南海緊張情勢：GDELT 時間序列數據之分析 - 政大學術集成

全文

(1) . . South China Sea Tensions:   State Involvement and Prediction Using GDELT Event Data Jonathan Spangler . 政治大 . 學. ‧ 國. 立. Doctoral Dissertation . ‧. International Doctoral Program in Asia-Pacific Studies National Chengchi University, Taipei, Taiwan . Nat. sit. . n. February 2018 . Ch. e n g c h i. er. io. al. y. . i n U. v. Dissertation Committee:  . 劉復國), Dissertation Supervisor Dustin Kuan-Hsiung Wang (王冠雄), Honorary Chair Mei-Shiu Chiu (邱美秀) Richard Hu (胡瑞舟) Ping-Yin Kuan (關秉寅) Fu-Kuo Liu (. . .

(2) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v.

(3) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . Abstract Discussions of the South China Sea maritime territorial disputes are rife with assertions that certain state actors escalate regional tensions and that it is only a matter of time before provocations trigger armed conflict. However, these claims are based primarily on incomplete evidence, inaccurate comparisons with historical conflicts, and country or individual biases. This dissertation questions these common assertions and uses empirical evidence to assess their validity. Using time-series event data from the Global Database of Events, Language, and Tone (GDELT), it analyzes (1) the relationship between state involvement and South China Sea tensions and (2) which forecast models can most accurately predict South China Sea tensions . 政治大 For RQ1, the analyses reveal that the involvement of certain countries 立. based on data from earlier time periods.  . ‧ 國. 學. corresponds with significantly higher tensions in the South China Sea, that state involvement and tensions are correlated at various positive and negative lags of . ‧. interest, and that these correlations go in both directions. These findings have important implications for policymakers and researchers in that they offer empirical . y. Nat. evidence that confirms or refutes assertions suggesting that certain countries’ . io. sit. actions lead to escalation or deescalation. They also provide a solid foundation for . n. al. er. future research, which could take specific countries as individual case studies to . i n U. v. further investigate the relationships between state involvement and South China Sea . Ch. engchi. tensions. Moreover, the results indicate that there may be even more interesting phenomena at play that merit attention in future research: evidence suggesting that certain countries may either contribute to lower tensions or avoid becoming involved when there are heightened tensions, and evidence that some countries may not be contributing to but instead reacting to tensions and volatility in the South China Sea.   For RQ2, two of the four forecast models perform better than the four benchmark models using both datasets. These findings also have important implications for policy and research. As governments become increasingly interested in using continuously updated global databases to facilitate policy-making, the results suggest that empirical data can help to inform conclusions about trends of 1 .

(4) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . escalation and deescalation in the South China Sea and be used to make relevant predictions. As a first cut at the data and a pioneering approach to analyzing South China Sea tensions, the analyses and findings of this dissertation represent a significant contribution to knowledge and a foundation for future research using time-series event data to understand the relationship between state involvement and tensions and the extent to which tensions can be forecasted in the South China Sea and around the world. Keywords: event data; GDELT; GKG; maritime territorial disputes; South China Sea; state involvement; tensions; time series . 立. . 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 2 . i n U. v.

(5) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . Acknowledgements The author would like express his appreciation to the many individuals that have contributed to this research. In particular, he offers his sincerest gratitude to the . 劉復國), . distinguished members of the dissertation committee – Drs. Fu-Kuo Liu (. 邱美秀), Richard Hu (胡瑞舟), Ping-Yin Kuan (關秉寅), and Dustin Kuan-Hsiung Wang (王冠雄) – for their enduring support and constructive feedback Mei-Shiu Chiu (. on the text and data analyses of this dissertation, without which completing it would not have been possible. He would also like to thank the researchers, coders, and authors that contribute countless hours of their own time to producing and releasing open-access so ware, code, and written guidance for the benefit of others. Above all, . 治政 – for providing the inspiration to accept new challenges 大and focus on seeing them 立 through to completion. . he is eternally grateful to his family members – Mom, Dad, Matt, YingChu, and Mesa . ‧. ‧ 國 io. sit. y. Nat. n. al. er. . 學. . Ch. engchi. 3 . i n U. v.

(6) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . . 立. 政治大. 學 ‧. ‧ 國 io. sit. y. Nat. n. al. er. . Ch. engchi. 4 . i n U. v.

(7) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . Table of Contents Abstract. 1 . Acknowledgements. 3 . Table of Contents. 5 . List of Tables. 9 . List of Figures. 11 . List of Abbreviations. 15 . 1. Introduction. 17 . 1.1 Background 1.2 Significance of the Study 1.3 Research Questions. 立. 17 . 政治大. 19 . ‧ 國. 學. 1.3.1 RQ1: State Involvement and Tensions. 17 . 1.3.2 RQ2: Predicting Tensions. ‧. 2. Literature Review. y. Nat. 2.1 Context of the South China Sea Disputes. n. al. Ch. 2.2.2 US Involvement and Tensions. engchi. er. io. 2.2.1 China’s Involvement and Tensions. sit. 2.2 Effects of State Involvement on Tensions. i n U. v. 19 19 21 21 25 26 29 . 2.2.3 Other Rival Claimants’ Involvement and Tensions. 33 . 2.2.4 Other Major Stakeholders’ Involvement and Tensions. 35 . 2.2.5 Ambiguous Statements Linking State Involvement and Tensions. 37 . 2.2.6 Empirical Evidence and the Relationship between State Involvement and Tensions 39 2.2.7 Theoretical Framework for RQ1 2.3 Predicting Tensions and Conflict. 40 42 . 2.3.1 Academic Predictions of Tensions and Conflict. 43 . 2.3.2 Policymaker Predictions of Tensions and Conflict. 46 . 2.3.3 Media Predictions of Tensions and Conflict. 48 . 5 .

(8) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . 2.3.4 Empirical Evidence and Predictions of Tensions and Conflict. 49 . 2.3.5 Theoretical Framework for RQ2. 51 . 2.4 Modeling Real World Processes. 53 . 2.5 Analyses with GDELT in Other Geographic Contexts. 54 . 2.6 Defining Relevant Terms. 60 . 3. Methodology. 63 . 3.1 Data Sources. 64 . 3.1.1 Global Database of Events, Language, and Tone (GDELT). 64 . 3.1.2 GDELT 1.0 Event Database. 65 . 3.1.3 GDELT 2.0 Global Knowledge Graph (GKG). 66 . 治政大 3.1.4 Database Comparison 立 3.2 Data Selection. 67 68 . ‧ 國. 學. 69 . 3.2.2 Linking GDELT 1.0 Event Database Data to Real-World Events. 73 . ‧. 3.2.1 Tensions Data Selection from GDELT 1.0 Event Database. y. 3.2.4 Tensions Data Selection from GDELT 2.0 GKG. io. sit. Nat. 3.2.3 State Involvement Data Selection from GDELT 1.0 Event Database for RQ1 78 . n. al. er. 3.2.5 Linking GDELT 2.0 GKG Data to Real-World Events. Ch. i n U. v. 79 83 . 3.2.6 State Involvement Data Selection from GDELT 2.0 GKG for RQ1. 86 . 3.2.7 Tensions Data Selection for RQ2. 88 . engchi. 3.2.8 Comparing Datasets. 89 . 3.3 Statistical Approach for RQ1. 90 . 3.3.1 Observed Data. 91 . 3.3.2 Difference Analyses. 104 . 3.3.3 Cross-Correlation Analyses. 106 . 3.3.4 Hypotheses for RQ1. 107 . 3.4 Statistical Approach for RQ2. 108 . 3.4.1 Observed Data. 109 6 .

