Objectives. To determine the impact of city-level cannabis decriminalization and medical marijuana laws (MMLs) on fatal traffic crashes in US cities. Methods. Using a census of fatal traffic crashes from the 2010 to 2017 Fatality Analysis Reporting System, we examined MMLs and cannabis decriminalization on fatal crashes by age and sex of driver. We used a Poisson difference-in-differences approach, exploiting temporal and geographic variation in marijuana decriminalization laws. Results. Cities experienced a 13% increase in fatal crashes involving 15- to 24-year-old male drivers following decriminalization (incidence rate ratio = 1.125; 95% confidence interval = 1.014, 1.249). This effect was immediate and strongest on weekend nights. We found no effect on female drivers or older males. Conversely, we found that MMLs were associated with fewer fatal crashes for both males and females, which was most pronounced in 15- to 24-year-old drivers. Conclusions. Unlike MMLs, which are associated with fewer fatal crashes, cities experienced a relative increase in fatal crashes involving young male drivers following marijuana decriminalization. Public Health Implications. MMLs stipulate consumption occurs at home, whereas decriminalization only lessens the penalty for marijuana possession. Therefore, travel incentives of such laws have heterogeneous effects on traffic safety.
Random effects probit and logit are nonlinear models, so we need predicted probabilities and marginal effects to communicate the economic significance of results. In these calculations, how one treats the individual-specific error term matters. Should one (i) set them equal to zero or (ii) integrate them out? We argue that (ii) is the quantity that most readers would expect to see. We discuss these in the context of the statistical package Stata, which changed its default predictions from (i) to (ii) in version 14. In Appendix 5, we illustrate how to calculate predictions and marginal effects using method (ii) in Stata 13 and earlier.
10 August 2023 in International Journal of Drug Policy
This study identifies the impact of Hurricane Matthew on school attendance in an agricultural community in rural Haiti. We conducted a survey of parents whose children attended a rural school prior to Hurricane Matthew to determine the mechanism by which hurricanes impact school attendance. We determined the marginal effect of family size and school enrollment using a probit model. Parents identified two primary causes for their children leaving school: a loss of income—through crop damage and livestock deaths—and requiring the children’s labor on the family farm. In our sample 96 children, 46% of the children enrolled in school, stopped attending because of the hurricane. No parent reported that their child(ren) left school because of illness or injury. Families with more children in school before the storm were 5% (p < 0.001) more likely to have a child remain in school. Families with some children not attending school before the hurricane were 7.6% (p < 0.001) more likely to leave school after the storm. The survey and probit model both suggest that an income constraint caused children to leave school. There is limited empirical evidence that students leave school to provide labor on family farms, and no evidence they leave school because of illness or injury.