Crime statistics and police numbers

Crime statistics and police numbers
March 2009
Ordered to be printed
Victorian government printer
Session 2006-09
P.P. No. 173
2
LETTER OF TRANSMITTAL
To
The Honourable the President of the Legislative Council
and
The Honourable the Speaker of the Legislative Assembly
Pursuant to section 25 of the Ombudsman Act 1973, I present to the Parliament
the report on my investigation into crime statistics and police numbers.
G E Brouwer
OMBUDSMAN
CONTENTS
ABBREVIATIONS.....................................................................................................6
EXECUTIVE SUMMARY.........................................................................................7
THE COMPLAINT ..................................................................................................15
INVESTIGATION...................................................................................................16
Scope.......................................................................................................................16
Methodology..........................................................................................................16
BACKGROUND ......................................................................................................17
How crime statistics are produced.....................................................................19
What crime statistics are available in Victoria?................................................22
What do the crime statistics show for Victoria?...............................................24
How crime statistics are used..............................................................................25
How many police are there in Victoria?............................................................26
Where are the police in Victoria?........................................................................29
CRIME STATISTICS ..............................................................................................30
Recording of crime................................................................................................30
(i) Comparing Victoria Police crime data with calls for service data........34
(ii) Comparing Victoria Police crime data with Victim of Crime
Assistance Tribunal data..................................................................................43
(iii) Comparing Victoria Police crime data with crime victim surveys.....45
Compliance with rules for recording crime......................................................50
Comparison with other Australian jurisdictions..............................................58
(i) Development of the National Crime Recording Standard.....................60
(ii) Victoria Police’s implementation of Rule 2.............................................61
Manual forms for recording crime .....................................................................64
Victoria Police Compstat and performance management..............................72
Deficiencies in police training.............................................................................76
Inadequacies of the LEAP database...................................................................77
Enhancing quality assurance practices..............................................................80
Organisational arrangements..............................................................................83
Reporting of crime statistics................................................................................86
4
POLICE NUMBERS.................................................................................................90
Understanding police numbers ..........................................................................90
Identifying police duties......................................................................................92
Using police resources to respond flexibly to demand...................................94
Changes in policing ............................................................................................102
The People Allocation Model (PAM) and its limitations..............................104
The Police Association’s Model for staff allocation ...........................................108
Other developments...........................................................................................109
Police rosters........................................................................................................111
SUMMARY OF RECOMMENDATIONS.........................................................118
ATTACHMENT 1 - List of 27 Victoria Police offence categories.................122
ATTACHMENT 2 - Maps of Victoria Police geographical regions.............123
ATTACHMENT 3 - CAD outcome categories and LEAP record matching
................................................................................................................................124
ATTACHMENT 4 - Victoria Police communication regarding the policy
change to crime recording .................................................................................125
ATTACHMENT 5 - Sample copies of LEAP forms .......................................126
ATTACHMENT 6 – Key Current Applications Architecture Overview
(Version 1.6) showing complexity of IT systems............................................127
ATTACHMENT 7 - Activity Return and Enforcement Analysis (AREA)
Form......................................................................................................................128
ATTACHMENT 8 - Sample copy of Patrol Duty Return..............................129
ATTACHMENT 9 - Safe Streets Taskforce mapping of CAD data for
weekend 15-17 February 2008...........................................................................130
ATTACHMENT 10 - CAD data relating to Victoria Police events for
dispatch and frequency of ‘Attempt to Dispatch’ code by priority level from
November 2007 to June 2008.............................................................................131
6
ABBREVIATIONS
ABS
Australian Bureau of Statistics
AIC
Australian Institute of Criminology
APMC
Australasian Police Ministers’ Council
AREA Form
Activity Return and Enforcement Analysis Form
BOCSAR
Bureau of Crime Statistics and Research
CAD
Computer Aided Dispatch
CBD
Central Business District
CDEB
Central Data Entry Bureau
CIU
Crime Investigation Unit
Compstat
COMParison of STATistics
CSU
Corporate Statistics Unit
DMSC
Data Management Steering Committee
DOJ
Department of Justice
DiRCS
Differences in Recorded Crime Statistics
ESTA
Emergency Services Telecommunications
Authority
FTE
Full Time Equivalent
HR
Human Resources
ICT
Information Technology Strategy
LEAP
Law Enforcement Assistance Program
LEDR
LEAP Electronic Data Recorder
LGA
Local Government Area
MAS
Members Activity Statement
NCRS
National Crime Recording Standard
NCSU
National Crime Statistics Unit
OPI
Office of Police Integrity
PAM
People Allocation Model
PSA
Police Service Area
SST
Safe Streets Taskforce
T & C
Tasking and Coordination
SBT
Scenario Based Testing
VOCAT
Victims of Crime Assistance Tribunal
VPIM
Victoria Police Intelligence Model
VPM
Victoria Police Manual
EXECUTIVE SUMMARY
In April 2008 I received a complaint from Mr Ted Baillieu MLA, Leader of the
Opposition, regarding whether crime statistics and/or police numbers have
been subject to manipulation. Upon receiving the complaint it was my view
that uncertainty regarding the validity of crime statistics and police numbers
needed to be resolved. Ongoing debate about the integrity and accuracy of
this data could undermine public confidence in Victoria Police and have a
deleterious effect on the quality of public debate regarding law and order
issues.
My investigation was concerned
with two
key
issues.
Firstly, do
Victoria
Police crime statistics accurately reflect the community’s experience of crime
as it is reported to police? Secondly, can the public have confidence that the
information
about
the
number
and
availability
of
police
in
Victoria
is
accurately reported?
My
investigators
interviewed
many
police
officers,
sought
advice
from
experts,
and
undertook
statistical
analysis.
They
also
reviewed
internal
reports, scrutinised police practices and compared
Victorian arrangements
with those in other jurisdictions.
While
my
investigation
did
not
find
corroborated
evidence
that
crime
statistics were subject to falsification (but see my concerns about clearance
rate
practices
below),
it
did
identify
poor
administrative
systems
and
historical Victoria Police practices which have led to some crime being under
reported, such as assaults and less serious offences. Victoria Police has stated
that there is no connection between crime being under reported and poor
administrative systems’. I disagree. Clearly there is a link.
Given the important role that crime statistics play for Victoria Police and for
public understanding of the crime problem, the issues I have identified during
my
investigation,
in
particular
in
relation
to
the
accurate
and
consistent
recording of crime by police, are of concern.
A
number
of
factors
are
identified
in
this
report
to
explain
how
under-
reporting of crime occurs.
8
Antiquated recording practices
Foremost
amongst
the
issues
identified
by
my
investigation
are
the
antiquated, time-consuming administrative practices for recording crime that
are followed by frontline’ police and the outdated information technologies
that
are
unsuited
to
a
twenty-first
century
approach
to
policing.
I
am
concerned that Victoria Police continues to miss opportunities to modernise
its practices in these areas.
For example, large numbers of handwritten forms
must be completed by police and then faxed to civilian support staff for data
entry
for
the
majority
of
reported
crime
to
be
captured
on
the
Law
Enforcement Assistance Program (LEAP) database.
Victoria
Police
agrees
that
there
are
good
opportunities
to
leverage
technology and improve recording practices. However, at this stage, the new
IT initiatives currently being pursued by Victoria Police will not eliminate the
fragmented, ad hoc administrative practices involved in recording crime data.
The
Victoria
Police
LEAP
database
has
attracted
criticism
in
the
past,
including in a report by the Office of Police Integrity in 2005. The list of
weaknesses of Victoria Police’s information technology systems includes:
·
incompatible systems
·
impeded information sharing
·
uncoordinated management information
·
variation in user interfaces
·
lack of common data standards
·
inability to link business intelligence with factual information about
incidents.
These
deficiencies
raise
serious
concerns
regarding
the
functionality
and
‘value for money’ aspects of the information technology environment within
Victoria Police. My investigation has also highlighted the need for significant
modernising
of
Victoria
Police’s
information
technology
to
achieve
efficiencies
and
improve
the
capture
and
analysis
of
data
for
measuring
demands on the organisation’s resources.
Victoria Police stated:
The administrative burden on police for data capture is significant.
The Data
Management Steering Committee (DMSC) with the LINK Project Team are
exploring whether any data captured is not relevant, with a view to reducing
the data capture burden on operational police.
Limited data capture
I am also concerned by the limited range of data that Victoria Police captures
to measure its objectives. In my view the data excludes many incidents of a
public order nature to which police are called to assist through the 000 calls
for
service
(CAD)
system.
Not
fully
using
this
information
undermines
Victoria
Polices
commitment
to
a
reform
agenda
that
emphasises
performance management, evidence-based and intelligence-led policing and
resource allocation which is flexible to respond to public demands.
Until 30 June 2008 when a policy change occurred in the way that crime is to
be recorded, police in Victoria needed to be satisfied that there was evidence
that a crime has occurred (the evidentiary approach), whereas most other
States and Territories have
been using a prima facie model, that is, where
police take the details of an alleged crime at face value and record its details. I
believe that the Victorian evidentiary approach has contributed to Victoria
Police crime statistics being lower than might be expected. The recent policy
change to move crime recording to a more prima facie approach was found to
be
poorly
implemented
and
inadequately
understood
by
police.
The
deficiencies in police training for accurate crime
recording have also been
highlighted in my investigation.
Statistics and community experience
The
gap
between
police
statistics
and
the
everyday
experience
of
the
community was borne out by an examination of calls to 000 seeking police
assistance.
I
found
that
in
many
instances
these
requests
for
assistance,
despite alleging sometimes serious offences, did not lead to reported crime
being
recorded
on
LEAP.
My
investigation
found
significant
differences
between the calls for service or Computer Aided Dispatch (CAD) data and
what was recorded on LEAP. During my investigation I was also advised that
a data matching exercise between CAD data and LEAP had been conducted
by an internal audit in Victoria Police some
years ago. The results found
significant differences
between
the
CAD data
and
what was recorded
on
LEAP.
It
identified
under
recording
of
crime
by
police
-
particularly
in
relation to assaults - and delays in recording crime on to the LEAP forms to
be of concern. I consider that closer examination is required to determine
why many crime events identified in the CAD system have no record on
LEAP, even though an offender may have been apprehended and an offence
reported.
10
Victoria Police has said:
While delays may occur in recording of crime by police, this does not have
any correlation with the subject of the inquiry whether the crime statistics
have been manipulated. Delays in recording do not affect the recording of
crime statistics.