(9) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . 3.4.2 Training Data and Test Data. 110 . 3.4.3 Benchmark Models. 111 . 3.4.4 Forecast Models. 120 . 3.4.5 Summary of Models. 131 . 3.4.7 Comparison of Forecast Errors. 132 . 3.4.8 Hypotheses for RQ2. 133 . 3.5 Limitations. 133 . 3.5.1 Proxy Variables. 134 . 3.5.2 Data Availability. 134 . 3.5.3 Definitions of State Involvement and South China Sea. 135 . 治政大 3.5.4 Correlation and Causation 立 4. Results and Discussion. 135 137 . ‧ 國. 學. 4.1 State Involvement and Tensions. ‧. 4.1.1 Difference Analyses. al. n. 4.2.1 Benchmark Models. er. io. 4.2 Predicting Tensions. 4.2.2 Forecast Models. sit. Nat. 4.1.3 Evaluation of Hypotheses for RQ1. y. 4.1.2 Cross-Correlation Analyses. Ch. 4.2.3 Model Comparison. engchi U. 4.2.4 Evaluation of Hypotheses for RQ2 4.3 Further Discoveries. v ni. 138 139 154 165 166 167 176 187 189 190 . 4.3.1 Trends of Escalation and Deescalation. 191 . 4.3.2 Changes in State Involvement. 192 . 4.3.3 Reacting to Tensions. 194 . 5. Conclusion. 197 . 5.1 State Involvement and Tensions. 197 . 5.1.1 Summary of Findings. 197 7 .

(10) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . 5.1.2 Implications, Policy Recommendations, and Future Research 5.2 Predicting Tensions. 199 201 . 5.2.1 Summary of Findings. 201 . 5.2.2 Implications, Policy Recommendations, and Future Research. 202 . 5.3 Further Discoveries. 203 . 5.3.1 Summary of Findings. 203 . 5.3.2 Implications, Policy Recommendations, and Future Research. 205 . Bibliography. 207 . Appendix I: GDELT Data Selection Code for Google BigQuery. 225 . Appendix II: GDELT Data Analysis Code for R. 249 . 治政大 Appendix III: Codebook Excerpts 立 ‧. ‧ 國 io. sit. y. Nat. n. al. er. . 學. . 331 . Ch. engchi. 8 . i n U. v.

(11) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . List of Tables Table 1: Excerpt of GDELT 1.0 Event Database showing select columns not containing null data . 66 . Table 2: Excerpt of GDELT 2.0 Global Knowledge Graph data showing select columns . 67 . Table 3: Comparison of the GDELT 1.0 Event Database and GDELT 2.0 Global Knowledge Graph . 68 . Table 4: Excerpt of queried South China Sea tensions data from GDELT 1.0 Event Database . 72 . Table 5: Excerpt of queried South China Sea tensions data from GDELT 2.0 GKG . 82 . Table 6: Excerpt of inclusion-based state involvement and tensions data (based on GDELT 1.0 Event Database) . 93 . 政治大 Table 7: Excerpt of inclusion-based state involvement and tensions data 立 (based on GDELT 2.0 GKG) . 學. 99 . Table 9: Excerpt of proportion-based state involvement and tensions data (based on GDELT 2.0 GKG) . 100 . Table 10: Overview of benchmark and forecast models . y. 132 146 . ‧. Table 8: Excerpt of proportion-based state involvement and tensions data (based on GDELT 1.0 Event Database) . sit. ‧ 國. 94 . Nat. n. al. er. io. Table 11: Difference in tensions with state involvement and statistical significance based on t tests (based on GDELT 1.0 Event Database) . v. Table 12: Difference in tensions with state involvement and statistical significance based on t tests (based on GDELT 2.0 GKG) . 153 . Table 13: Juxtaposition of findings for difference in tensions with state involvement and statistical significance based on t tests (based on GDELT 1.0 Event Database and GDELT 2.0 GKG) . 155 . Table 14: Summary of hypotheses and corresponding evaluations for RQ1 . 166 . Table 15: ARIMA model variants and AICc values . 185 . Table 16: Comparison of forecast accuracy of benchmark and forecast models (for analyses based on GDELT 1.0 Event Database) . 189 . Table 17: Comparison of forecast accuracy of benchmark and forecast models (for analyses based on GDELT 2.0 GKG) . 190 . Table 18: Summary of hypotheses and corresponding evaluations for RQ2 . 191 . Ch. engchi. 9 . i n U.


(13) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . List of Figures Figure 1: Conflicting claims to islands and other sea features in the South China Sea . 24 . Figure 2: Theoretical framework for RQ1 . 40 . Figure 3: Hypothetical illustration of South China Sea tensions data over time . 41 . Figure 4: Hypothetical illustration of South China Sea tensions over time with and without a state’s involvement . 42 . Figure 5: Theoretical framework for RQ2 . 51 . Figure 6: Hypothetical illustration of predicting South China Sea tensions in a given time period based on observed tensions in earlier time periods . 52 . Figure 7: Monthly records with action taking place in Paracel Islands or Spratly Islands from January 1979 to November 2017 (based on GDELT 1.0 Event Database) . 70 . Figure 8: Monthly records with action taking place in the Paracel Islands or Spratly Islands from February 2011 to November 2017 (based on GDELT 1.0 Event Database) . 71 . Figure 9: Monthly average tensions for events with action taking place in Paracel or Spratly Islands from February 2011 to November 2017 (based on GDELT 1.0 Event Database) . 73 . Figure 10: Monthly records of South China Sea events from March 2015 to November 2017 (based on GDELT 2.0 GKG)  . 81 . Figure 11: Monthly average tensions in the South China Sea from March 2015 to November 2017 (based on GDELT 2.0 GKG)  . 83 . Figure 12: Monthly average South China Sea tensions from February 2011 to November 2017 (based on GDELT 1.0 Event Database)  . 91 . Figure 13: Monthly average South China Sea tensions from March 2015 to November 2017 (based on GDELT 2.0 GKG)  . 92 . Figure 14: Tensions by month with and without Australia, Brunei, China, India, Indonesia, and Japan (based on GDELT 1.0 Event Database) . 95 . Figure 15: Tensions by month with and without Malaysia, Philippines, Taiwan, United States, and Vietnam (based on GDELT 1.0 Event Database) . 96 . Figure 16: Tensions by month with and without Australia, Brunei, China, India, Indonesia, and Japan (based on GDELT 2.0 GKG) . 97 . Figure 17: Tensions by month with and without Malaysia, Philippines, Taiwan, United States, and Vietnam (based on GDELT 2.0 GKG)  . 97 . 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi U. v ni. Figure 18: State involvement by month for Australia, Brunei, China, India, 11 . 101 .

