Citation

Xu J, Lu L, Wei J. Safety Sci. 2022; 152: e105789.

Copyright

(Copyright © 2022, Elsevier Publishing)

DOI

10.1016/j.ssci.2022.105789

PMID

unavailable

Abstract

Under the constraints of information flooding and asymmetry, the government attention has become a scarce resource, and appropriate allocation is a necessary condition for obtaining effective safety information and improving safety management. Focusing on the attention perspective, this paper studies the short-term dynamic difference and long-term stable relationship between the upper-level government and the lower-level government in the attention allocation of safety management under the Chinese bureaucratic system. Drawn on 3231 government work reports (2010-2020), we extract the topic contents of safety management through the LDA model. Then, with the help of machine learning and time series models, we calculated the intensity of safety management attention and analyzed their short-term and long-term relationship across different levels of governments. Empirical results identified the long-term consistency and short-term inconsistency in the safety management attention of provincial and prefecture governments. This phenomenon exists simultaneously between the central government and the provincial governments. We also found that increased short-term hierarchical differences in government attention significantly increase regional safety risk. In addition, the influence and stimulus factors, as well as the heterogeneity of hierarchical differences were detected in this study. We found that hierarchical differences of government safety attention affect access to safety-related information, thereby affecting regional safety risks. Meanwhile, the construction of a novel dataset based on government work reports provides a new pattern to explore government attention allocation. This study widens the horizon of safety management from the perspective of inter-government attention differences, and enriches the research on safety management and government attention.

Language: en

Keywords

Attention allocation; Hierarchical difference; Latent Dirichlet allocation; Local government; Safety management