Search
Browse By Day
Browse By Time
Browse By Person
Browse By Area
Browse By Session Type
Search Tips
ASC Home
Sign In
X (Twitter)
According to Campaign Zero’s Mapping Police Violence public dataset, from 2013-2023, police killed about 12,600 civilians in the US. However, localities across the US differ widely in the prevalence of these killings, ranging from 0.8 to 10.8 per million residents per year at the state level. Assessing simple trends of police violence over time, let alone the impact of interventions intended to curb violence, can be difficult, especially in localities with smaller populations. This is because police killings are discrete events that, in a given locality, can occur days, months, or years apart. We address these challenges by building a Poisson-based hierarchical model that represents the geographic variability within the US, groupings of cities and states within it, and across the cities and states themselves. We test different model structures, examine model accuracy, and estimate temporal trends in police killings at the different geographic levels, aiming to identify localities where violence has unequivocally increased or decreased. The models explore correlates across several administrative, socioeconomic, and demographic dimensions.