Crime Modelling

1

Modelling CrimeA Spatial Microsimulation
Approach
Charatdao Kongmuang School of Geography University
of Leeds
Supervisors Dr. Graham Clarke, Dr. Andrew Evans,
Dr. Dimitris Ballas
2
What is Crime?

• Crime is, first of all, a legal conception,
  human behaviour punishable under the criminal
  law
• (Mannheim 1965 22)

3
Why crime?

• It is one of the most important problems facing
  the UK today.
• Adds stress to people lives and impairs the
  quality of life of individuals and communities.

4
Study Of Crime
5
Geography of Crime

• Crime Mapping
• Spatial patterns of crime
• Ecological Analysis
• relationship between crime and socio-economic /
  environmental factors
• Spatial Analysis- using GIS
• Hot spot areas

6
Microsimulation

• A methodology aimed at building large-scale
  datasets on the attributes of individual units
  and analysing policy impacts on these micro
  units.
• 
  (Clarke, 1996)

7
Why Spatial Microsimulation?

• Criminal behaviour is related to current
  attributes of individuals.
• Can be used to conduct policy simulations and
  forecasting.
• Can generate spatial outcomes at a detailed level
  of resolution.
• It has not yet been applied to study
• crime.

8
Advantages of Spatial Microsimulation

• Data linkage ability
• Spatial flexibility
• Efficiency of storage
• Ability to update and forecast
• 
  (Clarke, 1996)

9
Drawbacks

• The difficulty to validating the model
• outputs
• Large requirements of computational
• power
• (Clarke, 1996)

10
Objectives

• Build a spatial microsimulation model
• for crime
• Use this model for forecasting crime
• - The effect on crime rates
• - What types of area tend to have high
  
• crime rates?
• - Estimate individuals propensity to
• commit crime and to be a victim.

11
Methodology

• 1. Construct a population microdata set.
• - A list of individuals along with associated
  attributes on the basis of Census and Survey data
  (e.g. British Crime Survey)
• - Conditional probabilities, calculated from
  available known data, will be used to reconstruct
  detailed micro-level populations.
• 2. Create the sample of individuals based on set
  of probabilities
• 3. Simulate
• Simulation of crime on the basis of
  individual propensities to commit crime
• 4. Validate
• Compare simulation outputs with actual data
• (e.g. from West Yorkshire Police)

12
Low Socio-Economic Status
13
Crime Data

• The official statistics do not represent the
  total crime.
• Only 27 of the total offences are recorded
  
• by the police (Home Office, 1995).
  
  
  
• 
  

Reported crime
Unreported crime
14
Sources of Data
15
Types of Crime

• Robbery
• Burglary
• - Burglary Dwelling
• - Burglary Other
• Vehicle Crime
• Theft
• Criminal Damage

16
Crime in Leeds

• In West Yorkshire, 40.9 of all crime committed
  takes place in Leeds
• Crime Rate 2000/2001
• Crimes/1000 pop.
• Leeds 146
• West Yorkshire 124
• England 102
• (Leeds Community Safety, 2001)
• Burglary and vehicle crime are the highest crimes
  in Leeds.

17
Offenders in Leeds

• Predominantly male, white
• 56 are unemployed
• Offender characteristics are related to
• drug, alcohol, financial problems, and
• unemployment
• 
  (Leeds Community
  Safety, 2001)

18
Victims in Leeds

• The most common age
• 30-39 (1999-2000)
• over 40 (2000-2001)
• The number of older people experiencing
• crime has been increased.
• Victims over 40 are most likely to be victims
• of burglary, criminal damage, theft, and
• vehicle crime.
• 
  (Leeds
  Community Safety, 2001)

19
Headingley
University
City and Holbeck
Burmantofts
20
(No Transcript)
21
(No Transcript)
22
Headingley
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
Sawasdee(sa-wat-dee)
33
Leeds ward