Counting Crimes — The Importance of Understanding Crime Concentration for the Design and Evaluation of Crime Reduction Strategies
| DOI | 10.1350/ijps.2008.10.4.097 |
| Date | 01 December 2008 |
| Published date | 01 December 2008 |
| Author | Michelle Rogerson |
| Subject Matter | Article |
Counting crimes — the importance of
understanding crime concentration for the
design and evaluation of crime reduction
strategies
Michelle Rogerson
Applied Criminology Centre, University of Huddersfield, Queensgate, Huddersfield, HD1
3DH, UK. Tel: +44 1484 473223; Fax: +44 1484 473760; email: m.rogerson@hud.ac.uk
Received 8 August 2007; accepted 31 January 2008
Keywords: crime reduction, crime statistics, crime data, victimisation data,
crime level analysis
Michelle Rogerson joined the Applied Crimino-
per victim (concentration) or a combination of
logy Group at the University of Huddersfield in
both. These underlying dimensions of a crime
1998 from the Data Archive, University of Essex.
problem must be understood in order to select the
She completed a BA in Sociology at Durham
most suitable crime prevention interventions, and
University and an MSc in Social Research
to target them appropriately.
Methods at the University of Surrey. Michelle is
currently registered for her PhD on the pro-
spective identification of emerging crime trends.
INTRODUCTION
Research interests include: technology, crime
The concentration of criminal victimisation
and crime prevention; crime, fear of crime and
is a phenomenon of which most police
quality of life; and regeneration and crime repeat
officers, crime-prevention practitioners and
victimisation
academic criminologists are all too aware.
However, this measure is rarely used to its
full potential. It is used to indicate a prob-
ABSTRACT
lem’s magnitude, that is, to stress that crime
Crime statistics are most frequently concerned
levels are not only high, but they are higher
with the incidence of crime (usually quoted as a
for particular victims or places. What is
rate per population), occasionally statistics are
neglected is a central tenet of this paper —
concerned with prevalence (number or proportion
that areas with low and high concentrations
of victims within the population) but the concen-
of crime experience a crime problem of a
tration of crime (number of crimes per victim) is
different nature which requires a very dif-
rarely quoted. This paper aims to demonstrate the
ferent response. In other words, crime
importance of all three indicators of crime, preval-
problems come in different shapes as well as
ence, concentration and incidence, for understand-
different sizes.
ing crime levels through an analysis of
An area may acquire a high level of crime
self-reported victimisation data from 39 high-
in one or both of two ways. First, a high
International Journal of Police
crime areas. The analysis illustrates that areas can
proportion of households, individuals or
Science and Management,
Vol. 10 No. 4, 2008, pp. 434–447.
have high crime either as a result of high levels of
properties may be victimised. Second,
DOI: 10.1350/ijps.2008.10.4.097
victimisation (prevalence), high numbers of crimes
properties or victims may be repeatedly
Page 434
Rogerson
targeted. The proportion of potential vic-
suggested that it was having a detrimental
tims who are victimised is referred to as
effect. In contrast, a more thorough assess-
crime prevalence. The number of crimes
ment, including both prevalence and
per victim (or property) is referred to as
concentration revealed not only a more
crime concentration. Together crime pre-
positive, but more importantly, a more
valence and crime concentration produce
accurate picture of the programme’s
the overall incidence (or level) of crime
effectiveness.
within a given area (often quoted as a rate
In the author’s experience the termino-
of crime per 1,000 population). To use the
logy of prevalence, concentration and
example of burglary, in an area of 100
incidence is not unknown amongst practi-
homes, an identical burglary rate will arise
tioners. However, the strategic direction
from one house being burgled ten times
that can only be gained by understanding
(high concentration, low prevalence) or 10
the relationship between prevalence and
houses being burgled once (high preval-
concentration is far from realised. In a
ence, low concentration).
