing Bureaucratic Behavior in an ‘Action Situation’ of Multilateral Aid Planning and Management This chapter presents another nested case. Multilateral aid policy is a selected action situation of foreign aid policy-making to further examine how the participatory state organizations redefine and perform their policy-making roles under the regulatory- legal institutions. In tandem with the previous chapter, the following analysis offers observations on the incentives and collective actions of the state organizations that directly and indirectly contribute to the long-standing gap between planning and practice in Korean ODA. 7.1.2 Aid Multilateralism Aid multilateralism refers to an ODA trend in which donor countries increase the share of multilateral funding to total ODA; this is an alternative policy choice of donor governments at the expense of disbursing development assistance through bilateral channels (Milner 2006). In the post-Cold War era, aid multilateralism has been increasingly observed within the DAC donor community (Addison, McGillivray, Odedokun 2003; Hawkins et al. 2006). This ODA trend gives rise to an interesting area of study: Some scholars have explored under what conditions the U.S. government scaled up its development funding to multilateral organizations (Milner 2006; Milner and Tingley 2013). There are potential benefits as well as costs to an increased share of multilateral funding in a nation-state’s provision of ODA. From the perspective of delegation theory, the invitation of international agencies into a country’s foreign aid delivery process can be cost-ineffective. Also, the delegation of ODA delivery to international organizations (IOs) can increase the possibility of agency slack. Despite the hypothesized demerits of disbursing ODA through IOs, some established donor countries have preferred multilateral funding due to the strategic- political benefits that multilateral funding can offer (Addison, McGillivray, Odedokun 2003). The competence and specialization of IOs—which are allegedly equipped with highly competent development experts, abundant information on local contexts, better access to some geographical regions and territories, and public credibility—serves as a 157 good reason for donor states to channel their overseas development assistance through IOs to maximize the developmental effects of ODA provision. Multilateral ODA is believed to be more pro-poor in nature (Rodrik 1996). In contrast to bilateral aid, which is significantly subject to donors’ geo-political and commercial considerations, multilateral aid is considered better for serving humanitarian/egalitarian interests because the allocation of the funds is decided by the joint donorship of multiple countries (Martens et al. 2002). Surprisingly, the positive perception of IOs—as development specialists and as honest brokers for Third World development—has been tested through few empirical studies (Addison et al. 2003), but is held prevalently in the academic and policy literature (Ufford 1988; Martens 2002; Easterly 2004). 7.1.3 Expansion of Multilateral Cooperation in the Korean Government With respect to aid multilateralism, Korea offers an interesting case. Korea’s annual multilateral contributions in absolute terms are constantly on the rise, but its annual shares of multilateral ODA have been gradually decreased (from 2000 to 2013). This pattern of multilateral funding is distinctive among the ODA trends in other DAC member countries. At the same time, multilateral aid-giving has been significantly preferred by emerging donorship in a recent decade (OECD 2015, p.178). Against the mixed trends of aid multilateralism across the global donor community, Korea, as a young donor, as well as a DAC member country, might provide in-depth knowledge about what bureaucratic incentives and collective actions are involved along with the scaleup of multilateral ODA. For the interest of the research, this nested case is instrumental in understanding the internal political dynamics among bureaucrats within the Korean aid administration. Like the case of knowledge aid, multilateral aid is the sub-policy area in which state organizations show non-passive behavior in initiating and managing multilateral aid projects. Despite the fact that the institutional guidelines and policy experience are largely missing in this field, the aid-managing ministries and implementing agencies have appeared to collectively promote multilateral cooperation. Although such opportunistic behavior of the aid bureaucracy at the individual and collective policy-making level can be praiseworthy in promoting development cooperation, this sustained phenomenon has resulted in a further fragmented, incoherent aid practice of the Korean government.

158 7.2 Multilateral Cooperation Activities in the Korean Government 7.2.1 Relative Multilateralization of Korean ODA Aggregate data on the performance of Korean multilateral aid (1990-2013) provide a mixed overview. In the past two decades, Korea’s annual contributions to multilateral systems have risen overall, with occasional ups and downs. At the same time, the annual share of multilateral funding to total ODA114 has displayed a downward trend overtime; in other words, Korea has impressively increased its multilateral ODA in absolute terms, but, in a relative sense, the importance of multilateral aid has decreased.

Figure 7.1 Annual Disbursements of Multilateral ODA in Korea (1990-2013): Relative Aid Multilateralism

500 90.0

450 80.0

400 70.0 350 60.0 300 50.0 250 40.0 200 30.0 150

100 20.0

50 10.0

0 0.0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012

MO volume (USD millions; left-axis) MO share, overall ODA (per cent; right-axis)

(Source: Author’s computation based on OECD QWIDS; Accessed Jul. 20, 2015) (Note: MO stands for multilateral ODA115)

In 2013, the Korean government disbursed a total of USD 445.8 million to IOs. The multilateral ODA ratio stayed at 25.4 percent (2013)—the average level is 27.4 percent during the period from 2000 to 2013. In a comparative sense, the ratio level is far below the DAC average (29.6 percent, 2000-13) but higher that of the non-DAC group116 (9.6 percent, 2000-13) (OECD 2015).

114 Total ODA is equivalent to a sum of bilateral and multilateral ODA. 115 The OECD makes a distinction between multilateral ODA and the use of the multilateral aid system. The ‘use of the multilateral aid system’ is a larger categorization that includes not only officially reported multilateral ODA contributions but also other non-ODA funding, and other kinds of support for multilateral organizations (OECD 2011). 116 27 non-DAC countries are included (OECD 2015). 159 By sectoral allocation117, Korea heavily focuses on MDBs, including regional development banks (RDBs). As shown in Table 7.1, the International Development Association (IDA) of the World Bank Group and the Asian Development Bank are top recipients of Korean ODA. Among a number of U.N. agencies and other IOs in partnership with the Korean government, UNDP receives the largest volume of non-MDB funding, but nonetheless remains at an insignificant scale.

Table 7.1 Accumulated Disbursements to Multilateral Organizations (2000-13)

Asian Multilateral United Nations Development World Bank ODA, Total Agencies Bank Volume 1135.7 720.2 2999.4 533.1 (USD millions) (IDA: 928.1) (UNDP: 51.5) Share 37.8 24.0 100.0 17.7 (Percent) (IDA: 30.9) (UNDP: 1.7)

(Source: Author’s reorganization of data available at OECD QWIDS; Accessed Jul. 20, 2015)

From 2007 to 2013, annual multilateral funding in the core contribution stayed between 25 and 30 percent. Non-core contributions (called multi-bi) amounted to USD 150 million in 2013, having increased from 2 to 9 percent of gross ODA since 2007. The final destinations of this non-core multilateral funding from Korea are the South and Central Asian regions (OECD Creditor Reporting System 2015). 7.2.2 Typology of Multilateral Cooperation As its traditional partnership with the U.N. agencies weakened in the post-Cold War milieu, Korea’s support for multilateral aid systems became concentrated on MDBs. UN-bound funding is managed in a more fragmented way than MDB contributions and, by and large, is provided in non-core funding. In an attempt to enhance the global role of Korea as a rising middle power, the Korean government has initiated a series of multilateral cooperation projects. Those funding initiatives arose in tandem with global events in Korea: The G20 summit meeting in 2009 and the High-Level Forum on Aid Effectiveness in 2011. Korea has a keen interest in promoting green growth and climate change mitigation.

117 The MDB-focused model of multilateral aid-giving is also found in Japan, France and the U.S. In contrast, small donors in the northern and western Europe tend to fund U.N. programs more. Also, the five new European DAC members (the Czech Republic, Poland, the Slovak Republic, Slovenia, and Iceland) focus their support on partnerships with EU institutions. 160 In the multilateral aid policy field, key organizational actors are listed below (See Table 7.2), reflecting accumulated budget allocations from 2006 to 2013: 17 organizations are de facto participants for managing multilateral ODA activities (KIEP 2013118). The divided feature of Korean ODA system—inherently divided under the dual supervision of the finance and foreign affair ministries—affects the political dynamics of multilateral aid policy-making. Likewise, the government’s multilateral outflows are legally and administratively divided between funding to MDBs and to non-MDBs. Multilateral ODA is disbursed in the following categories: contributions (similar to grants), capital subscriptions, and concessional lending. In fact, in-donor fragmentation of multilateral ODA is quite common in the DAC community (OECD 2015). By DAC standards, Korea was not recognized as a case with sporadic governance of multilateral aid. For the last decade, however, the number of participatory state organizations and the scale of Korea-funded U.N. programs have grown rapidly, and, therefore, the problem of fragmentation has been exacerbated in the process of policy-making and implementation (CIDC 2014; OECD 2012, 2015).

Table 7.2 The Top Ten Multilateral ODA Distributors in the Korean Government (2006-13)

Disbursements Government Nature of Disbursement (Accumulated, Agencies (USD million) USD million) Grant contribution (449.38); Capital 1 Bank of Korea 1288.79 subscription (871.87); Concessional lending (- 32.48) Ministry of Foreign 2 366.37 Grant contribution (U.N.: 325.58; others: 40.7) Affairs Ministry of Strategy Grant contribution (WBG: 135.75; RDBs: 60.37; 3 214.18 and Finance others: 18.05) Grant contribution (U.N.: 84.12; WBG: 6.34; 4 Ministry of Health 101.57 others: 11.11) Ministry of 5 Agriculture, Food 58.07 Grant contribution (U.N.: 58.07) and Rural Affairs Ministry of 6 42.03 Grant contribution (U.N.: 5.43; others: 36.62) Education Ministry of 7 Employment and 36.60 Grant contribution (U.N.: 36.60) Labour

118 The report (KIEP 2013) is titled Korea’s Multilateral Aid: Recent Developments and Future Challenges (다자 원조의 효과적 실행을 위한 통합추진전략 [in Korean]). 161

Ministry of 8 28.86 Grant contribution (U.N.: 15.5; others: 13.39) Environment 9 EDCF 20.00 Capital subscription (RDBs: 20) 1 KOICA 13.85 Grant contribution (U.N.: 12.74; others: 0.11) 0

(Source: Author’s compilation of data available at the EDCF website; Accessed Aug. 15, 2015) (Note: This figure is based on the following categorization of multilateral aid: (i) grant or contribution (출연 [in Korean]); (ii) capital subscription (출자 [in Korean]); and (iii) concessional loan (양허성 차관 [in Korean]); Except Bank of Korea and EDCF, other listed state agencies have disbursed grant contributions only)

7.2.3 Funding to Multilateral Development Banks In 2012, Korea contributed 63.3 percent (USD 262.3 million) of its multilateral ODA to IFIs (OECD QWIDS 2015). The Bank of Korea (BOK), MOSF, and Eximbank are primary contributors to MDBs. The Bank of Korea makes the largest contributions and exclusive capital subscriptions119 to MDBs. The BOK’s mandate dates back to the 1960s when Korea acquired membership to the IMF and the IBRD to secure lending from those IFIs. The scope of IFI partnership had been gradually expanded at the times of newly funding WB organizations (IDA (1961), IFC (1964), and MIGA (1988)). At the regional level, the BOK has funded regional development initiatives that are carried out by the Asian Development Bank120 (ADB), the African Development Bank (AfDB), the European Bank for Reconstruction and Development (EBRD), the Inter-American Development Bank (IDB, 2005), and the Asian Infrastructure Investment Bank (AIIB). The funding to multilateral development banks aims to achieve the dual objectives of supporting low-income countries and enhancing Korea’s status in the global financial community.

Table 7.3 Net ODA Disbursements to MDBs: Bank of Korea (2006-13) (Unit: USD million)

2006 2007 2008 2009 2010 2011 2012 2013 Total 23.0 139.3 202.4 153.6 143.7 176.5 228.0 221.9 WBG 0.0 0.0 6.0 13.5 10.5 1.5 0.0 0.0 (Contribution)

119 Given cumulative data from 2006 to 2013 (available at a national ODA data platform of the EDCF), its contributions and capital subscriptions account for 32.7 and 97.7 percent (of overall multilateral ODA), respectively. 120 Cumulative subscriptions have made Korea the 8th largest shareholder of the ADB and the Asian Development Fund (ADF). 162 RDBs 40.7 41.6 111.3 54.9 39.6 42.0 40.4 42.3 (Contribution) Other MDBs 1.6 0.7 0.8 1.1 0.1 0.1 0.1 0.0 (Contribution) WBG 0.0 82.0 69.6 71.3 79.1 83.3 124.1 128.4 (Subscription) RDBs 13.2 14.9 14.5 12.7 14.4 49.4 63.3 51.1 (Subscription) Concessional *-32.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 loans

(Source: Author’s compilation from EDCF Stats; Accessed Jul. 20, 2016) (Note: *Loan repayments)

In contrast, MOSF and EDCF121 have been emerging multilateral aid givers since 2009. Although MOSF’s accumulated multilateral contribution (2006-13) remains relatively small—15.5 percent of total multilateral funding, excluding capital subscriptions and concessional loans—its role as a budget setter and a policy coordinator for MDB funding activities is non-negligible. While the BOK procures MDB funding from foreign currency reserves, the finance ministry and EDCF rely on their ordinary government budgets. Capital subscriptions to selected IFIs122 are not necessarily made on a yearly basis since the capital replenishment cycles of IFIs come in multiple years, therefore leaving annual funding records more fluctuating 123 than the annual performances of bilateral ODA. 7.2.4 Funding to United Nations’ Agencies In 2013, the United Nations’ programs received an estimated USD 113.5 million 124 from Korea. Internally, those contributions are disbursed in a severely fragmented way: 21 national organizations performed as either occasional or small-scale funders. More than 80 global initiatives have been funded, including mandatory contributions and special funding to the U.N. (e.g. the United Nations Department of Peacekeeping Operations (UNDPKO125), WHO, and FAO) (KIEP 2013; OECD 2015; EDCF dataset 2015).

121 EDCF makes capital subscriptions (USD 20 million, 2010-13) to RDBs only. 122 They are listed in Article 2 [Investment-related Institutions and Investment Funds] of the IFI Act. As of August 2016, 15 IFIs are included. 123 This is exemplified by a recent case in which Korea decided to scale up its capital support for IDB, IBRD, and AfDB and those increases are attributable to sudden hikes in overall ODA. 124 This amount accounts for 25.6 percent of total multilateral ODA in the same year. 125 Korea joined the scheme in 1993 and its very first activity (i.e. a special army deployment to Somalia) took place in 1995. PKO support of Korea reached USD 9.5 million in 2011, topping as the largest U.N. contribution of the year. 163 MOFA performs a dual job as the largest U.N. funder and the policy coordinator with regard to non-MDB contributions. Other ministries—the Ministry of Health; the Ministry of Agriculture, Food and Rural Affairs; the Ministry of Labor; the Ministry of Environment; and the Prime Minister’s Office—also disburse multilateral aid on a regular basis, depending on their areas of specialty and interest. Below the ministerial level, KOICA and KDI have newly promoted multilateral cooperation in recent years, and their respective contributions per year have ranged between USD 1 and 5 million (2006-13). Among the 20 U.N. organizations funded by the Korean government, UNDP receives the largest non-MDB funding from the Korean government. UNICEF, UNHCR, WFP, UNFPA, and UNRWA come next. Compared to the MDB-destined ODA, no drastic increases are observed in non-MDB funding. On the other hand, there is a clear tendency to intensify earmarked practices among non-MDB funders (KIEP 2011, 2013).