(14) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . Indonesia, and Japan from February 2011 to November 2017 (based on GDELT 1.0 Event Database) Figure 19: State involvement by month for Malaysia, Philippines, Taiwan, United States, and Vietnam from February 2011 to November 2017 (based on GDELT 1.0 Event Database) . 102 . Figure 20: State involvement by month for Australia, Brunei, China, India, Indonesia, and Japan from March 2015 to November 2017 (based on GDELT 2.0 GKG) . 103 . Figure 21: State involvement by month for Malaysia, Philippines, Taiwan, United States, and Vietnam from March 2015 to November 2017 (based on GDELT 2.0 GKG) . 104 . Figure 22: Monthly average South China Sea tensions from February 2011 to November 2017 (based on GDELT 1.0 Event Database) . 109 . 政治大. Figure 23: Monthly average South China Sea tensions from March 2015 to November 2017 (based on GDELT 2.0 GKG) . 110 . Figure 24: Training and test datasets showing monthly average South China Sea tensions from February 2011 to November 2017 (based on GDELT 1.0 Event Database) . 111 . Figure 25: Training and test datasets showing monthly average South China Sea tensions from March 2015 to November 2017 (based on GDELT 2.0 GKG) . 111 . Figure 26: Predicted South China Sea tensions by month using random benchmark model (for analyses based on GDELT 1.0 Event Database) . 113 . Figure 27: Predicted South China Sea tensions by month using random benchmark model (for analyses based on GDELT 2.0 GKG) . 114 . 立. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. i n U. v. Figure 28: Predicted South China Sea tensions by month using fixed benchmark model (for analyses based on GDELT 1.0 Event Database) . 115 . Figure 29: Predicted South China Sea tensions by month using fixed benchmark model (for analyses based on GDELT 2.0 GKG) . 116 . Figure 30: Predicted South China Sea tensions by month using linear benchmark model (for analyses based on GDELT 1.0 Event Database) . 117 . Figure 31: Predicted South China Sea tensions by month using linear benchmark model (for analyses based on GDELT 2.0 GKG) . 118 . Figure 32: Predicted South China Sea tensions by month using average benchmark model (for analyses based on GDELT 1.0 Event Database) . 119 . Figure 33: Predicted South China Sea tensions by month using average benchmark model (for analyses based on GDELT 2.0 GKG) . 120 . Figure 34: Predicted South China Sea tensions by month using simple exponential smoothing model (for analyses based on GDELT 1.0 Event Database) . 122 . engchi. 12 .

(15) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . Figure 35: Predicted South China Sea tensions by month using simple exponential smoothing model (for analyses based on GDELT 2.0 GKG) . 123 . Figure 36: Predicted South China Sea tensions by month using autoregressive model (for analyses based on GDELT 1.0 Event Database) . 125 . Figure 37: Predicted South China Sea tensions by month using autoregressive model (for analyses based on GDELT 2.0 GKG) . 126 . Figure 38: Predicted South China Sea tensions by month using moving average model (for analyses based on GDELT 1.0 Event Database) . 127 . Figure 39: Predicted South China Sea tensions by month using moving average model (for analyses based on GDELT 2.0 GKG) . 128 . Figure 40: Predicted South China Sea tensions by month using four ARIMA model variants (for analyses based on GDELT 1.0 Event Database) . 130 . Figure 41: Predicted South China Sea tensions by month using four ARIMA model variants (for analyses based on GDELT 2.0 GKG) . 131 . 政治大 Figure 42: Difference in monthly tensions and mean tensions with 立 Australia, Brunei, China, India, Indonesia, and Japan (based on GDELT 1.0 . ‧ 國. 學. Event Database) . 141 . 143 . Figure 44: Difference in monthly tensions and mean tensions with Australia, Brunei, China, India, Indonesia, and Japan (based on GDELT 2.0 GKG) . 147 . Figure 45: Difference in monthly tensions and mean tensions with Malaysia, Philippines, Taiwan, United States, and Vietnam (based on GDELT 2.0 GKG) . 150 . Figure 46: Cross-correlations between state involvement and tensions for Australia, Brunei, China, India, Indonesia, and Japan (based on GDELT 1.0 Event Database) . 157 . Figure 47: Cross-correlations between state involvement and tensions for Malaysia, Philippines, Taiwan, United States, and Vietnam (based on GDELT 1.0 Event Database) . 159 . Figure 48: Cross-correlations between state involvement and tensions for Australia, Brunei, China, India, Indonesia, and Japan (based on GDELT 2.0 GKG) . 161 . Figure 49: Cross-correlations between state involvement and tensions for Malaysia, Philippines, Taiwan, United States, and Vietnam (based on GDELT 2.0 GKG) . 163 . Figure 50: Predicted South China Sea tensions by month using random benchmark model (for analyses based on GDELT 1.0 Event Database) . 169 . ‧. Figure 43: Difference in monthly tensions and mean tensions with Malaysia, Philippines, Taiwan, United States, and Vietnam (based on GDELT 1.0 Event Database) . n. er. io. sit. y. Nat. al. Ch. engchi. 13 . i n U. v.

(16) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . Figure 51: Predicted South China Sea tensions by month using random benchmark model (for analyses based on GDELT 2.0 GKG) . 170 . Figure 52: Predicted South China Sea tensions by month using fixed benchmark model (for analyses based on GDELT 1.0 Event Database) . 171 . Figure 53: Predicted South China Sea tensions by month using fixed benchmark model (for analyses based on GDELT 2.0 GKG) . 172 . Figure 54: Predicted South China Sea tensions by month using linear benchmark model (for analyses based on GDELT 1.0 Event Database) . 173 . Figure 55: Predicted South China Sea tensions by month using linear benchmark model (for analyses based on GDELT 2.0 GKG) . 174 . Figure 56: Predicted South China Sea tensions by month using average benchmark model (for analyses based on GDELT 1.0 Event Database) . 176 . Figure 57: Predicted South China Sea tensions by month using average benchmark model (for analyses based on GDELT 2.0 GKG) . 177 . 政治大 Figure 58: Predicted South China Sea tensions by month using simple 立 exponential smoothing model (for analyses based on GDELT 1.0 Event . ‧ 國. 學. Database) . ‧. 181 182 . Nat. Figure 60: Predicted South China Sea tensions by month using autoregressive model (for analyses based on GDELT 1.0 Event Database) . sit. er. io. Figure 61: Predicted South China Sea tensions by month using autoregressive model (for analyses based on GDELT 2.0 GKG) . Figure 62: Predicted South China Sea tensions by month using moving average model (for analyses based on GDELT 1.0 Event Database) . 183 . Figure 63: Predicted South China Sea tensions by month using moving average model (for analyses based on GDELT 2.0 GKG) . 184 . Figure 64: Predicted South China Sea tensions by month using an ARIMA(0,1,1) model (for analyses based on GDELT 1.0 Event Database) . 186 . Figure 65: Predicted South China Sea tensions by month using an ARIMA(0,1,1) model (for analyses based on GDELT 2.0 GKG) . 187 . al. n. . 179 . y. Figure 59: Predicted South China Sea tensions by month using simple exponential smoothing model (for analyses based on GDELT 2.0 GKG) . 178 . Ch. n engchi U. . 14 . iv.

(17) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . List of Abbreviations ADMM. ASEAN Defence Ministers Meeting . AR. autoregressive . ARF. ASEAN Regional Forum . ARIMA. autoregressive integrated moving average . ARFIMA. autoregressive fractionally integrated moving average . ASEAN. Association of Southeast Asian Nations . CCP. Chinese Communist Party . EEZ. exclusive economic zone . FIRST. Facts on International Relations and Security Trends . FON. freedom of navigation . G7. Group of Seven . GCP. Google Cloud Platform . GDELT. Global Database of Events, Language, and Tone . GKG. Global Knowledge Graph . ICEWS. Integrated Conflict Early Warning System . IHO. International Hydrographic Organization . LOS. Law of the Sea . MA. moving average . MAE. mean absolute error . MOFA. Ministry of Foreign Affairs . PLA. People’s Liberation Army . PLAN. People’s Liberation Army Navy . PRC. People’s Republic of China (China) . SQL. Structured Query Language . ROC. Republic of China (Taiwan) . RQ1. Research Question 1 . RQ2. Research Question 2 . SES. simple exponential smoothing . ‧. ‧ 國. n. engchi. 15 . sit. io. Ch. y. Nat. al. er. G1ED. 學. G2GKG. 治政 GDELT 1.0 Event Database 大立 Knowledge Graph GDELT 2.0 Global. i n U. v.

