Introduction
The way a
society chooses to approach issues of crime and justice is a mirror of
political, social, and cultural norms and institutions. As society slowly
changes and evolves, public demands and expectations of government change with
it. American policing has always walked a tightrope between maintaining order
through force and providing civil service. Since the 1980s the American public
has increasingly expected police to “provide safety and security in a way that
does not violate their privacy and freedoms.”[i]
The concept of Community Oriented Policing (COP) was developed to more
proactively combat crime by working with, rather than against, communities to
solve problems.[ii]
This study aims to assess the effect of the implementation of community
oriented policing programs on crime. I find that, in the assessed communities,
these programs have significant but complex impact.
Study Design
The most
effective study would focus on a single department that implemented the program
in selective districts. By selecting similar districts that did not implement
the program, a “difference in differences” method could be used to approximate
experimental conditions by identifying deviations from a common trend in the
policy group. Additionally, aggregate crime data only tells part of the story
of a community policing initiative; to that end, surveys of community attitudes
and perceptions would be useful to track effectiveness. A study like this was
conducted by the Northwestern University’s Institute for Policy Research on
community policing initiative in Chicago.[iii]
This study
attempts to approximate a similar quasi-experimental design using publicly
available crime data from the FBI’s Uniform Crime Reporting database to examine
the impact of community policing programs on a larger scale. Several county
level Sheriff’s Offices in California implemented community policing programs
simultaneously in 2007[1].
By comparing rates of reported and cleared[2]
crimes between counties with newly implemented programs and those that without,
insight can be gained into changes in the dynamics of community-police
relations on a larger scale. Data were collated from multiple sources were
compared for all counties in California to determine dates of community
policing initiatives.[3]
Those that had implemented policies prior to 2007 were eliminated from the
sample. The counties in the control group (those that had not implemented a
policy by 2010) represented a diverse range of urban and rural sections of the
state. The counties that implemented the policies (the treatment group) were
similarly diverse. Since there were significantly more counties in the control
group than the treatment group all statistics were standardized by aggregate
populations of the two groups.[4]
County Sheriff’s Offices were
chosen as the unit of comparison to draw attention to the wider populations
served by these units outside of metropolitan areas – this helps to eliminate
some of the concerns of biased results arising from specific differences
between municipalities. Table 1 provides summary statistics for the variables
used in the assessment.
Table 1. Summary Statistics
|
|||||
Mean
|
Std. Dev.
|
Min.
|
Max.
|
Description
|
|
Violent Crime Reports
|
115.35
|
85.7
|
5.41
|
536.86
|
Total reports of violent crimes by
county and year per 100,000 residents
|
Property Crime Reports
|
6148.16
|
405.62
|
42.36
|
2529.4
|
Total reports of property crimes by
county and year per 100,000 residents
|
Cleared Violent Crimes
|
57.92
|
43.07
|
0
|
190.72
|
Total violent crimes cleared by
officers by county and year per 100,000 residents
|
Cleared Property Crimes
|
77.84
|
72.86
|
0
|
515.51
|
Total property crimes cleared by
officers by county and year per 100,000 residents
|
Violent Difference
|
57.43
|
64.95
|
0
|
526.53
|
Difference in violent crimes reported
and cleared (per 100,000 residents)
|
Population
|
558423
|
623102
|
9686
|
3010232
|
Total population by County
|
Post
|
0.57
|
0.50
|
0
|
1
|
=0 if year < 2007, =1 if year >
2007
|
Treat
|
0.30
|
0.46
|
0
|
1
|
=0 if in control group, =1 if in
treatment group
|
Counties: 23
|
Treatment:
|
7
|
Control:
|
16
|
|
Observations: 161
|
|||||
Source: CA AG, FBI
UCR, BJS LEMAS
|
|||||
Results
In order
for a difference-in-differences study to be effective, the two groups (the
“control” and “treatment”) must exhibit common trends prior to policy
implementation. An analysis of trends over time for reports of property crimes
indicated that the control and treatment group both followed a common trend from
2004-2010. No statistically significant divergence from the trend occurred in
the treatment group after the COP policy was implemented in 2007. As a result,
no further analysis was conducted on property crimes. Figure 1 shows the time trends
of reported violent crimes of the treatment and control groups.[5]
A clear departure from the common trend is seen in 2008 when the treatment
group shifts towards a greater number of crimes reported for the next three
years.
Table 2 shows the results of the statistical analysis of
violent crime reports. It shows that the treatment group saw a statistically
significant increase of 12.82 violent crimes reported per 100,000 residents the
year after the COP policy was implemented.