7.3 Action Situations at Individual-Choice Level 7.3.1 President The Korean Presidency and Diplomacy at Multilateral Organizations Far before Korea’s full-scale disbursements of ODA in the early 1990s, multilateral contribution was made in an insignificant scale. During the Cold War era in Korea, the president was in the position of having a keen interest in multilateral cooperation, demonstrated in a bid to secure concessional lending from the World Bank and the IMF in the 1960s through 1980s. As stated in the IFI Act (Article 1-2, enacted in 1975), the president is allowed to decide on the initiation and scale of funding to IFIs upon the recommendations from the finance minister. Apart from IFI contributions, Seoul’s multilateral cooperation was traditionally centered on gaining regime legitimacy against North Korea. The United Nations was the strategic arena in which Korean diplomacy was most visible. Nevertheless, after the UN’s official recognition of the two Koreas in 1991, the importance of the U.N. in Korean diplomacy became diluted. It was almost the end of the 2000s by the time that the government’s funding to IOs started to diversify and scale up along with the rapid expansion of bilateral development aid. Noticeably, contemporary facilitation of multilateral ODA (from the 2000s onwards) appears largely different from that during the Cold War period, when Korea heavily depended on foreign capital for post-war reconstruction and experienced escalating diplomatic hostility with the north. Comparison between the Roh and Lee Administrations 164 Under the Roh and Lee administrations (2003-12), the growing use of multilateral ODA might be explained largely by the role of state bureaucracy, rather than by the role of the president. Quantitatively speaking, the two leaderships effectively accommodated scale-ups in multilateral aid to existing partner IOs, and further expanded the scope of international subscriptions—by newly funding the Inter-American Development Bank (IDB), the European Bank for Reconstruction and Development (EBRD), and the U.N. and other global initiatives.

Table 7.4 Components of Multilateral ODA: A Comparison between the Roh and Lee Administrations

Roh (2003-07, in office) Lee (2008-12)

Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Overall ODA 1173. 1324. 1597. 365.9 423.4 752.4 455.3 696.1 802.3 816.0 (USD million) 8 5 5 Multilateral ODA 120.7 92.6 289.0 79.2 205.6 263.1 234.9 273.1 335.0 414.3 (USD million) M-ODA Share 33.0 21.9 38.4 17.4 29.5 32.8 28.8 23.3 25.3 25.9 (Percent) ODA to MDBs 70.4 64.4 159.1 24.5 145.2 136.4 136.5 170.3 217.7 262.4 (USD million) MDB-destined Aid Share 58.3 69.5 55.0 30.9 70.6 51.9 58.1 62.3 65.0 63.3 (Percent) ODA to U.N. Agencies 25.1 21.6 38.3 42.1 46.7 45.7 54.8 75.4 80.0 112.5 (USD million) UN-destined Aid Share 20.8 23.3 13.2 53.2 22.7 17.4 23.3 27.6 23.9 27.1 (Percent)

(Source: OECD QWIDS; Accessed Aug. 2, 2015) (Note: In 2006, the government made a relatively low level of ODA disbursement compared to adjacent years. While its contribution to U.N. agencies stayed stable compared to previous and subsequent years, Korea recorded a negative contribution to IMF (debt repayment: USD 30.91 millions) and no contribution to IDA)

However, there might be no clear relationship between the use of multilateral ODA and political stances of the government. As far as the existing explanations126 of aid multilateralism are concerned (Addison et al. 2005; Milner 2006), the thesis of right-wing

126 Regarding the relationship between the political orientation of government and multilateral aid performance, the relevant study reveals a tendency for right-wing governments to give more multilateral aid than left-wing governments in DAC countries (in the 1980s through 1990s). 165 governments give more multilateral aid might be partially supported in the Korean case. Despite multilateral ODA trebling during this period, the ratios of multilateral aid (to overall ODA) are not significantly changed (rather, the figures stay around 25 percent annually) even after Korea experienced a drastic regime change from progressive to conservative government. Development of Multilateral Aid Initiatives at Presidential Level During the Roh government (2003-08), the scale of multilateral ODA significantly expanded at times of additional capital subscriptions127 to the IBRD and IMF (2004), new entry to the IDB, IIC, and MIF (2005), and additional subscriptions and contributions to IDA and AfDF (2007) (MOLEG 2016128). Korea’s subscriptions and contributions to IFIs, however, appear to be reactively made by the Korean government, less likely driven by pre-set plans or strategies. In 2005, ODA policy strategies were introduced in line with a newly set mid-term plan of the State Council to increase ODA to GNI/ODA of 0.1 percent by 2009 (0.06 percent in 2004) (PMO 2005 129 ). However, this document highlights the issues for the internal systemization of bilateral ODA delivery only, rarely delineating Korea’s objectives and strategies for the use of multilateral ODA. Roh’s diplomacy, which emphasized Korea’s role at the U.N. was not responsive to the multilateral aid performance of Korea. Despite Korea’s contributions to UN’s 14 peacekeeping operations—making Korea one of the top ten contributing countries, the shares of its U.N. contribution to overall multilateral ODA remained fluctuating and relatively low during the period. Under the Lee government, multilateral ODA continuously increased, albeit multilateral ODA ratios decreased as experienced in the previous government. The tendency of right-wing governments to prefer to fund MDBs than the U.N. among most of the DAC member counties (Addison et al. 2005) is not quite supported in the case of Korea. According to the government document that articulates the reasons for the amendments to the presidential decrees of the IFI Act (MOLEG 2016 130 ), the

127 Trust Fund for Korea Technology Consultation 128 Ministry of Government Legislation (MOLEG), Reason for enactment and amendments of Act on the Measures for the Admission to International Financial Institutions; Available at www.law.go.kr; Accessed Aug. 10, 2016 (국제금융기구에의가입조치에관한법률시행령 제정개정이유[in Korean]). 129 Office for Government Policy Coordination (PMO); the report submitted to the State Council, titled Strategy to Enhance Korea’s International Development Cooperation (국제개발협력 개선방안 [in Korean]) 130 The Ministry of Government Legislation (MOLEG); the report is titled Reason for enactment and amendments of Act on the Measures for the Admission to International Financial Institutions; Available at 166 amendments (2008-12) coincided with Lee’s growing interest in the following agendas: ‘Transfer of economic development knowledge’, ‘Resource diplomacy’, and ‘Green growth.’ With those newly highlighted slogans of the president, Korea’s funding to the RDB grew substantially (e.g. additional funding to the AfDB and the EBRD (2009) and to ADB’s Future Carbon Fund (2010)). Although a large portion of multilateral funding seems routinely disbursed, policy considerations at the executive policy-making level were slightly strengthened during this period. Korea’s entry to the OECD/DAC (2010) might also have served as a catalyst for the Korean government to carefully review OECD guidelines and recommendations (2008 131 ) and take some measures to address the problems raised with ODA management. However, as demonstrated in the mid-term ODA policy paper (CIDC 2010), substantial discussions across the government had yet to emerge regarding multilateral cooperation. At least, the Korean community had agreed that the bilateral-multilateral ODA ratio should stay at 70 to 30. Also, the importance of multilateral aid-giving was revisited in the context of increasing opportunities for Korean citizens and domestic industries to enter into international IO employment and procurement markets (CIDC 2010, p.14, p.99). Under the Park Geun-hye government, Korea made subscriptions to the AIIB (USD 3.73 billion) in 2015. 7.3.2 Political Parties In Korea, the role of political parties is considered largely insignificant in reshaping government’s multilateral funding. As commented by the previous member of the Strategy and Finance Committee132 at the National Assembly (AP 2010133), of a combined USD 6.18 billions of MDB funding in 2010, only USD 50 million required the approval of the Parliament; this means that less than 1 percent of MDB funding was actually under the deliberation of the National Assembly. However, in recent years134, there have been some noticeable actions taken by politicians to strengthen the monitoring role of the National Assembly over the multilateral funding of the government. In the context of growing concerns about the BOK’s discretionary use of foreign currency reserves, a recent amendment of the IFI Act

www.law.go.kr; Accessed Aug. 01, 2016 (국제금융기구에의가입조치에관한법률시행령 제정개정이유 [in Korean]) 131 In the report, one of the five recommendations was to prepare aid policy and strategy in multilateral cooperation. 132 This is a standing committee at the National Assembly (기획재정위원회 [in Korean]). 133 The Associate Press Interview with Yong-seob Lee (19th National Assembly Member, Former Minister of the Interior) 14th November 2010, [in Korean]; Available at http://news.naver.com/main/read.nhn?mode=LSD&mid=sec&sid1=100&oid=001&aid=0004763796. 134 The IFI Act and presidential decrees were partially amended during the period of the 18th and 19th National Assembly. 167 (2012) allows the National Assembly to receive an annual report from the government on year-by-year MDB contribution plans, and to have better monitoring and veto powers against the government’s performance in disbursing multilateral ODA. However, the National Assembly, after the 2012 amendment, has yet to perform as an effective control mechanism over the ever-growing ODA budget under the Park GH government. To exemplify, the government’s proposal to channel the new, large-scale subscription to the AIIB135 (USD 3.73 billion in 2015) was ratified without significant opposition from the political parties (NAM 2016136; Newsis 2015137). Given that multilateral aid to non-MDBs is secured from the ordinary government ODA budget in Korea, the Foreign Affairs, Trade and Unification Committee at the National Assembly is mandated to review and approve MOFA’s budget for bilateral and multilateral activities altogether. Since the multilateral ODA budget under the supervision of MOFA, compared to MOFAs’ bilateral ones, remains relatively very small, no substantial attention of the committee has been paid (NAM 2016138). 7.3.3 The Finance Ministry (Supervisory Agency) From the 2000s, what has been commonly agreed within the aid community is that Korea might have given too little multilateral aid by international standards. At the same time, state organizations altogether underpin the idea of national prestige139 in its multilateral aid-giving. Supposing that the scale of multilateral contribution determines a country’s political power within international financial institutions, the government’s expansionary policy stance towards multilateral ODA is well justified from the mid- 2000s onwards. However, this scaling-up motive has been interpreted by different ministries in varied ways. With the legal-regulatory frameworks for multilateral aid policy being loosely set, the following analysis reveals that the two managerial ministries, MOSF and MOFA, have championed distinctively different policy discourses and strategies. MOSF’s Reinterpretation of Mandate: The Obligatory Aid-Giving Discourse

135 Interview with a committee member, Newsway 2015, 27 March 2015, [in Korean]; Available at http://news.newsway.co.kr/view.php?tp=1&ud=2015032709325001331&md=20150327094201_AO). 136 Accessible at http://likms.assembly.go.kr/record/. 137 At the plenary meeting regarding a ratification bill on entry to AIIB, 196 MPs (out of 198) agreed to the bill, 30 November 2015, [in Korean]; Available at http://www.newsis.com/ar_detail/view.html?ar_id=NISX20151130_0010448398&cID=10301&pID=1030 0 138 Accessible at http://likms.assembly.go.kr/record/. 139 The term, or ‘national status’, is written and read in Korean as follows: 위상 [wi:-sang] (位相) or 국격 [kuk-gi:oak] (國格). Those terms are frequently used throughout ODA policy papers and mentioned as the primary motive and justification for ODA expansion. 168 Although MOSF’s multilateral cooperation is limited to 15 international financial institutions (as of August 2016), MOSF has long enjoyed its exclusive autonomy over overall multilateral funding decisions (Article 3 [Method of Investment]). It is generally the case that MOSF’s proposals of increasing contributions to IFIs have been approved. As previously mentioned, the Framework Act and the IFI Act do leave great room for supervising agencies to specify (i) the objectives of multilateral aid-giving, (ii) the reasons for joining IOs and expanding (or withdrawing) capital subscriptions or contributions, and (iii) strategies to achieve given objectives. According to the Mid-term ODA Plan 2011-2015 (2010, p.18), MOSF emphasizes that in the context of a significantly growing role of IFIs in supporting economic growth in developing countries, Korea might need to increase its contributions to IFIs—as Korea pledged greater contribution for global prosperity and peace at the G20 meeting140. In performing its expanded international duties, Korea’s national status and global responsibility were put forward as the central objectives in MOSF’s multilateral funding. Economic bureaucrats justified their decision to increase IO subscriptions on the basis that increased contributions would enhance the political prominence of Korea at IFIs. Conditioned by the prospective economic growth of Korea (i.e. 5 percent GNI growth, 2011-15), MOSF aimed to continuously increase multilateral funding. Such an expansionary policy stance has imbued the ministry’s strategies for every type and sector of multilateral ODA; aside from capital subscriptions and core contributions to IOs, other funding types, such as concessional loans and trust fund, are gradually expanding. Partly, this is a reflection of Korea’s own history of having benefited from concessional loans from IFIs. Apart from MOSF’s financial diplomacy, increased contributions to RDBs are self-justified under the objective of promoting business opportunities for domestic industries in emerging procurement markets, according to MOSF’s mid-term ODA plans. Interestingly, MOSF favors its increased MDB funding over the funding to the U.N. The ministry contends that the UN’s one-country-one-vote principle might discourage donors’ motivation to increase contributions for increased leverage in decision-making at IOs. More recently, the Second Mid-term ODA Plan 2016-2020 (MOSF 2015) confirms that MOSF’s multilateral ODA policy needs to be continuously expansionary and DAC standards-conscious141.

140 At the G20 Seoul summit meetings in 2010, Korea is committed to play a supportive role by indirectly financing the developing world through IFIs in the face of the global economic crisis. 141 For instance, a bilateral-multilateral ODA ratio becomes an important indicator for MOSF to assess its own policy. The standard of disbursing 70 (bilateral ODA) to 30 (multilateral ODA) is drawn from DAC’s 169 Evidence-Based Policy-Making?: Technical Considerations Behind the Chosen Policy Discourse and Strategy The finance ministry highlights business opportunities for domestic industries in the MDB’s procurement markets—which accounts for a third of global ODA transactions. Nevertheless, the bureaucratic expectation that enhanced national prestige at IFIs would turn into tangible benefits for domestic industries appears ill founded. Set against the ever-expanded scale of Korea’s subscriptions to MDBs, a multi-year record of winning MDB procurement contracts for Korean firms seems hardly responsive; from 2003 to 2007, Korean firms won 0.9 percent of total MDB procurements (KIEP 2010142).