I disagree.
Clearly delays in recording data affect the reliability of decisions
made on out-of-date information.
Victoria Police further stated:
CAD (ESTA) and LEAP were created for different purposes. They were not
designed to be linked, nor has the information in them been structured to
facilitate linking.
The CAD records calls for assistance under a broad range
of categories determined by the call taker actioning on information from a
member of the public. LEAP records crime under categories determined by
legislation. These categories often do not
match: a
person calling
000
to
report a robbery, for example, will not know that technically what they are
reporting
is theft
from
motor
vehicle’,
differences
that
cause
difficulties
when matching dataWhere other states link this type of data they do so
through an event number, rather than matching the classification in the CAD
and
crime
recording
systems.
Victoria
Police
has
identified
that
linked
systems
would
be
beneficial,
however
the
resources
to
do
this,
and
to
expand the CAD to a state-wide system are not currently available.
Similarly,
I
examined
crime
details
recorded
on
LEAP
for
a
sample
of
applications from victims of crime for assistance pursuant to the Victims of
Crime Assistance Act 1996. This also revealed a number of instances where
victims
had
reported
crime,
but
where
the
LEAP
database
appeared
to
contain no record of these and therefore would not be counted in Victoria
Police crime statistics.
Recording clearance rates
I also identified that some police misuse the procedures for recording cleared
crime to make it appear that more crime has been successfully solved than is
actually the case. The way this is done is that when an offender has been
apprehended and processed for certain offences, unrelated offences for which
no offender has yet been apprehended or interviewed are added on to the file
for the apprehended offender (without their knowledge) in order to clear up’
more crime. Victoria Police responded that ‘it would be disappointing if it was
found that this is a widespread practice’. I am referring this matter to the
Office of Police Integrity for investigation.
Inadequate auditing and data quality
I
found
that
the
audit
procedures
in
Victoria
Police
for
crime
data
are
focussed on the management of reported crime once it is entered on LEAP,
with no attention to the process for recording crime to ensure that officers at
police
stations
have
made
the
appropriate
decisions
regarding
whether
a
crime
incident
requires
a
LEAP
report.
I
consider
that
the
current
audit
processes
are
inadequate
as
they
fail
to
measure
compliance
with
crime
recording policies, which impacts on the accuracy of crime statistics. There is
evidence that crime recording practices are not compliant with the policies.
I
found
inconsistent
practices
in
recording
crime
across
Victoria,
reflecting
inadequate training and a lack of commitment to ensuring quality data at the
point of recording crime. The current quality assurance processes are applied
after the data-recording phase, which is too late. Overall, there appears to be a
fragmented, disconnected approach to ensuring high quality crime statistics.
The issue is that at the critical point where crime is recorded by police, the
quality assurance processes are lacking. At this point the cumbersome manual
forms
and
the
LEAP
database,
the
deficiencies
of
which
have
been
demonstrated, present significant obstacles to the effective recording of crime.
I therefore consider that the main focus for improvement must be on the
initial crime recording stage. Non compliance, for whatever reason, with the
policies for recording crime should be identified and addressed.
Generally, there appears to be a lack of clear vision and effective management
in
addressing
coordination
issues
in
relation
to
data
and
information
technology in Victoria Police, resulting in fragmented data systems, and more
broadly, in knowledge management. I note in this regard that according to
Victoria Police in May 2008:
The Corporate Committee endorsed the Data Management Strategy. The work of
the Data Management Steering Committee falls within the three key priorities of
that strategy:
·
Apply consistent data quality standards across all Victoria Police systems
·
Reduce complexity and duplication in data collection
·
Maximise accessibility to data to meet user needs.
However, I consider that the effectiveness to date of this strategy has not been
demonstrated.
12
Inevitably,
performance
of
a
large
organisation
requires
robust
measures
which allow achievement of benchmarks and targets to be assessed. Crime
data is an essential part of this measurement. I believe it is critical that there is
a commitment to ensuring that the data collected and used is of a high quality
and that a commitment to quality forms part of the performance management
system.
Enhancing public confidence
Public
confidence
in
the
integrity
of
crime
statistics
requires
independent
scrutiny and attention to the quality and integrity of the whole process - from
crime
recording
to
the
production
of
the
statistics.
The
public
debate
in
Victoria
regarding
crime
statistics
is
indicative
that
what
Victoria
Police
produces may not be understood or indeed trusted to reveal the nature and
extent of crime as the public experiences it and reports it to police.
I consider that the way in which crime statistics are reported by Victoria
Police to the public can be significantly improved. A number of issues were
identified
which
can
lead
to
distortion
and
misrepresentation
in
the
way
Victoria Police produce and present information on crime statistics. Unlike
some other Australian jurisdictions (NSW, SA, and WA), in Victoria there is
no agency external to Victoria Police which independently compiles, analyses
and publishes crime statistics. I am of the view that public confidence in crime
statistics is important enough to recommend to the Victorian Government
that
it
give
consideration
to
establishing
a
unit,
independent
of
Victoria
Police, responsible for the analysis and reporting of crime statistics.
Improving current practices
I have also made a number of recommendations in respect of crime statistics,
including that Victoria Police:
·
Better utilises the 000 emergency calls for service CAD data by linking
this information to the LEAP system
·
Reviews
training
in
crime
recording
and
provide
specialist
and
refresher training for police on the recording and use of crime data
·
Eliminates
the
use
of
manual
forms
for
recording
crime
data
by
progressing
the
LINK
project,
a
new
records
management
system
designed to replace the LEAP database
·
Reviews its organisational and management arrangements in relation
to the roles, responsibilities and resourcing of the collection, recording
and production of crime data and statistics.
Victoria Police stated that:
The strategic view of IT reform within Victoria Police aims to create effective
interfaces between systems such as CAD (ESTA) and LEAP/LINK. LINK will
replace LEAP in 2009-2010, creating a more effective platform upon which to
build these interfaces, however the resources to do this, and to expand CAD
to a state-wide system, are not currently available.
In
relation
to
police
numbers,
again,
while
I
did
not
find
evidence
of
manipulation of police numbers, I found archaic information technology and
confusing administrative practices in relation to police availability, which can
impact
on
the
under-recording
of
crime.
I
identified
that
Victoria
Police
closely monitors the number of police to ensure it meets the government’s
commitment.
It
also
has
developed
an
administrative
system
(the
People
Allocation Model) which enables allocation of police to the Police Service
Areas for operational deployment in a more transparent way. However, there
are a number of limitations to the model and I note that it excludes non-
operational police and civilian staff.
Victoria Police advised:
It is aware the Human Resources Management system used to administer
human resources requires replacement and it is misleading to indicate that
Victoria
Police
is
using
‘archaic
information
technology
without
acknowledging the work done over the past two years to address this.
In 2006 approval was given to set aside $28 million of internal funding to
build a new system. The findings in the supporting business case were the
result of extensive process audits conducted by Financial Services Division,
Corporate Management Review Division, the Auditor-General’s office and
recent reviews of current issues and business processes.
This project went to tender in late 2007 and the successful company to deliver
the new solution has been appointed. Work on building the new solution
commenced in 2008 and with Victoria Police have recently completed the
blueprint and are in the build stage. The solution is due to be implemented by
August 2009.
14
Measuring workload
I found the workload of operational police to be poorly measured. This has
implications for the effective management of the demand for police services.
It suggests that when allocating police, timely, accurate data on which to base
decisions
may
be
absent.
The
calls
for
service
CAD
data
reveals
how
frequently police are called to events which are not necessarily crime related.
Better use should be made of this data including measuring demand for
service and also for validating crime recording.
Victoria Police stated:
These
issues
have
been
previously
identified
by
Victoria
Police.
A
replacement for the AREA form a Member Activity Statement (MAS) - has
been trialed. It is in the process of being refined and will be
implemented
from 1 July 2009. It will take a broader approach to measuring operational
police workload. This work has been an initiative of the Data Management
Steering Committee over 2008.
The
changes
towards
more
proactive
policing
and
the
Victoria
Police
Intelligence Management approach require deployment of police in a flexible
way. However, there may at times be an imbalance in staffing because of
changing priorities and emerging issues. As a result, core general duties at a
number of police stations would appear to have been relegated, in my view,
to a lower priority; with fewer staff available to perform these tasks.
The administrative systems for filling police positions appear to be complex
and
slow.
This
has
an
impact
on
the
number
of
unfilled
positions
or
vacancies, which can create a shortage of staff.
Improvements to the roster
system,
greater
consistency
and
improved
terminology
would
assist
understanding of how police are appropriately allocated to where and when
they are needed in Victoria.
THE COMPLAINT
In April 2008, I received a complaint from Mr Ted Baillieu MLA, requesting
that I:
…initiate
an
investigation
into
whether
crime
statistics
and/or
police
numbers have been subject to manipulation...
In particular, the issues raised by Mr Baillieu and further described by Mr
Andrew McIntosh MP were that crime statistics and police numbers are being
distorted and misrepresented by Victoria Police.
While no specific allegations
were made of misconduct or fraudulent actions by individuals, there was
concern that administrative practices in relation to crime statistics and police
numbers
are
subject
to
manipulation
and
that
this
has
major
policy
implications and is of significant public interest.
A further concern expressed was that senior police may influence the way
crime is recorded so that police performance and crime reduction targets
are achieved.
It was further alleged by the complainant’s representatives
that
there
is
a
link
between
crime
statistics
and
police
numbers.
For
example,
because
of
police
As
a
matter
of
public
interest,
I
consider
staffing
shortages,
there
is
that uncertainty about the accuracy of crime
less time to record crime that
statistics
and
of
police
numbers
must
be
is brought to the attention of
resolved.
It has the potential to undermine
police hence not all crime
the
high
degree
of
confidence
Victorians
known to police is recorded,
have in their police.
making
crime
statistics
under-estimate
the
true
extent of crime.
It was also alleged that insufficient staff means fewer
police available at the front line’ to detect crime and to respond to calls
from the public, contributing to under-reporting of actual crime.
As
a
matter
of
public
interest,
I
consider
that
uncertainty
about
the
accuracy of crime statistics and of police numbers must be resolved. It has
the potential to undermine the high degree of confidence Victorians have
in
their
police,
as
demonstrated
by
surveys1
showing
a
majority
of
Victorians
were
generally
satisfied
with
their
police
and
their
effectiveness.
