Transcript Cure Violence - WordPress.com

Evaluating the Effectiveness of the
Cure Violence Model in New York City
Jeffrey Butts, Director
Research & Evaluation Center
John Jay College of Criminal Justice
City University of New York
November 2014
“Cure Violence utilizes a public
health approach. It considers gun
violence to be analogous to a
communicable disease that
passes from person to person
when left untreated. According to
the logic of Cure Violence, gun
violence is most effectively
reduced by changing the
behavior of individuals at risk to
participate in gun violence and
“denormalizing” violence by
working to change the community
norms that support and
perpetuate gun violence.”
— From the Summary, “Denormalizing Violence”
“The key question is whether the Cure
Violence model’s simpler and
potentially cheaper approach to
violence reduction can operate
alongside the more established
focused deterrence model. The Cure
Violence model is appealing in
theory, but does it work in practice in
a policy environment that is likely to
be less welcoming and less supportive
of any program whose employees
are former offenders and gang
members? Can the Cure Violence
model co-exist with programs that are
closely tied to law enforcement?”
— page 14, “Denormalizing Violence”
“Without the inclusion of former
offenders and formerly
incarcerated persons as staff, the
Cure Violence strategy would likely
not be as effective. Violence
interrupters and outreach workers
interact with social networks of
high-risk offenders and they are
able to obtain information that
would be unreachable by
conventional entities, such as law
enforcement authorities and social
service agencies. This feature,
however, is a serious obstacle for
evaluation.”
— page 15, “Denormalizing Violence”
“At its heart, Cure Violence is a
community-based violence
intervention—but part of its theory of
change is focused on changing the
behavior of high-risk individuals. …
[M]any questions [are]
unanswered… Is violence affected
at the community level because a
large number of individuals were
directly influenced by the program
to stop shooting, or were community
residents in general affected by
hearing or seeing the program’s
message? Or perhaps the primary
causal pathway to the larger
community is from individual
participants who then influence their
immediate social networks.
Evaluation research needs to identify
and test these hypotheses.”
— page 17, “Denormalizing Violence”
THREE DATA STRATEGIES IN THE EVALUATION OF “CURE VIOLENCE”
Question:
Were the Cure
Violence programs
implemented well,
with fidelity?
Central research question:
Is the presence of a Cure
Violence program in a community
related to the strength of antiviolence social norms expressed
by residents of the community or
to the incidence of actual violence
in the community?
Question:
Do young men in
areas with Cure
Violence
programs adopt
stronger antiviolence norms
than do young
men from similar
areas without
Cure Violence
programs?
POLICE & HOSPITAL DATA
Question:
Do violent crime and violent injury trends in areas
with Cure Violence programs improve over time
more than they do in similar areas without Cure
Violence programs?
THREE DATA STRATEGIES IN THE EVALUATION OF “CURE VIOLENCE”
Central research question:
Is the presence of a Cure
Violence program in a community
related to the strength of antiviolence social norms expressed
by residents of the community or
to the incidence of actual violence
in the community?
POLICE & HOSPITAL DATA
THREE DATA STRATEGIES IN THE EVALUATION OF “CURE VIOLENCE”
Resident Surveys: Brooklyn and the Bronx
3 “Matching” Neighborhoods in Each Borough
Cure
Violence
Area (N=200)
Comparison
Area 1
(N=200)
Comparison
Area 2
(N=200)
Same Survey Method in All 3 Areas
2014
2015
In-depth surveys of
attitudes and norms
about violence…
among 18-30 yearold men,
repeated 3 times to
detect change.
2016
THREE DATA STRATEGIES IN THE EVALUATION OF “CURE VIOLENCE”
Survey Approach
About 20 minutes of questions.
All questions administered via tablet (iPad).
Questions Focus on:
•
Personal experiences with violence
•
Attitudes toward the use of violence
•
Confidence in public institutions
•
Perceptions of personal safety
•
Collective efficacy
•
Knowledge of violence reduction programs
•
Contact with violence reduction initiatives
THREE DATA STRATEGIES IN THE EVALUATION OF “CURE VIOLENCE”
Central research question:
Is the presence of a Cure
Violence program in a community
related to the strength of antiviolence social norms expressed
by residents of the community or
to the incidence of actual violence
in the community?
POLICE & HOSPITAL DATA
Some of the survey crew
All interviews conducted in the
open in selected neighborhoods.
THREE DATA STRATEGIES IN THE EVALUATION OF “CURE VIOLENCE”
Central research question:
Is the presence of a Cure
Violence program in a community
related to the strength of antiviolence social norms expressed
by residents of the community or
to the incidence of actual violence
in the community?
POLICE & HOSPITAL DATA
THREE DATA STRATEGIES IN THE EVALUATION OF “CURE VIOLENCE”
DATA SOURCE:
Site visits, staff interviews, and
review of documents and
program records in Cure
Violence programs operated by
agencies across New York City.
Data gathered through these
methods to be used to assess
the quality and intensity of Cure
Violence programs
implemented in two
demonstration areas in New
York City relative to other
neighborhoods and other cities.
Central research question:
Is the presence of a Cure
Violence program in a community
related to the strength of antiviolence social norms expressed
by residents of the community or
to the incidence of actual violence
in the community?