Table 7.5 The Performance of Korean Businesses in the MDB Procurement Markets (2003-07, Accumulated)

Sector** WB* ADB IDB EBRD AfDB Total

Construction 168.0 777.0 0 0 21.0 967.0 & Commodity

Consultancy 0.6 4.5 0 0 0 5.1

Total 168.6 781.5 0 0 21.0 972.0 Market share 0.5 2.9 0 0 0.5 0.9 (Percent)

(Source: KIEP 2010, p.155) (Unit: USD million, Accumulated Contract Prices) (*Note: The World Bank Group (WBG) provides the largest MDB procurement market, consisting of almost half of overall MDB markets. The rest of the MDBs are listed by size of procurement market.) (**Note: Sectors in the MDB procurement markets are construction (76 percent), commodity (14 percent), and consulting (10 percent))

In 2010, Korea’s corporate procurement market share at the World Bank Group arose by 3.5 percent in contrast with 0.5 percent during the period between 2003 and 2007, but this hike turned out to be temporary given the sharp drops in the following years. In the case of ADB, in which Korean industries are considered to be winning relatively larger scales of contracts than in other MDB procurement markets, the performance of domestic firms appears also to be fluctuating. Although Korea held 5.06 percent of total

average level of bilateral-multilateral ratio. This aid goal does not have a particular value that is approved in terms of aid effectiveness or aid accountability. 142 Research Report 10-26, Enhancing the Effectiveness of South Korea’s Multilateral Aid (우리나라 다자원조 추진전략과 정책과제 [in Korean]) 170 shares (4.34 percent, voting weight) in 2014, annual ADB procurement market share is recorded at 6.46 percent in 2013; 2.86 percent in 2014; and 2.83 percent in 2015 (ADB 2015143, 2016144). The long-held bureaucratic notion of larger MDB contributions lead to greater business opportunities for domestic corporations is rarely supported, not only in Korea but also for in other major donor countries. For instance, DAC member country’s shareholding level at the ADB is not necessarily proportional to their shares of total procurements (e.g. Japan’s shareholding and cumulative procurement shares are 15.62 and 3.63 percent, respectively; the U.K.’s, 2.04, 1.35; Australia’s, 5.79, 1.17; the U.S.’s 15.51, 7.83) (ADB 2015). Although a recent study (McLean 2015) suggests that there is a tendency for donor governments to give more multilateral funding to MDBs where more potential benefits exist for domestic interest groups, other existing studies go against a direct link between multilateral aid and economic interest, due to the apolitical, independent nature of IOs (Fleck and Kilby 2001; Milner and Tingley 2010). When it comes to MOSF’s principles of joining and expanding MDB funding schemes, there are no specific ideas articulated in the MOSF’s policy papers—except that they mention that it is subject to domestic policy conditions (MOSF 2010). In other words, MOSF remains reactive in establishing its reasons and strategies, according to the conditions of the government budget, overseas aid demand, DAC recommendations, and the execution rate of multilateral aid projects. Capacity, Information, and Resource According to interviews, there might be an insufficient number of economic bureaucrats who specialize in MDB policy. A similar trend is found in the Japanese case, where a few economic bureaucrats led the expansion and diversification of MDB funding in the 1980s (Yasutomo 1997, p.181). Regarding MOSF’s behavior of collecting and disbursing policy information, there is no clear evidence that the ministry effectively controls and facilitates the information internally (within the aid administration) or externally (with domestic business interest groups). Also, there are weak incentives to exchange information among economic bureaucrats and domestic businesses, since both groups are skeptical about future public-private partnerships in this field. MOSF, Eximbank, and the Korea Trade- Investment Promotion Agency (KOTRA) increasingly publicize business opportunities

143 ADB’s Member Fact Sheet 2014; Available at https://think- asia.org/bitstream/handle/11540/5209/Korea_Fact%20Sheet%202014_web-ready.pdf?sequence=1 144 ADB’s Member Fact Sheet 2015; Available at https://www.adb.org/sites/default/files/publication/27792/rok-2015.pdf 171 in the MDB procurement markets for domestic industries. However, what has been raised by the private sector is that lack of information (particularly about the bidding procedures of MDB procurement) is considered one of the biggest issues (KIEP 2010, 2013b145). For the government’s multilateral funding, a combination of foreign currency reserves and ordinary government budget is used. In general, a large portion of the MDB contribution is paid by the BOK upon the request of MOSF (Article 2-(3) of the IFI Act). MOSF’s great autonomy in facilitating foreign currency reserves, when it needs extra financial resources for MDB funding, was legally and traditionally accepted without heavy opposition encountered at the National Assembly. Delegating Tasks to and Monitoring Implementing Agencies MOSF delegates the actual disbursement of multilateral funding to the BOK and EDCF. According to the IFI Act (Article 2-3), the Minister of Finance can request that the BOK pay for mandatory MDB subscriptions when the pre-allocated government budget is insufficient to cover the scheduled payment. For partial disbursement of RDB funding, EDCF has been mandated to make a contribution to the ADB since 2009. To address the associated problems of tarnishing the independence of the BOK, MOSF recently announces a strategy to secure ordinary government budgets for newly opening capital subscriptions and contribution to IOs. However, in general, the delegation between MOSF and its subordinate agencies takes place routinely, without observing the unconventional behavior of the BOK and EDCF. 7.3.4 The Foreign Affairs Ministry (Supervisory Agencies) MOFA’s Reinterpretation of Mandate: The Aid Effectiveness Discourse In 2014, while Korea allocated 69 percent of total multilateral ODA to IFIs, the rest went to the UN’s organizations and other IOs (CIDC 2015). MOFA positions multilateral aid as a key sub-policy sector that supports the achievement of the U.N.’s MDGs. As articulated in the Mid-term ODA Plan 2011-2015 (MOFA 2010), MOFA posits that Korea might have given insufficient multilateral contributions, incommensurate with Korea’s enhanced economic and social status. Through expanded cooperation with international development institutions, MOFA expects Korea to be recognized as an advanced development aid donor in the global community. UNDP, UNICEF, the WFP, the Global Fund, and U.N. Women are chosen as strategic IO partners.

145 Research Report 13-29; titled Korea’s Multilateral Aid: Recent Developments and Future Challenges (다자원조의 효과적 실행을 위한 통합추진전략 [in Korean]) 172 Explicitly, the promotion of multilateral cooperation is championed by MOFA since multilateral cooperation is expected to complement the existing and potential drawbacks of bilateral aid activities (e.g. lack of information on local contexts in partner countries, and insufficient technical expertise in implementing cross-national/cross- cutting development projects). Hence, for MOFA, multilateral aid is regarded as a supplementary tool to improve the aid effectiveness of Korea’s bilateral development activities. Evidence-Based Policy-Making?: Technical Considerations Behind the Chosen Policy Discourse and Strategy As for the more specific policy considerations of MOFA, the effective use of multi-bi and the set-up of evaluation/monitoring systems are identified as major areas of future multilateral commitment. Non-core funding, the so-called multi-bi, is a newly highlighted aid practice that makes earmarked contributions to specific IO projects rather than making a contribution to common financial pools of IOs. From the perspective of national aid provider, multi-bi results in better visibility of the national aid activities. On the other hand, this aid practice might harm the autonomy of IOs and complicate administrative procedures within IOs in digesting special funding. Speaking of aid harmonization, multi-bi can function as a factor to exacerbate the problem of fragmentation (OECD 2012, 2015). Despite an array of controversies over the use of multi-bi, bureaucrats at MOFA place a heavy emphasis on this fashionable practice. In line with that, MOFA made a plan to scale up overall U.N. funding by 10 percent (2011- 15)146. Such a plan is largely premised on the fact that Korea is giving multilateral aid just below the DAC level. With regard to the propriety of the multilateral ODA contribution level, the policy discourse rarely elaborates on the matter. Recently, the efforts to systemize the monitoring/evaluation systems of multilateral ODA are noteworthy in understanding the changed ministerial behavior (e.g. joining the MOPAN (Multilateral Organization Performance Assessment Network) in 2009). Yet, this better developmental aid stance is operationalized in a rather fragmented way. Compared to the MDB policy discourse, there is a far wider range of development goals imposed—ranging from conflict prevention, and gender equality, democratization to soft power—and partner MOs are listed without clear prioritization147. MOFA and

146 In 2011, Korea’s multi-bi ratio of total multilateral ODA stayed at 6 percent. In a comparative sense, France and Germany are recorded at 1 and 4 percent, respectively. Other small DAC donors—such as Belgium, Finland, and Ireland—stayed 9, 17, and 13 percent. 147 As articulated in relevant policy papers, prioritization among multiple multilateral ODA objectives is loosely made. 30-35 percent of total multilateral ODA is to be allocated for the MDG overall, and the 173 other ministries work with more than 80 multilateral organizations (KIEP 2013). UNDP, one of the three core strategic-partner MOs148, receives less than 9 percent of total non- MDB funding. This is the largest share that an individual non-MDB organization receives from the Korean government for the last decade. 7.3.5 Implementing Agencies In this section, the behavior of the following implementing agencies are observed and analyzed: the BOK, EDCF-Eximbank, and KOICA. Except for the BOK, the two ODA specialized agencies, which implement either loan or grant projects, perform the task of multilateral cooperation under the supervision of their respective supervisory ministries, MOSF and MOFA. From 2005 onwards, MOSF, in addition to its own practice of multilateral cooperation, has supervised EDCF’s multilateral activities of capital subscriptions to ADB. MOFA also delegates the implementation of multilateral cooperation to KOICA. In fact, the intra-governmental delegation in multilateral aid- giving is based on organizational laws (e.g. Article 7-2 [Projects] of the KOICA Act), but the details and scope of delegated tasks are seldom defined. Despite the loosely defined delegation, implementing agencies do not seem to have enough room for being strategically responsive to the growing demand of multilateral cooperation—mainly due to budgetary constraints and their lack of expertise in this area. Although their policy-making roles are considered indispensable in bilateral aid projects, rarely are they in this policy area. In recent years, an array of policy discourses and policy initiatives have been introduced by EDCF and KOICA. To some extent, their interest in multilateral cooperation has grown. To increase organizational influence in multilateral policy-making, those implementing agencies must start to show some strategic behavior towards their supervising agencies. The Bank of Korea The BOK, the largest multilateral ODA distributor in Korea, seems to perform its multilateral mandate in a relatively routine manner. The BOK appears to be in good coordination with MOSF in this particular field. With regard to the IFI Act (Article 2-3), in occasional cases, in which MOSF and EDCF-Eximbank have difficulties in paying IFIs from their own budgets, the BOK is legally bound to resolve the insufficiency.

rest is divided into sub-categories such as MDG sectoral, support for fragile states, Africa/Asia initiatives, and others. 148 Other core-strategic-partner MOs include: UNICEF, the WFP, the Global Fund, and U.N. Women (PMO 2010).

174 In general, the BOK traditionally and currently is on good terms with MOSF. No particular policy behavior has been observed, despite significant increases in capital subscriptions over the last decade. In line with the policy discourse of MOSF, the BOK performs its mandate of cooperating with IFIs in an appropriate manner (BOK 2016149). Nonetheless, the exclusive autonomy of the BOK in controlling foreign currency reserves is now significantly restricted by MOSF’s expansionary policy for IFI contributions. In response to such concerns, the IFI Act was amended to strengthen the watchdog role of the National Assembly in the management of foreign currency resources. However, the BOK is still under MOSF’s autonomous choice of expanding IFI memberships and subscriptions. The Economic Development Cooperation Fund Together with its core mandate of providing concessional loans to developing partner governments and domestic businesses, the EDCF also finances capital to IOs. EDCF’s actual execution of capital subscriptions to IFIs started in 2010. From its ordinary budget, the EDCF annually secures roughly USD 5 million. Despite its small-scale involvement in multilateral financing, the EDCF appears to have strengthened its role in the multilateral policy sector according to the organization’s post-2015 strategy (EDCF 2015150). As for EDCF’s approach to development financing through MOs, its major reference points are derived from the OECD/DAC policy guidelines. As delegation theory predicts (Hawkins and Jacoby 2006), the EDCF displays one of the strategic behaviors (as the agent), i.e. so-called policy permeability—inviting a third-party actor to its policy-making process in order to justify its policy choices—to a certain extent. However, buffering—creating barriers against the monitoring of supervising agency— and opportunistic (re)-interpretation of delegated tasks are rarely observed with EDCF. The EDCF has no systematic or in-depth understanding of multilateral development financing, as is demonstrated by MOSF’s multilateral cooperation strategy (2011-15). Cooperation between the EDCF and the private sector is rarely found. KOICA KOICA contends that bilateral aid activities can greatly benefit from multilateral cooperation (Park 2010). This is a new organizational emphasis at KOICA since the early 2010s. Reflecting on KOICA’s traditional and specialized areas of development intervention, multilateral cooperation justifiably remains as a marginalized subject for the

149 The BOK official website; Accessed Aug. 15, 2015; Available at http://www.bok.or.kr/broadcast.action?menuNaviId=2080. 150 EDCF Annual Report 2015; Available at www.edcfkorea.go.kr. 175 implementing agency. Nonetheless, in recent years, KOICA has shown a growing interest in expanding its scope of multilateral cooperation through more of multi-bi projects. Following its inception in 1991, KOICA’s multilateral activities were insignificant throughout the 1990s and 2000s. During this period, relevant projects were carried out in the forms of commodity support and expert dispatch to U.N. institutions. Noticeably, KOICA expanded the scale and scope of its activities since 2006 as it joined in partnerships with UNDP and UNICEF. Around that time, KOICA also joined trust fund schemes. In 2011, KOICA’s multilateral ODA ratio to overall ODA budget reached 9.2 percent: a total 26 U.N. institutions (including other IOs) work with KOICA; excluding emergency relief, its overall net multilateral grant contribution accounted for USD 37.6 million. As demonstrated in Table 7.6, KOICA’s core MO partners are UNDP, WFP, and UNESCO, and KOICA’s earmarked projects at IOs consist of USD 16.6 million (31 projects).

Table 7.6 KOICA’s Multilateral Funding in 2011

Amount IO in Theme Funding Project Period (USD Partnership million) Poverty UNDP UNDP MDGs Trust Fund 2010-12 6.5 Reduction

CAP humanitarian support 2011 1.0

Food Security WFP Korea-WFP Food for New Village 2011-12 2.0 Project Rural MP Korean Millennium Village MP* 2009-13 1.3 Development Project Humanitarian UNOCHA UNOCHA CERF Support 2011 2.1 Relief

(Source: Author’s compilation of data at the KOICA official website; Accessed Sep. 6, 2015) (Note: *MP is an abbreviated name for Millennium Project; Available at http://www.millennium- project.org/))

What is revealed by a review of policy documents and research reports from KOICA is that the positive complementarity between multi-bi and bilateral aid is central to the organization’s policy discourse. Despite the controversies about the benefits and costs of multi-bi, the effective use of multi-bi has become one of top agenda items for the organization. According to a recent internal report (Cho 2013), this hybrid type of multilateral cooperation is being newly highlighted by the organization, although there is 176 yet to be any relevant policy or strategy discussions on this at KOICA. Among the staff, the expansionary multilateral cooperation policy of KOICA is controversial given that multilateral ODA provision might be neither its major mandate nor its specialization. When it comes to KOICA’s working relationship with MOFA, a hypothetical conflict between the supervisory and implementing aid agencies is not observed at this early stage of multilateral aid policy development. Although KOICA is increasingly interested in multi-bi (non-core contribution) as a policy tool to enhance its visibility in the domestic and international aid community, its opportunistic behavior is observed only in a limited way against its organizational tradition and budgetary capacity.

7.4 Action Situations at Collective-Choice Level 7.4.1 Observed Behavior of MOSF and MOFA for Cooperation and Coordination To date, the provision of contributions to IFIs is the exclusive mandate of MOSF, whereas MOFA coordinates multilateral cooperation with non-MDB IOs (e.g. U.N.’s organizations and other global trust funds and funding initiatives). Comparing the group dynamics on the multilateral aid policy-making process with those of the knowledge aid process, the difference is that since their strategic MO partners do not overlap by law, no intensive bureaucratic competition is observed between the two ministries. No strong bureaucratic incentives for inter-agency cooperation among state agencies are found, either. While none of supervisory ministries takes proactive leadership in this field, the two ministries might have performed their mandates for policy agenda setting, policy coordination, and supervision in a relatively passive manner when compared to the knowledge aid policy-making situation.

