Economic Effect and Resolution Idea of the THAAD Political Conflict on South Korea’s Exports to China

Abstract

South Korea deployed the Terminal High Altitude Area Defense (THAAD) antimissile system in response to North Korea’s nuclear test, leading to the Chinese government’s opposition to THAAD, citing national security concerns. Chinese consumers held a large-scale “boycott for Korean products” campaign in 2014. In the same year, the China–South Korea free trade agreement (FTA) was signed. This study compares the double-difference (DID) and triple-difference (DDD) models using the impact of the THAAD political conflict on South Korean exports. In the DID model, THAAD reduced South Korean exports, while in the DDD model, the FTA effectively alleviated the impact of THAAD. The net difference of the FTA’s impact on the commodities boycotted in response to the THAAD conflict exists because some boycotted commodities were promoted by the FTA while others were not. The effect of the THAAD event shock was significant only in the first two years, with minimal subsequent changes in growth. THAAD and FTA’s dynamic effects provide evidence of how political conflict can eventually influence popular opinion and how the trade policy plays a significant role in the national conflict resolution. Finally, the study provides additional evidence on the effect of non-tariff barriers triggered by THAAD on the service industry.

KEYWORDS:

• THAAD political conflict
• free trade agreement
• double-difference estimation
• triple difference estimation

Disclosure statement

There are no conflicts of interest to declare.

Notes

1 The net difference in the DID model:

${\mathit{DID}}_{\mathit{thaad}}=\left\{E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{THAAD},t=1\right)-E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{THAAD},t=0\right)\right\}-\left\{E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{not }\mathrm{}\mathit{THAAD},t=1\right)-E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{not }\mathrm{}\mathit{THAAD},t=0\right)\right\}=\left(\left({\alpha }_0+{\gamma }_0+{\theta }_0+{\theta }_1\right)-\left({\alpha }_0+{\gamma }_0\right)\right)-\left(\left({\alpha }_0+{\theta }_0\right)-{\alpha }_0\right)=\left({\theta }_0+{\theta }_1\right)-{\theta }_0={\theta }_1$

2 The net difference on boycotted products:

${\mathit{DID}}_{\mathit{thaad}}=\left(E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{THAAD},t=1,\mathrm{j}=\mathrm{FTA}\right)-E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{THAAD},t=0,\mathrm{j}=\mathrm{FTA}\right)\right)-\left(E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{THAAD},t=1,\mathrm{j}=\text{not FTA}\right)-E\left(\mathrm{ln }{ex}_{jt}|j=\mathit{THAAD},t=0,\mathrm{j}=\text{not FTA}\right)\right)=\left(\left({\alpha }_0+{\gamma }_0+{\gamma }_1+{\gamma }_2+{\theta }_0+{\theta }_1+{\theta }_2+\rho \right)-\left({\alpha }_0+{\gamma }_0+{\gamma }_1+{\gamma }_2\right)\right)-\left(\left({\alpha }_0+{\gamma }_0+{\theta }_0+{\theta }_1\right)-{(\alpha }_0+{\gamma }_0)\right)={\theta }_2+\rho$

3 The net difference on products not boycotted:

${\mathit{DID}}_{\mathit{not }\mathit{thaad}}=\left(E\left(\mathrm{ln }{ex}_{jt}|j=\mathrm{not }\mathrm{}\mathit{THAAD},t=1,\mathrm{j}=\mathrm{FTA}\right)-E\left(\mathrm{ln }{ex}_{jt}|j=\mathrm{not }\mathrm{}\mathit{THAAD},t=0,\mathrm{j}=\mathrm{FTA}\right)\right)-\left(E\left(\mathrm{ln }{ex}_{jt}|j=\mathrm{not }\mathrm{}\mathit{THAAD},t=1,\mathrm{j}=\text{not FTA}\right)-E\left(\mathrm{ln }{ex}_{jt}|j=\mathrm{not }\mathrm{}\mathit{THAAD},t=0,\mathrm{j}=\text{not FTA}\right)\right)=\left(\left({\alpha }_0+{\gamma }_0+{\theta }_0+{\theta }_2\right)-\left({\alpha }_0+{\gamma }_0+{\gamma }_1+{\gamma }_2\right)\right)-\left(\left({\alpha }_0+{\gamma }_1\right)-{(\alpha }_0+{\theta }_0)-{\alpha }_0\right)={\theta }_2$

4 The net difference in the DDD model:

${\mathit{DID}}_{\mathit{thaad}}={DD}_{\mathit{thaad}}+{DD}_{\mathit{not} \mathit{thaad}}={(\theta }_2+\rho )-{\theta }_2=\rho$

5 The HS code's length ranges from 2 digits with the widest category to 10 digits with the most detailed category in the HS code standard. The more detailed the category, the longer the length of the code. For example, category HS 01 includes categories HS 0101, HS 0102, HS 0103, etc.; and category HS 0101 includes categories HS 010101, HS 010102, etc. According to the HS code, the range of tariff data corresponds to subdivided commodities, while the range of other variables' data is broader than tariffs. The data of other variables' HS code length determined selecting the tariff HS code's length. Therefore, this study uses the broader HS code, whose tariffs on all subdivided goods have been reduced, as treatment groups, while setting the rest as the control group, even if only one subdivided good has no tariff reduction.

Additional information

Funding

This study received no external funding.