THE DETERMINANTS OF THE DEMOCRATIC REPUBLIC OF CONGO’S BILATERAL TRADE: AN AUGMENTED GRAVITY = 콩고민주공화국의 양자무역 결정요인: 확장된 중력모형 분석

국문 초록 (Abstract)

본 논문은 확장된 중력모형（augmented gravity model）을 이용하여 콩고민주공화국（DRC）의 양자무역 결정요인을 분석한다. DRC 는 풍부한 광물자원과 아프리카 중심부라는 지리적 이점을 지니고 있음에도 불구하고, 경제 규모에 비해 훨씬 낮은 수준의 무역을 수행하고 있으며, 수출은 소수의 원자재와 일부 수출 대상국에 크게 편중되어 있다. 본 연구는 2000년부터 2025년까지 DRC의 주요 교역상대국 50개국으로 구성된 균형 양자 패널 자료（총 1,300개 관측치）를 대상으로 수출, 수입, 총무역에 대한 세 개의 중력방정식을 추정한다. 각 방정식은 DRC의 GDP, 교역상대국의 GDP, 양국 간 거리라는 표준 중력변수에 양자 환율, 정부 효과성（government effectiveness）, 무역 관련 인프라를 추가하여 확장하였다. 추정은 통합 최소자승법（Pooled OLS）에서 고정효과（Fixed Effects） 및 확률효과（Random Effects） 모형으로 순차적으로 진행되며, 각 방정식의 최적 추정량 선택은 하우스만 검정（Hausman test）에 따라 결정되고, 추론의 타당성은 HC1, 군집-강건（cluster-robust）, Driscoll–Kraay 표준오차를 통해 검증된다. 분석 결과, 교역상대국의 GDP 가 세 방정식 모두에서 가장 일관되게 강건한 결정요인으로 나타났다. DRC의 GDP는 수출과 총무역에서 강건한 것으로 확인되었다. 무역 관련 인프라는 DRC 수출에 있어 가장 강건한 정책 관련 결정요인 중 하나로 나타났으며, 정부 효과성은 수입과 총무역에서 양（+）의 방향으로 강건한 영향을 미치는 것으로 분석되었다. 거리는 수입과 총무역에서 강건한 음（−）의 효과를 보였고, 환율은 수입에 대해 강건한 양（+）의 효과를 나타냈다. 변수들이 서로 다른 척도로 측정되었기 때문에, 본 연구의 결과는 인과적 효과가 아닌 조건부 연관성（conditional associations）으로 해석된다. 연구 결과는 DRC 가 자원 부존, 외부 수요, 지역 통합의 기회를 실제 무역 성과로 전환하는 데 있어 지리적 요인만이 아니라 구조적 역량（structural capacity）이 결정적인 역할을 한다는 점을 시사한다. 본 논문은 무역 관련 인프라 투자와 제도 강화를 중심으로 한 우선순위별 정책 제언을 제시하고, 향후 연구 방향을 제안하며 마무리된다. 주제어（핵심어）: 중력모형; 양자무역; 콩고민주공화국; 무역 관련 인프라; 정부 효과성; 패널 자료

다국어 초록 (Multilingual Abstract)

This dissertation examines the determinants of the Democratic Republic of Congo’s （DRC） bilateral trade using an augmented gravity model. Despite exceptional mineral wealth and a central geographic position in Africa, the DRC trades far below the levels its economic size would predict, and its exports remain heavily concentrated in a few commodities and destination markets. The study estimates three gravity equations for exports, imports, and total trade on a balanced bilateral panel of the DRC’s 50 major trading partners over the period 2000–2025 （1,300 observations）. Each equation augments the standard gravity variables of DRC GDP, partner-country GDP, and bilateral distance with the bilateral exchange rate, government effectiveness, and trade-related infrastructure. Estimation proceeds sequentially from Pooled OLS through Fixed Effects and Random Effects models, with Hausman tests guiding the choice of preferred estimator for each equation and with inference verified under HC1, cluster-robust, and Driscoll–Kraay standard errors. The results show that partner GDP is the most consistently robust determinant across all three equations. DRC GDP is robust for exports and total trade. Trade-related infrastructure emerges as one of the most robust policy- relevant determinants of DRC exports, while government effectiveness is positive and robust for imports and total trade. Distance is robustly negative for imports and total trade, and the exchange rate is robustly positive for imports. Because the variables are measured on different scales, the results are interpreted as conditional associations rather than causal effects. The findings indicate that structural capacity, not geography alone, determines how effectively the DRC converts its resource endowment, external demand, and regional integration opportunities into realized trade. The dissertation concludes with prioritized policy recommendations centered on trade-related infrastructure investment and institutional strengthening and identifies directions for future research. Keywords: gravity model; bilateral trade; Democratic Republic of the Congo; trade-related infrastructure; government effectiveness; panel data.