Table 7.7 Multilateral Cooperation Strategies (2011-15): MOSF and MOFA

MOSF MOFA Policy Strategies for Multilateral ODA to the Strategies for Multilateral ODA to Document U.N. and Other International International Finance Institutes (title) Organizations Muskoka Initiative (with emphasis on Internation Global economic crisis and G20 role maternal, newborn, and child health) and al contexts the G20 Development Agendas Economic cooperation and Poverty reduction and sustainable Goal enhancement of Korea’s global development under MDGs political, economic status

Multilateral aid expansion Priority Multilateral aid expansion (e.g. Promotion of non-core funding) 177

(e.g. Expanding subscriptions to WBG and IMF)

To increase the number of Korean To offset the limitations of bilateral aid personnel at influential IOs; to activities; to learn lessons by national aid Purpose generate business opportunities for agencies from operational cooperation domestic industries in international with specialized IOs procurement markets

(Source: http://odakorea.go.kr/) (Note: this government document, entitled Sectoral Strategies for International Development Cooperation 2011-2015, was prepared by respective ministries and compiled by the Prime Minister’s Office. This 70-page-statement was put forward to the 8th meeting of the International Development Cooperation Committee in 2010.)

Table 7.7 demonstrates that MOSF and MOFA have embodied distinctively different rationales and strategies for multilateral cooperation. While they have commonly supported an expansionary policy for multilateral funding, their intentions behind multilateral ODA are different. For the MDB policy community, MOSF leads the policy discussion, partnership with Eximbank and KOTRA. As for the U.N. policy community, under the jurisdiction of the Inter-Agency Grants Committee (CIDC; chaired by the Prime Minister’s Office), a sub-committee is organized for non-MDB multilateral aid policy cooperation. However, this coordinating agency has yet to generate any comprehensive influence within the community, remaining nominal with the elusive leadership of MOFA. 7.4.2 The Outcomes of the Collective Actions: The Commitment Problem Multilateral ODA increased roughly six times from 2000 to 2013 (See Figure 7.2). The rapid growth in multilateral ODA has led to severe in-donor fragmentation and duplication of projects as acknowledged in the policy reports of CIDC. Noticeably, the degree of the commitment problem in this field appears less severe than overall ODA. As shown in Table 7.8, for the five years (2011-15), there are some occasions when annual disbursements exceed commitments from 2010 to 2012. Multilateral aid might have been disbursed under some pressure from IOs (the fear of losing face).

Table 7.8 Korea’s Performance of Making a Credible Aid Commitment: Multilateral Aid

Year Commitments Gross Performance Disbursements 178

2006 6.6 5.4 81.5 2007 15.5 16.1 104.1 2008 46.4 43.6 94.1 2009 76.6 49.7 64.8 2010 35.9 45.2 126.0 2011 74.7 80.9 108.4 2012 126.7 129.9 102.5 2013 218.0 164.7 75.5 2014 169.6 167.3 98.7 2015 195.1 182.6 93.6

(Source: OECD Data Set (DAC members’ total use of the multilateral system); Accessed Nov. 25, 2017) (Note: There is an alternative data set, The DAC Survey on Forwarding Spending Plans, available at OECD Statistics. However, specific data for the case of Korea are incomplete in this dataset)

There is no significant change in the number of participatory state organizations for MDB funding as an effect of having relatively restrictive boundary and scope rules with multilateral aid policy-making. Due to the feature of the IFI law that gives the largest decision-making and corrective powers to the president, the budgetary expansion of multilateral funding is expedited, hardly facing parliamentary scrutiny and opposition. Few legal-regulatory institutions have been developed in the categories of authority and scope rules, and, subsequently, the re-interpretation of their mandate in favor of organizational interests has been encouraged among state organizations. In particular, in the sector of U.N. funding, a group of opportunistically motivated implementing agencies manages multilateral funding in the most fragmented and unsystematic manner. Also, any rules for aggregation and information sharing in the action situation of multilateral aid policy-making are absent so that inter-agency coordination or operational cooperation in this sub-policy field has been largely neglected in the Korean aid administration.

7.5 Conclusion This multilateral aid policy-making case shows that the growth in multilateral funding in the Korean government is explainable by bureaucratic incentives and collective actions that promote multilateral activities as an opportunity to enhance their organizational autonomy. More than anything, the features of the legal-regulatory frameworks for the management of multilateral ODA provide some further explanations as the direct and indirect impacts on the bureaucratic incentives and collective actions create the gap between aid planning and practice. Nonetheless, compared to the political dynamics with knowledge aid policy-making, the opportunistic behavior of state organizations is far less frequently observed, and the commitment problem is less severe; 179 rather, at play there are the interests of the president who provides the strong momentum for the quantitative expansion of multilateral funding.

180 CHAPTER EIGHT CONCLUSION: EVALUATING THE IMPACTS OF RULE INSTITUTIONS ON THE COMMITMENT PROBLEM

8.1 Significance of the Research This thesis explores the question of how and to what extent rules determine state organizations’ incentives and collective actions on the policy-making process which are accountable for the planning-practice gap in the national system of foreign aid management and delivery. The thesis is intended to contribute to our better understanding of aid policy-making dynamics within donor governments in the following senses. 8.1.1 Evaluating the Logic, Design, and Performance of Legal-Regulatory Frameworks on Foreign Aid Policy Making Settings This study of the public policy-making process serves our genuine interest in rule institutions: In particular, responding to our interests in how and to what extent a set of rules determine ODA policy makers’ collective action. Existing literature still lacks systematic, meticulous measures to assess the logic, design, and performance of particular rule designs in donor countries’ aid management and delivery systems. To address the issue, this study turns to the merits of the Institutional Analysis and Development (IAD) approach. In particular, it uses the IAD’s rule typology, which identifies seven types of rules that are derived from the accumulated findings of IAD research. The rule typology represents the universal elements of rules-in-use in various social and political settings. Each of the rule types directs the disparate elements of the decision-making of individuals or groups. By adopting this rule typology in the study, this thesis is intended to provide a more systematic analysis to discern the impacts of diversely configured rule designs (as part of aid management design) on policy-makers’ behavior. 8.1.2 Introduction of a Refined IAD Framework: Taking Rules More Seriously This study is part of the new institutionalists’ effort to apply and advance Ostrom’s framework. The IAD framework is considered to be one of the most developed and sophisticated attempts to use institutional and stakeholder assessment in linking theory and practice (Aligia 2006, p.89). Although its usefulness for understanding variously configured institutional designs in both developed and developing countries, the language and actual application of the IAD framework remain quite complicated for potential readers and researchers. In this aspect, the IAD approach might need some recalibration in order to make pointed explanations and predictions given a selected set 181 of factors. Hence, I have suggested some analytical measures to enhance the benefits of the IAD framework. Above all, the refinement of the framework was designed to take rule institutions more seriously than the previous IAD research did. First, this research tried to recognize various layers (i.e. by hard and soft rules) and types (i.e. by IAD’s rule categories) of rule institutions in the public policy process. The classification of public policy-making rules appears necessarily to be incorporated into the framework; this is to discern the impacts of different rules on state bureaucrats’ behavior of goal-setting, strategy-building, performance of given mandates, and communication with other state actors. In doing so, the relationship between hard and soft rules was further examined in terms of what specific areas of decision-making are put under administrative discretion and to what extent informal rules fill the vacuum left by formal ones. Second, the refined framework re-emphasized the supplementary importance of non-rule contextual factors; the previous IAD studies in the development cooperation field largely dismissed the benefits of discussing non-rule factors in relation to the creation and evolution of the rule factors. To provide multifaceted evidence on the ‘rule- action’ nexus, this study offers an analysis of the earlier developments of rule institutions and the historical interplay between rule and non-rule factors. Third, the refined framework adopts a simplified model of decision-making. Originally, the previous framework adopted a decision-making model that conceptualized the public policy-making on the constitutional, collective-action, and operational levels. Such a three-level analysis was a source of complexity for the researchers and the readers. To minimize the associated complexity, this suggested refined framework focuses on the operational level, which is conceptualized into two arenas of choices. At the individual- choice level, there are a set of intra-organizational issues to be decided by organizational actors. At the collective-choice level, inter-organizational areas of choice (i.e. interaction with other state actors) are considered. 8.1.3 New Contextual Evidence on Young/Emerging Donors: What Bureaucratic Incentives and Collective Actions are Observed in Unconventional Donor Governments Given that the previous IAD research on foreign aid policy process are centered on the Western, established donor governments and multilateral organizations, it is attempted here to add knowledge about a group of new donor governments. This study highlights the issue of ‘big-plan-small-action’ with some donors’ aid giving behavior. Such behavior is equally shown with established but emerging donorship, albeit to varying degrees. As the global development community’s concerted efforts are being 182 refocused for making global development cooperation further sustainable, it is critical to understand the fundamental problems and dilemmas that are associated with donors’ discrepancy between planning and practice. It is also very important to discuss some feasible resolutions for donors’ sustainable supply and management of foreign aid. Distinctively, Korea has mixed features as a DAC member country and as an emerging donor. Those features are very useful for this research. Also, this research has benefited from the relatively high data availability of Korean ODA (when compared to that of other emerging donors); a good volume of data on Korean ODA, including data that capture policy activities and ODA performance at multiple levels of aid policy- making, were available through OECD statistics151. In addition, the county also represents a critical case to show how the nascent rule arrangements of a donor government might affect its policy making process and policy outputs.

8.2 Evaluation of the ‘Rule-Policy Output’ Nexus This section lays out the overall findings on the bureaucratic incentives and collective actions that are induced under the Korea’s unique set of rules, and explains which are directly and indirectly accountable for the commitment problem. In doing so, the following section highlights, once again, the rule configurations and the commitment problem in both cross-sectoral (overall ODA) and sector-specific manners (through knowledge and multilateral aid). 8.2.1 Identification of the Hard Rules for ODA Policy-Making In comparison with other DAC donors, Korea’s legal and regulatory frameworks display relative insufficiencies in content and definitional detail (OECD 2012). Currently, however, Korea’s foreign aid system is rapidly evolving through the frequent modifications of laws, and through restructuring. Considering an array of formal institutions for Korean ODA policy-making—the Framework Act on International Development Cooperation (Framework Act), the Enforcement Decrees of ODA-pertaining laws (Presidential Decrees), the Act on the Measures for the Admission to International Financial Institutions (IFI Act), the Foreign Economic Cooperation Fund Act (EDCF Act), the Korea International Cooperation Agency Act (KOICA Act), the Strategic Plan papers, the Mid-term Strategy papers, the Annual Implementation Plan papers, and the Country Partnership Strategy Papers (those

151 For further information with regard to this specific research agenda, OECD introduces a dataset, Survey on Donors Forward Spending Plans, that directly speaks to donor governments’ performance in planning and disbursing ODA. However, the dataset is yet to offer full information on Korea. 183 policy documents are available in English at www.odakorea.go.kr)—, those formal rules structure the incentives and deterrents for aid bureaucracy’s policy activities in a variety of ways. 8.2.2 Rule Configuration: The Centralized Decision Powers and the Non-restrictive Style and Broad Wording of the Laws With regard to the configuration of Korea’s ODA-pertaining laws, some particular categories of rules either have broad wordings or are largely absent, thereby leaving the given policy decisions/actions to the discretion of the president, and supervisory and implementing agencies. The bulk of the decision-making and corrective powers in ODA policy-making lies with the president. More specific and operational decisions are exclusively decided by MOFA and MOSF. The CIDC is allowed to wield only post ex powers over the performance of the supervising and implementing agencies. As for the boundary rule category, the country has non-restrictive laws in controlling the entry and exit of participants, and does not articulate the specific conditions for state agencies to initiate or terminate their operations of development intervention. No human/organizational attributes with relevance to international development cooperation are prescribed in the laws. The laws recognize a relatively small group of actors for international development cooperation (limiting the key actors to state agencies only). Regarding the rule category that defines positions and delegation (i.e. the organizational structure of ODA management), there is a relatively clear set of laws in place. A chain of command (forming a top-down structure from the President, to the CIDC, then to MOFA/MOSF, then to the implementing agencies) is prescribed. When it comes to authority rules, however, the overall and organization-specific objectives/mandates might be defined in a broad manner. In fact, the multiple ODA objectives for overall ODA policy include almost all objectives that are discussed at the international level. Some of the objectives are considered incompatible with one another. No priorities are set among those objectives. For the scope rules, on the other hand, there are a few legal institutions that effectively and substantially delimit the range of possible outcomes, leaving the matter, instead, to presidential discretion. For the division of labor and resources, there is a simple logic imposed in that MOSF supervises loans/IFI cooperation and MOFA supervises grants/non-IFI cooperation. Other types of ODA, besides grants and loans, are not defined in terms of coordination, monitoring, or evaluation. The leverage of MOFA on technical assistance has been diluted in recent years as an effect of the modifications of ODA laws 184 and regulations. The rules for aggregation (coordination) have been revised to gradually strengthen the autonomy of the CIDC. However, its decisions can still be vetoed effectively by the president and the two supervisory ministries. Noticeably, the payoff and information rules remain very weak. Those rules largely fail to place restrictions over the undesirable actions of implementing agencies, and they have failed to develop a nation-wide or government-wide information platform for information sharing and effective monitoring. 8.2.3 Soft Rules: The Fear of Losing Face and the Tradition of Aiming High With regard to the regulatory/normative frameworks that are assumed to determine bureaucratic behavior in the areas of administrative discretion, the empirical analyses (as explained in Chapters Five, Six, and Seven) reveal that two bureaucratic mindsets—the fear of losing face in the international community and the preference for setting ambitious policy goals—have exerted great influence among foreign affairs and economic bureaucrats in the aid planning phase and have been incorporated into their respectively-championed policy discourses, according to the interviews. Setting unrealistic aid targets has been clearly witnessed in the aid-managing ministries since Korea started its first small-scale development partnerships in the 1970s and 80s. Traditionally, in the multilateral diplomatic sphere, the Korean government has made great efforts to enhance the country’s image in a race for regime legitimacy against the north; Korea’s diplomatic strategies have often entailed overambitious pledges in the international settings. Even nowadays, these culturally and traditionally driven rationales for ODA expansion remain. These traditions are seen counter-productive when competent, knowledgeable leadership is missing in low-profile policy areas such as ODA. 8.2.4 How the Rules Have Shaped the Behavior of State Organizations All in all, while some areas of ODA management (e.g. formulation of Basic Plans and Annual Implementation Plans, and reporting and monitoring between supervising and implementing agencies) are relatively well-defined, other important areas (e.g. ODA scale, selection of funding sources, designated agencies, designated tasks, and information sharing) are left largely to the discretion of the president, or MOSF and MOFA, rather than being decided by solid policy guidelines. The existing legal framework, if understood in its configuration by the seven-rule- category system, tends to induce opportunistic-competitive behavior, rather than routinized behavior, among implementing agencies at their individual decision level. First of all, within the legal settings in which ODA scale and geographical allocation are decided by the discretion of the president (by issuing president decrees), the centralization 185 of aid decision-making power with political leadership generates greater uncertainty for the budgetary and political survival of implementing agencies at times of regime change. However, empirically speaking, regime change does not appear to serve as the primary factor in dramatic policy change or reorientation, during the period from 2000 to 2016. Furthermore, as the critical feature of the rule configuration, the division of labor and resources among state agencies is set simply by aid modality (grants vs. loans). That leaves great scope for bureaucratic reinterpretation and competition for the undefined areas of ODA planning and operations. For instance, the leading implementing agencies, KOICA and EDCF, both of which together absorb over 70 percent of the total ODA budget, show somewhat reactive and opportunistic patterns of behavior. In particular, as revealed in the analysis, KOICA tended to internally reallocate resources to high-profile or fashionable ODA practices (for instance, through its introduction of DEEP and multi- bi operations), as opposed to its areas of specialization and its organizational traditions. This pattern of behavior is confirmed through the two nested cases. At the collective-choice level, in which state agencies interact one another for information-sharing and policy/operational coordination, the legal structures do not provide a strong or coherent logic for the effective coordination and information management across the board. In 2008, although the coordinating body for ODA policy was established under the Prime Minister’s office, the CIDC, made up of the heads of ministries and civilian expert members, is legally given no proper authority or resources that are required for effective monitoring, evaluation, and enforcement of policy recommendations. The task of ODA statistics management is delegated to one of the offices at the finance ministry. The location of information-gathering prevents rival ministries and agencies from sharing or reporting information with the currently assigned agency. As a result, the opportunistically motivated organizational actors might collectively have led to policy outputs that were effectively expansionary in overall volume but overlapping aid projects overtime. 8.2.5 The Limitations of Finding the Institutional causes of the Commitment Problem Speaking of the policy patterns as the final, long-term outcomes of the ODA policy process, Korea has suffered from an intensive, persistent form of the commitment problem for decades, together with Japan, France, and Germany. However, the problem has not been seriously discussed in Korea since the importance of the issue might have been dwarfed by the country’s remarkable record of ODA scale-up. 186 Regrettably, the direct cause-and-effect relationship between the rules and the commitment problem might be reconstructed to a limited extent, through a few quantitative measures. To fill the knowledge gap, however, this qualitative analysis has highlighted the hidden mechanisms of how soft rules come into play to induce the particular patterns of collective action and explain the planning-practice gap in Korea’s development cooperation. The lack of the quantitative data in this study was supplemented by a great deal of qualitative data collected from a comprehensive review of government documents and person-to-person interviews with the policy elites and experts. Nonetheless, this study offers no explicit findings that further legal modifications (through imposing more rules) would lead to the expected results of alleviating the commitment problem in the Korean aid system. 8.2.6 Knowledge Aid and Multilateral Aid: Overall Evaluation To provide an overall evaluation of the rule-policy output nexus in the case of knowledge aid policy-making, Table 8.1 summarizes the findings from the analyses of the rule configuration (i.e. the location of decision/corrective powers, coordination, information and the contents and style of legal institutions), the patterns of bureaucratic behavior in the individual- and collective-choice situations, the level of coordination, and the outcomes of the policy process. Likewise, the following table (8.2) provides the overall evaluative findings from the analysis of the rule-policy output nexus in the case of multilateral aid policy-making. 187