1 Report on Government Services, Productivity Commission, Melbourne 2008.
16
INVESTIGATION
Scope
A number of administrative systems are involved in both the recording and
production
of
Victoria
Police
crime
statistics
and
in
determining
and
managing
Victoria
Police
staffing
numbers.
My
investigation
included
an
examination of these systems to assess whether they are adequate to ensure
the
accuracy
and
reliability
of
the
reported
information.
It
also
included
considering relevant policies and procedures and compliance with these.
My
investigation
specifically
sought
to
determine
whether
there
was
any
evidence of tampering, deliberate distortion or misrepresentation of crime
statistics and police numbers.
Methodology
For
the
purpose
of
the
investigation
I
reviewed
relevant
information
including:
·
annual
Victoria
Police
crime
statistics,
including
statistical
information
accessible
from
the
Victoria
Police
website
and
other
crime statistics produced internally for Victoria Police use
·
Recorded Crime Victims, Australia
statistics and Crime and Safety,
Australia surveys and related documents produced by the Australian
Bureau of Statistics (ABS)
·
Victoria Police policies and procedures relating to crime statistics and
the
numbers
of
police
officers,
including
allocation
and
rostering
practices
·
other
relevant
documents
relating
to
crime
statistics
and
police
numbers.
My investigators conducted interviews with:
·
police and civilian officers working at Victoria Police headquarters
who
manage
or
whose
responsibility
includes
the
recording
and
production
of
crime
statistics
and
associated
data
systems
and
reporting functions
·
more
than
60
police
officers
from
a
cross-section
of
metropolitan
Melbourne and regional police stations, including frontline‘ police,
police managers and supervisors
·
senior executives of Victoria Police, including the Chief Commissioner
·
senior executives of The Police Association
·
external persons with expertise in police statistics and crime data,
including the Manager of the National Crime Statistics Unit, ABS, the
Director
of
the
NSW
Bureau
of
Crime
Statistics
and
Research
(BOCSAR)
and
the
author
of
the
review
of
Victoria
Police
crime
statistics by the Australian Institute of Criminology (AIC)
·
other relevant persons, including senior managers of the Emergency
Services Telecommunications Authority (ESTA), and police officers
and police statisticians in other Australian jurisdictions.
In addition, my investigators conducted:
·
visits to observe and examine crime recording practices and police
rosters at a number of police stations across Victoria
·
visits to the Central Data Entry Bureau (CDEB) where data is entered
on to the Law Enforcement Assistance Program (LEAP)
database at
the Victoria Police Centre
·
visits to the ESTA call dispatch centre
·
analysis
of
Victoria
Police
crime
database
records
and
their
comparison with other sources of information, such as police running
sheets/diaries, calls for service data and victim data
·
examination of Victoria Police policies and procedures in relation to
the recording of crime data, including definitions and counting rules
and those required for the production and release of crime statistics
and for the national Recorded Crime –Victims ABS statistical collection.
BACKGROUND
Crime
statistics
and
police
numbers
go
to
the
heart
of
long-standing,
perennial debates about law and order and are central to the crime conscious
society we live in.
The importance of these issues for the Victorian public is
highlighted by the recent media interest2 in violence in the Melbourne Central
Business District (CBD) and actions taken by police .
According to official statistics produced by Victoria Police, the overall crime
rate
in
Victoria
has
fallen
consistently
since
2001.
However,
there
are
indications that the public does not believe or accept that the crime problem is
decreasing.
Surveys show that there is a high level of concern; and that there
has been an increase in the number of Victorians who agree that:
2 Herald Sun, 20 July 2008; Herald Sun, 1 September 2008; Herald Sun, 8 September 2008; Herald
Sun, 14 September 2008; The Age, 14 September 2008.
18
…crime is rising and that [they] personally feel at risk…and that violent
crime is getting worse (70 per cent).
And that:
78 per cent [of Victorians] are concerned/very concerned about alcohol-
related violence in Melbourne at night3.
Fear of crime, associated with perceptions
about the increasing incidence of crime, has
been identified as a widespread social
problem, in spite of a reduction in official
crime rates.
Fear of crime, associated with
perceptions about the increasing
incidence of crime, has been
identified as a widespread
social problem, in spite of a
reduction in official crime rates.
Such fear can influence individual health and
wellbeing
and,
at
a
broader
level,
diminish
the
quality
of
life
in
local
communities
by
restricting
interaction
and
trust
among
people,
and
influencing
people’s
willingness
to
undertake
various
activities,
such
as
visiting the inner city after hours. When official statistics do not appear to
reflect public perceptions or the direct experience of individuals, or when
official
categories
and
definitions
used
do
not
agree
with
the
public’s
interpretation, cynicism or indeed confusion and misunderstanding can occur
about what the official crime statistics show.
When official statistics do not appear to
reflect public perceptions or the direct
experience of individuals, or when official
categories and definitions used do not agree
with the public’s interpretation, cynicism or
indeed confusion and misunderstanding can
occur about what the official crime
statistics show.
Concerns about police numbers
also feature prominently in
public debate about law and
order and are of significant
public interest. It is argued that
more police will lead to less
crime; that more police on the
streets will deter potential
offenders; and that their visible
presence will reassure the public that they are safe and secure to go about
their daily lives, although the evidence is far from clear-cut.
Information relating to crime statistics and to police numbers should be as
accurate, consistent and as timely as possible. This is critical, not only to
police and to the public they serve, but also to others, such as local and state
3 AustraliaSCAN, Perceptions of Crime and Safety Survey 2005-08; The Age/Nielsen Poll,
24 November 2008.
governments
who
rely
on
crime
statistics
to
develop
crime
reduction
strategies and to assess their effectiveness.
How crime statistics are produced
Crime statistics produced by Victoria Police are arrived at from initial reports
by the public or by police detecting crime themselves. The main ways in
which the public can report crime in Victoria are:
·
by telephoning a police station4
·
by going to a police station
·
by calling Crime Stoppers on 1800 333 000
·
by calling 000.
For many crimes where no victim is involved, such as for drug crime, crime
recording
is
more
dependent
on
police
priorities
and
deployment
of
resources.
Police
can
also
become
aware
of
crime
during
their
policing
activities, for example, through patrols, proactive policing initiatives, or by a
suspect admitting to a crime.
Crime statistics have long been recognised as understating the true extent of
crime and beset with significant limitations due in particular to non-reporting
by
the
public
and
non-detection
by
police5.
Some
crimes
such
as
family
violence and sexual assault are less likely to be reported or become known to
police, as victims are reluctant to come forward. For other types of crime, it
may be too trivial or inconvenient for victims to report the incident to police.
The extent of unreported crime is therefore not known; it has been referred to
by criminologists and statisticians as the ‘dark figure’ of crime.
Crime statistics are further limited by whether crime is recorded or not by
police. Not all crime incidents that police become aware of will be recorded.
Until recently in Victoria, police needed to be satisfied that there is evidence
that a crime has occurred. Police also need to determine whether the incident
meets standard definitions and classifications for one of the more than 4,000
crimes.
4 Many police stations have a recorded telephone message advising the caller that if the
matter is urgent, they should hang up and call 000.
5 Morris, N. and Hawkins, G. The Honest Politician’s Guide to Crime Control, University of
Chicago Press, 1979; Victoria Parliamentary Drugs and Crime Prevention Committee Inquiry
into Crime Trends, 2002; Coleman, C. and Moynihan, J. Understanding crime data: haunted by the
dark figure, Open University Press, 1996; Van Dijk, J. The World of Crime, Sage Publications,
2008.
20
In
deciding
whether
an
incident
should
be
recorded,
a
police
officer
is
required
to
follow
operational
procedures
outlined
in
the
Victoria
Police
Manual (VPM).
Up until 30 June 2008, the instruction (VPM 108-3) stated
that:
Any police officer receiving a report of an incident must make sufficient
initial inquiries to satisfy themselves that a crime has been committed or
not.
[emphasis
added]
Where
the
facts
indicate
that
a
crime
has
been
committed, complete and submit all relevant LEAP reports, containing the
best quality information available at the time.
The
instruction
was
recently
amended
by
Victoria
Police,
following
a
Memorandum
of
Understanding
between
the
Australian
police
commissioners and the ABS to adopt ten rules for the consistent recording of
crime across all jurisdictions.
Of these rules Rule 2 is significant as it shifts the
balance
in
the
discretion
excercised
by
police
in
recording
crime
-
from
recording an incident when satisfied about the evidence that a crime has
occurred, to now recording an incident unless there is evidence that a crime
has not occurred.
Since 1 July 2008, the instruction reads:
All criminal incidents reported to police must be recorded as offences unless
there is credible evidence available at the time of reporting to suggest that a
crime has not occurred. [emphasis added]
The reporting officer is required
to complete and submit all relevant LEAP/LEDR reports, containing the best
quality information available at the time. A comprehensive record of the
investigation must be kept on LEAP.
When the reported or detected incident is recorded and entered on to the
LEAP database, it can then be counted for statistical purposes. To record a
crime, police write the details of the incident manually on to paper forms,
which are checked by a station supervisor and sent by fax to the Central Data
Entry Bureau (CDEB) for data entry into the LEAP database. Exceptions to
this
practice
occur
for
some
property
crimes
in
many,
but
not
all,
police
stations where police officers are able to enter the information electronically
into LEDR (LEAP Electronic Data Recorder) instead of filling in the manual
forms. The data from LEDR is electronically added to the LEAP system at a
later time after checking by the Crime Desk supervisor.
Once data relating to the crime has been entered on to the LEAP database at
CDEB, there are a number of quality control measures which are applied to
ensure the completeness and integrity of the data. Following the various data
audits, extracts are taken from LEAP and categorised, according to statistical
rules and offence classifications, for use within Victoria Police, as well as for
the publication of annual official crime statistics which are available to the
public. A data extract is also sent to the ABS for the national crime statistics
collection.
There are different counting rules for different types of crime as part of the
process of compiling statistics. The main ones are:
·
For crime against the person and most property crime, the counting
unit is the number of principal victims for each separate occurrence of
the crime - if 3 alleged offenders assault 2 victims, the count is 2
·
For
crimes
against
statutes
(usually
where
there
is
no
individual
victim) the number of alleged offenders is the counting unit for 3
alleged offenders found in possession of cannabis, the count is 3
·
For some other crimes, the incident itself is the count.
