"Black Lives Matter...?" Naw. How About America Might Murder Me and Replace My Innocent Life With a Bullshit Catch Phrase Slogan, drawing by B. Culpepper

Identifying the drivers of the exceptionally high rates of lethal and racially biased violence by police in the United States requires geographically resolved data and analysis

Drawing on the Fatal Encounters database [Burghart, 2014], Hirschfield [2015] presents a brief review of plausible drivers for the exceptionally-high rate at which police officers in the United States kill American civilians. Hirschfield [2015] outlines how police violence in America is likely to be driven by the interaction of a variety of structural factors (e.g. institutionalized racism, neoliberal cuts in the social safety net, and a booming weapons industry—for both civilians and police), canalized cultural norms (e.g. hyper-masculine and racist norms among police officers, and the American ideology of rugged individualism and minimal government), and proximate causes (e.g. police fear of possibly armed civilians and high crime rates). Future research on police violence will likely transition from such broadscale surveys of theoretically possible drivers of killings by police to locally-resolved studies that address the principle empirical drivers of police violence operating in a given county or department. Especially apposite will be locally-resolved qualitative studies that contrast the roles of racist norms in a specific police department with other structural drivers operating in that locale.

The first step in realizing such locally-resolved studies is identifying the police departments in which rates of racial bias in police killings is highest. While unarmed black Americans are—on average—more than twice as likely to be killed by police than unarmed white Americans [Hirschfield, 2015], such statistics overlook the massive spatial heterogeneity in racial bias in police shootings. For example, in Cook, IL (Chicago), Philadelphia, PA, Clark, NV (Las Vegas), Travis, TX (Austin), Baltimore, MD, and Orleans Parish, LA (New Orleans), unarmed black American civilians are more than seven times as likely to be shot by police as unarmed white Americans [Ross, 2015]; in Los Angeles, CA, Pulaski, AR (Little Rock), Allegheny, PA (Pittsburgh), and Miami-Dade, FL, unarmed black American civilians are more than ten times as likely to be shot by police as unarmed white Americans [Ross, 2015]. Furthermore, in Miami-Dade, FL, Allegheny, PA, Pulaski, AR, Travis, TX, and Los Angeles, CA, police shot more unarmed black civilians than armed black suspects in 2011-2014, problematizing the claim by Hirschfield [2015, pp. 1] that police “generally [kill] someone holding a dangerous weapon.” While at the national level a person shot by police is about three times as likely to be armed than unarmed, there is significant heterogeneity in this rate geographically, and by the race of the victim [Ross, 2015]. Contrary to the claim that racism by police “cannot explain the fact that police lethality is greatest in states where African-Americans are least prevalent” [Hirschfield, 2015, pp. 1], county-level analysis of police shootings shows that racial bias in police shootings is actually strongest in large metropolitan counties with low median incomes and an elevated portion of black residents, especially when there is high financial inequality in that county [Ross, 2015]. However, even if this wasn’t the case empirically, racist norms and the violence they provoke could theoretically be strongest in counties with predominately white populations.

In addition to the Fatal Encounters database [2014], the Stolen Lives Project [1999], the Killed by Police database [2015], the U.S. Police-Shooting Database [2014], Wikipedia.org [2015], and the Guardian [2015] have begun rigorous documentation of police violence. Future research should draw on and compile geo-located data from all of the above databases. By analyzing such data as a function of other geographically-resolved variables, such as rates of weapons ownership by civilians (and weapons budgets for police), local-level racist norms, generalized social values (individualist vs collectivist spectrum), race-specific rates of crime, poverty and inequality levels, and population size, rigorous statistical methods can be used to evaluate the theoretical predictions outlined by Hirschfield [2015].

Read the paper online here or download the preprint here.


  title={Identifying the drivers of the exceptionally high rates of lethal and racially biased violence by police in the United States requires geographically resolved data and analysis},
  author={Ross, Cody T},
  booktitle={Sociological Forum},

Dialogue & Discussion