review of the strategies of the Crime and
The crime reduction strategies appro-
Disorder Reduction Partnerships, 35 per
priate to high prevalence and high concen-
cent of 376 partnerships included targets to
tration areas are different. In the first case,
reduce burglary repeat victimisation, 43 per
precautions should primarily be distributed
cent included targets to reduce domestic
(by housing providers, police and others)
violence repeat victimisation, and 13 per
amongst the not-yet-victimised. In the
cent included targets to reduce repeat vic-
latter, substantial effort should be allocated
timisation of other crime types (Deakin &
to the already victimised as the strongest
Chenery, 2002). Measuring repeat victim-
predictor of future victimisation is prior
isation requires a count of the number of
victimisation (Pease, 1998). In order to
crimes per victims (concentration), and yet
select strategies for targeting crime reduc-
the Crime and Disorder Audits that precede
tion interventions it is important to know
and support Crime and Disorder Strategies
whether an area has a high rate of burglary
do not contain this supporting analysis of
because of high prevalence, high concentra-
crime concentration.
tion or both.
This paper aims to demonstrate the
Assessments of crime prevention impact
importance of both crime prevalence and
should also consider concentration and
concentration to the understanding of
prevalence jointly. Farrell and Buckley’s
crime levels through an analysis of data
(1999) evaluation of a Merseyside domestic
from 39 high crime areas. By examining the
violence programme provides a case in
nature of crime distribution in parallel with
point. Following the implementation of the
the overall number of crimes in these areas,
initiative, reports of domestic violence
the paper aims to demonstrate how areas
increased. At face value this would suggest
with similarly high crime levels can have
that the programme may have exacerbated
remarkably dissimilar crime problems. The
the problem. However, a more detailed
paper will also demonstrate how crime
analysis, informed by both prevalence and
change (both increases and reductions) can
concentration, identified that the number
be brought about either by changes in pre-
of repeat victims reporting domestic viol-
valence, changes in concentration or a
ence incidents had reduced. In other words,
combination of both. The paper is divided
restricting the focus to prevalence, an all too
into three parts. The first section examines
common approach, would have pointed to
the distribution of prevalence and concen-
an unsuccessful programme and, worse,
tration in 2002. The second section looks at
Page 435
Counting crimes
how changes in the distribution of crime
assault, threats, racial assault, theft of motor
affected the overall incidence of crime in
vehicles, theft from motor vehicles and van-
2004. The final section considers how ana-
dalism of motor vehicles.1 Residents may
lysis of prevalence and concentration can
have experienced other types of crime:
help to select the most appropriate crime
however, the analysis in this report is
reduction strategy.
restricted to those crime types included in
the survey. Crimes such as theft from the
person, assault and vehicle crimes can be
DATA AND METHODOLOGY
committed against individuals visiting or
This report analyses responses from a
working in an area. Area-based household
household survey conducted in New Deal
surveys exclude these groups and only pro-
for Communities (NDC) areas by MORI
vide estimates of crimes committed against
in 2002 and then repeated in 2004 (MORI.
residents. Crimes against business are also
Social Research Institute, 2006). The
excluded.
Neighbourhood Renewal Unit’s NDC pro-
Research into crimes as diverse as
gramme aims to tackle problems including
domestic burglary (Pease, 1998), domestic
housing, education, employment, health
violence (Hamner, Griffiths, & Jerwood,
and crime. Regeneration is to take place
1998), crime on industrial estates (Johnston,
over 10 years. While this report has used
Leitner, Shapland, &Wiles, 1994) and racial
self-reported victimisation, the approach
attacks (Sampson & Phillips, 1994) have
taken is as important when considering
found that large proportions of crime are
police recorded crime data (as demonstrated
the consequence of high levels of concen-
by Sunder & Birks, 2004).
tration. Pease identified crime concentra-
Survey measurements of victimisation
tion as the primary reason why high-crime
have some advantages over police recorded
areas suffer much crime. The central focus
data. Only a small proportion of crime
of the current paper is to demonstrate that
comes to police attention and methods of
...
To continue reading
Request your trial