Skip to main content
Fig. 2 | Injury Epidemiology

Fig. 2

From: Re-prioritizing traffic stops to reduce motor vehicle crash outcomes and racial disparities

Fig. 2

Treatment – Control Trends and Placebo Tests, pre- and post-intervention period. All permutations of non-intervention law enforcement agencies were treated as if they had enacted the intervention during the period, even though they hadn’t (placebo treatment), then likewise matched on pre-intervention period metrics using the same synthetic control process. The estimated change (treatment – control, ideally zero during the pre-intervention period) is graphed for both actually treated and placebo treatments. Some placebo comparisons produce outlier trend lines because the control pool was selected for Fayetteville, and placebos may be inadequately matched

Back to article page