(18) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . SIPRI. Stockholm International Peace Research Institute . SLOC. sea line of communication . TAC. ASEAN Treaty of Amity and Cooperation . UCDP. Uppsala Conflict Data Program . UN. United Nations . UNCLOS. United Nations Convention on the Law of the Sea . USS. United States Ship . VPA. Vietnam People’s Army . WEIS. World Event/Interaction Survey . . 立. 政治大. 學 ‧. ‧ 國 io. sit. y. Nat. n. al. er. . Ch. engchi. 16 . i n U. v.

(19) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . 1. Introduction 1.1 Background The South China Sea maritime territorial disputes have become a nexus of intense state interaction. In the contemporary Asia-Pacific political landscape, there are few other issues that involve such a diversity of vested interests. South China Sea claimants and stakeholders have much to gain or lose depending on the direction and outcome of the disputes, and the potential benefits and costs of future trends in escalation and deescalation are political, economic, and strategic in nature.   Because of the significance of South China Sea issues, there has been a surge . 治政大 analyses, the vast majority sensitivity and complexity of the topic has clouded many 立 of which are grounded in incomplete evidence, historical analogies, or the policy . of interest from policymakers, academics, and political pundits of all stripes. Yet the . ‧ 國. 學. perspectives of a single or incomplete subset of actors. Few studies have managed to take a quantitative approach to analyzing international political relations in the . ‧. South China Sea, not in small part due to the relative scarcity of such data.  . y. Nat. Now, more than ever, there is a need for research based on empirical evidence . io. sit. that focuses on the role of state involvement in the disputes and the level of tensions . n. al. er. in the maritime area. Studies of this nature can shed light on the relationship . i n U. v. between state behavior and tensions, historical trends in escalation and deescalation, . Ch. engchi. and the likelihood of tensions increasing or decreasing in the future. Moreover, they can confirm or contradict the findings or claims of existing studies, thereby imbuing South China Sea research with an empirical clarity that is sorely lacking in the field.  . 1.2 Significance of the Study This dissertation aims to contribute to academic research on the South China Sea disputes by using empirical, quantitative data on state involvement and tensions in the maritime area to conduct its analyses. In this way, it differs from most of the mainstream commentary and existing academic literature on the South China Sea, which remain predominantly conjectural and theoretical. Policy perspectives based on incomplete evidence or biased information remain the norm in discussions of 17 .

(20) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . South China Sea issues, and remarkably few have sought to resolve this issue through empirical research.   One issue that many observers have made claims or engaged in discussions about is the relationship between state involvement and South China Sea tensions. For policymakers, there is a natural tendency to view rival claimants and other countries as engaging in behavior that leads to escalation and their own countries’ actions as cooperative and conducive to decreasing tensions.1 Media outlets, academic researchers, and members of the general public for their own reasons have also joined in making similar conclusions that the involvement of certain states increases or decreases South China Sea tensions. Needless to say, it would be a . 治政 been the topic of much mainstream commentary and 大 academic analysis is the level of 立 tensions over time. These discussions relate to historical and current trends as well logical fallacy for all of these observers’ assertions to be true. Another issue that has . ‧ 國. 學. as future predictions of tensions, with many even flirting with ominous prophecies of armed conflict and war.  . ‧. Yet virtually all observations related to these issues rely on incomplete . sit. y. Nat. evidence or simplistic assumptions. In contrast, this research uses empirical . io. evidence in order to gain deeper insights into the complex issue and bring new . n. al. er. clarity to discussions about the South China Sea disputes. The dissertation is . i n U. v. motivated by the widespread disagreement highlighted above, and it focuses . Ch. engchi. specifically on two aspects of the maritime territorial disputes: (1) the relationship between state involvement and tensions and (2) the extent to which tensions can be modeled and predicted. The analyses use time-series event data from the Global Database of Events, Language, and Tone (GDELT). These are supplemented by reviews of official documents and news media coverage to confirm the validity of the data and connect the analyses to real-world events, highlighting the usefulness of the data and implications of the findings. While GDELT event data has been used to analyze and predict different forms of conflictive events, such as protests, violent uprisings, armed conflict, and genocides, in Africa, Eastern Europe, and the Middle See {2.2 Effects of State Involvement on Tensions} for an in-depth discussion and further examples of relevant commentary. . 1. 18 .

(21) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . East, no studies to date have used them in the South China Sea context. Through its analyses, the dissertation intends to fill this gap in the growing body of South China Sea research; provide complementary evidence and counterevidence relevant to the claims of policymakers, media analysts, and academic researchers; and as a first cut at the data, serve as a foundation for future research in the field. . 1.3 Research Questions The dissertation seeks to answer two primary research questions, both of which are related to South China Sea tensions. In subsequent sections, the first and second research questions are referred to as RQ1 and RQ2, respectively. 1.3.1 RQ1: State Involvement and Tensions . 政治大. RQ1 focuses on the relationship between the involvement of different state actors . 立. and South China Sea tensions. Concisely stated, RQ1 asks: . ‧ 國. 學. What is the relationship between state involvement and South China Sea tensions? . ‧. For readability, this concise form of RQ1 will be used throughout the text. In-depth . sit. y. Nat. discussion of the relevant theoretical framework is covered in {2.2.7 Theoretical . er. io. Framework for RQ1}; the methodology for the its two sets of analyses can be found in {3.3 Statistical Approach for RQ1}; and the corresponding results and discussion are . n. al. provided in {4.1 State Involvement and Tensions}. 1.3.2 RQ2: Predicting Tensions . Ch. engchi. i n U. v. RQ2 focuses on the prediction of South China Sea tensions based on the level of tensions in earlier time periods. Concisely stated, RQ2 asks: What forecast models most accurately predict South China Sea tensions? As with RQ1, this concise form of RQ2 will be used throughout the text. In-depth discussion of the relevant theoretical framework is covered in {2.3.5 Theoretical Framework for RQ2}; the methodology for the benchmark and forecast models can be found in {3.4 Statistical Approach for RQ2}; and the corresponding results and discussion are provided in {4.2 Predicting Tensions}. 19 . .


(23) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . 2. Literature Review 2.1 Context of the South China Sea Disputes The waters and coastal areas of the South China Sea have been host to human activity for thousands of years. However, the concept of political territory as it is understood today is a much more recent phenomenon, and it was not until the 1800s that incompatible maritime territorial claims in the contemporary sense of the term began to take shape in the South China Sea. Even then, littoral states’ interest in South China Sea territory was relatively muted while European governments and merchants started to invest heavily in surveying, mapping, claiming, and eventually . 治政大 later returned to the on Pratas Island were confronted by a Chinese ship, which 立 mainland to report the news. As the earliest documented confrontation in the South controlling its waters. In March 1909, a team of Japanese explorers harvesting guano 2. ‧. ‧ 國. things to come.  . 學. China Sea between regional powers, the event and its a ermath were an early sign of . Throughout the twentieth century, littoral states gradually gained interest in . sit. y. Nat. staking their claims to maritime territory in the South China Sea. In the early years, naval landings and military occupations were temporary and mostly symbolic in . io. n. al. er. nature. As political boundaries shi ed around the world in the years following World . i n U. v. War II, territorial grabs in the South China Sea involving permanent occupations . Ch. engchi. became increasingly common. In 1956, Taiwanese (Republic of China) forces garrisoned troops on Itu Aba (Taiping) Island. The Philippines’ territorial ambitions shi ed from concept to reality in 1970–1 when it occupied five sea features in the Spratly Islands.3 In the mid-1970s, tensions over territorial disputes between China and (South) Vietnam escalated with the latter taking control of six features in the Spratlys in 1973, its first such offensive. In early 1974, the military engagement between the two sides, later referred to as the Battle of the Paracel Islands, resulted Bill Hayton, The South China Sea: The Struggle for Power in Asia, New Haven, CT: Yale University Press, 2014. . 2. M. Taylor Fravel, Strong Borders, Secure Nation: Cooperation and Conflict in China’s Territorial Disputes, Princeton, NJ: Princeton University Press, 2008, p. 278. . 3. 21 .