Table 8.1 The Rule-Policy Output Nexus on the Knowledge Aid Policy Process (by author)

Rule Category Individual Action Hard Soft Collective Action Situation [Directly Affected Situation Impacts on Policy Outputs Rule Rule & Pattern of Collective Action Elements of [Directly Affected Actors] Action Situations] [Implementing Agencies] Boundary Rules Non- Discretio Easy to join and exit; No. of Reallocate resources to initiate and manage more of knowledge Proliferation of aid operations [Actors] restrictive nary participants grows projects Position The relatively flat [Supervisory & Implementing Agencies] No particularly non-passive collective Rules Fixed X organizational/delegation structure Routinely report and monitor the performance of state agencies actions observed [Delegation] streamlines aid management [State Executives] Authority/Scope Favor expansionary knowledge aid policies in his/her rules diplomatic/industrial interests overseas Aid scale for knowledge aid Bureaucratic competition escalates [Division of Non- Discretio [Supervisory Agencies] constantly expands; Non- to take the lead in the knowledge aid Labor and restrictive nary Justify their exclusive authority over knowledge aid operations conventional aid practices widely sector Resource & [Implementing Agencies] introduced into the ODA system Planning] Reallocate resources and reorient strategies for further promotion of knowledge aid No systematic platforms to Aggregation [Coordinating Body] Little incentives for state agencies to evaluate and monitor the Rules Partial Hard to monitor and evaluate due to insufficient resource and reflect the recommendations from performance of overall ODA [Coordination] exclusive leverage over supervisory ministries CIDC operations in the long term Information Rules [Implementing Agencies] Little incentive to share Overlapping operations of Non- Discretio [Reporting & Develop their own information platforms; report their performance to information/knowledge between knowledge projects; No integrated restrictive nary Knowledge the principal supervisory agencies information management Management] Payoff [All Agencies] Potential for corruption and other Rules Non- Discretio Discretionary use of resources Be loosely vigilant at monitoring internal affairs of sectoral allocation non-accountable ODA operations [Oversight & restrictive nary encouraged of budget, employment, and dispatch of staff to field offices increased Penalty]

188

Table 8.2 The Rule-Policy Output Nexus on the Multilateral Aid Policy Process (by author)

Rule Category [Directly Affected Individual Action Situation Collective Action Situation Hard Rule Soft Rule Impacts on Policy Outputs Elements of Action [Directly Affected Actors] & Pattern of Collective Action Situations] MDB funding is confined to a specific group of actors, Boundary Rules No. of state agencies and projects for Fragmentation observed Partial whereas U.N. funding is widely available to state [Actors] multilateral cooperation grows particularly with U.N. funding agencies [Supervisory & Implementing Agencies] The relatively flat delegation Position Rules No particularly non-passive collective Fixed X Routinely report and monitor the performance of state structure streamlines aid [Delegation] actions observed agencies management [State executives] Favor expansionary knowledge aid policies in his/her diplomatic/industrial interests overseas Authority/Scope Rules Presidential initiatives for [Supervisory & Implementing Agencies] No bureaucratic competition observed [Division of Labor and Partial expansionary multilateral funding For MDB funding, agencies perform their roles with multilateral funding Resource & Planning] are routinely reflected routinely; But for U.N. funding, some agencies increase their multilateral cooperation to enhance their prominence No consensual policy discourse Aggregation Rules Non- [A Coordinating Body] Little coordination and cooperation among Discretionary in the provision of multilateral [Coordination] restrictive No effective coordinating role performed state agencies observed funding Information Rules [Implementing Agencies] Weak incentive to share [Reporting & Non- No integrated information Discretionary Develop their own information platforms; report their information/knowledge between Knowledge restrictive platform available performance to the principal supervisory agencies Management] [All Agencies] Potential for corruption and other Payoff Rules Be loosely vigilant at monitoring internal affairs of Discretionary use of resources Partial non-accountable ODA operations [Oversight & Penalty] sectoral allocation of budget, employment, and encouraged increased dispatch of staff to field offices

189 8.3 Extended Discussion: Implications of Bureaucratic Discretion and Incentives Beyond the Case of Korea This section includes an extended discussion on how the findings of the Korea- specific study can interact with policy process research on a wider range of aid modalities and country cases. As far as a diverse set of rule configurations, aid modalities, and the national experiences of laws and development are concerned, this section provides some points for future study. 8.3.1 Broader Rule Contexts and Commitment Problems to be Considered How do rule configurations matter in order for a donor country to make credible commitments to global development cooperation? This issue is relevant not only to Korea but also to other donor countries152. To further understand the relationship between rules and policy products, a cross-country analysis of rule configurations is required. A study on Regulation of Foreign Aid in Selected Countries 2011/12 (GLRC 2012) that delves into the legal foundations of 18 donor countries’ ODA management and delivery153 provides a good starting point for future study. Also, in order to draw in-depth implications from the Korean case, a comparison with other small donors in terms of their rule configurations and the performances of making a credible commitment—such as, Finland (87.8), Belgium (88.8), Denmark (106.8), Poland (90.4), and Ireland (100)— would be useful. 8.3.2 A Wider Range of Aid Modalities to be Considered Different types of aid might induce varied levels of the commitment problem. As oppose to initial expectations (about the hypothesized effects of the dramatic expansion in ODA resources), the chosen nested cases of knowledge and multilateral aid were not the cases that were particularly problematized by a gap between planning and practice. Although the problem of fragmentation has been exacerbated substantially with the two aid modalities (i.e. knowledge aid and multilateral aid), a bigger commitment problem does not occur. A possible reason is that those ODA activities are carried out on a relatively small scale compared to other major development interventions (in the forms of concessional loans and grants). Also, this indirectly implies that the commitment problem seems severe in the ODA sectors that do not require any coordination.

152 Donors with severe commitment problems (when considering the last decade records of DAC donors): Japan (49.8), Germany (76.0), and France (78.3). 153 The 18 countries include the following seven non-DAC countries: China, Brazil, Kuwait, Israel, Russia, Saudi Arabia, and South Africa. 190 The commitment problem is highly relevant to the size of aid projects. Compared to other types154 of ODA, project-type ODA, which is operated at the largest scale, shows the most severe commitment problem (See Table 8.3). Between grant and loan, loan projects tend to show a larger gap between planning and practice across the DAC and Non-DAC donor countries (See Table 8.4).

Table 8.3 Commitment Problem with Project-Type Aid in Korea

Year Commitments Gross Disbursements Performance

2006 453.1 188.0 41.5

2007 742.0 261.9 35.3

2008 1,109.3 279.8 25.2

2009 1,166.4 366.2 31.4

2010 1,585.9 699.3 44.1

2011 1,331.6 724.5 54.4

2012 1,334.6 812.1 60.9

2013 1,688.5 884.9 52.4

2014 1,860.6 964.6 51.8 2015 1,800.3 1,033.6 57.4

(Source: OECD CRS; Accessed Nov. 25, 2017) Table 8.4 Ratios of Net Disbursements to Total Commitments (Bilateral Aid to All Developing Countries)

Performance 2011 2012 2013 2014 2015 DAC Grants 105.0 105.7 104.4 101.6 97.0

Countries, Loans 86.0 64.9 71.7 67.0 61.2 Total Non-DAC Grants 96.8 38.1 67.6 347.0 473.4

Countries, Loans 133.0 68.2 73.8 45.4 69.7 Total Korea Grants 100.9 111.4 77.9 80.0 83.2

Loans 43.6 46.6 47.2 46.6 52.6

154 OECD’s Creditor Reporting System recognizes nine types of aid: budget support; core contributions and pooled; project-type interventions; experts and other technical assistance; scholarships and student costs in; debt relief; administrative costs; other in-donor expenditures; and not-applicable. In 2015, global ODA is disbursed in the following types (share): budget support (0.6); core contributions and pooled (8.9); project-type interventions (66.7); experts and other technical assistance (13.6); scholarships and student costs (4.5); debt relief (0); administrative costs (4.2); other in-donor expenditures (1.3); and not applicable (0). 191 (Source: OECD CRS; Accessed Nov. 25, 2017)

8.3.3 Regional/National Experiences of Laws and Development to be Considered This study is relevant to the national experience of laws and development for donor countries, as well. East Asia, in particular, is the region with countries whose economic development has been successfully mobilized without substantial developments of public and private laws. In particular, with regard to the East Asian contexts of the public policy process (with respect to public laws and administrative laws), the region is considered distinctive because the heads of state and the state bureaucracy in the region tend to be given great legal autonomy (Pistor, Wellons, and Sachs 1999; Ginsburg 2000; Jauasuriya 2006; Ohnesorge 2016). The highly centralized structure of public policy-making along with the broad wording of rule statements, has been continuously sustained in the region. Through future IAD research, the unique legal- regulatory contexts in East Asia can be further explored to highlight the relationships between legal institutions and policy performance on public policy process.

8.4 Concluding Remarks This IAD study sheds light on the collective action problems that arise in the management of public resources. The legal-regulatory arrangements for Korea’s ODA policy-making have shaped incentives and collective actions among state organizations that are inducive to the expansionary but highly fragmented ODA practice over time. The Korean ODA governance has a dualistic structure, which might not be particularly unique across the young and traditional donor community. Through the in- depth analysis of the organizational structure and legal foundations—featured with the exclusive powers of the President, the shared supervision between MOFA and MOSF, and —it is revealed that this distinctive configuration of rule institutions serves as the major explanatory factor not only for Korea’s impressive aid expansion records but also for the relatively high level of in-donor fragmentation and policy incoherence in comparison with other DAC donors. Above all, the legal and social mechanisms of coordination have been absent from the early years of the aid policy system onward, so that the gap between aid planning and aid implementation has been continuously substantial in Korea. The Korean ODA system appears to be the policy-making environment, one which hardly fosters reciprocity among state organizations. Without the intervention of strong leadership, the coordination and 192 cooperation between ministerial and other state agencies has rarely been induced not only in the ODA policy domain but also across the public policy areas. Given the complexity associated with the institutional causes for the sustained policy behavior of a donor government, this research might be revealing the political dynamics of ODA policy-making only in a limited way. From the new institutionalist perspective, however, this research pinpoints that to understand the commitment problem with Korean ODA, due attention should be given to the unique institutional arrangements of Korea’s public policy-making, in which the decision and corrective powers are highly centralized with the President, and inter-agency coordination is left largely to his/her discretion and capacity. Reflecting on the recent scandal of the impeached former president, Park GH, and her confidante’s leverage over various public policy areas, including ODA, this argument might be no longer so far-reaching regarding the issue of the heavily centralized public policy system in Korea. This research speaks to the wisdom of neo-institutionalists: Rational choices of individuals make an irrational choice of all. I have little doubt that there is a group of highly committed policy makers and experts within the Korean government putting great efforts to make their development intervention more meaningful. Despite the good intentions of those development professionals and organizations, this research captures that the Korean ODA system is also vulnerable to making unintended, ineffective outcomes of ODA management due to its structural and motivational constraints. Although this study is intended to highlight the specific areas for potential institutional restructuring that are likely to create positive incentives for inter-agency cooperation and long-term reciprocity within the system, the remaining research question would be whether further imposition of restrictive legal institutions might be the best way to solve the problem to mitigate the unintended or even negative effects of such a centralization of power in the management of public resources.

193 Appendix

A. Other delegation models (with reference to Chapter Two) Analysis of Micro-Level Incentive Problems First of all, Martens and his colleagues (2002), a group of institutional economists, carried out a collaborative study of human/organizational incentives that harm aid effectiveness within a bilateral aid agency. Their theoretical discussion was motivated by a request from the European Commission (EC). Through the funded study, the EC intended to evaluate its development interventions, the Phare and Tacis Projects, which were implemented in several transitional economies in Central and Eastern Europe in the 1990s.

