한미 FTA 개정과 한국의 대미 해외직접투자 이중차분법(DiD) 분석

This study evaluates the effect of the Korea–U.S. FTA amendment, which entered into force on January 1, 2019, on Korea’s outward FDI to the United States. Using an annual destination-country panel for 2013–2023, the analysis treats the United States as the policy(treated) unit and Germany, France, Australia, the Netherlands, Sweden, Italy, Spain, and Denmark as the control group. The empirical strategy is a Difference-in-Differences(DiD) framework with destination fixed effects, controlling for destination GDP(USD billions), the KRW/USD exchange rate, the U.S. average tariff, the destination corporate tax rate, and the destination policy rate. To address zeros, negatives, and heteroskedasticity common in FDI data, the paper reports both a levels specification(in dollar terms) and a Poisson pseudo-maximum likelihood(PPML) specification, and traces the dynamic path of effects via an event-study design. First, in the levels specification, Korea’s FDI to the United States increases by approximately USD 8.21 billion per year after the amendment(p<0.001), indicating a sizable expansion in investment scale. Second, PPML estimates indicate a statistically significant rise of about 184% (p=0.015), complementing the levels result by providing a proportional effect robust to zeros and heteroskedasticity. Robustness exercises—including an alternative control set (Japan, the United Kingdom, Canada), leave-one-out tests, and wild cluster bootstrap p-values—reinforce the substantive pattern of larger post-amendment FDI while underscoring identification caveats. Overall, the evidence points to economically meaningful post-amendment increases in Korea’s FDI to the United States, accompanied by prudent qualifications related to pre-trends and model sensitivity.