Detecting DUI (Non) deterrence: A macro-methodology to uncover "restrictive v permissive" county jurisdictions in...

This paper builds a method to detect the apparent restrictiveness or permissiveness of communities towards drunk-driving. A framework of three mutually interacting community domains is used to motivate a minimum set of DUI patterns to be expected from an appropriately deterrent environment. Based on the (simplified) system dynamics model, an empirical estimation strategy and scoring methodology is outlined. This "macroscopic" approach is demonstrated using results from time-series panel analyses of California's 58 counties for the years 1990 to 2010 (Van Vleck et al., 2017). The process successfully classified three-quarters of California counties, encompassing almost 90% of the state population. The paper demonstrates a potential tool to classify communities' systemic behavior toward drinking-and-driving and other enforcement-sensitive subjects.