(24) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . in China’s first seizure of South China Sea territory when the People’s Liberation Army Navy (PLAN) gained control of the Crescent Group of islands and reefs.4 Malaysia came into the fold in the 1980s, newly occupying sea features in 1983 and again in 1986.   As countries jostled to assert their claims, territorial expansion through military means during these years was for the most part politically motivated. Economic motivations remained limited given the lack of advanced technology for the exploration and exploitation of living and non-living maritime resources in the waters of the South China Sea. It was only in the late-1980s and 1990s that the disputes emerged as one of the key policy issues in Asia-Pacific regional relations. . 治政大 these was the shi in strategic significance of the South China Sea. First among 立 global political order towards the end of and a er the Cold War. The collapse of the Several key factors contributed to the mainstream recognition of the economic and . ‧ 國. 學. Soviet Union and resulting dissolution of the bipolar global power structure paved the way for a diversification of relevant actors and focus on territorial issues that had . ‧. previously been overshadowed by the struggle between the two great powers.5 . y. Nat. Second, the unilateral actions of several claimant states led to a reorganization of . sit. territorial claims and triggered responses from regional actors affected by the . er. io. actions. The nineties marked the first period of regular military confrontations on . al. iv n C resources led to a widespread realization h e nofgthe i U potential hidden beneath c heconomic n. and around the islands. Third, the initial exploration and discovery of energy the surface of the South China Sea.  . Into the twenty-first century, the further evolution of these three factors continued to amplify the significance of the South China Sea. Power shi s led to a reconfiguration of the Asia-Pacific geopolitical landscape, particularly with the emergence of China as a pivotal economic and political actor. Unilateral actions, maritime incidents, and military engagements continued to take place on a regular John W. Garver, “China’s push through the South China Sea: the interaction of bureaucratic and national interests,” The China Quarterly 132: 999–1028, 1992, p. 1001, <http://www.jstor.org/stable/654191>. . 4. Timo Kivimäki (ed.), War or Peace in the South China Sea?, Copenhagen: Nias Press, 2002. . 5. 22 .

(25) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . basis, resulting in intense but fleeting periods of escalation in bilateral relations. Meanwhile, oil and natural gas exploration capabilities continued to improve, bringing the exploitation of energy resources into sharper focus. Today, in addition to the political significance of maintaining territorial integrity, this drive for securing potential resources remains one of the most crucial factors fueling tensions over the South China Sea issue.6 As the site of intense interaction and trigger of dramatic fluctuations in tensions between states with incompatible maritime territorial interests, the South China Sea is naturally primed for analyses of state involvement and tensions over time. Conflicting claims among seven claimants compounded with the economic and . 治政大media outlets around the myriad instances of interaction that are covered by news 立 world that, in turn, provide an abundance of data for analysis. Figure 1 shows the political interests of rival claimants and major non-claimant stakeholders offer . ‧ 國. . ‧. . 學. complexity of overlapping territorial claims. . n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. Ralf Emmers, Resource Management and Contested Territories in East Asia, Basingstoke: Palgrave Macmillan, 2013, <https://doi.org/10.1057/9781137310149>. . 6. 23 .

(26) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . Figure 1: Conflicting claims to islands and other sea features in the South China Sea7. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. i n U. v. Note: Solid lines indicate complete claims over the territory; dashed lines indicate partial claims. . . engchi. Although academic and non-academic analyses of maritime territorial disputes in the region have proliferated in recent years, few studies have sought to apply scientific models based on empirical evidence to explain the dynamics of state interaction in the South China Sea. It is this significant gap in knowledge that this dissertation seeks to fill. . Author, 2017. . 7. 24 . .

(27) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . 2.2 E ects of State Involvement on Tensions Recent discussions about the South China Sea typically highlight the ongoing tensions in the maritime area. In many of these discussions, there is mention of the relationship between state involvement and tensions, particularly in terms of its role in leading to escalation and, albeit less frequently, deescalation. Policymakers, academics, and media analysts have all shown an eagerness to express their views, and relevant dialogues, statements, and publications clearly reveal key points of agreement and disagreement.   The primary point of agreement on the issue is that state involvement in the South China Sea has an effect on tensions in the maritime area. The relationship . 政治大 observers from all relevant stakeholder countries. Among these diverse individuals, a 立 general consensus has emerged that there is a correlation between state involvement between state involvement and tensions has been suggested in many fora and by . ‧ 國. 學. and tensions and, importantly, that certain states’ involvement is positively correlated with tensions. Despite the broad agreement that state involvement affects . ‧. tensions, there are several major points of disagreement related to the issue. In . y. Nat. particular, disagreement remains about (1) which country’s involvement influences . n. al. er. io. has an effect.  . sit. tensions, (2) whether it increases or decreases tensions, and (3) the extent to which it . Ch. i n U. v. Two major camps have formed regarding the first two of these three issues . engchi. (i.e., different countries’ relative effects and the direction of those effects). One side claims that US involvement is a cause of escalation and potential conflict while the other side asserts that China is the aggressor and whose actions, commonly framed as provocations, are the source of increased tensions. Although both countries have key interests in the future progression of the maritime territorial disputes, their roles and interests differ in significant ways, the most important being that China is a claimant and the US, a major non-claimant stakeholder. In addition to those two major camps, there have also been some other, albeit less prominent, voices asserting that the involvement of other countries, including both rival claimants and major stakeholders, is related to the escalation or deescalation of tensions. Moreover, it has become standard practice for many government and military officials to explicitly 25 .

(28) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . oppose state involvement that can escalate tensions while cautiously avoiding mention of any specific state in particular. The following subsections examine these views in greater detail and highlight relevant examples as supporting evidence. 2.2.1 China’s Involvement and Tensions China’s interests lie in the maintaining or reclaiming of sovereignty over islands and sea features considered to be historically Chinese and the regulation of South China Sea maritime territory and airspace, which would allow it to manage resource exploration and exploitation and ensure that actors from outside the region are unable to interfere in what is regarded as a domestic issue. As an extension of these interests, Beijing is also concerned with restoring its place in the global system . 治政大 displays of strength Communist Party (CCP) legitimacy domestically through 立 internationally, and reinforcing its role as the leader in the East Asian regional order. following an extended period of foreign incursions,8 maintaining Chinese . ‧ 國. 學. In one camp, policymakers, scholars, and pundits assert that China’s . ‧. involvement in the disputes is the major source of tensions. Chinese involvement and actions are o en framed by commentators as aggressive provocations that disrupt . sit. y. Nat. the status quo and run up against established norms or regulations governing . io. regional relations.9 In July 2011, then US Secretary of State Hillary Clinton cautiously . n. al. er. accused China of escalating tensions without explicitly naming the country, . i n U. v. declaring, “all of us have a stake in ensuring that these disputes don’t get out of . Ch. engchi. control, and in fact, the numbers have been increasing of intimidation actions, of ramming [and] cutting of cables.”10 Other officials in the administration have further advanced this line of reasoning, one example being when “a senior US State Department official said that while ‘no claimant is solely responsible for the state of tensions … a pattern of provocative and unilateral behaviour by China has raised . 百年國恥). . This era is commonly referred to as the “century of humiliation” (. 8. Jonathan Spangler, 2015, “Let’s Get Real about the South China Sea ‘Status Quo’,” Issue Briefings 9, Taipei: South China Sea Think Tank, 2015, <http://scstt.org/issue-briefings/2015/360/>. . 9. Michael T. Klare, “The United States Heads to the South China Sea,” Foreign Affairs, February 21, 2013, <http://www.foreignaffairs.com/articles/139023/michael-t-klare/the-united-states-heads-to-the-sout h-china-sea>. . 10. 26 .

