作者：
Yu-Yuan Shih
;
Chin-Wen Yang
;
Chung-Huang-Huang
;
Shin-Cheng Yeh
摘要：
Despite green jobs being well defined and case studies abundant in literature, insufficient data deters long-run forecasts. To resolve this problem, we designed a procedure to integrate ARDL estimates with data collected from in-depth interviews and surveys of 1,500 companies across 38 industries. This integrative approach allows us to map changes in employment to the demand for green jobs in terms of job opportunities, supply sources, and position expertise throughout the transition periods. To estimate the demand for green jobs for the 2050 NZE scenario, we employed a vector ARDL system that incorporates time-varying coefficients, synergy effects, financial variables, and sudden occurrences. Based on the deterministic phased regulations and the national long-run reduction targets, two scenarios of emission paths were specified and impacts on employment estimated using existing time-series data, followed by a log-run baseline projection up to 2050 using two forecasting approaches. Our empirical results are consistent with theoretical expectations and practical experiences and provide adequate evidence to reinforce the importance of modeling specifications. More interestingly, our empirical results of ARDL estimation, cointegration regression and error-correction model, revealed that the comparative advantages of ARDL model and time-varying coefficients specification are likely to be state-dependent.