In addition, only the most serious offence which best describes a distinct
course of criminal conduct (called a sub-incident in LEAP) is counted, even
though there may be multiple charges laid against the alleged offender for
that course of action. The example given in the annual Victoria Police Crime
Statistics states:
…an
offender
carrying
a
firearm
commits
an
armed
robbery
only
the
offence of armed robbery is recorded although the offender would be charged
with armed robbery and possession of a firearm.
In relation to a distinct course of criminal conduct’, the following example is
given:
…if an offender presents three valueless cheques to a [bank] teller only one
offence
would
be
recorded
but
if
the
three
cheques
were
presented
at
different
times
or
at
different
branches
then
three
offences
would
be
recorded.
The crime statistics derived from the database relate to the month and year
when the information was recorded by police, not when the crime actually
occurred. The complexity of the counting rules and the classification of crimes
into categories as well as the impact of this on accuracy are further discussed
below.
22
The steps in producing crime statistics by police are broadly summarised in
the following diagram:
Crime comes toattention of police
Police follow policy instruction to
determine whether to record the crime
or not(policy change from 1 July 2008).
Police enter data manually on to forms
which are faxed to a central point for
processing on to the LEAP database (for
some high volume property crime, data
can be entered electronically via LEDR
on to LEAP at those stations with a
crime desk).
Data extracted from LEAP for police to
produce statistical reports
What crime statistics are available in Victoria?
Victoria Police crime statistics are made available to the public following the
end of the financial year. They are derived from a July extract of the LEAP
database. This is the principal source of reporting crime statistics to the public
in Victoria and includes a comprehensive range of information
It shows types
of recorded crimes by 27 categories (Attachment 1), divided into four broad
groups:
·
offences against the person
crimes agai
st
roperty
·
r
g crime
·
other crime.
These annual statistics also include characteristics such as age, sex, and racial
appearance of victims and alleged offenders, and the location of the offence.
Comparisons with the previous two years are also presented, as well as crime
rate per 100,000 population of recorded offences and clearance rates.
Victoria Police provides some electronic access to crime statistics to the public
through downloadable reports in PDF format and an online selection tool to
obtain fixed one-page reports for police Regions, Divisions, Police Service
Areas
(PSA),
Local
Government
Areas
(LGA)
and
postcodes.
The
online
selection tool allows an online user to select one police region, division, PSA
or LGA and receive a one-page report for the selected area. Victoria Police is
also able to respond to requests for crime statistics by external users at a cost.
For internal use Victoria Police produces a number of statistical reports from
LEAP, such as the Monthly Crime Management Report for senior management.
A second source of crime statistics available is data from the ABS national
Recorded Crime Victims statistics collection. These statistics are compiled to
national standards and classifications developed by the ABS in conjunction
with
jurisdictions.
They
are
produced
on
a
calendar
year
basis
and
are
usually released in May/June. In Victoria they are derived from the same
database (LEAP) as the Victoria Police crime statistics.
Victoria Police provides the ABS with data on the number of offences in the
categories of the national crime statistics collection, with rules being applied
to map the Victoria Police codes to the ABS codes. The ABS conducts regular
quality checks on the data provided by jurisdictions. However, there are a
number of differences in the two sets of crime statistics—the national crime
statistics use different counting rules and classifications from Victoria Police
statistics and are designed to ensure comparability across the jurisdictions. A
notable difference from Victoria Police crime statistics is that the ABS national
statistics include fewer categories of crime (covering approximately 70 per
cent of total recorded crime by Victoria Police). Also for the ABS statistics the
count of crime is victim based – only the number of victims for each offence is
counted6 - whereas for Victoria Police statistics, there are counts for victims as
6 The ABS advises that the counting rules are complex -a person reporting a crime with
multiple offences in the same incident may either be counted multiple times, or may be
counted only once depending on the types of offences committed during the incident. If
multiple offences within the same incident fall within the same Australian Standard Offence
Classification then the victim will be counted only once.
24
well as alleged offenders and incidents. As discussed further below, due to
concerns about a lack of comparability for assault and sexual assault offences,
the ABS no longer provides figures for these offences, although for the other
offences the level of comparability is satisfactory.
A
less
frequent
but
significant
source
of
crime
statistics
is
crime
victim
surveys. The surveys have been conducted by the ABS7 and other survey
organisations on representative samples of adults who are directly asked if
they have been a victim of selected serious criminal incidents during the
previous
12
months.
The
surveys
provide
information
about
incidents
of
crime experienced by victims, regardless of whether the incident is reported
to police. Survey results are unaffected by changes in public reporting to
police or police recording practices. Although they have known limitations,
they are able to shed some light on the ‘dark figure’ of unreported crime.
Each source of crime statistics provides only a partial picture of crime; each
has limitations and care needs to be taken when interpreting figures and
trends from these three sources.
What do the crime statistics show for Victoria?
Crime in Victoria, as measured by Victoria Police crime statistics, has shown
significant reductions 17.0 per cent in total crime and 24.5 per cent in the
rate per 100,000 population since 2000-01, although the downward trend has
slowed recently. In 2007-08 the number of offences recorded declined by 0.4
per
cent
compared
with
the
previous
year.
Total
Crime
against
the
person
increased by 1.4 per cent; decreases occurred for Homicide and Sex offences but
increases were recorded for Robbery (up by 15.7 per cent), Assault (up by 0.7
per
cent)
of
which
around
a
quarter
are
attributed
to
family
violence.
In
relation to total Property offences, these decreased by 0.4 per cent, although
Thefts
from
motor
vehicles
increased
by
12.3
per
cent.
Total
Drug
offences
recorded a decrease of 2.5 per cent compared with the previous year.
I note at the national level, the comparative results of the ABS Recorded Crime -
Victims statistics presented in the Report on Government Services show Victoria
to have lower levels of crime in most crime categories over a number of years,
compared
to
other
Australian
jurisdictions.
Also
in
terms
of
crime
7 The most recent ABS Crime and Safety survey was released in 2005; results from the 2008-
2009 survey will be released in 2010. The ABS has commenced running the survey from July
2008 with a ‘rolling’ collection throughout the year, subject to ongoing review.
victimisation rates, as measured by the ABS Crime and Safety surveys, Victoria
had lower rates for household and personal crime, which is significant as the
surveys are not affected by public reporting to police or by police detection
My investigation found evidence of under-
reporting of some crime over a long period,
as a result of poor crime recording practices
and deficient administrative processes.
and recording practices. In this
context, Victoria is also lower on
a range of criminal justice
measures (court cases, prison
population and young people in
detention)
both
in
numerical
terms
and
per
capita,
compared
to
other
Australian jurisdictions8.
Crime
levels
in
Victoria,
compared
to
other
Australian
jurisdictions,
are
shown to be lower and may be attributable to a range of socio-demographic
factors such as age and education. Notwithstanding this, my investigation
found evidence as presented below, of under-reporting of some crime over a
long
period,
as
a
result
of
poor
crime
recording
practices
and
deficient
administrative processes.
How crime statistics are used
Crime statistics serve the following main purposes:
·
They provide key measures of criminal activity for police intelligence
purposes, enabling police to identify priorities and emerging crime
issues and to develop crime control strategies.
·
They assist in the allocation of police resources for various reactive
and proactive activities.
·
They are extensively used to assess the performance of police and
provide accountability to government in relation to agreed targets,
such as those published in the Victoria Police Annual Report and
Business Plan.
·
They keep the public informed about the nature, level and extent of
crime in the community.
·
They represent a form of accountability for government spending on
police.
·
Use is made of them by the media to communicate information about
crime to the public.
·
Specific
groups
such
as
victims
of
crime
may
require
more
information about crime trends.
8 The Victorian Parliamentary Drugs and Crime Prevention Committee in its Inquiry into
Crime Trends 2002 provided a historical analysis of Victorian and Australian crime trends.
26
Crime statistics are central to Victoria Police for setting performance targets,
measuring their achievement, and for identifying priorities for police action.
They are also used for planning, managing and allocating police resources
across
Victoria.
Victoria
Police
analysts
use
crime
statistics
extensively
to
identify crime trends and to produce assessments of crime problems on a
regular basis to enable management to direct resources to priority areas. For
instance, an analyst from the Safe Streets Taskforce based in Region 1 (see
Case
Study
p96)
will
use
crime
data
from
the
previous
weekend,
in
association with other relevant information, to assist management in making
appropriate staffing decisions for the forthcoming weekend.
Victoria Police’s
Corporate
Performance
Unit
also
relies
heavily
on
crime
statistics
for
the
Compstat forums held at executive, divisional and regional levels to evaluate
police performance against targets.
There
are
many
other
users
of
crime
statistics.
These
can
range
from
government departments to assist policy development and implementation to
local governments with an interest in community safety. Individuals such as
applicants (and opponents) for a liquor licence in a local neighbourhood, may
use crime statistics to support their case. Insurance premiums for car theft and
burglary
are
also
determined
with
reference
to
crime
levels
in
specific
localities.
How many police are there in Victoria?
In the discussion of police numbers, it is useful to clarify some key terms.
While the term Victoria Police is not defined or used in the Police Regulation
Act 1958, it is generally understood to refer to a statutory body. The Public
Administration
Act
2004
(Vic)
defines
Victoria
Police
as
a
‘special
body’.
Another term commonly used to denote Victoria Police is the force’, again a
term not defined in the Police Regulation Act.
The Chief Commissioner is empowered by the Police Regulation Act and the
Public Administration Act 2004 to employ staff. Victoria Police employees make
up three groups.
‘Sworn’ police officers have the power of arrest, have taken
the oath of constable and can exercise police powers including the use of force
in
the
course
of
their
duty.
‘Sworn’
police
are
appointed
by
the
Chief
Commissioner in accordance with the Police Regulation Act. A second group of
employees of Victoria Police are also appointed under the Police Regulation Act
and consists of police recruits, considered as employees on probation until
they complete training, as well as protective service officers and reservists.
‘Unsworn’ civilian officers are the third group.
They are public servants
employed
by
the
Chief
Crime statistics are central to Victoria
Police for setting performance targets,
measuring their achievement, and for
identifying priorities for police action. They
are also used for planning, managing and
allocating police resources across Victoria.
Commissioner under the Public
Administration Act. They are not
‘officers of the force’ in
accordance with the Police
Regulation Act. They generally
undertake support functions,
such
as
intelligence
and
data
analysis,
financial
and
human
resource
management,
forensic
science
and
information technology.
It is the Victorian government that determines the number of police officers or
any
increases
to
the
number.
The
number
of
civilian
officers
working
in
operational areas is determined by Victoria Police.