Figure A-1. Foreign Aid Delegation Chain in the Study of Martens and others: What Informational/Motivational Problems Occurred at Respective Delegations with Donor Agents

(Adapted from Martens et al. (2002, p.76))

To address underlying issues associated with aid effectiveness, this group of scholars argue that a fundamental determinant of aid performance might lie in the donors’ systems. In part, the research also highlights a striking absence of incentive analysis of donors’ organizations in existing aid performance studies. The study reveals that EC’s policy-makers and project managers had shared views on some routinized patterns in their practices, regardless of geographical and thematic distributions of aid. Such observed similarities in aid outcomes allowed the scholars to infer the working mechanisms of incentives and constraints in the process of aid policy formation and implementation. The study is primarily focused on incentive problems that exist inside donor organizations and that directly harm overall aid performance. The analysis looks at micro-level decision-making processes, paying attention to behavior (incentives and constraints) of individual human agents. According to this research, the dominant problems with EC’s aid agencies and staff were highlighted in three forums: (i) the characteristics of public organizations; (ii) a broken information feedback loop; and (iii) types of aid project. First, the nature of public organizations per se contributed to inherent agency problems with EC’s aid delivery process. Compared to private or non-profit organizations, public organizations tended to work on multiple objectives that were supported from multiple constituencies. The objectives for foreign aid giving, in particular, are relatively hard-to-measure and mutually exclusive. Therefore, such a mixed set of aid goals was hypothesized to put significant constraints on aid practitioners. On this point, Seabright (2002) demonstrates that the multiplicity of aid goals and tasks is highly likely to allow aid practitioners to disproportionately 194 focus on a particular set of aid activities that could better serve their budgetary and resource-use interests. Second, it is the imperfect feedback loop between final principals and final beneficiaries in the foreign aid delivery setting that inherently prevented the EC policy-makers from making better policy choices. Given the unique context, in which the funder and the beneficiary of foreign aid are geographically and politically isolated, it is less likely that effective information-sharing takes place between the two. Hence, a suggestion made to solve the informational and incentive problems was to equip donors’ aid delivery systems with reinforced monitoring and evaluation mechanisms. Third, as foreign aid is disbursed in various forms, the research reveals that different types of foreign aid induced different degrees of incentive problems. For instance, technical assistance projects, compared to infrastructure building projects, are prone to produce less tangible outputs so that a greater degree of moral hazard and adverse selection problems was observed with EC aid practitioners. Overall, this micro-level incentive research allowed us to turn to the so-far marginalized issue of incentive problems in donors’ organizations. However, this type of research appears hard to conduct or reproduce. Without substantive determination and efforts of aid organizations to self-evaluate, relevant data are rarely collected or openly available. Also, the research, since largely based on qualitative data collected from interviews and surveys, is often challenged in terms of objectivity. The study also has a weakness in not providing a holistic understanding of incentive problems within the system.

Analysis of ‘Donor Government-International Organization’ Relationships Hawkins and others (2006) suggest another delegation/incentive analysis that is inspired by the game-theoretical, international relations (IR) notion. The work focuses on the creation and dynamics of a delegation between donor governments and international organizations—in other words, to explain why governments fund/join multilateral organizations and why some governments actually prefer to give development aid through multilateral organizations at the expense of disbursing bilateral aid (called, aid multilateralism). More specifically, those IR theorists raised the following questions of under what conditions donor governments delegated aid delivery to international organizations (IOs); through what mechanisms and apparatuses funding governments, as the principals of delegation, controlled the IO agents; and what strategic responses were observed with IOs against the control of the principals. This research initiative arose in the late 1990s, coinciding with a rise in anti-globalist sentiment around the world. The anti-globalization movement largely questioned the democratic legitimacy of global institutions, and criticized the neo-liberal ideology championed by the World Trade Organization, the World Bank, and the International Monetary Fund. Responding to the public skepticism, this study arose in an attempt to delve into state governments’ rationale and control over the performance of IOs. This issue was importantly to discuss because donors’ decision to transfer resources to IOs is held accountable by its taxpayers. In this regard, the significance of this study lies in its orientation toward aid accountability.

195 Figure A-2. Delegation Chain in the Study of Hawkins and Others: A Delivery of Multilateral ODA

(Sources: Hawkins et al. 2006, p. 379)

To explain and predict the interactive behavior of nation-states and IOs in delegation, both governments and IOs are equally assumed to be strategic-rational actors. Regarding the power relationship between donor governments and IOs, the study lays out the potential benefits of state-IO delegation in terms of enhancing policy credibility, utilizing expertise of IOs, and avoiding domestic criticism over its non-altruistic ODA operations. Major implications from the study are boiled down to two points: (i) redefining multilateral funding from a delegation perspective and (ii) highlighting IO agents’ strategic behavior and incentive problems. First, this research newly identifies a series of delegation-ships for national governments’ multilateral membership/funding. The identified PA chains that stretch between private actors (end principals) and IO (end agents) form a rather non-hierarchical structure, as shown in the above figure. This unique power structure, in contrast to other hierarchically structured delegation types for bilateral aid management and delivery, might induce disparate types of rationales and interaction between the principal and agents. In the specific case of the influence of US domestic politics over IMF policy-making, the voters and interest groups exercised their influence simultaneously over the Congress, political leadership, state bureaucrats, and IOs. Importantly, the multilateral aid delegation has distributional effects, which benefit domestic principals back through IOs according to the collection of empirical analyses in Hawkins et al. (2006). Under this model, foreign aid is understood as a tool that delivers the preference of end principals (private actors) and generates benefits in return. Also, the nation-states turn out to be the key decision-makers for potential contracts with IOs, and monitoring actors on IO performance. Their strategic decisions are made in reflection of their preferences. Therefore, states’ preferences (including strategic calculations of the costs and benefits of initiating and expanding a delegation relationship with an IO) are the critical factor in this model. To the second point, while conventional ways of understanding PA relationships and delegation outcomes are largely based on the characteristics of principals, this collaborative work adds new knowledge about the behavioral characteristics of agents and the resulting impacts on delegation outcomes through several IO cases. In contrast to the traditional perception of agents who are considered principal-dependent, their IO cases suggested counter-evidence. IOs’ proactive-strategic reaction to their member countries—called agent opportunism—was well observed (Hawkins and Jacoby 2006; Lyne et al. 2006; Martin 2006). In the work of Hawkins et al. (2006), such opportunism of IOs was motivated to increase their autonomy. One of the frequent strategies taken by IOs is emphasizing their specialization. This strategy appeared to work in effectively influencing governments’ decision regarding the joining or funding of multilateral schemes. As far as the overall implications of the proposed model are concerned, the research represents a new attempt to adopt agency theory in the discipline of international relations. What 196 is mainly argued is whether governments’ decisions about IO delegation depend not only on preferences of state principals, but also on opportunistic behaviors of IO agents. To curb the IO opportunism that goes against the interest of taxpayers in donor countries, governments’ control mechanisms matter. Although a focused examination of the power interplay between nation-states and IOs through delegation theory enables us to understand the dynamics of non-hierarchical delegation at the global level, the presented evidence and theorization might have limitations in their application for other types of delegation.

B. Sample Questionnaire for Interviews (with reference to Chapter Four)

Part 1: Overall Perspective on Korea’s Aid System: Motivations and Structure

1. [Aid motivation] In your opinion, what are the top priorities and principles of Korea’s ODA policy? Based on what sources do you prioritize or define these aid motives?

2. [Momentum or explanation for aid expansion] In your opinion, what are the main motives or explanations for Korea’s drastic aid growth? What do you think of Korea’s transformation into a recognized donor? Do you think what aspects of Korea’s aid policy system allowed ODA expansion? Did Korea’s aid expansion lead to significant changes to aid management system, and vice versa?

3. [Aid Management Design] What do you think of the extent to which Korea’s aid system is optimized to deliver the aid motives; or in other word, how well the stated aid motives are reflected into the aid management system?

4. [Aid Management: the role of state agencies] Who do you think are most active or influential actors (or state agencies) in the aid policy arena and in what sense and in what degree? How is foreign aid policy related to other fields of international policy in terms of state bureaucrats’ areas of autonomy and organizational interest?

5. [Aid Management: involvement of non-state actors in the policy making process] With regard to domestic debate on ODA, what are the major issues? What is the role of non-governmental actors (e.g. the media, the academics, NGOs, businesses, etc.)?

Part 2: Inter-Organizational Interactions within the Korean Aid System

6. [Inter-organizational interaction] When do you think inter-agency cooperation is necessary and critical (or when it is mandatory?) For policy or operational cooperation (including information sharing and organizational learning), how and how often does it happen among state agencies or with international agencies?

7. [Bureaucratic competition] Regarding Korea’s dualistic structure of the current aid systems, what possibly explains the emergence and persistence of the dualistic structure?

8. [Role of a coordinating body] What do you think of the contributions and limitations of an aid policy coordinating body, the Committee for International Development Cooperation (CIDC)?

Part 3: Intra-Organizational Perspective (In-House Discourse)

9. [Aid agencies’ reinterpretation of their organizational objectives and autonomy] How does your affiliated agency define the overall objective of aid 197 giving and its assigned mandates? Do you think the given (official) mandate is well suited to the organizational talent, capacity, and priority?

10. [Aid agencies’ strategic interest in knowledge and multilateral aid] What can explain the increasing interest of aid agencies in the promotion of knowledge and multilateral aid in recent years?

11. [Aid agencies’ perception and management of uncertainty with regard to aid management design] What are the key concerns or agendas for your organization (or department) in terms of budgetary constraint, employment, overseas operations, public relations, monitoring & evaluation, communication with development partners, etc.? What kind of strategies are discussed and implemented at the organizational (departmental) level?

12. [Aid agencies’ performance of making reliable aid delivery] (with relevant figures if possible) To what extent your organization meets its annual aid target (commitments), if not, why is that?

13. [Internal policy-making processes in aid agencies] Who are the aid target setter and implementing bodies within your organizations? How should one communicate with other organizations in the processes of procuring, designing, delivering, evaluating and reporting aid operations? (e.g. ministries, other governmental agencies, aid implementing agencies, the private sector, NGOs, international development partners, etc.)

14. [Alternative aid management design?] What reforms and structural changes might be being discussed within your organization?

C. Korean ODA Allocation by (a) Income and (b) Region (Average, 2010-11) (with reference to Chapter Five)

(a)

(b) 198

(Source: Adapted from OECD/DAC (2012))

D. Top Recipients of Korean ODA (2011) (Unit: USD million) (with reference to Chapter Five)

Rank Concessional loans Grants Total 1 Vietnam (109.4) Mongolia (32.6) Vietnam (139.5) 2 Bangladesh (70.8) Vietnam (30.1) Bangladesh (80.0) 3 Cambodia (39.1) Afghanistan (28.0) Cambodia (62.2) 4 Sri Lanka (31.9) Philippines (25.8) Sri Lanka (43.4) 5 Jordan (26.9) Cambodia (23.1) Philippines (35.7)

(Source: Author’s computation based on ODA Statistics (EDCF 2014))

E. Laws and Regulations Pertaining to ODA Management and Delivery in Korea (with reference to Chapter Five)

Subject of Law Content clause Poverty reduction; making a better human rights environment for women, children and the disabled; gender equality; sustainable Objectives development; humanitarianism; and promotion of global peace and prosperity and economic cooperation Definition of Identified aid actors are Korean central, local governments and public International agencies; Developing countries and strategic partner countries for Framework Act Development aid beneficiaries are defined and decided by the government and the on International Cooperation Committee (CIDC). Development (i) respect the U.N. pledges; (ii) support self-help and capacity Cooperation155 building within recipient countries; (iii) respect developmental needs (Enacted in Principles in developing countries; (iv) share Korean experience with the 2009) developing world; and (v) boost international partnerships and harmonize with the global develop cooperation community CIDC is composed of 12 ministries, heads of Korea Eximbank and KOICA, and non-governmental members (9 personnel) and chaired Coordination by the Prime Minister’s Office; as a coordinating platform, the CIDC sets a comprehensive plan for ODA and decides on evaluation details

155 Available at http://www.law.go.kr/lsInfoP.do?lsiSeq=160744&efYd=20150416#0000 199

A loan-grant division; cooperation with international financial Division of institutions is in charge of MOSF and other global cooperation in Labor charge of MOFA Scope of Overall volume of ODA; selection of strategic partner countries for Presidential mid-term strategy; and other critical issues that are dealt with CIDC Intervention Presidential Annually planned respectively by implementing agency and reported Decrees on the Planning to supervisory ministries Framework Act Evaluation (Enacted in and Criteria and methods of ODA evaluation, reporting details to the 2010) Information National Assembly, and the scope of information disclose to the Disclose public are decided exclusively by presidential decrees policy Law on the Defines procedures for the Korean government to join international Objectives Measures for finance institutions and to perform its duties the Admission Scope of to International 15 international finance institutions (IFIs) Partnership Financial 156 Details on planning and performance with regard to capital Institutions Scope of subscriptions to IFIs are decided by presidential decrees; MOSF and (Enacted in Authority the Governor of the Bank of Korea are delegated to perform the 1963) capital subscriptions to IFIs, report to the National Assembly Support industrial development and economic stability of developing EDCF Law Organizational countries and enhance the economic cooperation between partner (Enacted in Objectives countries and South Korea; report the performance of loan 1987) operations to the Minister of Strategy and Finance Enhance bilateral relationships with developing countries and KOICA Law Organizational economic, social development in those countries and international (Enacted in Objectives partnerships; report the ODA performance to MOFA; subsequently 1991) reports to the National Assembly (Source: Korea Ministry of Government Legislation; Summarized by author)

156 Available at http://www.law.go.kr/lsInfoP.do?lsiSeq=165465&efYd=20141230#0000 200 Bibliography

Abbott, A., 1992. An old institutionalist reads the new institutionalism. Contemporary Sociology, 12(6), pp.754-756.

ADB, 2014. Asian Development Bank & Republic of Korea: Fact Sheet. [Online] Available at: http://www.adb.org/publications/republic-korea-fact-sheet [Accessed Dec. 22, 2017].

Addison, T., McGillivray, M., & Odedokun, M. 2003. Donor funding of multilateral aid agencies: Determining factors and revealed burden sharing. WIDER Discussion Papers.

Aghion, P. and Tirole, J., 1997. Formal and real authority in organizations. Journal of political economy, 105(1), pp.1-29.

Aldasoro, I., Nunnenkamp, P. and Thiele, R., 2010. Less aid proliferation and more donor coordination? The wide gap between words and deeds. Journal of International Development, 22(7), pp.920-940.

Alesina, A. & Dollar, D., 2000. Who gives foreign aid to whom and why? Journal of economic growth, 5(1), pp.33-63.

Aligica, P. D., 2006. Institutional and stakeholder mapping: frameworks for policy analysis and institutional change. Public Organization Review, 6(1), pp.79-90.

Aligica, P. D., 2013. Institutional Diversity and Political Economy: The Ostroms and Beyond. Oxford University Press.

Allison, G., & Zelikow, P. 1999. Essence of Decision: Explaining the Cuban Missile Crisis (2nd ed.). Longman.

Amsden, A.H., 1992. Asia’s Next Giant: South Korea and Late Industrialization. Oxford University Press.

Andersson, K., 2009. Motivational dilemmas in collaborative learning activities: The case of the Swedish International Development Cooperation Agency (SIDA). Public Administration and Development, 29(5), pp.341-351.

Antons, C. ed., 2005. Law and Development in East and South-East Asia. Routledge.

Aranson, P.H., Gellhorn, E. and Robinson, G.O., 1982. Theory of legislative Delegation. Cornell L. Rev., 68, p.1. 201 Barder, O., 2009. Beyond planning: markets and networks for better aid.

Bendor, J., Glazer, A. and Hammond, T., 2001. Theories of delegation. Annual review of political science, 4(1), pp.235-269.

Bennett, A. and Checkel, J.T. eds., 2014. Process tracing. Cambridge University Press.

Bennett, A., 2010. Process tracing and causal inference.

Birkland, Thomas A. 2010. An Introduction to the Policy Process. Ed. Armonk, NY: M.E. Sharpe.

Boden, D., 1994. Agendas and arrangements: Everyday negotiations in meetings. In The Discourse of Negotiation (pp. 83-99).

Brady, H.E. and Collier, D. eds., 2010. Rethinking social inquiry: Diverse tools, shared standards. Rowman & Littlefield Publishers.

Bräutigam, D., 2011. Aid ‘With Chinese Characteristics’: Chinese Foreign Aid and Development Finance Meet the OECD-DAC Aid Regime. Journal of international development, 23(5), pp.752-764.

Breuning, M., 1995. Words and deeds: Foreign assistance rhetoric and policy behavior in the Netherlands, Belgium, and the United Kingdom. International Studies Quarterly, 39(2), pp.235-254.

