Street Gangs, Gun Violence, and Focused Deterrence: Comparing Place-based and Group-based Evaluation Methods to Estimate Direct and Spillover Deterrent Effects

• Author: Anthony A. Braga,Rod K. Brunson,Lisa Barao,Andrew V. Papachristos,Greg Zimmerman,Chelsea Farrell
• Published date: 01 July 2019
• Date: 01 July 2019
• DOI: http://doi.org/10.1177/0022427818821716
• Subject Matter: Articles

Article

Street Gangs, Gun

Violence, and

Focused Deterrence:

Comparing Place-based

and Group-based Evaluation

Methods to Estimate Direct

and Spillover Deterrent Effects

Anthony A. Braga

1

, Greg Zimmerman

1

, Lisa Barao

1

,

Chelsea Farrell

1

, Rod K. Brunson

2

,

and Andrew V. Papachristos

3

Abstract

Objectives: Focused deterrence seeks to ch ange the violent behavi or of

gangs not directly targeted by the intervention by leveraging network ties

between groups. This study appraises group-based and place-based meth-

odologies in detecting direct and spillover crime reduction impacts of a

focused deterrence strategy implemented in Oakland, California. Method:

Quasi-experiment al designs and panel regr ession models were us ed to

compare shooting trends for directly treated gangs and census block groups

1

School of Criminology and Criminal Justice, Northeastern University, Boston, MA, USA

2

School of Criminal Justice, Rutgers University, Newark, NJ, USA

3

Department of Sociology, Northwestern University, Chicago, IL, USA

Corresponding Author:

Anthony A. Braga, School of Criminology and Criminal Justice, Northeastern University,

360 Huntington Avenue, Boston, MA 02115, USA.

Email: a.braga@northeastern.edu

Journal of Research in Crime and

Delinquency

2019, Vol. 56(4) 524-562

ªThe Author(s) 2018

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0022427818821716

journals.sagepub.com/home/jrc

to shooting trends for comparison gangs and block groups that did not

directly receive the treatment. The design further analyzed whether vicar-

iously treated gangs and untreated block groups immediately surrounding

treated block groups experienced spillo ver shooting reductions. Results:

The focused deterrence program generated statistically significant direct

reductions in shootings in treated block group areas and by treated gangs

relative to shootings in untreated block group areas and by untreated gangs,

respectively. Spillover deterrent effects of varying magnitudes were found

when shooting trends by vicariously treated gangs and in surrounding block

groups were compared to shootings trends in comparison units. Conclu-

sions: Group-based evaluation methodologie s were better positioned to

detect changes in violent gang behaviors after treatment and to shed light

on deterrence mechanisms supporting program efficacy. Future focused

deterrence studies should be theory driven rather than black box

evaluations.

Keywords

deterrence, gun violence, gangs, spillover effects, violence prevention

Focused deterrence strategies seek to change offender behavior by under-

standing underlying crime-producing dynamics and conditions that sustain

recurring crime problems and by implementing a blended strategy of law

enforcement, community mobilization, and social service actions (Kennedy

1997, 2008). Direct communications of increased enforcement risks and the

availability of social service assistance to target groups and individuals are

defining characteristics of focused deterrence. The available scientific evi-

dence suggests that focused deterrence programs are effective at controlling

gang violence. A recently updated Campbell Collaboration systematic

review found that focused deterrence programs were associated with sig-

nificant crime reduction impacts, with gang and criminally active group

violence reduction strategies (GVRS) generating the largest crime reduction

effects (Braga, Weisburd, and Turchan 2018). Further, the National Aca-

demies’ Committee on Proactive Policing concluded that “evaluations of

focused deterrence programs show consistent crime-control impacts in

reducing gang violence” (Weisburd and Majmundar 2018:175).

While the available evaluation evidence is generally supportive of deter-

rence principles, it is difficult to know whether observed gang violence

reductions represent a true deterrent impact, a selective incapacitation effect

Braga et al. 525

generated by the apprehension and prosecution of targeted chronic offen-

ders, or the influence of some other crime prevention mechanism at work in

focused deterrence programs (Braga 2012; Brunson 2015; Corsaro and

Engel 2015). In essence, many GVRS program evaluations suffer from

construct validity concerns (Meehl and Cronbach 1955) in associating gang

violence reductions with the focused deterrence regime established by

implemented strategies. There seem to be two methodological shortcom-

ings in most GVRS evaluations seeking to establish deterrent impacts

(Braga and Weisburd 2014). First, changes in the violent behavior of treated

gangs are often inferred from violence reductions observed at larg er areal

units such as cities and neighborhoods rather than directly observed at the

group unit of analysis, an ecological inference fallacy (Robinson 1950).

As such, it is difficult to know whether treated gangs actually modified

their violent behaviors as a result of the focused deterrence strategy.

Second, the violent behavior of untreated gangs who are socially con-

nected to treated gangs is scarcely examined for evidence of “spillover”

deterrent impacts. Establishing the existence of indirect or spillover

effects of the GVRS intervention on gun violence represents a more com-

plete test of deterrence theory than finding direct impacts, which include

selective incapacitation effects.

The updated Campbell review found that only 2 of the 12 identified

GVRS program evaluations conducted supplemental analyses to determine

whether the intervention generated spillover violence reductions for

untreated gangs that were connected to treated gangs through ongoing

rivalries and alliances (Braga et al. 2018). These two program evaluations

used different place-based and group-based approaches when measuring

spillover deterrent impacts. The Los Angeles Ceasefire GVRS analyzed

serious violent crime trends in proximate census block groups with

untreated gangs socially tied to tr eated gangs in targeted block gr oups,

relative to serious violent crime trends in matched untreated block groups

(Tita et al. 2004). In contrast, an evaluation of a revitalized Boston Cease-

fire GVRS analyzed shootings by and against matched untreated gangs that

were socially connected to treated gangs, relative to matched untreated

gangs not socially connected to tre ated gangs (Braga, Apel, and Welsh

2013). Both evaluations reported significant spillover violence reductions

generated by the respective GVRS programs.

This study seeks to advance the methodological approaches used to

evaluate GVRS programs and to improve our understanding of the deter-

rence mechanisms supporting the adoption of these violence reduction stra-

tegies. Drawing on place- and group-based quasi-experimental evaluation

526 Journal of Research in Crime and Delinquency 56(4)