(29) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . serious concerns about China’s intentions and willingness to adhere to international law and standards.’”11 In a joint press conference with US Secretary of Defense Ash Carter and his Philippine counterpart Voltaire Gazmin on April 14, 2016, Carter asserted, “in the South China Sea, China’s actions in particular are causing anxiety and raising regional tensions.”12 In an official statement by US Senator John McCain, Chairman of the Senate Armed Services Committee, and Senator Dan Sullivan, following the Award of July 12, 2016, issued by the Tribunal in the Philippines v. China arbitration case, they are forthcoming in their accusations, stating that “China faces a choice. China can choose to be guided by international law, institutions, and norms. Or it can choose to reject them and pursue the path of intimidation and coercion. Too o en in recent years, China has chosen the latter. The world will be . 政治大. watching to see the choice China makes.”13 . 立. These statements by US government officials have continued under the . ‧ 國. 學. current administration. More recently, at the Shangri-La Dialogue in July 2017, US Secretary of Defense James Mattis criticized China’s militarization efforts as . ‧. “undermin[ing] regional stability,” arguing that its “construction activities in the . y. Nat. South China Sea differ from those in other countries [due to] the nature of its . sit. militarization, China’s disregard for international law, its contempt for other nations’ . er. io. interests, and its efforts to dismiss non-adversarial resolution of issues. … [The US] . al. iv n C U the most vocal in their extent, US government and militaryhofficials e n g chave h ibeen n. cannot and will not accept unilateral coercive changes to the status quo.”14 To a large . Rowan Callick, “Taiwan escalates contest of the seas,” The Australian, July 14, 2014, <http://www.theaustralian.com.au/news/world/taiwan-escalates-contest-of-the-seas/story-e6frg6so1226987445836>. . 11. Ash Carter and Voltaire Gazmin, “Joint Press Conference by Secretary Carter and Secretary Gazmin in Manila, Philippines,” US Department of Defense, April 14, 2016, <https://www.defense.gov/News/Transcripts/Transcript-View/Article/721687/joint-press-conference -by-secretary-carter-and-secretary-gazmin-in-manila-phili/>. . 12. John McCain and Dan Sullivan, “Statement by Senators McCain and Sullivan on South China Sea Arbitration Award,” US Senate, July 12, 2016, <https://www.mccain.senate.gov/public/index.cfm/2016/7/statement-by-senators-mccain-and-sulliva n-on-south-china-sea-arbitration-award>. . 13. James Mattis, “News Transcript: Remarks by Secretary Mattis at Shangri-La Dialogue,” US Department of Defense, June 3, 2017, <https://www.defense.gov/News/Transcripts/Transcript-View/Article/1201780/remarks-by-secretary -mattis-at-shangri-la-dialogue/>. . 14. 27 .

(30) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . assertions about the relationship between Chinese state involvement and South China Sea tensions.   East Asian leaders, particularly those of ASEAN member states, have been relatively cautious in public, but more outspoken in private, about their accusations that Chinese involvement increases tensions in the South China Sea. Nevertheless, there have been various instances of vocal criticism. For example, in the midst of the Haiyang Shiyou 981 standoff between China and Vietnam, the Permanent Mission of Vietnam to the United Nations protested Beijing’s actions in a letter sent to UN Secretary-General Ban Ki-moon, which stated, “China’s activities, in violation of international law, escalate tension in the East Sea. … [The two countries’ sovereignty] . 治政 current tension is caused by China’s illegal placement大 of Haiyang Shiyou 981 oil rig 立 within the exclusive economic zone and continental shelf of Viet Nam”. The dispute is not the cause of the on-going escalating tension in the East Sea … The 15. ‧ 國. 學. Philippines, in its arbitration case against China, repeatedly accused China of escalating South China Sea tensions. These accusations can be found in many of the . ‧. documents that its legal team prepared throughout the arbitral proceedings. Its . y. Nat. Memorial submitted to the Tribunal on March 30, 2014, for example, states, “[t]he . sit. dispute between the Parties concerning their maritime entitlements in the South . er. io. China Sea escalated significantly following the official espousal of the nine-dash line . al. iv n C complicated and escalated the disputes China Sea, and imperilled the h e inn the hi U g cSouth n. claim to the United Nations in 2009.”16 It further asserts, “China’s conduct has . peace and stability of the region.”17 Although rival claimants have in the tensest cases explicitly denounced Beijing’s involvement and actions for being the primary cause of increased tensions, government officials have o en avoided mentioning China . “Viet Nam continues opposing China’s illegal rig placement at UN,” Vietnam Ministry of Foreign Affairs, July 3, 2014, <https://www.mofa.gov.vn/en/nr040807104143/nr040807105001/ns140704040726/view>. . 15. “Memorial of the Philippines: Volume I,” Government of the Republic of the Philippines, March 30, 2014, p. 8, <https://files.pca-cpa.org/pcadocs/Memorial%20of%20the%20Philippines%20Volume%20I.pdf>. . 16. “Memorial of the Philippines: Volume I,” Government of the Republic of the Philippines, March 30, 2014, p. 242, <https://files.pca-cpa.org/pcadocs/Memorial%20of%20the%20Philippines%20Volume%20I.pdf>. . 17. 28 .

(31) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . specifically in making these statements. Instead, they have opted to use cautious, diplomatic wording that calls upon all parties to avoid involvement that might lead to increased tensions.18   The US and rival claimants have not been the only ones asserting that China’s involvement escalates South China Sea tensions. In an interview on July 13, 2016, Australian Foreign Minister Julie Bishop stated, “Australia has been calling on China for some time to halt reclamation work and not to militarise its structures and we certainly urge all parties to take steps to ease tensions, to refrain from provocative actions that would escalate tensions and lead to greater uncertainty,” clearly implying a link between Chinese involvement and heightened tensions.19 . 治政 concluded in their Award of July 12, 2016, that China 大 had aggravated the disputes 立 over “the status of maritime features in the Spratly Islands” as well as those about Furthermore, the five arbitrators in the Philippines v. China arbitration case . ‧ 國. 學. the countries’ “respective rights and entitlements” and “the protection and preservation of the marine environment” at Mischief Reef.20 As the examples above . ‧. demonstrate, assertions about the relationship between Chinese state involvement . y. Nat. and South China Sea tensions have become a common theme for policymakers and . io. n. al. er. 2.2.2 US Involvement and Tensions . sit. analysts in both claimant and non-claimant states.21 . i n U. v. In the other camp are those asserting that US involvement has an effect on tensions . Ch. engchi. in the South China Sea. As a non-claimant stakeholder, US interests rest in the securing of sea lines of communication (SLOCs) through what is one of the most vital shipping routes globally and the maintenance of the status quo in regional relations, 18. See {2.2.5 Ambiguous Statements Linking State Involvement and Tensions}. . 19. Julie Bishop, “ABC AM, Perth - Interview with Kim Landers,” Minister for Foreign Affairs, Government of Australia, July 13, 2016, <https://foreignminister.gov.au/transcripts/Pages/2016/jb_tr_160713.aspx>. . 20. Award, Permanent Court of Arbitration, July 12, 2016, X, 1203, B, 16, <https://www.pcacases.com/web/sendAttach/2086>. . 21. In a sense, competing schools of diplomatic thought on the roles of China and the US in regional relations have emerged. Perceptions of China are divided between viewing it as the aggressor or as a partner in peaceful development. Likewise, views of the US are split between viewing it as the hegemon or as a guarantor of regional security and stability. 29 .

