The influence of inferred traffic safety culture on traffic safety performance in U.S. States (1994–2014).

• This study replicated the method proposed by Page (2001) to infer the effect of traffic safety culture on traffic safety performance for the United States. • The results suggest that a positive traffic safety culture can have a protective effect by producing a lower-than-expected fatality rate. • Conversely, a negative traffic safety culture can have an exacerbating effect by producing larger-than-expected fatality rate. • The derived metric for estimating traffic safety culture had strong concurrent validity by correlating with the ranking of states based only on total crash fatality rate. • Consistent with Page (2001), the analysis also identified common risk factors across states including per capita alcohol consumption and unemployment rate. Introduction: Traffic safety performance (crash fatalities per billion vehicle miles traveled) is influenced by many factors related to the physical and social environment. The traffic safety culture in the local environment can influence behaviors that influence the risk of a fatal crash. However, if traffic safety culture is defined as "shared beliefs," it is not possible to directly observe the effect of culture on traffic safety performance. Method: This study replicated the method proposed by Page (2001) to infer the effect of traffic safety culture on traffic safety performance for U.S. states between 1994 and 2014. This method infers the influence of traffic safety culture from the error between actual and predicted performance based on observable variables that measure the physical and social environment as well as behavioral hazards. Results: The results suggest that a positive traffic safety culture can have a protective effect by producing a lower-than-expected fatality rate. Conversely, a negative traffic safety culture can have an exacerbating effect by producing a larger-than-expected fatality rate. Conclusion: The derived metric for estimating traffic safety culture had strong concurrent validity by correlating with the ranking of states based only on total crash fatality rate. Practical Implications: Consistent with Page (2001), the analysis also identified common risk factors across states including per capita alcohol consumption and unemployment rate. [ABSTRACT FROM AUTHOR]