목차 (Table of Contents)

• CHAPTER ONE: INTRODUCTION 1
• 1.1 Background and Global African Trade Context 1
• 1.2 The DRC Trade Paradox and Structural Context 3
• 1.2.1 Export and Import Structure 3
• 1.2.2 Structural and Institutional Constraints 4
• 1.3 Statement of the Research Problem 6
• 1.4 Research Gaps and Positioning of the Study 7
• 1.5 Research Questions, Objectives, Hypotheses, and Analytical Proposition 9
• 1.6 Scope and Delimitations of the Study 13
• 1.7 Significance and Contributions of the Study 14
• 1.7.1 Identification Contribution 14
• 1.7.2 Measurement Contribution 14
• 1.7.3 Diagnostic Contribution 15
• 1.7.4 Estimation Contribution 15
• 1.7.5 Coverage Contribution 15
• 1.7.6 Policy Significance 15
• 1.8 Overview of the Empirical Strategy and Main Findings 16
• 1.9 Organization of the Dissertation 17
• CHAPTER TWO: LITERATURE REVIEW AND THEORETICAL FRAMEWORK 19
• 2.1 The Gravity Model: Origins, Theoretical Development, and Estimation 19
• 2.1.1 Early Empirical Formulations 19
• 2.1.2 Theoretical Foundations: Anderson (1979) and Bergstrand (1985, 1989) 20
• 2.1.3 Multilateral Resistance: Anderson and van Wincoop (2003) 20
• 2.1.4 Firm Heterogeneity and Zero Trade Flows 21
• 2.1.5 Estimation: From OLS to PPML and the Choice of This Study 21
• 2.2 International Trade Theories: Connections to the Gravity Model and DRC Variables 24
• 2.2.1 Comparative Advantage and the Gravity Model 25
• 2.2.2 The Heckscher-Ohlin Model: Justification for lnGDP_it 25
• 2.2.3 New Trade Theory: Micro-foundation for the Log-Linear Gravity Specification 26
• 2.2.4 Structuralist Theory: Justification for GEPR_it 27
• 2.3 Empirical Gravity-Model Evidence on Bilateral Trade Determinants 29
• 2.3.1 Partner GDP (lnGDP_jt): Market Size and Mineral Demand 29
• 2.3.2 DRC GDP (lnGDP_it): Domestic Economic Size and Production Capacity 29
• 2.3.3 Bilateral Distance (lnDist_ij): Transport Costs and Trade Corridor Constraints 30
• 2.3.4 Infrastructure Quality (Infra_it): The Primary Binding Constraint 31
• 2.3.5 Governance Effectiveness (GEPR_it): The Institutional Trade Cost 32
• 2.3.6 Exchange Rate (lnExch_ijt): Currency Depreciation and Trade Competitiveness 33
• 2.3.7 Regional Trade Agreements, Colonial Ties, Common Language, and Contiguity 34
• 2.4 Gravity-Model Evidence on African Trade and the DRC 37
• 2.4.1 Infrastructure and Institutional Constraints on African Bilateral Trade 38
• 2.4.2 Regional Integration and the AfCFTA 38
• 2.4.3 DRC-Specific Gravity Evidence and the Gaps This Study Addresses 39
• 2.5 Research Gaps and Positioning of This Study 39
• 2.5.1 Identification Gap: Bilateral Panel Data versus Time-Series Data 40
• 2.5.2 Measurement Gap: Governance Variable 40
• 2.5.3 Diagnostic Gap: Model Diagnostic Tests 40
• 2.5.4 Estimation Gap: Consistent Research Method 40
