How do individual supply-chain integration dimensions independently and collectively affect operational performance? A study of manufacturing industries in Thailand

Purpose Following counterarguments from previous studies, overwhelming evidence from meta-analyses shows that supply-chain integration (SCI) is not universally associated with improved performance. This study aims to clarify all the mechanisms purported to reach the combined effects of SCI dimensions. Design/methodology/approach In this study, the authors first conducted a literature review and identified the effects of different individual SCI dimensions and their interactions and mediation, or arcs of integration, as reported by past studies. Data from a survey of 348 firms were analyzed using multiple regressions, structural equation modeling and latent class analysis. Findings The results showed that the positive independent associations between SCI dimensions and operational performance change as they interact. Larger combinational effects were found in a second-order SCI construct and a mediation model. Smaller effects and fewer significant paths (and even negative signs) were also found. This shows that more SCI is not always better. Finally, configurational effects varied: the latent class analysis showed that a more inward (internal integration) gestalt performed better for some firms. Research limitations/implications Some variables that could affect operations performance were missing from the models, and this paper did not specify the directions of relationships among the five performance dimensions. The authors are aware of cumulative models, but they are not within the scope of this paper. In addition, this study did not consider other sociocultural factors that may influence different dimensions of SCI. Originality/value This study shows SCI dimensions play different roles. Therefore, it is needed to theorize the distinct roles of internal and external integration and how they somehow compensate or substitute for each other. The results suggest that future studies need to reexamine the universal predictions especially from resource-based, resource-advantage and relational views. SCIs are not inimitable and rare resources for all firms. Furthermore, implementing SCI does not guarantee resource and relational advantages for all firms.