(32) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . in which international waters and airspace remain under the governance of international regimes, the regional balance of power remains multipolar, and Washington continues to enjoy its role as a leader and key decision maker in the Asia-Pacific region.22 Its interests in maintaining the status quo in the region further depend on the continued relevance of middle powers, including ASEAN, Australia, India, Japan, and South Korea; no abrupt changes to the de facto independent status of Taiwan; and ongoing support from regional allies for its leadership role. Chinese policymakers and media have been at the center of the camp arguing that US involvement, and that of foreign powers in general, increases tensions in the South China Sea. Internally, Beijing’s rationale is sound. Because the maritime . 治政大of sovereignty, even more so definition. For outside forces to get involved is a breach 立 than it is for those states with overlapping claims to assert their own. territorial disputes concern Chinese sovereignty, they are a domestic issue by 23. ‧ 國. 學. Chinese leaders’ statements clearly reflect its approach to the South China Sea . ‧. disputes and the notion that foreign involvement leads to escalation. Following US President Barack Obama’s suggestion that the South China Sea disputes be . sit. y. Nat. addressed at the ASEAN Summit in Bali in November 2011, then Chinese Premier . io. Wen Jiabao warned him against the involvement of “external forces,” declaring that . n. al. er. the issue would not be up for discussion at the meeting.24 In August 2014, Chinese . i n U. v. Foreign Minister Wang Yi, participating in the ASEAN Regional Forum (ARF) . Ch. engchi. meeting, remarked, “Some countries outside the region are restless, and stir up tensions … might their intention be to create chaos in the region? … China and ASEAN [can] safeguard well the peace and stability of [the] South China Sea, [and] . 22. Jonathan Spangler, “Claimant and Non-Claimant Views on SLOCs and Freedom of Navigation,” Issue Briefings 7, Taipei: South China Sea Think Tank, 2015, <http://scstt.org/issue-briefings/2015/11/>. . 23. Ralph A. Cossa, “Security Implications of Conflict in the South China Sea: Exploring Potential Triggers of Conflict,” Honolulu: Pacific Forum, Center for Strategic and International Studies, 1998, p. C-2, <http://www.geocities.ws/kweibohuang/SChinaSea.pdf>. . 24. Alex Spillius, “China warns Barack Obama over South China Sea ‘interference’, The Telegraph, November 18, 2011, <http://www.telegraph.co.uk/news/worldnews/asia/china/8899244/China-warns-Barack-Obama-over -South-China-Sea-interference.html>. 30 .

(33) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . relevant disputes should be addressed by countries directly concerned.”25 Following US Navy acknowledgement that it had flown a P-8 Poseidon, its latest and most advanced scout aircra , over the South China Sea in February 2015, Chinese Foreign Ministry spokeswoman Hua Chunying responded, “third countries should say less and stop stirring up trouble.”26 In October 2015, interactions between the PRC and the US further exacerbated perceptions that US involvement in the South China Sea increased regional tensions. On October 27, the USS Lassen, a guided-missile destroyer, passed within 12 nm of Chinese-occupied Subi Reef and Mischief Reef.27 The operation was part of the US Freedom of Navigation (FON) Program intended to “exercise and . 治政大reflected in the Law of the Sea manner that is consistent with the balance of interests 立 (LOS) Convention.” In response to the high-profile move, Beijing accused the US of assert its navigation and overflight rights and freedoms on a worldwide basis in a 28. ‧ 國. 學. acting “illegally” and summoned Max Baucus, the Beijing-based ambassador to China, to a meeting with Chinese vice-foreign minister Zhang Yesui. The Chinese . ‧. embassy in Washington warned that the US should “refrain from saying or doing . y. Nat. anything provocative and act responsibly in maintaining regional peace and . sit. stability”.29 US officials’ guarantees that more freedom of navigation operations . al. er. io. would take place in the future and the timing of the operation – just two days before . v. n. the Arbitral Tribunal constituted under the United Nations Convention on the Law . Ch. engchi. i n U. 25. “Beijing hits out at US South China Sea proposal,” AFP/Yahoo! News, August 11, 2014, <http://news.yahoo.com/china-hits-us-proposal-south-china-sea-063321948.html>. . 26. “US Navy flies new plane over disputed South China Sea,” RT, February 26, 2015, <http://on.rt.com/7nt5ae>. . 27. Keith Johnson and Dan de Luce, “Washington’s Muddled Message in the South China Sea,” Foreign Policy, November 5, 2015, <http://foreignpolicy.com/2015/11/05/washingtons-muddled-message-in-the-south-china-sea-lassen -fon-subi/>. . “Maritime Security and Navigation,” Bureau of Oceans and International Environmental and Scientific Affairs, US Department of State, <https://www.state.gov/e/oes/ocns/opa/maritimesecurity/>. . 28. Tom Phillips, “Beijing summons US ambassador over warship in South China Sea,” The Guardian, October 27, 2015, <https://www.theguardian.com/world/2015/oct/27/us-warship-lassen-defies-beijing-sail-disputed-so uth-china-sea-islands>. . 29. 31 .

(34) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . of the Sea (UNCLOS) in the Philippines’ arbitration case against China released its Award on Jurisdiction and Admissibility – only fueled perceptions in Beijing that US involvement in the South China Sea increases the level of tensions.   Since then, these assertions by government officials have continued. As an article written by Chinese Ambassador to the European Union Yang Yanyi on May 28, 2016, claims, “it is the United States that has been sending more and more military vessels and aircra s in close proximity to China’s coastal waters, constituting grave threats to China’s security on land and sea, and threatening to escalate tensions in Asia. More and more Chinese people have good reasons to believe that the U.S. is no longer just an invisible hand behind the rising tension in . 治政 transit of the USS Stethem, a US Navy destroyer, near 大 the Paracel Islands, China’s 立 Foreign Ministry Spokesperson Lu Kang commented, “China and ASEAN member the South China Sea.”30 More recently, in remarks on July 2, 2017, following the . ‧ 國. 學. states have cooled down and improved the situation in the South China Sea. The US … deliberately stirs up troubles in the South China Sea … [China] strongly urges the . ‧. US side to immediately stop such kind of provocative operations that violate China's . sit. y. Nat. sovereignty and threaten China’s security.”31 . io. Yet it is not only Chinese officials that have suggested a correlation between . n. al. er. US involvement and escalation. US professor and defense analyst Michael T. Klare, for example, has commented,  . Ch. engchi. i n U. v. In China’s eyes, [the US] is partisan in the island disputes and … Washington is actively spurring Japan and the Philippines to assume a more assertive stance … feeding distrust and resentment of the United States, and increasing the likelihood that future incidents at sea – however they are provoked – could . Yang Yanyi, “The European Sting publish an article by H.E. Ambassador Yang Yanyi, Head of the Chinese Mission to the EU, entitled Some Prevailing Arguments and Perceptions over the South China Sea Issue Are Simply Wrong,” Ministry of Foreign Affairs, People’s Republic of China, March 28, 2016, <http://www.fmprc.gov.cn/mfa_eng/topics_665678/nhwt/t1367333.shtml>. . 30. Lu Kang, “Foreign Ministry Spokesperson Lu Kang’s Remarks on US Missile Destroyer Stethem Trespassing China’s Territorial Waters off Xisha Islands,” Ministry of Foreign Affairs, People’s Republic of China, July 2, 2017, <http://www.fmprc.gov.cn/mfa_eng/xwfw_665399/s2510_665401/t1474745.shtml>. . 31. 32 .

