A Multilevel Study On User Resistance At Multi-site Enterprise System Post-implementation Stage

User resistance has become a serious problem for enterprise system (ES) implementation. After a few years of implementation, ES is at its post-implementation stage in current companies. Given the complexity of the system, ES are usually adopted by large-scale companies which have many subsidiaries. Thus, investigating user resistance in the context of multi-site ES post-implementation is becoming critical for both academic and practical communities. Based on the literature of user resistance and the characteristics of ES post-implementation, user resistance at the ES post-implementation stage is to be expected. With the characteristics of multi-site ES implementation ( i.e., top-down implementation ) , it is believed that both individual-level and subsidiary-level factors affect user resistance. Therefore, in order to better understand the phenomenon in the multi-site ES post-implementation context, this dissertation investigates user resistance from a cross-level perspective.Based on the literature, this study proposes a research model that consists of several individual and subsidiary-level factors that may work together in affecting user resistance. Specifically, the model suggests that at the individual-level, IT-induced losses (IT-induced loss of power such as authority, resource control, and network centrality, as well as IT-induced role overload) can result in user resistance. At the subsidiary level, organizational power context (headquarters'coercive and reward power over subsidiaries) can moderate the relationships between the individual-level identified antecedents and user resistance. To test the hypotheses, a field study is conducted in a large organization in China. Partial least square (PLS) and hierarchical linear modeling (HLM) are used for data analysis.The results strongly confirm the hypotheses developed from the model. Some of the key findings are: 1) IT-induced loss of resource control, loss of network centrality and role overload positively affect user resistance; the effect of IT-induced loss of authority on user resistance is fully mediated by IT-induced loss of resource control, and the effect of IT-induced loss of network centrality on user resistance is partially mediated by IT-induced loss of resource control. 2) Headquarters' reward power and coercive power over subsidiaries have distinct moderating effects on individual-level relationships, i.e., headquarters'reward power over subsidiaries significantly attenuates the relationship between IT-induced role overload and user resistance, and headquarters'coercive power over subsidiaries significantly strengthens the relationship between IT-induced loss of resource control and user resistance. Limitations and implications for research and practical fields are discussed, and directions of future studies are suggested.