Expert Commentary

Using public health methods to investigate the diffusion of homicide

2012 study from Michigan State University published in Justice Quarterly on the diffusion of homicide clusters in a low-income city.

Man with gun (iStock)
(iStock)

Research suggests that homicides tend not to be randomly situated, and they typically cluster in lower-income areas. To understand this phenomenon better, a 2012 study from Michigan State University published in Justice Quarterly, “Homicide as Infectious Disease: Using Public Health Methods to Investigate the Diffusion of Homicide,” examined the spatial diffusion of criminal activities over time. Guided by a hypothesis that “homicide would diffuse in a similar process to an infectious disease with firearms and gangs operating as the infectious agents,” the researchers — April M. Zeoli, Jesenia M. Pizarro, Sue C. Grady and Christopher Melde of Michigan State University — tracked homicides in Newark, New Jersey, from January 1982 through September 2008.

Approximately one-quarter of Newark residents live under the federal poverty line; 38% of those 25 years and older have completed less than a college degree. During the study period, Newark’s homicide rate was 30.5 per 100,000 residents, nearly four times the national average. Gang activity dramatically escalated in the mid-1990s, and nearly a third of all murders were categorized as gang-related from 1997 to 2008.

Study findings include:

  • Homicides were initially confined to the Central Ward section of the city but spread to Newark’s West and South Wards. Firearm homicides spread throughout the West Ward and the western portion of the South Ward, and gang-related homicides from the lower section of the Central Ward and outwards to the upper Central Ward as well as the West and South Wards.
  • Gang-related homicides were concentrated in areas of the city already suffering from higher-than-average homicide rates before they spread to adjacent areas.
  • Approximately two-thirds of homicide victims and over three-quarters of offenders were African-American males from 26 to 30 years old.
  • “Of the 20 overall homicide clusters… 75% of them have a population that is over 60% African-American and, in the census tract [center] of 70% of the clusters, over 20% of the residents live below the poverty line.”
  • In Newark there is a small census tract flanked by areas with high levels of gang activity that reported no gang-related homicides during the study period. The North and East Wards, which have remained relatively immune to the elevated homicide rates of adjacent wards, are comparatively wealthy and home to a lower percentage of African-Americans.
  • Homicide is more likely to encourage future homicides when citizens are armed, out to exact revenge or bolster one’s pride, and when formal legal channels are viewed as inadequate or discriminatory. “Interactions, such as drug transactions or disputes aimed at saving face, that do not lend themselves to resolution through the legal system are disproportionally related to homicide in areas with high homicide rates…. The lethality of these interactions increases in the presence of crime facilitators such as firearms, drug markets and gangs.”
  • “Gang members and those who live criminal lifestyles may be uniquely susceptible [to homicide perpetration],” the researchers state, citing previous research that indicates that gang homicide rates are up to 100 times that of the broader population.

The researchers conclude that crime in Newark indeed spread in ways similar to the patterns of infectious disease, and they suggest that this insight may help improve intervention strategies: “Public health offers a focus on primary prevention within populations, which can be developed from information gained through tracking trends, patterns, and epidemiologic characteristics of the problem under study. This research is among the first attempts to use spatio-temporal clustering techniques from the field of medical geography to track the movement of homicide clusters through an urban city. The pattern of movement identified suggests that social structural variables, such as economic disadvantage and racial isolation, foster the spread of homicide, and that there exist barriers to the spread of homicide to some communities.”

Tags: crime, guns, drugs, policing

About The Author