• 2.5.5 Coverage Gap: Country-Specific Study of the DRC 41
• CHAPTER THREE: RESEARCH METHODOLOGY 43
• 3.1 Research Design and Data Structure 43
• 3.2 Data Sources, Variable Definitions, and Data Construction 44
• 3.3 Model Specification 50
• 3.3.1 Equation 1: DRC Export Gravity Equation 50
• 3.3.2 Equation 2: DRC Import Gravity Equation 51
• 3.3.3 Equation 3: Total Bilateral Trade Equation 51
• 3.3.4 Research Hypotheses 52
• 3.4 Treatment of Zero Trade Flows 53
• 3.5 Estimation Strategy and Model Selection 55
• 3.5.1 Pooled OLS 56
• 3.5.2 Fixed Effects 56
• 3.5.3 Random Effects 57
• 3.6 Diagnostic Tests and Robustness Checks 58
• 3.7 Methodological Limitations 61
• 3.7.1 Log-Linear Specification versus PPML 61
• 3.7.2 Infrastructure Variable: Limited Variation and Measurement Error 61
• 3.7.3 Informal Trade Exclusion 62
• 3.7.4 Governance Variable 62
• 3.7.5 Endogeneity 62
• 3.7.6 Data Vintage: 2024–2025 Observations 63
• 3.7.7 Selection of Trading Partners 63
• 3.8 Chapter Summary 64
• CHAPTER FOUR: EMPIRICAL RESULTS AND DISCUSSION 67
• 4.1 Introduction 67
• 4.2 Descriptive Statistics 68
• 4.3 Correlation Analysis and Multicollinearity Diagnostics 74
• 4.3.1 Pearson Correlation Matrix 75
• 4.3.2 Variance Inflation Factor (VIF) Diagnostics 76
• 4.4 Model-Selection Tests 77
• 4.5 Preferred Estimation Results 80
• 4.6 Robustness Checks 81
• 4.7 Discussion of Main Findings 84
• 4.7.1 Infrastructure Quality 84
• 4.7.2 Economic Size and GDP Effects 86
• 4.7.3 Governance Effectiveness 87
• 4.7.4 Distance and Trade Costs 89
• 4.7.5 Exchange Rate Effects 91
• 4.7.6 Evaluation of Research Hypotheses 92
• 4.8 Chapter Summary 93
• CHAPTER FIVE: CONCLUSIONS, POLICY RECOMMENDATIONS, AND FUTURE RESEARCH 96
• 5.1 Summary of Empirical Findings 96
• 5.2 Conclusions and Contributions 98
• 5.2.1 Core Gravity Predictions Are Supported, but Not Uniformly Across Trade Flows 98
• 5.2.2 Structural Constraints Amplify the Geographic Costs of DRC Trade 99
• 5.2.3 Governance Quality Is an Important but Asymmetric Factor in Trade 100
• 5.2.4 Structural Constraints Are Consistent with Persistent Unrealized Trade Potential 101
• 5.2.5 Contributions of the Study 101
• 5.3 Evaluation of Hypotheses 102
• 5.4 Policy Recommendations 104
• 5.5 Limitations of the Study 108
• 5.5.1 Log-Linear Model versus PPML 108
• 5.5.2 Exclusion of Informal Trade 108
• 5.5.3 Limited Variation and Measurement of the Infrastructure Variable 109
• 5.5.4 Endogeneity 109
• 5.5.5 Projected Trade Values for 2024–2025 109
• 5.5.6 Selection of Major Trading Partners 110
• 5.6 Directions for Future Research 110
• 5.6.1 Methodological Priorities 110
• 5.6.2 Substantive Extensions 112
• 5.7 Overall Concluding Statement 113
• REFERENCES 114
• APPENDIX A: LIST OF THE 50 PARTNER COUNTRIES INCLUDED IN THE STUDY 118