Without Progress, There Is Only Regression: A Study of the Spatial Spillover Effects of Work Safety Target Assessments

Chinese-style modernization requires enhancement of the national emergency management system and improvements in public safety governance. Guided by a comprehensive national security strategy, actively advancing modernization of the emergency management system and governance capabilities are essential for achieving such modernization. Work safety constitutes the foundation of emergency management and is a crucial area for applying the principle of “coordinated development and safety.” The work safety target assessment system introduced in 2004 is regarded as a key measure to control various workplace safety accidents and to put an end to the passive approach to safety regulation. However, the spatial interaction patterns it has engendered among regions require in-depth analysis and validation. Although previous studies have explored the institutional effects and pathways of work safety target assessments, they often overlook the interactive behavior among spatially correlated regions resulting from implementation of these policies. Furthermore, there is an ongoing potential for methodological improvements in the estimation approaches. This article treats the “one-vote veto” system for work safety, implemented by provinces in 2005, as a quasi-experiment, constructing a nested Difference in Difference model and a spatial Durbin model to investigate the spatial spillover effects and the regional heterogeneity of target assessments on safety governance in related areas. Additionally, it aims to decompose these spillover effects to uncover the differing impacts of target assessments on safety governance in spatially correlated regions, analyzing both within-group and between-group variations and exploring distinct potential spatial interaction patterns. The study reveals several key findings. First, target assessments significantly enhance safety governance outcomes; provinces implementing the “one-vote veto” system saw a notable reduction in accident fatalities and death rates compared to those that did not implement such a system, indicating improved local governance. Second, there is a positive spatial spillover effect on safety governance in related regions; results from a spatial Durbin model show that the benefits of target assessments extend beyond the implementing region. Third, the spillover effects differ among spatially correlated areas: there is positive intra-group spillover in regions with a “one-vote veto” system, while those without such a system experience negative inter-group spillovers. Fourth, spatial spillover effects vary across industries and regions. In mining and industrial trade accidents, geographic factors play a dominant role in shaping the spillover effects, whereas economic factors have a minimal impact. In the eastern and central regions, the governance effects are largely contained within the regions, with limited spillover effects. In contrast, the central and western regions exhibit competitive effects, while in the eastern regions, the governance outcomes are more influenced by economic factors, demonstrating a clear demonstration effect. These conclusions provide valuable insights for improving future work safety target assessment systems and they effectively expand our understanding of intergovernmental strategic interaction models at local levels.