The daily operation and maintenance work of railways are very dangerous. Railway employees often have safety accidents while working, and the safety citizenship behavior (SCB) of railway employees can effectively reduce the accident rate. Therefore, it is of great significance to identify the main safety constructs affecting the SCB of railway employees to minimize accidents. This paper puts forward a supposed model of the influence mechanism of safety leadership (SL) on employee SCB through the mediation of safety attitude (SA). A questionnaire was used with railway employees, and 238 valid responses were finally collected. A structural equation model (SEM) was used to explore the relationship between SL, SA and SCB. The results showed the SL is positively related to the SA, and it can further promote the SCB of railway employees. In addition, SA has a positive impact on employee SCB.

In recent years, the safety issue of construction workers has become a research hotspot, and many researchers have achieved results in the impact of safety behavior regarding China’s construction industry. However, the existing research about the driving factors of safety citizenship behavior is insufficient. To fill this gap, this paper explores the driving factor of safety citizenship behavior from the perspective of social capital theory. A cross-sectional questionnaire survey, involving 311 Chinese construction workers, was conducted to verify the influence of Social Safety Capital on Safety Citizenship Behavior. The results showed that safety citizenship behavior made by workers was significantly related to social safety capital. Autonomous safety motivation mediated the relationships between social safety capital and safety citizenship behavior. Further, this research supports the differences between social safety capital and autonomous safety motivation. Specifically, the paper found that social safety capital had the largest regression coefficient for participation of suggestion-making, and autonomous safety motivation had the largest regression coefficient for the relationship between superior and subordinate by multiple regression analysis.