危机学习是否有效——一项基于SDID的实证研究
Is Crisis Learning Effective—An Empirical Study Based on SDID

在新发展理念稳步推进、总体国家安全观深入落实的时代背景下，发展与安全成为当代社会的两大关键命题，如何统筹二者关系是亟待解决的重要时代难题。发展进程中频发的生产安全事故危害巨大，不仅造成了巨额经济损失，还引发了社会各界的高度关注，使政府面临严峻的舆论压力。危机学习作为针对性策略，能回溯事故全貌、剖析问题根源、弥补监管短板，优化生产安全政策制度。当前，危机学习研究在理论构建和案例分析等方面取得了一定成果，但受限于文本数据定性分析的局限，对其效能的量化研究仍显不足，缺乏广泛的实证支撑。基于此，本研究聚焦于探究影响危机学习效果的因素、衡量危机学习效果的方式以及危机学习成效的时效跨度，即其是仅具有短暂警示效应，还是能产生持久的正向驱动作用。为深入解答这些问题，本研究选取广东“12·20”特别重大滑坡事故作为案例，利用省级面板数据，采用合成差分法(SDID)进行实证分析。实证结果显示，广东省在事故发生后持续开展危机学习，取得了良好的效果，显著降低了年生产安全事故死亡人数。
Against the backdrop of the steady advancement of new development concepts and the in-depth implementation of the overall national security concept, development and security have become two key propositions in contemporary society. How to coordinate the relationship between the two is an important and urgent issue that needs to be addressed. Frequent production safety accidents during the development process cause huge harm. They not only lead to huge economic losses, but also attract the high attention of all sectors of society, putting the government under severe public opinion pressure. As a targeted strategy, crisis learning can trace back the entire picture of the accident, analyze the root causes of problems, make up for regulatory shortcomings, and optimize the policy and institutional framework of production safety. At present, crisis learning research has achieved certain results in aspects such as theoretical construction and case analysis. However, due to the inherent limitations of qualitative analysis of text data, quantitative research on its effectiveness is still relatively scarce, lacking extensive empirical support. Based on this, this study focuses on exploring the factors that affect the effectiveness of crisis learning, the ways to measure the effectiveness of crisis learning, and the time span of the effectiveness of crisis learning, that is, whether it only has a short-term warning effect or can generate a lasting positive driving force. To deeply answer these questions, this study selects the “12？20” especially major landslide accident in Guangdong as a case, uses provincial panel data, and conducts empirical analysis using the Synthetic Difference-in-Differences (SDID) method. The empirical results show that Guangdong Province has continuously carried out crisis learning after the accident and achieved good results, significantly reducing the annual number of deaths in production safety accidents.