Organizational Context Matters: Interdepartmental Collaboration and Performance in Law Enforcement vs. Scientific...

This study challenges the assumption that collaboration beyond work-unit boundaries consistently enhances performance and demonstrates that the impacts of inter-unit collaboration are not uniform across different government agencies. Using multi-group analysis of Federal Employee Viewpoint Survey (FEVS) data, we compare two U.S. federal agencies with contrasting institutional logics: a bureaucratic law enforcement agency (ICE) and an entrepreneurial science research center (GSFC). In doing so, we propose a model explaining how inter-unit collaboration influences performance through knowledge acquisition and resource availability as mediating variables. The analysis revealed that inter-unit collaboration had a significantly greater impact on work unit performance in the bureaucratic ICE than in the entrepreneurial GSFC. Specifically, the total effect of inter-unit collaboration on unit performance was 0.166 for ICE and 0.087 for GSFC, whereas the total effect of intra-unit collaboration reached 0.525 and 0.686, respectively. These findings contradict our hypothesis that inter-unit collaboration will be more beneficial in entrepreneurial agencies. Looking closely at the mediating paths, we argue that high technical complexity and rigid task compartmentalization in scientific agencies create barriers to lateral sharing, whereas law enforcement tasks are often case-based and emergency-driven and require more effective real-time coordination among units. Paradoxically, this unexpected result reinforces our initial assumption that inter-unit collaboration can produce diverse outcomes by showing that the nature of tasks and the resulting collaboration structures-factors we initially overlooked-serve as contingencies for performance gains.