“Neighborhood Factors, Individual Characteristics, and the Lethality of Violence,” by Irshad Altheimer and four others from the Department of Criminal Justice at the Rochester Institute of Technology.
This was a very interesting presentation that begins with some empirical observations about the nature of gun violence in Rochester, NY from 2010 to the middle of 2013. Rochester, apparently, is sometimes thought of as the “murder capital” of New York.
First, gun violence is not randomly distributed in social and geographic space. Some people and places are far more dangerous than others. In this case, although the population of the Rochester metropolitan area is about 11% Black, 80% of gunshot victims in this time period were Black. This pattern has been found elsewhere, such as in a study of over 300,000 emergency room visits at a Level I Trauma Center in Middle Tennessee from 2004 to 2009 found that Black patients were 3 times more likely to present to this ED for gunshot wounds than non-black patients, after controlling for age, sex, and insurance status.
The victimization in Rochester was also concentrated in just a few geographic areas, with a single zip code accounting for 30% of all shootings. In some areas, gun violence is even more concentrated. As summarized by prominent researcher Anthony Braga (Rutgers University and Harvard), 5% of street blocks accounted for 74% of all shootings in Boston from 1980 to 2008. Just 60 locations experienced 1,000+ shootings.
Second, a lot of attention is paid to the number and rate of gun homicides, but violence is not limited to homicide. For every gun homicide there are two non-fatal shootings. In this case study of Rochester, there was 595 shootings in the 3.5 years studied, but only 76 homicides – accounting for just 13% of shootings.
The question this paper seeks to answer, then, is what distinguishes the 76 gun homicides from the 519 non-fatal gunshot injuries? What factors determine the lethality of violence?
Altheimer and his colleagues use a number of independent variables in a logistic regression analysis to predict lethality of shootings. These include situational factors such as whether the victim was the aggressor, intoxicated, or had a weapon; whether the shooting was in a drug location or there was a gang affiliation; and if the shooting was dispute related or committed during a robbery or drug sale. The statistical models also looked at the location of the gunshot wounds: head, pelvis, upper extremities, spine/back, thorax, abdomen, and lower extremities.
The results of the analysis show:
- The only situational factor that is a statistically significant predictor of lethality is if the victim had a weapon.
- Even stronger predictors of lethality than the victim being armed (3 times stronger, in fact) were if the victim was shot in the head or the thorax.
As interesting as these two findings are in themselves, the results are not amenable to any simple conclusions.
It makes sense that getting shot in the head or throat makes the gunshot injury more likely to be lethal, but can we also infer some INTENT on the part of the shooter when the victim is shot in these locations? Are head/throat shots more likely to be “execution style” shootings, as opposed to spray and pray? Were the gunshot victims who did not die just lucky that their attackers were bad shots?
In terms of the victims being more likely to die of their gunshot wounds if they have a weapon themselves, does this mean that the attacker fires more shots or more intentionally if their victim has a weapon? How many of these cases were instances of justifiable homicide in self-defense against an armed perpetrator? It makes more sense to me that someone who has a weapon may be more likely to GET SHOT, but why more likely to DIE?