The information in Table 1 was provided to me by Victoria Police as at 3
November 2008. It shows that there were a total of 11,090 police officers,
including
167
recruits,
8,168.63
operational
police
officers
(Full
Time
Equivalent
(FTE))
located
in
the
five
police
regions
(75
per
cent),
2,277
operational police in specialist areas (21 per cent); and 477.36 (FTE) police in
support areas (4 per cent).
In addition, there were 2,512.7 civilian staff (FTE)
employed by Victoria Police on that date.
28
Table 1:
Victoria Police numbers as at 3 November 2008
REGION ONE TOTAL POLICE RESOURCES
1517.52
REGION TWO TOTAL POLICE RESOURCES
1742.23
REGION THREE TOTAL POLICE RESOURCES
1671.06
REGION FOUR TOTAL POLICE RESOURCES
1771.18
REGION FIVE TOTAL POLICE RESOURCES
1466.64
Subtotal 8168.63
OTHER OPERATIONAL POLICE
Counter Terrorism and Emergency Management
45.00
Covert and Intel
311.61
Crime
545.92
Ethical Standards
151.37
Forensic
82.89
Legal Services
205.80
Operations Co-ordination
61.89
Probationary Constables
81.00
Specialist Support
435.45
Traffic and Transport
356.07
Subtotal 2277
SUPPORT POLICE
Airlie Leadership
8.00
Business Information Technology
30.58
Business Management Dept
53.58
Corporate Communications
72.38
Corporate Management Review
15.00
Corporate Strategy and Performance
36.16
Education Dept
189.90
Human Resource
47.26
Licensing Services
11.50
Office of Chief Commissioner
13.00
Recruits
167.00
Total
11,090.00
Source: Victoria Police
Where are the police in Victoria?
The number of police officers in operational positions and where they are
located
is
a
relevant
issue
for
my
investigation.
Operational
police
are
described as:
·
police who are first responders’ to the community and work within
the five geographical police regions in Victoria (Attachment 2)
·
police
who
attend
and
provide
operational
support
to
‘first
responders’
·
specialist
police
who
provide
support
to
operational
police
in
achieving their objectives, for example, legal and forensic services.
Victoria
Police
provides
a
24
hour
seven
days
a
week
response
to
the
Victorian
community
through
a
network
of
‘first
responders’
who
work
within the five regions. The regional policing services are provided through
operational police based at police stations, traffic management units, Crime
Desks and crime investigation units, and are grouped into 56 Police Service
Areas (PSA), generally aligned with Local Government Areas within each of
the regions. Each PSA is headed by a manager of Inspector rank. The diagram
below summarises where Victoria Police staff are located as at September
2008.
Source: Victoria Police
30
While the Victorian government determines the number of police officers, it is
Victoria Police that determines their allocation across the state and how they
are to be used operationally.
The majority of police officers are assigned to the five police regions.
Within
each region most police work in the general duties uniform section at police
stations or in investigation teams, such as Crime Investigation Units (CIU)
and the Crime Desks. A number of police officers work in regional offices.
Victoria Police also has nine specialist operational departments, including the
Crime Department.
A smaller number of police are distributed among non-
operational
departments,
including
the
Human
Resource
Department,
the
Office of the Chief Commissioner and the Corporate Strategy & Performance
Department.
My investigation is specifically concerned with the allocation of police officers
to the regions and, in particular, those working as uniform officers in police
stations.
These officers as first responders’ are at the direct interface between
Victoria Police and the public and are generally the first to become aware of
crime and other incidents.
They are often the first point of contact for people
requiring
police
services.
One
of
their
key
responsibilities
is
to
record
incidents and information about crime for entry on to the LEAP database, so
that data can be extracted to enable crime statistics to be produced.
CRIME STATISTICS
Recording of crime
My
investigation
found
a
number
of
concerns
about
the
recording
and
collection of crime data, which could impact on the accuracy of the statistics.
In this regard, my investigation identified a number of concerns about the
quality of the data and the administrative systems used to produce crime
My investigation found a number of
concerns about the recording and collection
of crime data, which could impact on the
accuracy of the statistics.
statistics. Issues were identified
which could lead to distortion
and misrepresentation in the way
information about crime statistics
is produced and presented, and
also in relation to police numbers.
While there was no corroborated evidence
that these instances involved manipulation to the extent that they could be
considered
to
constitute
systematic
falsification
of
statistics,
they
were
of
concern.
Police officers and civilian staff of Victoria Police were given opportunities to
provide me with any information relevant to the investigation. I was provided
with
examples
of
what
was
believed
to
be
manipulation
of
crime
data.
However, in none of the examples examined was it possible to draw firm
conclusions.
The following examples are indicative of those received:
One Senior Sergeant stated that:
…we have been told that they have too many burglaries and weve got
Compstat coming up in 3 or 4 months, if it’s clear someone broken a screen
door and tried to break the front door, the normal category of offence is
‘attempted burglary; but you dont make it an attempted burglary’ if you
cant catch the crook; so you make it a wilful damage’. There was one
particular division where attempted burglary’ went down by 11 per cent and
‘wilful damage went up by about 12 per cent.
Another police officer described an incident involving fraud where several
persons had been charged with more than 1,000 offences over a four-year
period.
…I charged A with [] counts and B with [] counts… [and] ...forms were
sent to LEAP data entry for recording. Naturally all these offences have been
cleared. I have eagerly been awaiting for the stats to appear on the monthly
statistics. This hasn’t happened.
When the officer contacted LEAP data entry to establish why they had not
been recorded he was informed that:
…the incidents and sub-incidents have all been recorded as one incident/sub-
incident as they are all committed by the same offenders and on the same
victim. I expressed my concern as to why this would be, as it appears that the
figures produced by LEAP are false and misleading. The person I spoke to at
LEAP was in agreement and said that they are false. They also informed me
that their computers are not capable of recording crimes in excess of 999. Of
course my total offences are [more than 1,000].
What disturbed me also is
that my team has put in months of work on this file, to be rewarded with the
clean up of one crime. This is not fair on me or this office.
Victoria Police advised in respect of the above example:
It appears the concern shown by this police officer relates to the fact that
many hours may have been spent investigating this offence, only to be given
32
credit for solving one incident rather than the 1000+ offences committed.
Hence, the offences are of course recorded, but there is an issue with the
police officer’s understanding of the sub-incident count.
Many examples confirmed the central role that crime statistics play. They are
used to identify ‘spikes’ when there may be a sudden increase in some crimes,
which may need explanation and require strategies to address them. The
statistics also enable comparisons of trends across the PSAs to assess how well
they are performing and to assist in staff allocations.
However, in a review of Compstat prepared by the Corporate Management
Review Division (February 2006), inaccurate crime data was identified as a
significant issue.
The report stated:
…Examples were given where inconsistent interpretation of data entry fields
had led to inaccuracies in the data presented to Compstat, and on occasion
resulting in unnecessary resources being utilised to cleanse the data.
The review team noted that some of the concerns about crime statisticsare
symptomatic
of
uncoordinated
legacy
data
systemswhich
do
little,
if
anything to address the deeper issue of rectifying our data integrity issues…
Police interviewed as part of the review expressed frustration with data
integrity,
with
the
resultant
need
for
data
cleansing…others
interviewed
pointed the blame at sergeant level regarding quality assurance of data that is
entered at the front line.
The review further queried:
…to what extent any deficiencies which are embedded in our data collection
systems contribute to this issue [of data integrity]to do this one should look
at the design of our data collection systems in a holistic manner; the context of
their usability, their ability to minimise the risk of data entry error, their
capacity to be integrated, interrogated and developed to meet the changing
needs… [of Victoria Police].
Also:
…Given that a key element of the Compstat process is data driven analysis
which are made available to identify and analyse problems and to track and
assess
police
responses;
…to
reduce
crime,
police
need
to
know
about
crimespecifically police need to know: what crime is happening, where
crime
is
happening,
when
crime
is
happening
and
why
crime
is
happeninggood fast crime analysis is vital to crime reduction.
I recognise the difficulties faced in attempting to validate the crime recording
process as one officer put it….’if it isn’t there, it can’t be audited...’
When
my investigators sought information about any internal validation reviews or
audits of the recording of crime practices, particularly at the station level, no
details were able to be provided. Despite the lack of auditing information
about crime recording, I examined how the validity of LEAP crime statistics
could be tested, in view of the concerns expressed about the quality of crime
data.
Victoria Police advised
Validating crime statistics can also be performed by comparing recorded
crime with activity on member running sheets.
This methodology was used
by the Australian Institute of Criminology Review of Victoria Police Crime
Statistics
2002
when
it
was
found
that
the
crime
statistics
published
by
Victoria Police accurately reflect the counting rules and crime classifications
on LEAP.
Furthermore, part of the standard duties of a section sergeant has
always
been
to
review
running
sheets
and
ensure
that
there
is
a
corresponding crime report with any entry that indicates that one would be
required as well as checking all LEAP reports for accuracy and completeness.
In my view, comparing running sheets with
LEAP records does not validate whether a
LEAP record should have been required, as
there is no check on whether the running
sheets are accurate.
However, on running sheets, crime
can often be categorised by police
as NOD no offence detected;
NPH no person home; GOA
gone on arrival; no complaint; or
victim did not wish police
involvement.
These instances would not be recorded on LEAP. In my view,
comparing running sheets with LEAP records does not validate whether a
LEAP record should have been required, as there is no check on whether the
running sheets are accurate.
I conducted the following exercises to illustrate aspects of the Victoria Police
crime recording processes about which I have most concern.
I compared:
1.
recorded
crime
data
on
LEAP
with
calls
for
service
(CAD)
data
provided by ESTA
2.
recorded crime data on LEAP with data on victims’ applications to the
Victims of Crime Assistance Tribunal (VOCAT)
3.
Victoria Police annual recorded crime statistics with estimates of crime
derived from crime victim surveys.
34
The results of these are dealt with in the following sections.
(i) Comparing Victoria Police crime data with calls for service data
One way to identify any gap between crime that police are made aware of and
crime that is recorded on the LEAP database is to examine data on emergency
calls
made
by
dialling
000.
The
Emergency
Services
Telecommunications
Authority
(ESTA)
manages
the
triple
zero
call
service
and
dispatches
emergency services across the Melbourne metropolitan area and Geelong
for Victoria Police, the Country Fire Authority, the Metropolitan Fire Brigade,
Ambulance Victoria and the State Emergency Service. The calls are received
and assessed by call centre staff who then dispatch the calls to the relevant
emergency
service
through
the
CAD
system
via
Telstra.
