No abstract

Recently, several states have revisited their laws governing cannabis possession and consumption. Whereas some states have legalized cannabis use at different levels (i.e., legal recreational use, decriminalized possession, or legal medical use), others have zero tolerance. This study used 5 years of data (2010 to 2014) from the Fatality Analysis Reporting System to assess differences across states with various laws for cannabis use. Random parameter logistic regression models were estimated to examine differences in the prevalence of positive cannabis tests across states while controlling for other salient factors, including driver and roadway characteristics. The findings demonstrate that drivers in states with more lenient laws for cannabis possession and consumption were more likely to test positive compared with drivers in states with zero tolerance. In addition, ordinal logistic regression models were estimated to assess the relationships between injury severity, cannabis laws, and other factors associated with crash injury outcomes. The results suggest driver injuries were more severe when the driver tested positive for cannabis. A strong association was also observed between blood alcohol content and cannabis test results. Ultimately, the results provide evidence in support of continuing research to better understand the public health impacts of cannabis use as more states consider changes to the laws governing possession and consumption.

No abstract

No abstract

Work zone temporary traffic control strategies generally affect both traffic safety and operations. However, there is a substantial gap in the knowledge base with respect to the safety impacts of various work zone characteristics. The Highway Safety Manual provides crash modification functions that account for the effects of project length and duration on crash frequency as compared with normal road operations. However, these methods do not allow for explicit comparisons of expected safety performance among different work or closure types. This research examined the safety impacts of various temporary traffic control strategies on freeways, including shoulder closures, lane closures, and lane shifts. Data were collected for the periods during which these treatments were in effect and during similar nonconstruction periods from the preceding year. Safety performance functions were estimated that account for segment length, duration, traffic volume, and closure type. Random parameter count data models were estimated to accommodate segment-specific temporal correlation and unobserved heterogeneity. Crash rates were shown to vary roughly in proportion to traffic volumes. In contrast, segment length and project duration showed inelastic effects; this finding implies that crash rates increase more rapidly in work zones that are shorter in length or duration. Single-lane closures, multilane closures, and lane shifts were associated with an increase in crashes, whereas shoulder closures did not show a significant difference compared with similar, non-work-zone conditions. Ultimately, the study results provide important information that can be used to assess the crash risk for various temporary traffic control strategies.