25
Intelligence - Led Policing:
How the Use Of Crime Intelligence Analysis Translates in to the Decision-Making
model addresses a simple but broad conceptual framework for intelligence-led policing that
is likely to be applicable to most agencies” (p. 114).
There are analysts who interpret the criminal environment and influence the decision-
makers, who in turn make decisions and set policies that impact the criminal environment.
It is, in one sense, and integrated theory of problem-oriented policing (Goldstein, 1990)
and compstat (Bratton, 1998). Different than problem-oriented policing, intelligence-led
policing focuses on serious and prolific offenders, not incidents or problems only. Unlike
community-oriented policing, hierarchy (and centralization) is supported, while solving
crime problems with the community is not. However, Ratcliffe (2003) acknowledges that
a policing philosophy with a broader approach may help create the appropriate context for
intelligence-led policing to be successful.
Figure 1. The 3-i model showing the interaction of interpretation, influence, and impact on
policing.
Note. Copied and adapted from Intelligence-Led Policing (p. 110) by J. H. Ratcliffe, 2008, Cullompton, Devon,
UK: Willan Publishing.
Ratcliffe (2008) contends that the 3-i model provides a big picture and aims to apprehend
serious and prolific offenders in order to decrease crime. Crime analysts have a critical role
to influence the decision-makers in their decisions for achieving efficient outcomes in the
criminal environment. This model, as argued, should be applied as a whole. All of the three
processes (i.e., interpret, influence, and impact) should be included. The absence of any of
these processes may result in unsuccessful, inefficient, and undesired outcomes.
The 3-i Model
Criminal Environment
Crime Intelligence
Analysis
Decision-Maker
Interpret
Impact
Influence
26
International Journal of Security and Terrorism • Volume: 4 (1)
2. Method
2.1. Data
2.1.1. Unit of Analysis
The unit of analysis in this study is organizational level, with 544 cases (police departments)
in a comprehensive, merged data-set. All of the law enforcement agencies in this study are
large municipal departments with 100 or more sworn officers nationwide in the U.S.A. The
smallest department has 100 sworn officers, while the largest department has 13,271 sworn
officers.
2.1.2. Data Sets
1
The data used in this study were constructed by combining several comprehensive,
secondary data-sets in order to create a richer and replicable data set for exploring the
research questions. The following data sets were used:
• O’Shea and Nicholls’ Crime Analysis Survey (CAS
2
) Data Set (2000)
• U.S. Bureau of Justice Statistics’ Law Enforcement Management and
Administration Statistics (LEMAS) (2000)
• U.S. Bureau of Justice Statistics’ Law Enforcement Agency Identifiers Crosswalk
(2000)
• Federal Bureau of Investigation Uniform Crime Reports (UCRs) (1999)
The data sets were merged by using common key identifiers, such as the Originating
Reporting Agency Identifiers (ORIs) that police agencies use to report crime statistics to the
FBI’s UCR program. Each reporting agency has its own unique identification number, which
typically is referred to as an ORI number.
By using these ORI codes, one can either merge some new variables to the main data
set, or combine cases with it. In this study, the former function was performed.
2.2. Variables
2.2.1. Dependent Variables: Decision-Makers/Decision-Making
With the three decision-making variables (i.e., command-level manager, detective, and
patrol officer), one can get an idea of which decision-makers are perceived to use a given
type of crime analysis in their organizational decision-making. In Ratcliffe’s (2008) 3-i
model, decision-makers are not clearly specified and can be any person or any institution.
1
Except the first data set, all of the data sets were derived or downloaded from the Interuniversity Consortium
for Political and Social Research. The first data set was requested directly from O’Shea, the principal
investigator, because it was not available on the Interuniversity Consortium Web site or any other Web site.
The data were received via e-mail in SPSS format. SPSS is a statistical-analysis software program. The
data-merging process was replicated three times in order to decrease potential mistakes and therefore
increase the consistency of the merged data.
2
The name and abbreviation for the Crime Analysis Survey was determined and used by the authors, based
on the permission and approval granted by the survey’s principal author, T. C. O’Shea, via e-mail and phone
on March 3, 2009.
27
Intelligence - Led Policing:
How the Use Of Crime Intelligence Analysis Translates in to the Decision-Making
In this study, however, three decision-making levels are used as indicators of the dependent
variables: command-level manager, patrol officer, and detective. It is critical to note that
in most cases, a single individual reported on the level of crime analysis usage by each
of the three user levels. For example, the manager of a crime analysis unit (the most
likely organizational respondent) reported on how well crime analysis products were used
by command-level managers, patrol officers, and detectives. While it would have been
preferable to have each of these users report on their own perspectives, the crime analysis
unit manager is thought to be the most knowledgeable single reporter.
One of the reasons for choosing these three positions is that the crime analysis
unit provides the results of using crime data to support different levels of law enforcement
personnel, particularly these three distinct levels. By “providing the police officer, detective
or administrator with crucial information helps them make better decisions” (Baker, 2009:
6). Put differently, the collected, collated, and analyzed data are disseminated “primarily
to patrol officers, investigators, and command staff” (Osborne & Wernicke, 2003: 36).
Therefore, it is assumed that checking the association at the highest level (command-
level manager), the lowest level (patrol officer) and a different functioning level (detective)
would provide an opportunity to see difference in perspective when exploring the research
questions. In addition, these three levels (i.e., command-level manager, patrol officer and
detective) are vital information sources for the law enforcement agency’s crime analysis unit
(Buck, 1973). At the same time, these three groups of officials interact heavily with the crime
analysis unit. The extent of that interaction depends on their position within the organization.
In the CAS data set, the researchers asked to what degree the results of crime analysis efforts
were thought to be utilized by command-level managers
3
, detectives, and patrol officers.
The measurement level of those variables is ordinal as the answering scale is categorized
as “not utilized”, “utilized some”, and “highly utilized”. Contrary to the O’Shea and Nicholls’
study (2002, 2003), where they used this group of variables as independent variables, these
variables will be used as dependent variables in the current study to represent decision-
making. In their discussion of study findings, O’Shea and Nicholls stated that “We caution
the reader that the opposite may also be true; that is, the higher the levels of crime analysis,
the greater the appreciation. Our analysis cannot say which way these are related” (2002:
35, footnote).
2.2.2. Independent Variables: Main (Explanatory)
There are two groups of independent variables in this study: main (explanatory) and
control (organizational/internal). The main independent variables are the analysis types (i.e.,
statistical analysis, crime analysis, intelligence analysis, survey analysis, patrol strategy
analysis, and displacement/diffusion analysis), which are used as latent variables. Control
variables are organizational (internal) and environmental (external). The two types of control
variables are discussed in the following two sections. In contrast to O’Shea and Nicholls
(2002, 2003), the crime analysis types in the current study were used as independent
variables rather than dependent variables. Another difference is that the authors created
3
Command-level managers are chiefs and deputy chiefs, the ones who are able to make policies (telephone
consultation with T. C. O’Shea on February 27, 2009).