本研究以2012年07月至2021年12月間，以2018年7月中美貿易戰前後比較，觀察製造業採購經理人指數(PMI)國家別臺灣、中國、南韓、日本、美國及台灣六大產業別是否有影響。本研究在分析上區分成基本統計分析與迴歸模型分析，在基本統計分析將區分成中美貿易戰前後進行比較分析。在迴歸模型分析則設定中美貿易戰前後的虛擬變數做探討，在控制變數上則加入出口金額變數，以單根檢定、一階差分檢定、OLS模型與共整合檢定分析。 以國家別觀察，由基本統計分析發現，中美貿易戰後，美國、台灣、韓國指數平均值為上升，中國指數平均值下降，與預期相符，意謂著受到美國關稅制裁因素造成採購經理人指數影響。日本平均值為下降，則不符合本研究之預期。由迴歸模型估算結果，中美貿易戰變數對於各國PMI指數的影響均呈現負向影響，在日本與臺灣的PMI指數上，呈現高度統計顯著負面影響，在中國、美國與南韓則未達到10%統計顯著水準。各國出口值對於指數的影響，除了在中國的模型估計外，其餘各國都出現正向而且高度統計顯著的結果。就共整合檢定，中美貿易戰與出口值對於PMI指數變動分別為負向與正向顯著影響，但在非定態一階差分模型估算，則在日本與臺灣的PMI指數解釋上，則不具解釋能力。 以臺灣產業別觀察，由迴歸模型估算結果，中美貿易戰對電子暨光學產業、交通工具產業、電力暨機械設備產業具有明顯負面影響且具統計顯著性。在出口值的影響估算上，所有產業出口值對於產業別PMI指數都具有統計顯著正向影響。;This study examines the impact of the Purchasing Managers′ Index on t the US-China trade war across six major industries in Taiwan， China， South Korea， Japan， and the United States， comparing the period from July 2012 to December 2021. The analysis is divided into basic statistical analysis and regression model analysis. The basic statistical analysis compares the period before and after the US-China trade war. The regression model analysis introduces dummy variables related to the trade war and includes export value as a control variable， employing single-root tests， first-difference tests， and OLS model tests.
By country， the basic statistical analysis reveals that after the US-China trade war， the average PMI for the United States， Taiwan， and South Korea increased， while the average PMI for China decreased， consistent with expectations. This indicates that the US tariff sanctions affected the PMI. The decrease in the average PMI for Japan is contrary to the study′s expectations. The regression model estimates show that the US-China trade war has a negative impact on the PMI indices of all countries. For Japan and Taiwan， the negative impact is highly statistically significant， while it does not reach the 10% statistical significance level for China， the US， and South Korea. Regarding the impact of export values on the indices， except for China， the model estimates show a positive and highly statistically significant impact for all other countries. While by using the cointegration test， the explanatory variables of US-China trade war and export values have the significant negative and positive effects on the PMI indices， however， both the statistical significant effects do not exist by using the non-stationary first-difference estimation.
About the estimation results of Taiwan′s industries， the regression model estimates show that the US-China trade war has a significant negative impact on the electronics and optics industry， the transportation equipment industry， and the power and machinery equipment industry. Regarding the impact of export values， all industries have a statistically significant positive impact on their respective industry PMI indices.
