Does Taiwan's entry into the WTO truly reduce its agricultural output values?

Abstract

This study has used the time series data of Australia, China, Denmark, India, Japan, the Netherlands, Norway, the Philippines, South Korea, Spain, and the United States from the World Bank to estimate the agricultural economic shock of Taiwan's accession to the World Trade Organization (WTO). We have adopted the recently developed panel data approach for policy evaluation to construct the annual agricultural growth path of Taiwan, which is mainly based on the cross-sectional correlations between the domestic and international agricultural trade markets in the absence of Taiwan's entry into the WTO. Our results have not only revealed the importance of ex post counterfactual analysis, but also provided empirical evidence that the agricultural economic shock is not as severe as predicted by many ex ante studies. Based on these results, we have concluded that the outcome may have arisen from the slow progress of achieving agricultural trade liberalization under the WTO and the Taiwanese Government's effective adoption of phase-in periods and relative adjustment policies.

Keywords:

• WTO
• panel data
• treatment effect
• counterfactual analysis
• agriculture
• Taiwan

Jel Classification:

• Q17
• Q18

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. According to Hsiao's Monte Carlo studies (Hsiao et al. 2012), there are a number of model selection criteria one can use to select the best approximating model. But by comparing the post-intervention mean square prediction error (PMSE), they have shown that when the number of countries included (p) increases, the pre-intervention MSE decreases, whereas PMSE decreases initially and then raises when p increases. In addition, they have also presented that between 4 and 6 cross-sectional countries are chosen in terms of AICC and 6–12 ones are chosen in terms of AIC on average. Thus, when T is small, say 25 or 40, the PMSE in terms of models chosen by AICC and AIC are significantly smaller than the models using all cross-sectional countries.