As
part
of
the
process,
CAD
data
is
collected
by
ESTA
for
reporting
and
performance
requirements.
Calls for police via the CAD system have been increasing by around 5 per
cent per year and in 2008 represented about 55 per cent of all emergency calls
received. While Victoria Police receives regular CAD data from ESTA and can
request data checks, there is no direct interface between the CAD data and the
LEAP
database9.
I
understand
that
the
other
emergency
services
have
a
broader
interconnection
with
CAD,
which
assists
them
in
resource
management and allocation and in event location and response notification.
During my investigation I was advised that a data matching exercise between
CAD data and LEAP had been conducted by an internal audit in Victoria
Police some years ago. The results found significant differences between the
CAD data and what was recorded on LEAP; it identified under recording of
crime by police and delays in recording the crime on to the LEAP forms to be
of concern.
I understand that an internal audit report at that time showed that
crime statistics were understated, particularly in relation to assaults.
During my investigation, I obtained CAD data from ESTA for events which
had been assigned to 12 metropolitan police stations in the five regions for a
randomly selected day in November 2007 and a day in April 2008. A sample
of 572 events from the CAD system was taken for further analysis.
The events
were
broadly
classified
by
ESTA
into
the
crime
and
public
disturbance
categories
used
by
the
CAD
dispatch
operators.
Crime
events
included
9 The exception is a link that provides CAD event information to the Victoria Police Media
Unit to ensure that timely and accurate information is communicated to the general media.
personal
crime
categories,
such
as
assault,
and
property
crime
categories
included burglary, wilful damage and theft of motor vehicle were examined. I
would expect to find LEAP records for these events in many instances.
LEAP
records would be unlikely for public disturbance events, which included pub
brawls,
people
causing
trouble,
noise
complaints
and
drunken
behaviour.
With the assistance of staff at the CDEB, each of the 572 CAD events was
examined to determine whether a LEAP record existed. Figures 1-3 show the
results obtained in relation to days of the week, the types of event category
and by police station.
Figure 1: Percentage of CAD crime and public disturbance events by days
of the week and whether a LEAP record was identified
Figure 1 shows whether there was a matching LEAP record for CAD crime or
public
disturbance
events
relating
to
each
day
of
the
week.
The
analysis
identified that there was no difference in the matching of LEAP records for
different days of the week. Crime events were more likely to have a LEAP
Overall, the majority of CAD events had no
LEAP record.
record around 20 per cent,
compared to almost none for
public disturbance events. For
about 10 per cent of CAD crime events it was not clear whether a LEAP
record existed. Although I would not necessarily expect there to be a direct
correlation between CAD events and LEAP, it was nevertheless surprising to
36
find such a low level of relationship between the CAD crime events and
LEAP. Overall, the majority of CAD events had no LEAP record.
Victoria Police advised:
CAD (ESTA) and LEAP were created for different purposes.
They were not
designed to be linked, nor has the information in them been structured to
facilitate linkingVictoria Police has identified that linked systems would be
beneficial, however the resources to do this, and to expand the CAD to a
state-wide system are not currently available.
Figure 2: Percentage of CAD crime events by type and whether a LEAP
record was identified
Figure 2 shows the crime events from the CAD data by the type of crime (as
categorised by the ESTA call taker) and whether a matching LEAP record was
found.
The results10 show that for most burglary and shop theft events, there
was a LEAP record (73 and 70 per cent respectively).
However, only about 40
per cent of motor vehicle theft and wilful damage events were found to have a
corresponding LEAP record. For crime events categorised as family disputes
and suspected loitering, slightly less than 10 per cent had LEAP records.
10 For 16 per cent of crime events, police recorded the outcome as ‘cancelled’ or duplicate’.
Figure 3: Percentage of CAD crime events by police station and whether a
LEAP record was identified
When the results for CAD crime events assigned to the 12 selected police
stations were examined (Figure 3), they showed marked differences in the
frequency of matching LEAP records.
The smallest percentage of matching
LEAP records occurred in Frankston (7 per cent) and the highest percentage
in Dandenong (33 per cent).
I examined more closely the CAD data for
Dandenong and Frankston by type of event which showed differences in the
nature and frequency of some of the crime events (Figure 4).
It is unclear
whether
these
could
account
fully
for
the
marked
difference
in
the
proportions of LEAP records found between these two stations.
38
Figure 4: CAD crime and public disturbance events by type for Dandenong
and Frankston
Part of the explanation for not finding greater matching between CAD crime
event data and LEAP records is that in terms of outcomes, that is, when police
arrive, over 50 per cent were categorised as either All apparently correct’ or
‘Gone on arrival’.
Due to the previous Victoria Police policy to only record on
to LEAP where there was sufficient evidence that a crime has occurred (the
evidentiary approach), no LEAP entry would be expected and thus would not
be counted in crime statistics.
Further analysis of the outcomes was conducted to attempt to clarify the lack
of LEAP records for so many CAD crime events by comparing LEAP record
matches
with
Victoria
Police
CAD
outcome
categories.
The
results
(Attachment 3) showed that where an offence was detected’ or an offender
apprehended’, 23 out of 31 crime events (74 per cent) had a LEAP record.
Where enquiries were pending’, there were 30 out of 55 events (55 per cent)
with
a
LEAP
record.
For
other
outcomes
however,
including
‘no
offence
detected’, gone on arrival’, cancelled event/duplicate’, and all apparently
correct’, the number of LEAP record matches for CAD crime events dropped
to 4 out of 191 events (2 per cent).
These results are significant in that they show for CAD crime events where
there was a tangible outcome for police an offender apprehended’ or an
‘offence detected’ - a quarter were not recorded on LEAP; and where the
These results are significant in that they
show for CAD crime events where there was
a tangible outcome for police an offender
apprehended’ or an offence detected’ - a
quarter were not recorded on LEAP; and
where the outcome was enquiries pending’,
around half did not have a LEAP record.
outcome was enquiries
pending’, around half did not
have a LEAP record. Out of a
total of 285 CAD crime events
examined, only 11 per cent had
an outcome of offender
apprehended’ or an offence
detected’ and 19 per cent were
categorised as enquiries pending’.
For a majority of CAD crime events (70
per cent), the police outcome was noted as unable to locate, duplicate or
cancelled event, no offence detected, gone on arrival or all apparently correct’.
Table
2
shows
a
summary
of
the
results
for
the
CAD
crime
and
public
disturbance events and the extent of LEAP database matches.
Overall, it was
found that for all CAD events which had been assigned to police only 10 per
cent had a matching LEAP record - 57 crime events and 2 public disturbance
events.
40
Table 2:
CAD crime and public disturbance events and whether a LEAP
record was identified
Matching LEAP record identified
CAD event type
Yes
No
Unclear
Total CAD
events
Crime
57
201
27
285
(20%)
(70 %)
(10%)
(100%)
Public disturbance
2
277
8
287
(<1%)
(97%)
(3%)
(100%)
Total
59
478
35
572
(10%)
(84%)
(6%)
(100%)
I consider that closer examination is required in this area to determine factors
influencing the recording of crime events on LEAP and in particular, why
many CAD crime events have no record on LEAP and thus are not counted in
crime
statistics.
Further
The
lack
of
correlation
between
the
examination
is
also
required
to
classification
of
events
and
outcome
clarify why so many CAD public
descriptions used by ESTA and by Victoria
disturbance-related
events,
Police
is
problematic
and
needs
to
be
which may touch on crime and
reviewed.
public
safety
matters,
are
not
recorded
on
LEAP.
The
lack
of
correlation between the classification of events and outcome descriptions used
by ESTA and by Victoria Police is problematic and needs to be reviewed.
In a larger exercise of this kind conducted by the NSW Bureau of Crime
Statistics and Research in 2000, similar comparisons were made using crime
data
and
data
from
the
computerised
incident
dispatch
system
for
two
offences - Break and Enter and Motor Vehicle Theft - for each of the NSW
Local Area Commands11. A comparison between NSW and Victoria is possible
by using the Victorian burglary and motor vehicle theft calls for service data
11 Chilvers, M. & Doak, P. Validation of NSW Police Crime Statistics: A Regional Analysis, NSW
Bureau of Crime Statistics and Research. General Report Series. Sydney, 2000.
Validation of NSW police crime statistics: a regional analysis, BOCSAR, Attorney-General’s
Department, 2000. Last accessed October 2008 at
http://www.lawlink.nsw.gov.au/lawlink/bocsar/ll bocsar.nsf/vwFiles/r48.pdf/$file/r48.pdf
which resulted in a LEAP record for the selected stations in Victoria in 2007
and 2008. Although the comparison is limited by the 7-year difference in time,
which
may
include
variations
in
recording
practices,
improved
use
of
technology, as well as the prima facieapproach used by police in NSW to
record crime, the results show a lower level of matching for Victoria than for
NSW.
For Victoria 73 per cent of calls for service relating to burglary events had
clearly identifiable matching LEAP records (for another 7 per cent of burglary
events it was unclear whether a LEAP record existed). In the NSW study the
average match across all Local Area Commands was 87 per cent in 1997 and
1998,
and
94
per
cent
in
1999
for
burglary.
For
motor
vehicle
theft,
the
Victorian calls for service data show that only 40 per cent of calls for service
had a matching LEAP record. This is much lower than would be expected
given the requirement of insurance agencies for a victim to report the theft to
police. By comparison, the NSW study found the ratio of the relationship
between recorded crime to calls for service in relation to motor vehicle theft
ranged between 62 per cent and 75 per cent.
I
note
the
difficulties
encountered
in
the
matching
exercise
where
it
was
unclear for many CAD events whether a LEAP record had been created or
not.
Greater harmonisation of event classification and outcome disposition
categories between the CAD system and LEAP to accurately describe events
reported and the circumstances found on arrival would be of benefit.
When asked about discrepancies between CAD events and LEAP records,
operational police said that much crime is written off’12 and not recorded on
Mapping analysis for a region where CAD
data had been overlaid with LEAP offences,
showed that LEAP records only revealed a
fraction of what was occurring in terms of
public safety and crime incidents to which
police are called out via the CAD system.
