Effects of Proactive Environmental Strategy on Environmental Performance: Mediation and Moderation Analyses
Shanshan Zhang,Zhiqiang Wang,Xiande Zhao
DOI: https://doi.org/10.1016/j.jclepro.2019.06.220
IF: 11.1
2019-01-01
Journal of Cleaner Production
Abstract:Contradictory findings on the relationship between proactive environmental strategy and performance are reported in the existing literature. The way that proactive environmental strategy is defined, the omission of crucial mediating variables and the misfit between strategy and business structures may help explain these conflicting findings. Drawing upon the logic of strategy-action-performance, this study tests the effects of two implementation variables of proactive environmental strategy—green human resource management (GHRM) and environmental legitimacy (EL)—on environmental performance as well as the mediating roles played by the implementation of green operational practices (GOP). Further, based on contingency theory, this study proposes that the impacts of GHRM and EL on the implementation of GOP might be moderated by operational barriers (OB) and managerial barriers (MB). The proposed hypotheses are empirically tested using a dataset of 260 sample plants in 11 countries/regions (Finland, Sweden, Germany, Taiwan, Israel, Mainland China, Spain, Brazil, Italy, South Korea and Japan). The results show that the effect of GHRM on environmental performance is fully mediated by the implementation of GOP (the path coefficient b for the direct link between GHRM and environmental performance changes from 0.436, p < 0.001 to 0.180, p > 0.1). EL has no statistically significant direct or indirect effects on environmental performance (direct effect: b = −0.08, p > 0.1, indirect effect: b = 0.083, p > 0.1) but has a marginally significant effect on the implementation of GOP (b = 0.218, p < 0.1). The results also indicate that OB reduce the effectiveness of EL in facilitating the implementation of GOP (b = −0.103, p < 0.1), and MB reduce the effectiveness of GHRM in supporting the implementation of GOP (b = −0.11, p < 0.1).