Broz, J.L. and Hawes, M.B., 2006. U.S. domestic politics and international monetary fund policy. Delegation and agency in international organizations, pp.77-106.

Bulíř, A. and Hamann, A.J., 2008. Volatility of development aid: From the frying pan into the fire?. World Development, 36(10), pp.2048-2066.

Cairney, P. and Heikkila, T., 2014. A comparison of theories of the policy process. Theories of the policy process, 3.

Cairney, P., 2011. Understanding public policy: Theories and issues. Palgrave Macmillan.

Calder, Kent E., 1988. Japanese Foreign Economic Policy Formation: Explaining the Reactive State. World Politics 40 (4): 517-541. 202 Celasun, O. and Walliser, J., 2008. Predictability of aid: Do fickle donors undermine aid effectiveness?. Economic Policy, 23(55), pp.546-594.

CGD, The commitment to development index 2017. [Online] Available at: https://www.cgdev.org/commitment-development-index-2017#CDI [Accessed Dec. 30, 2017]. Chang, H., 2014. ODA and Korean assembly: oversight accountability of Korean assemblymen in ODA budget policy’ [in Korean]. The Korean Association of Party Studies, 13(2), pp. 255–85.

Checkel, J.T. and Bennett, A., 2015. Beyond Metaphors: Standards, Theory, and the “where next” for Process Tracing. Process tracing: from metaphor to analytic tools, pp.260-275.

Choi, D. and Kim, P.S., 2014. Promoting a policy initiative for nation branding: the case of South Korea. Journal of comparative Asian development, 13(2), pp.346-368.

CIDC, 2010. Evaluation report on Saemaul movement ODA projects (meeting report of the sixth meeting of the sub-committee for evaluation (pp. 38–71)) by the Committee for International Development Cooperation) [in Korean]. [Online] Available at: http://www.odakorea.go.kr/fileDownLoad.xdo?f_id=13872718066091921681217R2D7 2FXKK5YPQKZHWEQ [Accessed May 31, 2015].

Cole, D.H., 2014. Formal Institutions and the IAD Framework: Bringing the Law Back In.

Collier, D., 2011. Understanding process tracing. PS: Political Science & Politics, 44(4), pp.823-830.

Dietrich, S., 2013. Bypass or engage? Explaining donor delivery tactics in foreign aid allocation. International Studies Quarterly, 57(4), pp.698-712.

Dietrich, S., 2016. Donor political economies and the pursuit of aid effectiveness. International Organization, 70(1), pp.65-102.

DFID, 2014. Economic development for shared prosperity and poverty reduction: strategic framework (corporate report, by Department for International Development). [Online] Available at: https://www.gov.uk/government/publications/economic- development-for-shared-prosperity-and-poverty-reduction-a-strategic-framework [Accessed May 31, 2015]. 203 Doucette, J., 2010. The terminal crisis of the “participatory government” and the election of Lee Myung Bak. Journal of Contemporary Asia, 40(1), pp.22-43.

Dreher, A., Nunnenkamp, P. and Thiele, R., 2011. Are ‘new’donors different? Comparing the allocation of bilateral aid between nonDAC and DAC donor countries. World Development, 39(11), pp.1950-1968.

Dubash, N.K. and Morgan, B. eds., 2013. The rise of the regulatory state of the south: Infrastructure and development in emerging economies. Oxford University Press.

Easterly, W. and Pfutze, T., 2008. Where does the money go? Best and worst practices in foreign aid. The Journal of Economic Perspectives, 22(2), p.29.

Easterly, W. and Williamson, C.R., 2011. Rhetoric versus reality: the best and worst of aid agency practices. World Development, 39(11), pp.1930-1949.

Easterly, W., 2002. The cartel of good intentions: the problem of bureaucracy in foreign aid. The Journal of Policy Reform, 5(4), pp.223-250.

Easterly, W., Levine, R. and Roodman, D., 2004. Aid, policies, and growth: comment. American economic review, 94(3), pp.774-780.

EDCF, 2014. ODA statistics of the Economic Development Cooperation Fund. [Online] Available at: http://www.edcfkorea.go.kr/edcf/info/oda/system.jsp [Accessed Sep. 4, 2015].

Eichenauer, V.Z. and Hug, S., 2018. The politics of special purpose trust funds. Economics & Politics, 30(2), pp.211-255.

Eichenauer, V.Z. and Reinsberg, B., 2017. What determines earmarked funding to international development organizations? Evidence from the new multi-bi aid data. The Review of International Organizations, 12(2), pp.171-197.

Eisenhardt, K.M., 1989. Agency theory: An assessment and review. Academy of management review, 14(1), pp.57-74.

Epstein, D. and O'halloran, S., 1994. Administrative procedures, information, and agency discretion. American Journal of Political Science, pp.697-722.

Epstein, D. and O'halloran, S., 1999. Delegating powers: A transaction cost politics approach to policy making under separate powers. Cambridge University Press. 204 Evans, P.B., 1995. Embedded autonomy: States and industrial transformation. Princeton University Press.

Evans, Peter B., 2010. Chapter 2. Constructing the 21th century developmental state: potentialities and pitfalls. In Constructing a Democratic Developmental State in Africa: Potentials and Challenges. eds. Omano Edigheji. Cape Town: HSRC Press.

Fischer, A. and Petersen, L., 2004, October. How incentives matter–a conceptual framework for natural resource governance in German Development Cooperation. In Conference on International Agricultural Research and Development, Berlin, Germany (pp. 5-7).

Fligstein, N., 2008. Fields, power and social skill: a critical analysis of the new institutionalisms. International Public Management Review, 9(1), pp.227-253.

Forster, J. and Stokke, O.S. eds., 2013. Policy coherence in development cooperation (Vol. 22). Routledge.

Frey, B.S., Horn, H., Persson, T. and Schneider, F., 1985. A formulation and test of a simple model of World Bank behavior. Review of World Economics, 121(3), pp.438- 447.

George, A.L. and Bennett, A., 2005. Case studies and theory development in the social sciences. MIT Press.

Gibson, C.C., 2005. The Samaritan's dilemma: the political economy of development aid. Oxford University Press on Demand.

Ginsburg, T., 2000. Does law matter for economic development? Evidence from East Asia. Ginsburg, T., 2001. Dismantling the" Developmental State"? Administrative Procedure Reform in Japan and Korea. The American Journal of Comparative Law, 49(4), pp.585- 625.

Ginsburg, T., 2002. Comparative administrative procedure: evidence from Northeast Asia. Constitutional Political Economy, 13(3), pp.247-264.

GLRC, 2012. Regulation of foreign aid in selected countries 2011/12 (By Global Legal Research Center, The Law Library of Congress, USA).

Goertz, G. and Mahoney, J., 2005. Two-level theories and fuzzy-set analysis. Sociological Methods & Research, 33(4), pp.497-538. 205 Graham, A. and Philip, Z., 1971. Essence of decision. Little, Brown and Company.

Gulrajani, N. and Honig, D., 2016. Reforming donors in fragile states.

Gulrajani, N., 2015. Dilemmas in donor design: organisational reform and the future of foreign aid agencies. Public Administration and Development, 35(2), pp.152-164.

Gulrajani, N., 2016. Bilateral Versus Multilateral Aid Channels: Strategic Choices for Donors. odi Report.

Gulrajani, N., 2017. Bilateral Donors and the Age of the National Interest: What Prospects for Challenge by Development Agencies?. World Development, 96, pp.375- 389.

Gulrajani, N., 2018. Merging development agencies.

Haggard, S., 2004. Institutions and growth in East Asia. Studies in comparative international development, 38(4), pp.53-81.

Hahm, S.D. and Choi, Y., 2009. An Early Assessment of the Lee Myung Bak Presidency: Leadership Style and Qualities. APSA 2009 Toronto Meeting Paper.

Hahm, S.D. and Plein, L.C., 1995. Institutions and technological development in Korea: The role of the presidency. Comparative Politics, pp.55-76.

Hall, P.A. and Taylor, R.C., 1996. Political science and the three new institutionalisms. Political studies, 44(5), pp.936-957.

Handley, G., 2010. Effects of aid predictability on service delivery in Uganda. London: ODI. Hawkins, D.G. and Jacoby, W., 2006. How agents matter. Delegation and Agency in International Organizations. Cambridge University Press.

Hawkins, D.G., Lake, D.A., Nielson, D.L. and Tierney, M.J., 2006. Delegation under anarchy: states, international organizations and principal-agent theory. Delegation and agency in international organizations, 3, pp.27-31.

Hay, C. and Wincott, D., 1998. Structure, agency and historical institutionalism. Political studies, 46(5), pp.951-957.

Herring, R.J., 1999. Embedded particularism: India’s failed developmental state. The developmental state, pp.306-334. 206 Hill, R., 2005. Assessing rhetoric and reality in the predictability of aid (No. HDOCPA- 2005-25). Human Development Report Office (HDRO), United Nations Development Programme (UNDP).

Hirsch, P.M. and Lounsbury, M., 1997. Putting the organization back into organization theory: Action, change, and the" new" institutionalism. Journal of Management Inquiry, 6(1), pp.79-88.

Honig, D. and Gulrajani, N., 2018. Making good on donors’ desire to Do Development Differently. Third World Quarterly, 39(1), pp.68-84.

Honig, D., 2014. Navigation by judgment: Organizational autonomy and country context in the delivery of foreign aid. Cambridge, MA: Kennedy School of Government, Harvard University.

Honig, D., 2018. Navigation by Judgment: Why and when Top Down Management of Foreign Aid Doesn't Work. Oxford University Press.

Hook, S., 1995. National Interest and Foreign Aid. Lynne Rienner Publishers.

Hook, S.W. and Zhang, G., 1998. Japan's aid policy since the cold war: Rhetoric and reality. Asian Survey, 38(11), pp.1051-1066.

Hwang, S. and Lim, L., 2003. Restructuring Revisited: Flexible Korea and Rigid Japan. Development and Society, 32(2), pp.229-252.

Im, T., Campbell, J.W. and Cha, S., 2013. Revisiting Confucian bureaucracy: Roots of the Korean government's culture and competitiveness. Public Administration and Development, 33(4), pp.286-296.

Inoguchi, T., 1990. Four Japanese scenarios for the future. International Affairs, 65(1), pp.15-28.

Inoguchi, Takashi and Purnendra Jain. 2011. Japanese Politics Today: From Karaoke to Kabuki Democracy. Palgrave Macmillan. Jayasuriya, K. ed., 2006. Law, capitalism and power in Asia: the rule of law and legal institutions. Routledge.

Jayasuriya, K., 2000. Authoritarian liberalism, governance and the emergence of the regulatory state in post-crisis East Asia. Routledge as part of the Taylor and Francis Group. 207 Jayasuriya, K., 2001. Globalization and the changing architecture of the state: the regulatory state and the politics of negative co-ordination. Journal of European Public Policy, 8(1), pp.101-123.

Jayasuriya, K., 2005. Beyond institutional fetishism: From the developmental to the regulatory state. New Political Economy, 10(3), pp.381-387.

Jerve, A.M. and Selbervik, H., 2009. Self-interest and global responsibility: Aid policies of South Korea and India in the making. CMI Report, 2009(9).

Jerve, A.M., 2007. Asian Models for Aid: Is There a Non-Western Approach to Development Assistance? Summary record of seminar held in Oslo, December 2006. Chr. Michelsen Institute.

Johnson, C., 1982. MITI and the Japanese miracle: the growth of industrial policy: 1925-1975. Stanford University Press.

Johnson, C., 1999. The developmental state: Odyssey of a concept. The developmental state, pp.32-60.

Jones, B.D., 2003. Bounded rationality and political science: Lessons from public administration and public policy. Journal of Public Administration Research and Theory, 13(4), pp.395-412.

Kalinowski, T. and Cho, H., 2012. Korea's search for a global role between hard economic interests and soft power. The European Journal of Development Research, 24(2), pp.242-260.

Kang, D.C., 2002. Crony capitalism: Corruption and development in South Korea and the Philippines. Cambridge University Press.

Kangoye, T., 2013. Does aid unpredictability weaken governance? Evidence from developing countries. The Developing Economies, 51(2), pp.121-144.

Katada, S.N., 2002. Japan's two-track aid approach: the forces behind competing triads. Asian Survey, 42(2), pp.320-342.

KCOC, 2017. OECD DAC Peer Review 2017 Korean Civil Society Report [in Korean] (by Korea NGO Council for Overseas Development Cooperation). [Online] Available at: https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=8&ved=0ahU KEwjvkZbx3M_XAhVFfxoKHZKMCTMQFghYMAc&url=http%3A%2F%2Fwww.n 208 gokcoc.or.kr%2Fbbs%2Fdownload.php%3Fbo_table%3Dpaper01%26wr_id%3D116% 26no%3D1&usg=AOvVaw3zrbo_gUB5sw0KpJZIkDn8 [Accessed Dec. 22, 2017].

KIEP, 2012. Study on public awareness of ODA in Korea [in Korean] (by Korea Institute for International Economic Policy). [Online] Available at: http://odakorea.go.kr/fileDownLoad.xdo?f_id=135780420318917220161799C83VIQC WXF3ZDMIGJSS [Accessed May 31, 2015].

KIEP, 2013. Korea's multilateral aid: Recent developments and future challenges.

Kim, E.M. and Oh, J., 2012. Determinants of foreign aid: The case of South Korea. Journal of East Asian Studies, 12(2), pp.251-274.

Kim, E.M., 1993. Contradictions and limits of a developmental state: with illustrations from the South Korean case. Social Problems, 40(2), pp.228-249.

Kim, H.S. and Potter, D.M. eds., 2012. Foreign aid competition in Northeast Asia (p. 15). Sterling, VA: Kumarian Press.

Kim, H.S., 2012. DAC membership and the development of Korean Official Development Assistance. Foreign Aid Competition in Northeast Asia, pp.37-58.

Kim, J., 2011. Foreign aid and economic development: The success story of South Korea. Pacific Focus, 26(2), pp.260-286.

Kim, J., Chun, S. and Roberts, D., 2011. Using knowledge exchange for capacity development-what works in global practice? Three case studies in assessment of knowledge exchange programs using a results-focused methodology. Knowledge Sharing Project.

Kim, S., 2000. The politics of democratization in Korea: The role of civil society. University of Pittsburgh Press.

Kim, S., 2011. Bridging troubled worlds? An analysis of the ethical case for South Korean aid. Journal of International Development, 23(6), pp.802-822.

Kim, S.T. and Seddon, D., 2005. ODA policy and practice: Japan and the Republic of Korea. Japan’s Foreign Aid: Old Continuities and New Directions, pp.152-190.

Kim, Y. and Tcha, M., 2012. Introduction to the Knowledge Sharing Program (KSP) of Korea. Policy Brief. 209 King, G., Keohane, R.O. and Verba, S., 1994. Designing social inquiry: Scientific inference in qualitative research. Princeton university press.

King, K. and McGrath, S.A., 2004. Knowledge for Development? Comparing British, Japanese, Swedish and World Bank Aid. Zed Books.

Kohli, A., 2004. State-directed development: political power and industrialization in the global periphery. Cambridge University Press.

KOICA, 2013. Study on the Strategies to Improve DEEP Program [in Korean] (by ODA Research Team, Korea International Cooperation Agency). Available at: http://lib.koica.go.kr/bbs/content/4_1864 [Accessed May 31, 2015].

