The Propensity Score Matching for the Treatment Effects of Employee Training

• Main Authors: Yu-Heng Chen, 陳昱亨
• Other Authors: 張嘉玲
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/cgi-bin/gs32/gsweb.cgi/login?o=dnclcdr&s=id=%22107NCHU5412015%22.&searchmode=basic

碩士 === 國立中興大學 === 應用經濟學系所 === 107 === Since the “Knowledge-Based Economy” began to receive extensive attention and discussion in 1996, human capital has become an essential asset to the sustainable development of enterprises. In addition to recruiting new employees, employee training has become an important channel to accumulate human capital. In this study, the enterprise was divided into two groups: the investment employee training (experimental group) and the non-investment employee training (control group). The main purpose is to compare the effect of the two groups on the treatment effects of labor productivity and output. Using Industry and Service Census data from The Directorate General of Budget, Accounting and Statistics (DGBAS) in 2006 and 2011 as well as taking manufacturing as the research samples. Moreover, we used propensity score matching as a research method to avoid the bias of the estimation results caused by the differences of enterprise characteristics (for example: R&D density, proportion of technical employees, etc.) between the experimental group and the control group. The empirical results show that enterprises effectively improve labor productivity and output after training employees. Further estimation by 2-digit industries can also find that most industries have a positive relationship in the average treatment effect (ATE) and the average treatment effect on treated (ATT). Among them, manufacture of pharmaceuticals and medicinal chemical products、manufacture of electrical equipment、manufacture of beverages and manufacture of other non-metallic mineral products have the best performance on the treatment effect of output. However, manufacture of plastics products has a negative relationship in the average treatment effect (ATE) and the average treatment effect on treated (ATT).