Developing an Offender-Based Tracking System: The Western Australia Inois Project
| Author | Anna Ferrante |
| Published date | 01 December 1993 |
| Date | 01 December 1993 |
| DOI | http://doi.org/10.1177/000486589302600305 |
232
AUST
&
NZ
JOURNAL
OF
CRIMINOLOGY
(December
1993) 26 (232-250)
DEVELOPING AN OFFENDER-BASED TRACKING SYSTEM:
THE
WESTERN AUSTRALIA INOIS
PROJECT
Anna
Ferrante"
In 1989 the Crime Research Centre (University
of
WesternAustralia) and criminal justice agencies
in WestemAustralia became involved in the
INOIS
(Integrated Numerical Offender Identification
System) project. The principalaim
of
this project was to introduce acommon, unique identifier for
offenders so that a longitudinal database could be established that could "track" offenders through
the criminal justice system. This paper explains reasons for establishing this kind
of
database
and
exploressome
of
the conceptual
and
technical issues which arise. Central to the development
of
the
database and, therefore, to the
INOIS
project, was the underlying need to uniquely identify each
offender. This paper focuses on
how
record-linking techniques were used to achieve this. The
record-linking component
of
the
INOIS
project is discussed at length, and results from trials and
from preliminary runs
of
the linkage system are presented
and
discussed. Additional research and
operational usages
of
such a system are also identified.
The Need for a Database
In 1989 the Crime Research Centre moved to establish alongitudinal database of
offenders that would enable "tracking" of offenders through the criminal justice
system in Western Australia and over time. Construction of the database required
the collaboration of the Western Australia Police Department,
the
Department
for
Community Services,
the
Crown Law
Department
and the Department of
Corrective Services, each of which maintained its own independent
and
autonomous
information system(s). Traditionally, as with most jurisdictions, these information
systems had been developed in ways which
met
the operational and administrative
needs of each agency
rather
than from the viewpoint of establishing an integrated
database relevant to
the
overall operation of the criminal justice system.
Consequently, these systems were often incompatible with each other, and
numerous information gaps existed. A database that amalgamated data from all of
these agencies would be
the
first ever constructed in the State and would explicitly
acknowledge the inter-relatedness of the criminal justice system.
Adatabase was designed to prospectively collect cross-sectional
data
(ie offender
records collected over a short period such as three or six months) from a number of
criminal justice agencies
and
place them together into a single
data
system. With
appropriate links between the various records, the resultant database could be used
to "track" offenders both through the criminal justice system cross-sectionally (ie
from arrest to charge
and
court appearance and finally to conviction, possible
detention and subsequent release) and over time (ie longitudinally).
Traditional sources of
data
on offenders and the operation of criminal justice have
typically relied on aggregate crime statistics, which are often in summary form, or
specific-purpose data collections governed by
the
time and resources available to the
researcher. Such data impose limitations on the nature and form of research
undertaken by those investigating the state of crime and the effects of interventions
and policies. Most importantly these traditional sources do not usually (or only at
great cost) enable the adequate description of the fundamental dimensions of crime,
that is, the prevalence, frequency, duration and severity of offending. Describing and
monitoring all these dimensions of offending will constitute a significant advance on
*Senior Analyst/Programmer,
Crime
Research
Centre,
University of Western Australia.
DEVELOPING AN OFFENDER-BASED TRACKING SYSTEM 233
existing sources
of
knowledge
about
crime
on a scale similar to
the
impact of
the
science
of
epidemiology
upon
the
problems
of
public
health
(Blumstein,
Cohen,
Roth
&Visher, 1986).
These
dimensions
of
crime
cannot
be easily
or
accurately
estimated
unless
the
individual
unit
records
of
offenders
containing
all
the
relevant
and
salient events
(arrest, convictions, imprisonment,
etc)
are
collected
and
integrated
across agencies
and
through time. Since
the
advent
of
electronic
data
systems
and
their
rapid
adoption by many criminal justice agencies,
the
capacity to develop comprehensive
longitudinal
offender
records has
become
affordable
and
feasible.
Among
individual
criminal justice agencies such longitudinal case records
are
commonplace
but
few
are
linked to
other
agencies. Taking advantage of
these
technological advances,
somejurisdictions (notablyin
the
United
States)
have
begun
to develop cross-agency
information systems.
One
such example is discussed below.
Considerable technical (eg
data
collection
and
exchange protocols),
bureaucratic
(eg
the
preference
for secrecy
and
information control)
and
political
(eg
privacy
laws) barriers exist in
the
development
of
such
"birth
to
death"
offender
records.
Resource efficient
methods
of
the
kind described
here
address
many
of
these
difficulties
and
bring
the
prospect of a
combined
longitudinal
database
from
the
realm of
the
fanciful to
the
feasible.
Offender-Based
Transaction
Statistics
The
US
Department
of
Justice,
Bureau
of
Justice Statistics (BJS)
and
the
California
Department
of
Justice,
Bureau
of
Criminal Statistics (BCS) have
each
developed
offender-based tracking databases, known as Offender-Based
Transaction
Statistics
(OBTS).
These
operate
on
national
and
state
levels respectively
(US
Department
of
Justice, 1986a;
Bureau
of
Criminal Statistics, 1991)
and
rely on local criminal justice
authorities to
send
offender-based event
data,
which
are
collected
either
manually
or automatically, to
central
State
repositories.
The
national
OBTS
solicits
State
repositories
each
year
to extract
and
submit
data
from
their
master
records,
following
OBTS
guidelines.
Data
submitted
by
the
States consist
of
details
about
the
offender (age, race, sex, ethnic origin),
the
arrest
(arresting agency,
date
arrested,
arrest offence,
and
date
and
type
of
police disposition), any
prosecutor
action,
pre-trial actions,
court
activities (dates, disposition offence,
court
type, trial type,
counsel type, final
plea),
the
outcome
and
the
sentence,
if convicted.
The
Californian system collects similar details. In short,
each
OBTS
record
consists
of
selected facts
about
an offender, starting with an
arrest
and
following
through
with
the
actions
taken
by police, prosecutors
and
the
courts.
A key
feature
of
OBTS
data
is
that
records
are
structured
so
that
one
can
distinguish
among
arrestees, incidents
and
charges.
The
arrestee
is identified
through fingerprints
and
assigned a
unique
number
which can be linked to
earlier
records. An incident
number
identifies
each
incident so
that
records in
multiple-charge cases
can
be condensed.
For
incidents involving multiple charges,
multiple dispositions,
and/or
multiple sentences,
the
information
written
to
the
condensed
OBTS
record
is selective.
That
is, a
condensed
record
consists of
the
most serious
arrest
charge as well as
the
most serious disposition by
the
police,
the
prosecutor
and
the
court. Sentence information is selected on
the
basis of
the
sentence imposed for
the
most serious conviction offence.
These OBTS systems have become useful tools for
understanding
and
assessing
the performance
of
the
adult criminal justice system in
the
United
States
(eg US
Department
of Justice, 1983, 1990 &1991;
Bureau
of Criminal Statistics, 1990).
The
statistics provide
comparable
information
about
the
outcomes
of
judicial processes
To continue reading
Request your trial