KOICA, 2014. Statistics System. [Online] Available at: http://stat.koica.go.kr:8077/komis/jsptemp/ps/stat_index.jsp [Accessed Aug. 30, 2016].

Kondoh, H., 2012. Taiwan’s Foreign Aid: Seeking Statehood. Foreign Aid Competition in Northeast Asia, pp.81-104. Kumarian Press

Kondoh, H., Kobayashi, T., Shiga, H. and Sato, J., 2010. Diversity and transformation of aid Patterns in Asia’s “Emerging Donors”. JICA-RI working paper, 21.

Kuhonta, E.M., Slater, D. and Vu, T. eds., 2008. Southeast Asia in Political Science: Theory, Region, and Qualitative Analysis. Stanford, Calif.: Stanford University Press.

Laffont, J.J. and Martimort, D., 2009. The theory of incentives: the principal-agent model. Princeton university press.

Lancaster, C., 2008a. Foreign aid: Diplomacy, development, domestic politics. University of Chicago Press.

Lancaster, C., 2008b. The Chinese aid system (No. id: 2278).

Lee, H. K., 1999. Globalization and the emerging welfare state—the experience of South Korea. International Journal of Social Welfare, 8(1), pp.23-37.

Lee, H., and Kang, H., 2013. Roles and responsibilities of actors in development aid: Applying the principal-agent theory. Korea Social Policy Review, 20(4), pp.77-116.

Lee, S. K., 2009. Revisiting Korean Diplomacy [in Korean]. Guiparang.

Lensink, R. and Morrissey, O., 2000. Aid instability as a measure of uncertainty and the positive impact of aid on growth. The Journal of Development Studies, 36(3), pp.31-49. 210 Lim, H. C., 1985. Dependent Development in Korea 1963-1979. Seoul National University Press. Lim, H., 2010. The transformation of the developmental state and economic reform in Korea. Journal of Contemporary Asia, 40(2), pp.188-210.

Lim, H.S. and Lim, J.K., 2013. Performance Evaluation on the Gyeongsangbuk-do Provincial Government’s Global Saemaul-undong Project. National Policy Studies, 27(4), pp.139-65.

Lim, W., 2014. Beyond One-Time, One-Sided Knowledge Transfer: Knowledge Sharing Program with the Dominican Republic1. Analysis of Development Policy and International Cooperation, pp.95-120.

Lumsdaine, D. and Schopf, J.C., 2007. Changing values and the recent rise in Korean development assistance. The Pacific Review, 20(2), pp.221-255.

Lumsdaine, D.H., 1993. Moral vision in international politics: the foreign aid regime, 1949-1989. Princeton University Press.

Mahoney, J., 2010. After KKV: The new methodology of qualitative research. World Politics, 62(1), pp.120-147.

Manning, R., 2006. Will ‘Emerging Donors’ Change the Face of International Co- operation? Development policy review, 24(4), pp.371-385.

Martens, B., 2005. Why do aid agencies exist? Development policy review, 23(6), pp.643-663.

Martens, B., Mummert, U., Murrell, P. and Seabright, P., 2002. The institutional economics of foreign aid. Cambridge University Press.

Mawdsley, E., 2012. From recipients to donors: emerging powers and the changing development landscape. Zed Books Ltd.

McCubbins, M.D. and Page, T., 1986. The congressional foundations of agency performance. Public Choice, 51(2), pp.173-190.

McCubbins, M.D., Noll, R.G. and Weingast, B.R., 1989. Structure and process, politics and policy: Administrative arrangements and the political control of agencies. Virginia Law Review, pp.431-482.

McDonnell, I., Lecomte, H.B.S. and Wegimont, L., 2003. Public Opinion Research, Global Education and Development Co-operation Reform. 211 McGinnis, M.D., 2011. An introduction to IAD and the language of the Ostrom workshop: a simple guide to a complex framework. Policy Studies Journal, 39(1), pp.169-183.

Milner, H.V. and Tingley, D., 2013. The choice for multilateralism: Foreign aid and American foreign policy. The Review of International Organizations, 8(3), pp.313-341.

Milner, H.V., 2006. Why multilateralism? Foreign aid and domestic principal-agent problems. Delegation and agency in international organizations. Cambridge University Press.

Miyashita, A. and Sato, Y. eds., 2001. Japanese foreign policy in Asia and the Pacific: domestic interests, American pressure, and regional integration. Springer.

MOFA, 2010. Strategies for Multilateral ODA to the U.N. and International Organizations [in Korean]. Ministry of Foreign Affairs, Republic of Korea.

Mok, K.H. and Forrest, R. eds., 2008. Changing governance and public policy in East Asia. Routledge.

Moore, M., 1984. Mobilization and disillusion in rural Korea: the Saemaul movement in retrospect. Pacific Affairs, 57(4), pp.577-598.

MOSF, 2010. Strategies for multilateral ODA to international finance institutes [in Korean]. Ministry of Strategy and Finance, Republic of Korea.

Nee, V., 2005. The new institutionalisms in economics and sociology. The handbook of economic sociology2, pp.49-74.

Nielsen, K., 2003. Review of institutionalist approaches in the social sciences: typology, dialogue and future prospects. Research Paper, 7(01).

Nielson, D.L. and Tierney, M.J., 2003. Delegation to international organizations: Agency theory and World Bank environmental reform. International organization, 57(2), pp.241-276.

ODA Watch, 2012. Parallel report on Korea’s international development cooperation (by ODA Watch) [in Korean]. [Online] Available at: http://eurodad.org/files/integration/2012/10/finalParallel-Report_ODAWatch_0614.pdf [Accessed May 31, 2016]. 212 OECD, 1999. A comparison of management systems for development cooperation in OECD/DAC members.

OECD, 2008. Republic of Korea DAC special review.

OECD, 2009. Managing aid: Practices of DAC member countries. [Online] Available at: http://www.oecd.org/dac/peer-reviews/35051857.pdf [Accessed May 26, 2014].

OECD, 2010. Beyond the DAC: The welcome role of other providers of development co-operation.

OECD, 2013a. Multilateral aid report.

OECD, 2013b. Peer review 2012 Korea.

OECD, 2015a. Development co-operation report 2015: Making partnerships effective coalition for action.

OECD, 2015b. Multilateral aid 2015: Better partnerships for a post-2015 world.

Ohnesorge, J., 2010. Administrative law in East Asia: a comparative-historical analysis. Comparative Administrative Law: Research Handbooks in Comparative Law. Cheltenham/Northampton: Edward Elgar.

Ohnesorge, J., 2016a. East Asia and the study of law and development, Routledge Handbook of Asian Law. Routledge.

Ohnesorge, J., 2016b. The regulatory state in East Asia. Comparative Law and Regulation: Understanding the Global Regulatory Process, Edward Elgar Publishing.

Ohnesorge, J.K., 2003. The rule of law, economic development, and the developmental states of northeast Asia. Law and Development in East and South-East Asia, pp.91-130.

Ohnesorge, J.K., 2007. Developing Development Theory: Law and Development Orthodoxies and the Northeast Asian Experience. U. Pa. J. Int'l Econ. L., 28, p.219.

Orr, Robert M., 1990. The Emergence of Japan's Foreign Aid Power. Columbia University Press.

Ostrom, E. and Basurto, X., 2011. Crafting analytical tools to study institutional change. Journal of institutional economics, 7(3), pp.317-343.

Ostrom, E., 1995, October. The institutional analysis and development framework: An application to the study of common-pool resources in sub-Saharan Africa. In Paper, 213 Bloomington Indiana: Workshop in Political Theory and Policy Analysis, Indiana University (Vol. 10).

Ostrom, E., 2002. Aid, incentives, and sustainability: an institutional analysis of development cooperation. Main report. Swedish International Development Agency (SIDA).

Ostrom, E., 2005a. Doing institutional analysis digging deeper than markets and hierarchies. Handbook of new institutional economics, pp.819-848.

Ostrom, E., 2005b. Understanding institutional diversity. Princeton University Press.

Ostrom, E., 2011. Background on the institutional analysis and development framework. Policy Studies Journal, 39(1), pp.7-27.

Ostrom, E., 2015. Governing the commons. Cambridge university press.

Ostrom, E., Gardner, R. and Walker, J., 1994. Rules, games, and common-pool resources. University of Michigan Press.

Otter, M., 2003. Domestic public support for foreign aid: does it matter?. Third World Quarterly, 24(1), pp.115-125.

Page, J. and Campos, E., 1993. The East Asian Miracle: economic growth and public policy. World Bank, Washington DC.

Park, H., 2012. A neo-institutional approach to aid effectiveness: A comparative study of Swedish and Korean aid. PhD Thesis. Ewha W. University.

Park, H., 2013. What’s the matter with development aid? A new institutional Approach. International and Regional Studies, (22)1, pp.39-69.

Park, M., 2010. Strategies for enhancing the effectiveness of multilateral ODA [in Korean].

Pistor, K., Wellons, P.A. and Sachs, J., 1999. The Role of Law and Legal Institutions in Asian Economic Development, 1960-1995. Oxford University Press.

PMO, 2012. The strategic plan for international development cooperation (by Prime Minister’s Office, Republic of Korea). [Online] Available at: http://odakorea.go.kr/eng.policy.StrategicPlan.do [Accessed May 31, 2015].

Pollack, M.A., 2003. The engines of European integration: delegation, agency, and agenda setting in the EU. Oxford University Press. 214 Posner, E. A., 2000. Law and social norms: The case of tax compliance. Virginia Law Review, pp.1781-1819.

Potter, D. and Sueo, S., 2003. Japanese foreign policy: no longer reactive?. Political Studies Review, 1(3), pp.317-332.

Powell, W.W. and DiMaggio, P.J. eds., 2012. The new institutionalism in organizational analysis. University of Chicago Press.

Quarles van Ufford, P., 1988. The hidden crisis in development: development bureaucracies in between intentions and outcomes. The hidden crisis in development: Development bureaucracies, pp.9-38.

Radelet, S., 2006. A primer on foreign aid (Working paper 92). Center for Global Development.

Riddell, R.C., 2008. Does foreign aid really work?. Oxford University Press.

Rix, A., 1989. Japan's foreign aid policy: a capacity for leadership?. Pacific Affairs, pp.461-475.

Rix, A., 2010. Japan's foreign aid challenge. Routledge. Reinsberg, B., Michaelowa, K. and Eichenauer, V.Z., 2015. The rise of multi-bi aid and the proliferation of trust funds. Handbook on the Economics of Foreign Aid. Cheltenham: Edward Elgar, pp.527-554.

Rodrik, D., 1995. Why is there multilateral lending? (No. w5160). National bureau of economic research.

Round, J.I. and Odedokun, M., 2004. Aid effort and its determinants. International Review of Economics & Finance, 13(3), pp.293-309.

Sato, Jin and Yasutami Shimomura. eds. 2012. The Rise of Asian Donors: Japan's Impact on the Emerging Donors. Routledge.

Schaler, E., 2014. An assessment of the institutional analysis and development framework and introduction of the social-ecological systems framework. Theories of the policy process, 267.

Schlager, E., 1999. A comparison of frameworks, theories, and models of policy processes. Theories of the policy process, 1, pp.233-260. 215 Schraeder, P.J., Hook, S.W. and Taylor, B., 1998. Clarifying the foreign aid puzzle: A comparison of American, Japanese, French, and Swedish aid flows. World Politics, 50(2), pp.294-323.

Seabright, P., 2002. Conflicts ofobjectives and task allocation in aid agencies. The institutional economics of foreign aid, p.34-68.

Shapiro, S.P., 2005. Agency theory. Annual review of sociology, 31.

Shin, S.O., 2016. South Korea's elusive middle-power-man-ship: regional or global player?. The Pacific Review, 29(2), pp.187-209.

Siddiki, S.N., Carboni, J.L., Koski, C. and Sadiq, A.A., 2015. How policy rules shape the structure and performance of collaborative governance arrangements. Public Administration Review, 75(4), pp.536-547.

Snidal, D., 1985. Coordination versus prisoners' dilemma: Implications for international cooperation and regimes. American Political Science Review, 79(4), pp.923-942.

Söderberg, M., 2010. Challenges or Complements for the West: Is there an ‘Asian’Model of Aid Emerging?. In Challenging the Aid Paradigm (pp. 107-137). Palgrave Macmillan UK.

Swedlund, H.J., 2014. Ducey donors: How donor commitment problems can explain changing foreign aid practices. [Online] Available at: http://aiddata.org/blog/foreign- aids-commitment-problem.

Swedlund, H.J., 2017. The development dance: How donors and recipients negotiate the delivery of foreign aid. Cornell University Press.

Tansey, O., 2007. Process tracing and elite interviewing: a case for non-probability sampling. PS: Political Science & Politics, 40(4), pp.765-772.

Thiele, R., Nunnenkamp, P. and Dreher, A., 2006. Sectoral aid priorities: Are donors really doing their best to achieve the millennium development goals?

Townley, B., 2008. Reason's neglect: rationality and organizing. Oxford University Press.

U.S. Law Library of Congress, 2012. Regulation of foreign aid in selected countries 2011/2012 [Online] Available at: https://www.loc.gov/law/help/foreign-aid/index.php [Accessed Dec. 28, 2017]. 216 Van der Veen, A.M., 2011. Ideas, interests and foreign aid. Cambridge University Press.

Vennesson, P., 2010. Process Tracing and Historical Inquiry: Policy-Making Sequences and 'Possibilism'.

Watson, I., 2011. Global Korea: Foreign aid and national interests in an age of globalization. Contemporary Politics, 17(1), pp.53-69.

Watson, I., 2014. Foreign aid and emerging powers: Asian perspectives on Official Development Assistance. Routledge.

WB, 1993. The East Asian miracle: Economic growth and public Policy.

WB, 1999. Knowledge for Development.

WB, 2015. Republic of Korea Overview. [Online] Available at: http://www.worldbank.org/en/country/korea/overview

Weible, C., Heikkila, T. and Gerlak, A.K., 2010. Upscaling from Individual to Collective Learning in Policy Process Research.

Weible. Christopher M. 2014. Introducing the scope and focus of policy process research and theory. Westview Press.

Welch, E. and Wong, W., 1998. Public administration in a global context: bridging the gaps of theory and practice between western and non-western nations. Public Administration Review, pp.40-49.

Williamson, C.R., 2010. Exploring the failure of foreign aid: The role of incentives and information. The review of Austrian economics, 23(1), pp.17-33.

Wong, J., 2011. Betting on biotech: Innovation and the limits of Asia's developmental state. Cornell University Press.

Woo, J. E., 1991. Race to the Swift: State and Finance in Korean Industrialization. Columbia University Press. Woo-Cummings, M., 1999. Introduction: Chalmers Johnson and the Politics of Nationalism and Development. The Developmental State. Ithaca and London: Cornel UP.

Woods, N., 2008. Whose aid? Whose influence? China, emerging donors and the silent revolution in development assistance. International Affairs, 84(6), pp.1205-1221. 217 Yasutomo, D.T., 1989. Why Aid? Japan as an" Aid Great Power". Pacific Affairs, pp.490-503.

Yi, J., 2015. Re-mapping foreign aid literature: How do we understand East Asian emerging donorship? International Development and Cooperation Review7(1), pp.1-40.

Yi, J., 2016. The ‘reactive state’ as foreign aid giver: A historical analysis of aid policy regime (1953-99) in South Korea. Asian International Studies Review, 17(1), pp.55-75.

Yin, R.K., 2009. Case study research: Design and Methods. SAGE publications.