DATA SOURCE:
In-person, structured surveys
with 18-30 year-old male
residents in six high-violence
areas of New York City — two
areas with new Cure Violence
programs and four areas without
Cure Violence programs. Each
neighborhood sample includes
approximately 200 subjects who
are asked about their
experiences, attitudes, and
norms related to interpersonal
violence. Each neighborhood to
be surveyed three times over 24
months to detect any changes in
norms and attitudes.
Everything measured at
the level of census tract
POLICE & HOSPITAL DATA
DATA SOURCE:
Violence indicators for all New York City neighborhoods, including:
1) police data since 2003 on all violent crimes reported to NYPD,
geocoded to the census tract in which the crime occurred; and
2) hospital data since 1995 about all patients appearing at emergency
rooms with injuries due to violence, geocoded to the census tract
of the patient’s home address.
■ About 2,200 census tracts in all five boroughs of NYC
■ Each census tract typically has 2,000 – 10,000 residents
■ Half of all tracts either have 0 shootings or too few residents to calculate rates
■ 198 census tracts (9%) account for half of all shootings across all 5 NYC boroughs
■ We are analyzing violence trends in about 1,000 census tracts
Cure Violence
Catchment
Areas
Typically
Affecting 4-6
Census Tracts
HOSPITAL-REPORTED
VIOLENT INJURIES BY
PATIENT HOME ADDRESS:
1995-CURRENT
POLICE-REPORTED
VIOLENT CRIMES BY
LOCATION OF INCIDENT:
2003-CURRENT
For every census tract, we
have 3 sources of data
DEMOGRAPHICS,
EMPLOYMENT, POVERTY:
2000 & 2010
DATA ANALYSIS PLAN
So, every census tract can be categorized by:
The presence and/or duration of violence reduction
programming.
e.g.
0 = at least 2 tracts away from a program area
1 = adjacent to, or 1 tract away from a program area
2 = tract is actually in a program catchment area
e.g.
0 = never even adjacent to a program area
1 = in, or adjacent to a program area less 2 years ago
2 = in, or adjacent to program more than 2 years ago
DATA ANALYSIS PLAN
Tracts can also be ranked relative to all other tracts:
Our analysis will compare each census tract by how it
ranks citywide before and after program operation
e.g.
shooting rate 2003-2015
violent crime rate 2003-2015
rate of increase in crime 2003-2015
rate of E.R. admissions for violent injuries 1995-2015
rate of increase in E.R. admissions 1995-2015
Say this was census tract “20742” and it ranked as the 20th
most violent tract in 2014, but by 2016 it had dropped to
30th most violent, or 40th, or 100th? What would this mean?
DATA ANALYSIS PLAN
.
.
.
2014
2015
2016
2017
th
65033
46984
40744
94852
_____________________________________________________________
20
st
75020
94852
20742
32919
_____________________________________________________________
21
nd
98731
32919
46984
40744
_____________________________________________________________
22
rd
58001
13075
94852
20742
_____________________________________________________________
23
th
40744
45983
32919
46984
_____________________________________________________________
24
20742
65033
13075
13075
25th
_____________________________________________________________
If we establish
the
amount of change
in ranking
by
46984
75020
45983be expected
45983
26ththat would
_____________________________________________________________
chance alone, we can establish the
94852
98731
65033 of the actual
65033
27th
_____________________________________________________________
statistical
significance
change_____________________________________________________________
inthone census
32919
40744
75020 track relative
75020
28
to all other census tracts, and then
th
13075
20742
98731
29infer
perhaps_____________________________________________________________
a program
effect. 98731
th
45983
58001
58001
58001
_____________________________________________________________
30
Research Questions
Data/Analysis Plan
Summarizing…
Research Questions
Were the programs done right?
Data/Analysis Plan
Site visits, staff interviews, and
review of documents and
program records in Cure Violence
programs across New York City.
Data gathered through these
methods are being used to assess
the quality and intensity of Cure
Violence programs implemented
in Brooklyn and the Bronx,
compared with other sites in New
York City and other U.S. cities.
Research Questions
Were the programs done right?
Can we detect changes in the
anti-violence attitudes of 18-30
year-old men in high-risk
neighborhoods?
Data/Analysis Plan
Conducting in-person, structured
surveys with young male residents
of six high-violence areas of New
York City — two areas with and
four without Cure Violence
programs. Each sample includes
approximately 200 subjects who
are asked about experiences,
attitudes, and norms related to
interpersonal violence. Each
neighborhood is surveyed three
times over 24 months to detect
changes in norms and attitudes
about violence.
Research Questions
Were the programs done right?
Can we detect changes in the
anti-violence attitudes of 18-30
year-old men in high-risk
neighborhoods?
Can we measure improvement
in violent crime and violent
injuries at the census-tract level
across New York City.
Data/Analysis Plan
Compiling indicators of violence
and violent injuries for all New
York City neighborhoods using:
1) police data since 2003 on all
violent crimes reported to
NYPD, geocoded in the census
tract in which the crime
occurred; and
2) hospital data since 1995 about
all patients appearing at
emergency rooms with injuries
due to violence, geocoded in
the census tract of patient’s
home address.
The Research & Evaluation Center’s Study of
Cure Violence in New York City is funded by the
New York City Council and by the Robert Wood
Johnson Foundation of Princeton, NJ
Keep Up with Study Developments:
http://JohnJayREC.nyc