Public service motivation (PSM) is widely treated as a resource that strengthens engagement, yet excessive motivation may yield diminishing returns. Integrating the job demands-resources model, conservation of resources theory, and the too-much-of-a-good-thing logic, this study tests a curvilinear relationship between PSM and job exhaustion, and the moderating role of collaborative culture. Using a two-wave survey of 408 Korean public servants with a one-month lag, hierarchical regressions identify a U-shaped pattern: exhaustion declines at low-to-moderate PSM but rises beyond a threshold. Collaborative culture significantly flattens this curve, indicating that supportive, collective norms mitigate PSM-related strain at very high levels of PSM. The findings challenge the assumption that PSM is uniformly beneficial, recasting PSM as a conditional resource that can operate as a demand at high levels. The study clarifies when PSM is sustainable and positions collaboration as an actionable boundary condition, yielding implications for motivation theory and for the collaborative organization of public service.
Street-level bureaucrats (SLBs) often face a dilemma between complying with formal rules and helping citizens, especially in enforcement contexts. While prior research on public services has focused on prosocial rule-breaking (PSRB), this study conceptualizes pro-citizen rule-breaking (PCRB) as SLBs' intentional violation of rules to benefit citizens within enforcement contexts where rule-breaking carries professional and institutional risks. Factors influencing PCRB were examined using a vignette-based survey experiment with 733 Chinese traffic police officers, who operate under a dual leadership system shaped by both formal bureaucratic rules and local responsiveness. The results reveal three main findings. First, formalized written rules significantly suppress PCRB compared to informal verbal directives. Second, PCRB cues from peers strongly increase officers' willingness to break rules in order to help citizens. Third, these peer influences moderate the effect of rule formalization on PCRB, weakening its inhibitory influence. Robustness checks using alternative measures confirm these results. By introducing PCRB into the study of enforcement contexts, this study highlights factors shaping street-level bureaucrats' decision-making.
This article empirically critiques the mobility of participatory governance models when transferred to hybrid institutional contexts. In so doing, it critically examines Jordan's 2017 decentralization reforms through elite interviews with 14 senior policymakers to unpack the structural conditions that impede reform. Through a blame avoidance lens, the study identifies three competing narratives on the reforms' shortcomings, which are subsequently analyzed using the Institutional Analysis and Development framework to highlight the institutional incompatibilities that shaped reform outcomes. Despite its aims to strengthen local democracy through networked governance, the reforms produced a dysfunctional model that amplified centralization and patronage in Jordanian public administration. The study problematizes assumptions that Western governance frameworks can inherently enhance public sector performance, highlighting that successful policy transfer of participatory models in hybrid settings like Jordan is contingent on contextualization and co-design of transferred reform programs in order to synthesize external solutions with local institutional configurations.
The rapid advancement of generative AI is expanding the scope of leadership research, enabling systematic historical text analysis for psychometric assessment of heads of state-a domain largely overlooked in traditional public management research. In this study, we utilized multiple generative AI tools to evaluate the leadership of 452 Chinese emperors spanning two millennia (Study 1) and of U.S. presidents across 46 presidencies spanning two centuries (Study 2) to examine the strengths and limitations of generative AI in this context. Across both studies, our historiometric analysis showed that generative AI exhibited high test-retest reliability, strong interrater agreement across different tools, and good convergent validity. Furthermore, generative AI-generated leadership scores were highly correlated with ratings provided by human experts and demonstrated robust predictive validity for criterion variables. However, generative AI demonstrated limitations in distinguishing between leadership styles that were conceptually similar but substantively distinct, suggesting relatively poor discriminant validity. These findings suggest the potential of generative AI to advance research on the leadership of heads of state and other historical figures, while also supporting contemporary public-sector leadership assessment by organizing and comparing evidence from available records.