Table 2. Violent Crime Reports
DV: Violent Crime Reports
|
Pre-2007
|
Post-2007
|
Difference
|
COP Policy
|
111.14
|
115.84
|
4.7
|
No Policy
|
90.22
|
82.1
|
-8.12
|
Difference
|
20.92
|
33.74
|
12.82**
|
Table 3 shows the results of a similar analysis conducted on
the effect of implementing a COP policy on the difference between reported
violent crimes and cleared violent crimes. This shows that, while both reported
and cleared violent crimes increased in the treatment groups after policy
implementation, the gap between them decreased by 1.64 per 100,000 residents in
2008.
Table 3. Difference in Violent Crimes
Reported vs. Cleared
DV: Reports - Clearances
|
Pre-2007
|
Post-2007
|
Difference
|
COP Policy
|
55.19
|
53.15
|
2.04
|
No COP Policy
|
58.75
|
58.35*
|
-0.397*
|
Difference
|
3.56
|
5.2
|
-1.64*
|
Discussion
The
implications of the results of this study are not straightforward. At first
glance, the analysis seems to suggest that community policing initiatives lead
to more violent crime. The treatment group demonstrated a clear upward
departure from the common trend in both reported and cleared violent crimes.
Should we then simply scrap the whole notion of policing by engaging the
community?
Not at all. In fact, the results here are exactly what we would
expect to see given the goals of these programs. An uptick in reporting is
indicative of improved relationships between police and the community. It
is likely that a larger number of crimes go unreported when the community does
not trust their police force. Though this study does not take into account
events that could have occurred between 2007 and 2008 that could have affected
crime reports, the number and variety of agencies in the treatment group
suggest causality more than a simple cross-sectional analysis.
Of additional interest is the result
of the analysis of the difference between cleared and reported violent crimes.
After policy implementation, both reported and cleared violent crimes increased
but there was a decrease in the difference between them. This suggests that not
only are citizens reporting more of the crimes that are occurring but that
those reports are also more substantial. There appears to be less of a divide
between what citizens see as a crime and what the police see as a crime. This
also suggests that there is an improving relationship between police and the
community.
Conclusion
Of course, this study is severely
limited in its ability to draw far-reaching conclusions about the effectiveness
of these policies. It does not examine how
or to what degree these programs
were implemented in each agency. Were they simply throwaways on paper that led
to business as usual or a concerted effort to change operations? Without
in-depth individual studies we don’t have that information available. It is also unclear why a change was seen in
violent crime activity but not property crime activity. One possible
explanation assumes that residents were more likely to fully report property
crimes than violent crimes (with lower associated investigation and arrest
rates) already before the COP policy was implemented. Evolving policing norms
will have a complicated impact on the community; this study showed one aspect
of a significant effect of a popular policy choice.
[1] See Appendix A for a list
of included counties.
[2] Police departments around
the country separate crimes into two categories: reported and cleared. Crimes
reported represent the total number of calls for service for a particular type
of crime. Crimes cleared are those that are substantiated by a responding
officer who notes a disposition for the contact. Dispositions can be official
reports, an arrest, the opening of an investigation, etc.
[3] Sources include: the FBI
Uniform Crime Reporting data, the State of California’s Attorney General’s
Office, the Law Enforcement Management and Administration Survey.
References
[i] Bush, Michael and Dodson,
Kimberly. Police Officers as Peace Officers: A Philosophical and Theoretical
Examination of Policing from a Peacemaking Approach. Journal of Theoretical and
Philosophical Criminology. Sept. 2014 pp.194-204
[ii] Cordner, G. W. (2010).
Community Policing Elements and Effects. In R. G. Dunham, & G. P. Alpert,
Critical Issues in Policing (pp. 432-449). Long Grove, IL: Waveland Press Inc.
[iii] Skogan, Wesley (1996).
Evaluating Problem Solving Policing: The Chicago Experience. Institute for
Policy Research, Northwestern University
Appendix A
El Dorado*
|
Placer*
|
Fresno*
|
Sacramento
|
Humboldt
|
San Bernardino
|
Imperial
|
San Luis Obispo
|
Kern*
|
San Mateo
|
Lassen
|
Santa Barbara*
|
Marin
|
Santa Cruz
|
Merced
|
Shasta
|
Modoc
|
Stanislaus
|
Monterey*
|
Tulare*
|
Orange
|
Ventura
|
Yolo
|
|
*indicates part of treatment group
|
|
No comments:
Post a Comment