Barriocide: Investigating the Temporal and Spatial Influence of Neighborhood Structural Characteristics on Gang and Non-Gang Homicides in East Los Angeles

• Published date: 01 November 2017
• DOI: 10.1177/1088767917726807
• Author: George E. Tita,Shannon E. Reid,Matthew Valasik,Michael S. Barton
• Date: 01 November 2017
• Subject Matter: Articles

https://doi.org/10.1177/1088767917726807

Homicide Studies

2017, Vol. 21(4) 287 –311

© 2017 SAGE Publications

Reprints and permissions:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1088767917726807

journals.sagepub.com/home/hsx

Article

Barriocide: Investigating

the Temporal and Spatial

Influence of Neighborhood

Structural Characteristics

on Gang and Non-Gang

Homicides in East Los

Angeles

Matthew Valasik1, Michael S. Barton1,

Shannon E. Reid2, and George E. Tita3

Abstract

This study explored how changes in neighborhood structural characteristics predicted

variation in gang versus non-gang homicides in a policing division of the Los Angeles

Police Department (LAPD). Longitudinal negative binomial models were examined

to test the relationship between-neighborhood structural covariates with gang and

non-gang homicides over a 35-year period. This study highlights the potential to

estimate temporal effects not captured by cross-sectional analyses alone. The results

underscore a unique feature that distinguishes gang homicides from other forms of

non-gang violence, its tenacious clustering, and spatial dependence over time.

Keywords

street gangs, homicide, structural covariates, spatial analysis, violent crime trends

Introduction

The United States has experienced an overall decline in violence, with rates not seen

since the 1960s. More recent concerns about shifting homicide rates highlighted

1Louisiana State University, Baton Rouge, USA

2University of North Carolina at Charlotte, USA

3University of California, Irvine, USA

Corresponding Author:

Matthew Valasik, Louisiana State University, 17B Stubbs Hall, Baton Rouge, LA 70803-2804, USA.

Email: mvalasik@lsu.edu

726807HSXXXX10.1177/1088767917726807Homicide StudiesValasik et al.

research-article2017

288 Homicide Studies 21(4)

post-2000 crime patterns that were “largely specific to individual cities” (Wallman &

Blumstein, 2006, p. 343) or specific to certain neighborhoods or streets within a city

(Weisburd, Groff, & Yang, 2012). Yet gang-related homicide remains unabated in

many distressed communities and continues to be a serious concern for police depart-

ments and policymakers (Papachristos, 2014; Petersen, 2016). The overwhelming

majority of gang-related homicides in the United States, approximately 85%, occurred

in large cities with a population over 100,000 or in spatially adjacent suburban coun-

ties (National Gang Centre, 2017). Howell, Egley, Tita, and Griffiths (2011) demon-

strated this trend by analyzing gang-related homicide trends from 1996 to 2009 in 247

cities with a population over 100,000 residents. They found gang-related homicides

accounted for anywhere between 20% and 40% of the total number homicides docu-

mented in the majority of the cities sampled (70%; Howell et al., 2011). Even though

overall murder rates have declined, the tenaciousness of gang-related homicides in

urban areas increased the likelihood that gang members would be involved in a homi-

cide (Papachristos, 2013). That said, a gap remains in understanding how gang homi-

cide is influenced by long-term changes in the structural characteristics of an area.

Prior research examining gang homicide focused much attention on distinguishing

differences between the micro-level characteristics of gang and non-gang events, such

as the characteristics of the incident or the participants (e.g., Maxson, Gordon, &

Klein, 1985). Pyrooz (2012) recently addressed “a notable void” (p. 494) in the gang

homicide literature by conducting a macro-level analysis that examined the relation-

ship of structural covariates with gang homicide in large U.S. cities. While both

macro- and micro-level studies have expanded our knowledge of gang homicide, the

utilization of cross-sectional designs limited researchers’ ability to account for changes

in the prevalence of gang violence over time. This is especially salient in gauging the

impact of structural characteristics on different types of homicide (e.g., gang or non-

gang), given that inequality was often considered a durable feature of neighborhoods

(Papachristos, 2014; Sampson, 2013). To better understand the relationship between

an area’s structural environment on gang homicide, it is important to examine these

influences over time.

Decker, Melde, and Pyrooz (2013) contended that neighborhood-level gang

research is “‘stuck in time,’ meaning that it is overwhelmingly cross-sectional” and

argue that “descriptive and predictive longitudinal analyses” (p. 393) are worthy

investments in advancing gang research. The present study answers this call by exam-

ining the temporal changes in gang homicides over a 35-year period in a Los Angeles

Police Department (LAPD) Division, a region with a protracted history of gang vio-

lence. This analysis builds upon current research by investigating the influence of

changes in structural characteristics on gang and non-gang violence in several ways.

First, we control for spatial autocorrelation because gang violence tends to be clus-

tered in particular areas (Blasko, Roman, & Taylor, 2015; Brantingham, Tita, Short, &

Reid, 2012; Papachristos, Hureau, & Braga, 2013; Tita, Cohen, & Engberg, 2005;

Zeoli, Grady, Pizarro, & Melde, 2015). Second, we examine how longitudinal changes

in a neighborhood’s structural covariates were associated with either gang or non-gang

homicide. Third, the confined, delimited, and local nature of gang violence makes it