LEAP even though it appears on
the CAD system. This is
particularly the case for public
order incidents, categorised in
CAD data as public disturbance’
such as brawls and young people
causing trouble. I was informed
that for a mapping analysis for a
region where CAD data had been overlaid with LEAP offences, the results
12 The term written off’ was used, not in the sense of finalised, but to indicate it was
dismissed and not reported for one of the following reasons: No offence detected; No Person
Home; Gone on Arrival; No Complaint; Did Not Want Police Involvement.
42
showed that LEAP records only revealed a fraction of what was occurring in
terms of public safety and crime incidents to which police are called out via
the CAD system.
Many
of
the
matters
that
police
attend
in
response
to
000
calls
are
not
categorised as crime events and, as noted, these would not be expected to lead
to the creation of a LEAP report. Such events however do provide valuable
information about the broad nature and extent of demand for police services.
I understand that Ambulance Victoria uses a computer application (Siren Live)
on
CAD
historical
data
to
I consider that the differences found between
recommend
where
ambulances
the
LEAP
records
and
CAD
data
require
should be moved to so patients
further
investigation
as
to
the
level
of
are treated as quickly as possible.
compliance
by
police
with
policies
for
It takes into account the expected
recording crime.
caseload
distribution,
travel
speeds
and
the
likely
time
for
ambulances
to
complete
their
cases,
and
plots
where
and
when
calls
for
service are likely to be required to assist Ambulance Victoria in its demand
and
resource
management.
CAD
data
is
clearly
valuable
for
workforce
planning
and
resource
deployment
in
Victoria
Police
and
also
provides
information for police intelligence gathering.
I consider that the differences found between the LEAP records and CAD
data require further investigation as to the level of compliance by police with
policies for recording crime. It was not possible to investigate the reasons why
no LEAP record existed for crime events as recorded by the CAD system
within the scope of my investigation. While exact matching between CAD and
LEAP
data
is
not
possible,
the
CAD
data
nevertheless
provides
a
good
auditing framework for checking the data on to LEAP.
Such auditing would
be
facilitated
if
the
CAD
event
number
is
entered
on
the
police
running
sheets.
Victoria Police advised:
ESTA is a separate entity, a multi agency approach would be required to link
a LEAP incident number with an ESTA Event ID (EID); which is the matching
process used in other states.
Recommendation 1
I
recommend
that
Victoria
Police
conduct
regular
audits
where
LEAP
records are audited against data from the CAD system for validation.
The
validation should form part of permanent data quality and data integrity
processes for crime recording.
Victoria Police response:
As part of the implementation and quality assurance program of the ABS
National
Crime
Recording
Standards,
Victoria
Police
in
November
2008
agreed with other jurisdictions to explore comparisons between CAD data
and criminal record data. Such comparisons may have difficulties, as, in
Victoria,
not
all
CAD
data
is
state
wide,
however,
Victoria
Police
has
committed to doing this work.
While Victoria Police advised that the CAD system for Victoria Police calls
only
covers
metropolitan
Melbourne,
I
understand
that
ESTA
coverage
already includes Geelong, Victoria’s second largest city. I note that there is
also an Estimates Review Committee submission led by the Department of
Justice (DOJ) under consideration by the Department of Treasury and Finance
which includes extension of the CAD system to regional and rural Victoria.
(ii) Comparing Victoria Police crime data with Victim of Crime Assistance
Tribunal data
My
investigation
also
examined
data
provided
by
the
Victims
of
Crime
Assistance Tribunal (VOCAT) to determine whether a corresponding LEAP
record existed for a sample of victims’ applications.
The Tribunal, established
under the Victims of Crime Assistance Act 1996 came into operation on 1 July
1997, provides assistance to victims of crime by paying them for expenses
incurred as symbolic expression by the state of the community’s sympathy for
the suffering caused or where compensation for injury cannot be obtained
from the offender or other sources, as a direct result of the crime.
My
investigators
considered
data
from
applications
of
victims
of
crime
received by VOCAT between November 2007 and April 2008. It could be
expected, given the likely serious nature of the crime experienced by these
victims, that a high percentage of applications would have a corresponding
LEAP
record.
It
is
also
a
requirement
by
VOCAT
that
the
victims
have
reported the crime to the police.
A random sample of 166 applications was
44
examined
by
my
investigators
for
matching
LEAP
records
at
the
CDEB,
representing 21 per cent of the total applications for that period.
As shown in Table 3, out of the sample of 166 applications, 130 were found to
have LEAP records (78 per cent). Of those with LEAP records, the majority
were found to be related to assaults (59 per cent), followed by sex offences (22
per cent).
Table 3: Comparing VOCAT applications with LEAP records
VOCAT applications by
Number of
Per cent
crime type as recorded
LEAP record
on LEAP
matches
Assault
76
59
Sex offences
28
22
Burglary
7
5
Robbery
8
6
Other
11
8
Sub total
130
100
Unable to determine
36
Total examined
166
With respect to the 36 applications which CDEB staff were not able to locate
For some serious crimes there is no LEAP
record or a record cannot be found which
may indicate under-recording of crime by
Victoria Police. I would have expected that
for most crime where victims apply to the
VOCAT, a LEAP record should exist.
and match to any LEAP record,
further investigation by VOCAT
found that in terms of alleged
crime involved these consisted of
Assaults (13), Sexual assaults (7),
Culpable driving causing death
(5), Threats (5) and Others (6). In
terms
of
outcomes
for
these
36
applications,
7
were
struck
out,
5
were
awarded while 24 were still pending.
The analysis demonstrates that for some serious crimes there is no LEAP
record or a record cannot be found which may indicate under-recording of
crime by Victoria Police. I would have expected that for most crime where
victims
apply
to
the
VOCAT
(and
allowing
for
old
crimes
only
recently
reported, for example, a sexual assault), a LEAP record should exist. I believe
further investigation is required to determine why no record could be located
for the 36 cases (22 per cent) identified in this exercise.
I also consider that as
part of auditing the recording of crime by police, use should be made of
VOCAT data for validation.
(iii) Comparing Victoria Police crime data with crime victim surveys
Another way of examining crime statistics produced by Victoria Police is to
compare
them
with
crime
victim
surveys.
Such
comparisons
have
been
conducted elsewhere, in NSW and internationally13, and are able to show the
extent of under-reporting of crime by the public and under recording of crime
by police themselves.
My investigation was able to conduct such a comparison using estimates of
crime victimisation from the available Victorian survey data with figures from
Victoria Police’s annual crime statistics on five types of crime.
ABS Crime and Safety and DoJ survey data on three household offences (Break
and Enter, Attempted Break and Enter/Burglary, Motor Vehicle Theft) and three
personal offences (Assault, Sexual Assault, Robbery) from 1996 to 2007 were
analysed alongside Victoria Police statistics for these offences. It should be
noted
that
because
of
small
numbers,
the
survey
estimates
relate
to
the
frequency of the victimisation incidents only, not to the proportion of these
incidents reported to police. The proportion of victims reporting the incident
to police varied depending on the type of crime.
In its 2005 survey, the ABS
found that 74 per cent of Burglaries/Break and Enter offences were reported (31
per
cent
of
Attempts),
90
per
cent
of
Motor
Vehicle
Thefts,
38
per
cent
of
Robberies and 31 per cent of Assaults.
In Figure 5, for the offences of Break and Enter/Attempted Break and Enter, the
downward trend in survey data is consistent with the trend in the number of
offences reported by Victoria Police. However, survey estimates of incidents
reported by victims in the sample appear considerably higher than police
figures and could be attributable to varying degrees to non-reporting by the
public and non-recording by police of these incidents.
13 Van Dijk, J. The World of Crime. Sage Publications, 2008; Coleman, C. and Moynihan, J.
Understanding crime data: haunted by the dark figure, Open University Press, 1996.
46
Figure 5: Comparison between crime victim survey estimates and Victoria
Police crime statistics for Burglary/Break and Enter and Attempts
For
Motor
Vehicle
Theft,
shown
in
Figure
6,
the
survey
estimates
and
the
numbers from Victoria Police correspond closely. It is likely that insurance
requirements
encourage
the
majority
of
victims
to
report
this
offence
to
police.
Figure 6: Comparison between crime victim survey estimates and Victoria
Police crime statistics for Motor Vehicle Theft
For Robbery, the trend for police statistics shown in Figure 7 appears flatter
compared to the number of victims estimated by the surveys. This can be
partially attributed to small numbers of victims and sampling errors in the
surveys and the variability in reporting rates.
Figure 7:
Comparison between crime victim survey estimates and Victoria
Police crime statistics for Robbery
Trends for Assaults (Figure 8), as estimated from the victim surveys, would
suggest
a
reduction
in
recent
years.
In
contrast,
the
number
of
crimes
recorded by police has gradually increased. This could be due in part to
improved recording by police of Assaults which occur in the context of family
violence,
where
a
code
of
practice
for
police
introduced
in
2004
had
a
significant
impact
on
both
victim
reporting
and
police
recording
of
these
incidents. The difference in levels between the survey estimates and Victoria
Police figures relates partly to the reluctance of some victims to report such
incidents to police, as well as to police practices regarding the recording of
Assaults.
48
Figure 8: Comparison between crime victim survey estimates and Victoria
Police crime statistics for Assault
Figure 9 shows Sexual Assault as reported in the crime victim surveys to be a
decreasing trend to 2002 before rising again in 2005, although small numbers
make the data less reliable. Recorded crime numbers were at their lowest level
in 2003 increasing to 2006, although remaining below survey estimates. For
Sexual
Assault
victims,
barriers
to
reporting
these
incidents
to
police
may
again impact on police figures as well as police recording practices14.
14 A further difficulty in comparing this data is that crime recorded by police is counted in the
year in which it was reported to police, whereas the surveys relate to victimisation in the
previous 12 months. There can be delays, often of more than 10 years, in reporting sexual
assaults to police.
Figure 9: Comparison between crime victim survey estimates and Victoria
Police crime statistics for Sexual Assault
The results from the comparisons between the crime victim survey data
and
the Victoria Police official crime figures for the five types of crime show that
The results from the comparisons between
the crime victim survey data and the
Victoria Police official crime figures for the
five types of crime show that with the
exception of Motor Vehicle Theft, there are
wide differences between victim survey
estimates and what is reported to and
recorded by Victoria Police.
with the exception of Motor
Vehicle Theft, there are wide
differences between victim
survey estimates and what is
reported to and recorded by
Victoria Police. While the non-
reporting of crime to police by
the public is a factor in
explaining the gap, police
recording
practices
are
also
relevant.