(35) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . spark a clash between Chinese and U.S. vessels. … [R]egardless of whether it was Obama’s intention when he pivoted to the Pacific, he has surely increased the chances that rash and potentially incendiary behavior by any one of the countries … could lead to war.32 Such assertions that US involvement in the South China Sea is a cause of increased tensions are common.33 While those coming from a Chinese perspective demonstrate a particular tendency towards making these claims, they are by no means alone in their suggestions, as those from other backgrounds have also forwarded similar claims. 2.2.3 Other Rival Claimants’ Involvement and Tensions . 政治大 China increases tensions, there 立 are several other less prominent voices on the issue In addition to the two major camps asserting that involvement by either the US or . ‧ 國. 學. that attribute the involvement of other claimants besides China to escalation. Such . assertions are o en made by South China Sea claimant countries against other rival . ‧. claimants in response to specific incidents or actions causing diplomatic friction.  . y. Nat. One example of such voices are those arguing that state involvement by the . sit. Philippines has been positively correlated with increased tensions, particularly in . er. io. terms of its unilateral initiation of arbitral proceedings against China in 2013. In a . al. press conference on July 23, 2016, Chinese Ambassador to the UK Liu Xiaoming . n. iv n C argued that the Philippines’ arbitration case against U China had “intensified tension hen gchi. in the region,” further stating that “[t]he purpose of UNCLOS is to ensure the stable and peaceful environment of oceans and also relations between states. Yet this tribunal, instead of serving the purpose of UNCLOS, caused further tension between . 32. Michael T. Klare, “The United States Heads to the South China Sea,” Foreign Affairs, February 21, 2013, <http://www.foreignaffairs.com/articles/139023/michael-t-klare/the-united-states-heads-to-the-sout h-china-sea>. . 33. It deserves mention that some analysts have also suggested the reverse―that heightened tensions in the South China Sea are the cause of U.S. involvement. See, for example, M. Taylor Fravel, Policy Report: US Policy Towards the Disputes in the South China Sea since 1995, S. Rajaratnam School of International Studies, Nanyang Technological University, March 2014, p. 2, <https://www.rsis.edu.sg/wp-content/uploads/2014/11/PR140330_US_Policy_Towards_Disputes.pdf> . 33 .

(36) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . countries.”34 Government officials have also voiced concerns that Vietnamese involvement has led to escalation. In a statement on May 10, 2012 responding to earlier remarks by its Vietnamese counterpart agency, Taiwan’s Ministry of Foreign Affairs concluded, “[t]he government of the Republic of China (Taiwan) calls on the Socialist Republic of Vietnam to resolve disputes related to the South China Sea through dialogue and peaceful means, and avoid raising tensions in the area.”35 Chinese government officials have also blamed Vietnamese involvement for escalation, describing its actions during the Haiyang Shiyou 981 standoff with the statement that “the Vietnamese side takes advantage of China’s exercise of restraint to complicate the situation and escalate the tension, and further intensifies their actions without any restriction.”36 Likewise, the Vietnamese government has accused . 政治大. Taiwan of escalating tensions. Following naval exercises by the ROC Navy in the . 立. South China Sea, Vietnam’s Ministry of Foreign Affairs Spokesperson Le Thi Thu . ‧ 國. 學. Hang remarked, “Repeated live-fire drills by Chinese Taipei in the waters around Ba Binh in the Truong Sa Archipelago of Viet Nam have seriously violated Viet Nam’s . ‧. territorial sovereignty over the archipelago, threatened peace, stability, maritime safety and security, caused tension and complicated the situation in the East Sea.”37  . y. Nat. sit. Over the years, there have been many other instances of government officials . al. er. io. asserting that rival claimants’ involvement has increased tensions in the South China . v. n. Sea. However, like similar claims about China and the US, these statements are . Ch. engchi. i n U. Liu Xiaoming, “Ambassador Liu Xiaoming Holds a Press Conference on the Award of the Arbitral Tribunal in the South China Sea Arbitration,” Ministry of Foreign Affairs, People’s Republic of China, July 23, 2016, <http://www.fmprc.gov.cn/mfa_eng/wjb_663304/zwjg_665342/zwbd_665378/t1383890.shtml>. . 34. “The Ministry of Foreign Affairs of the Republic of China (Taiwan) reiterates its absolute sovereignty over Taiping Island,” Ministry of Foreign Affairs, Republic of China (Taiwan), May 10, 2012, <https://www.mofa.gov.tw/en/News_Content.aspx?n=1EADDCFD4C6EC567&sms=5B9044CF1188E E23&s=94F81457A4E53CA9>. . 35. Hong Lei, “Foreign Ministry Spokesperson Hong Lei’s Regular Press Conference on June 5, 2014,” Ministry of Foreign Affairs, People’s Republic of China, June 5, 2014, <http://www.fmprc.gov.cn/mfa_eng/xwfw_665399/s2510_665401/2511_665403/t1162678.shtml>. . 36. Le Thi Thu Hang, “Remarks by MOFA Spokesperson Le Thi Thu Hang on live-fire drills carried out by Chinese Taipei in the waters around Ba Binh in the Truong Sa Archipelago of Viet Nam,” Vietnam Ministry of Foreign Affairs, August 24, 2017, <http://www.mofa.gov.vn/en/tt_baochi/pbnfn/ns170824163224>. . 37. 34 .

(37) South China Sea Tensions: State Involvement and Prediction Using Gdelt Event Data . almost always based on incomplete evidence or biased interpretations of regional diplomatic relations rather than empirical evidence grounded in reality. 2.2.4 Other Major Stakeholders’ Involvement and Tensions In addition to these assertions that involvement by rival claimants has increased South China Sea tensions, there have also been assertions that involvement by major non-claimant stakeholders besides the US leads to escalation in the maritime area. Chinese government officials have been the most vocal in this regard, particularly targeting countries expressing preferences about diplomatic negotiation mechanisms for regional dispute resolution and those with navies operating in the South China Sea. . 政治大 increased tensions. In a regular 立 press conference on January 16, 2017, Chinese . Chinese officials have, for example, asserted that Japan’s involvement has . ‧ 國. 學. Foreign Ministry Spokesperson Hua Chunying stated, “the South China Sea situation is improving and parties concerned have come back to the right track of negotiation . ‧. and consultation thanks to the joint efforts of China and relevant ASEAN countries. However, the Japanese leader [Prime Minister Shinzo Abe] is still going to extremes . sit. y. Nat. to sow discord and play up regional tension. Such practice harbors ulterior motives . io. and such mentality is extremely unhealthy.”38 Addressing questions at a regular press . n. al. er. conference on July 28, 2017, about remarks by UK Foreign Secretary Boris Johnson . i n U. v. and Defence Minister Michael Fallon that the British government would send . Ch. engchi. aircra carriers and warships to the area the following year, Chinese Foreign Ministry Spokesperson Lu Kang responded, “not a single regional country is involved in hyping up tension in the South China Sea. At present, countries in the region are working together to safeguard and promote regional peace, stability and prosperity, yet we see some countries outside the region who insist on stirring up trouble while the situation is trending towards calm in the South China Sea. Whoever they are, under whatever pretexts and whatever they say, their precedents of interfering in other regions on high-sounding reasons but only leaving behind chaos and Hua Chunying, “Foreign Ministry Spokesperson Hua Chunying’s Regular Press Conference on January 16, 2017,” Ministry of Foreign Affairs, People’s Republic of China, January 16, 2017, <http://www.fmprc.gov.cn/mfa_eng/xwfw_665399/s2510_665401/t1430903.shtml>. . 38. 35 .