As
Victoria
Police
has
been
using
an
evidentiary
approach
to
recording crime up to 30 June 2008, non-recording of crime could occur if
police were not satisfied that there was sufficient evidence for the crime.
Recommendation 2
I recommend that Victoria Police use VOCAT records and crime victim
surveys to enable recorded crime statistics to be compared and validated.
50
Victoria Police response:
The Australian Bureau of Statistics runs an annual household crime and safety
survey, which collects information about individual and household experiences of
selected crime and other relevant topics such as feelings of safety or problems in the
neighbourhood. Interviews are conducted by telephone with respondents as part of an
omnibus multi purpose household survey run from the ABS labour force sample
frame.
The
survey
is
being
run
annually
from
2008-09,
with
rolling
collection
throughout the year.
However, this response does not address the issue of VOCAT records.
Compliance with rules for recording crime
The rules for recording crime should be clear and easy to follow.
They should
be regularly reviewed in the light of legislative and practice changes and they
should be effectively communicated to police to enable compliance. I found a
number of deficiencies in this regard.
In Victoria, police must determine if an event or incident that they become
aware of should be recorded on to the LEAP database as a crime. The decision
whether to record is influenced by a number of factors, which impact on the
gap between the number of crimes police became aware of and the number of
crimes actually recorded on LEAP.
Defining ‘crime’ on the LEAP database
Police are required to determine firstly whether an incident brought to their
attention meets the definition of a crime as referenced by legislation. For the
purpose of producing crime statistics, crime is categorised into 27 categories
as shown in Attachment 1.
Over
time
the
definition
of
what
is
and
is
not
a
crime
will
vary15
and,
depending
on
police
understanding
and
awareness
of
these
changes,
will
impact on recording practices. In addition, some crimes do not require LEAP
records.
Subsection 4.2 of VPM 108-3 states that offences not requiring LEAP
reports are:
15 What is defined as a crime will also vary between police jurisdictions. Although the crime
categories used in the ABS national statistics are considered comparable, except for assault
and sexual assault.
·
drunkenness and drunk and disorderly offences committed by adults where
these are the only offences alleged
·
traffic offences, except for child cautioning incident
·
any offence where an Infringement Notice has been issued and this is the only
action taken.
Although
the
public
may
report
these
incidents
to
police
or
police
may
become aware of them in other ways, they will not be recorded as a crime on
LEAP unless they are associated with other offences. For instance, between 1
July 2007 and 30 June 2008, police officers in Division 1 of Region 1 made
approximately 2,700 arrests for drunkenness, which is still a criminal offence
in Victoria. While a record of these incidents is kept and trends monitored for
‘tasking’ reasons, they are not included in Victoria Police crime statistics.
The notion of ‘crime’ for statistical purposes can also depend on which agency
has responsibility for enforcement. Thus crime which falls within one of the
Victoria
Police
27
crime
categories
(Attachment
1)
and
occurs
on
public
transport represents only 2.5 per cent of the total recorded in the Victoria
Police crime statistics (Victoria Police Crime Statistics Official release 2007-08
October
2008,
p14).
However,
the
majority
of
offences
relating
to
public
transport and many of which touch on public safety issues are dealt with by
infringement
notices
issued
by
the
Department
of
Transport.
That
Department also deals with more serious non-infringeable offences on public
transport,
such
as
Assault
or
Resist
Authorised
Officer,
Fraud
in
relation
to
Concession Ticket. The Victoria Police Crime Statistics publication shows crime
on public transport that police deal with to be relatively small.
It should be
noted that other government agencies, such as WorkSafe, the Environment
Protection Authority and the Department of Primary Industry, can also detect
and prosecute criminal offences under their respective legislation. Crime dealt
with by agencies other than Victoria Police and not brought to their notice are
not
included
in
Victorian
crime
statistics.
This
issue
is
further
discussed
below.
Changes to legislation also impact on what is a ‘crime’ and how it is classified
and recorded.
One example is the Infringements and Other Acts Amendment Act
2008.
The
purpose
of
the
Act
is
to
expand
the
infringement
system
by
allowing
some
criminal
offences
to
be
dealt
with
by
way
of
an
official
warning or penalty notice.
Some of the offences included are Shop Theft (under
$600),
Wilful
Damage
(under
$500),
Indecent/Obscene
language
and
Offensive
52
Behaviour.
Some police officers interviewed were of the understanding that
since these offences are dealt with by infringement notices, they will not be
recorded as crime’ on LEAP.
The impact of this change could be a decrease
in these minor but high volume offence categories, which contribute to the
total recorded crime figure used to determine whether crime has increased
from year to year.
Victoria Police stated:
It acknowledged the inconsistency of the issue surrounding Infringement
Notices and subsequent LEAP reports, as per VPM 108-3, and have passed
this on to the Corporate Policy team to make the appropriate amendments.
Another example relates to Aggravated Burglary, where a change in legislation
in 1997 broadened the definition, leading to an increase in Aggravated Burglary
statistics and a decrease in Burglary.
Evidentiary vs prima facie practices of recording crime
Whether police record an incident as crime on to the crime data recording
system can be based broadly on:
1.
The prima facie approach whereby police take the details of alleged
crime
at
face
value
and
record
the
crime,
prior
to
any
further
investigation - now implemented in Victoria since 1 July 2008.
2.
The evidentiary approach requires police to have sufficient evidence to
support an allegation of crime before it is recorded - used in Victoria
up to 30 June 2008.
While police must exercise discretion about whether to record an incident
under both models, the evidentiary model gives police officers a greater level
of discretion. The more closely police adhere to the evidentiary model, the
greater
the
potential
for
The
more
closely
police
adhere
to
the
inconsistencies and the wider the
evidentiary model, the greater the potential
gap between what is reported to
for
inconsistencies
and
the
wider
the
gap
police
and
what
is
recorded
as
between
what
is
reported
to
police
and
crime.
Other jurisdictions, such
what is recorded as crime.
as the UK, have found that the
use of the evidentiary recording
rules was a major source of inconsistency in crime statistics and have moved
to prima facie rules, which also better measure demand for police service.
The Australian Institute of Criminology (AIC) was commissioned by Victoria
Police to conduct a review of its crime statistics in 200116 to assess data quality
and the processes applied to counting crime, following concerns (Herald-Sun
22
November
1998;
Victorian
Parliamentary
Drugs
and
Crime
Prevention
Committee) about the accuracy of Victorian crime statistics. The report found
that there were inconsistencies in the way some crimes were recorded; in
particular that when police attended a domestic dispute or an assault, they
were less likely to lead to a LEAP record. However, it found that once crime
was recorded on the LEAP database, the policies and procedures for quality
assurance of the data were effective and the level of error in the records used
to produce the crime statistics was negligible.
The review included a validation exercise in which running sheets, listing
activities undertaken by police, were matched to LEAP records (the AIC was
not given access to calls for service data).
The review analysed 2,264 activities
generated from a sample of 153 running sheets and found that 26 per cent of
the activities resulted in the creation of a LEAP record. A similar study cited
by the AIC and conducted in Queensland (1996)17 found that about one third
of all matters coming to the attention of police were recorded as crime.
The
AIC further found that activities listed on the running sheets and not recorded
as crimes, but which might have been recorded as such, represented only 1.5
per
cent
of
the
total
activities
on
the
running
sheets.
The
AIC
report
concluded that the amount of under-recording overall is low and that the
crime statistics published by Victoria Police accurately reflect the counting
rules
and
crime
classifications
applied
to
those
incidents
recorded
on
the
LEAP
database
as
crimes.
However,
some
concerns
have
been
expressed
about
the
methodology
used
in
the
AIC
study,
including
that
the
under
recording may be much higher for specific offences, for example, assaults.
One of the recommendations made in the AIC report was that Victoria Police
should
move
from
an
evidentiary
model
to
a
prima
facie
model
of
crime
recording, as the latter is considered best practice for improving consistency
in crime recording. This recommendation was only implemented in July 2008,
under the impetus of ABS developments discussed below, and demonstrates
a general slowness by Victoria Police to move towards best practice in crime
16 Carcach, C. and Makkai, T. Review of Victoria Police Crime Statistics, Australian Institute of
Criminology Research and Public Policy Series, No. 45, 2002.
17 Criminal Justice Commission, The nature of general police work. Research Paper series, vol.3,
no. 2, Brisbane, 1996.
54
recording. On 1 July 2008 Victoria Police revised its crime recording policy
(VPM 108-3) to a prima facie model.
All police were advised on 27 June 2008
that:
Under
the
new
policy,
all
criminal
incidents
reported
to
police
must
be
recorded as one or more offences unless there is credible evidence available at
the time of reporting to suggest that a crime has not occurred.
My investigation of the way this change in policy was implemented is further
discussed below.
Inconsistencies in crime recording practices
During
my
investigation,
I
identified
inconsistencies
in
crime
recording
practices and a lack of systematic auditing at the station level to address such
inconsistencies.
The following example illustrates how some allegations of crime made by the
public
are
handled
by
police.
My
investigators
were
advised
that
at
ski
resorts, theft of ski/board equipment is frequently reported to police by the
public:
…. the figures go through the roof the Senior Sergeant says wait a week
before recording them as theft and completing the LEAP forms.
The reason for deferring the decision to record the crime reports is to see if the
equipment is found. It may have been misplaced or taken by mistake and
returned. While this practice may have been compliant with the evidentiary
approach used up to 30 June 2008, police were uncertain how these matters
should now be recorded under the new prima facie policy commencing 1 July
2008.
Another example relates to petrol drive-off’ where a customer fills their car
up
with
petrol
and
leaves
without
paying.
Should
the
theft
of
petrol
be
recorded
on
LEAP
if
the
I
identified
inconsistencies
in
crime
alleged offender says it was a
recording practices and a lack of systematic
genuine
mistake
and
was
auditing at the station level to address such
coming
back
to
pay?
It
inconsistencies.
depends,
according
to
some
police
at
what
point
police
speak to the person; yes or perhaps not, according to others. Similar examples
were cited for car theft and other thefts.
The example of family violence incidents was cited by a number of police to
illustrate how crime recording practices can be improved and made more
consistent
and
compliant
with
policy.
It
also
demonstrated
how
large
increases in assault statistics can occur because of changes in police practices.
A Victoria Police code of practice and related strategies were implemented on
31 August 2004 to improve the police response to family violence incidents
and encourage public confidence in reporting these offences to police.
Since
these initiatives, there has been a significant rise in the number of assaults
recorded
which
are