GIS in the analysis of small areas

1
Spatial microsimulation for urban, regional and
social policy analysis
Dimitris Ballas Centre for Computational
Geography School of Geography University of Leeds
2

Outline

• Traditional spatial modelling approaches to
  policy analysis and socio-economic impact
  assessment
• What is microsimulation?
• Spatial microsimulation for socio-economic impact
  assessment - modelling a plant closure in Leeds
• Spatial microsimulation for social policy
  analysis
• Simulating the city
• Spatial microsimulation research agenda

3
Spatial modelling approaches to socio-economic
impact assessment

• Regional Keynesian multiplier analysis
• Input-output models
• Regional Econometric Models
• Spatial Interaction Models (modelling TTW flows)

4
Spatial modelling approaches to socio-economic
impact assessment

• Which regions will suffer most? Which towns will
  be most affected?
• (Armstrong and Taylor, 19935)
• BUT
• regions and cities comprise of smaller areas,
  which differ considerably in population size,
  demographic structure, etc.

5
Spatial modelling approaches to socio-economic
impact assessment

• To move to better dynamic representations of
  urban processes suggests that individuals rather
  than groups or aggregates must form the elemental
  basis of these simulations
• (Batty, 1996261)
• Governments need to predict the outcomes of
  their actions and produce forecasts at the local
  level.
• (Openshaw, 1995 60)
• Which neighbourhoods will suffer/benefit most?
• Which households will be most affected?
• What will be the intra-region, intra-urban and
  intra-ward impact of a possible plant
  closure/development?

6
What is microsimulation?

• A technique aiming at building large scale data
  sets
• Modelling at the microscale
• A means of modelling real life events by
  simulating the characteristics and actions of the
  individual units that make up the system where
  the events occur

7
What is spatial microsimulation? An example

• sex by age by economic position (1991 UK Census
  SAS table 08)
• level of qualifications by sex (1991 UK Census
  SAS table 84)
• socio-economic group by economic position (1991
  UK Census SAS table 92)

8
What is microsimulation? An example

• p(xi ,S,A,Q,EP,SEG)
• given a set of constraints or known
  probabilities
• p(xi ,S,A,EP)
• p(xi ,Q,S)
• p(xi ,SEG,EP)

IPF-based microsimulation, CO-based
microsimulation
9
Microsimulation tenure allocation procedure
After Clarke, G. P. (1996) , Microsimulation an
introduction, in G.P. Clarke (ed.) ,
Microsimulation for Urban and Regional Policy
Analysis, Pion, London.
10
Advantages and drawbacks of microsimulation

• Advantages
• Data linkage
• Spatial flexibility
• Efficiency of storage
• Ability to update and forecast

• Drawbacks
• Difficulties in calibrating the model and
  validating the model outputs
• Large requirements of computational power

11
Microsimulating the local labour force
12
SimLeeds a spatial microsimulation model for
Leeds

• Object oriented framework
• Households or individuals can be viewed as
  objects - e.g. a Household class describes the
  features of all households (e.g. age, sex and
  marital status of head of household, employment
  status, tenure, etc.)
• Different approaches to the estimation of
  household attributes

13
SimLeeds microsimulated attributes (variables of
micro-unit)
Stage 1
Stage 2
Stage 3
14
Using SimLeeds for impact assessment - modelling
a plant closure in Leeds
15
The hypothetical plants workforce structure
16
Journey to work to Seacroft
17
Spatial distribution of SIC3 Managerial and
Technical workforce
18

Estimated spatial distribution of total income
loss in Seacroft, Halton and Whinmoor
19

Estimated spatial distribution of Job Seekers
Allowance (JSA) new recipients in Seacroft,
Halton and Whinmoor
20
Estimate the change in the demand for groceries
21

Estimated spatial distribution of change of
demand for Food non-alcoholic drinks in
Seacroft, Halton and Whinmoor
22
Modelling a change in income tax
23
Current estimated distribution of tax paid
24
Estimated spatial distribution of change in tax
paid under scenario 1
25
Estimated spatial distribution of change in tax
paid under scenario 2
26
Modelling the Income and Substitution effect

• A substitution effect making leisure more
  attractive than work
• An income effect, encouraging people to work more
  to make up the loss of income
• Different taxes have different effects, and
  affect people at different levels of income or in
  different household circumstances in different
  ways.
• (Hill and Bramley, 1986 85)

27
Towards a microsimulation-based local multiplier
impact analysis

• Multiplier effects to different localities
• Increase/Decrease of consumption of goods and
  possible changes of consumer preferences
• Further employment and income effects (caused by
  increase/decrease in consumption etc.)
• Third and fourth round local multiplier effects
  (further job/income gains/losses generated)

28
Modelling the budget changes

• Use SimLeeds to model the government budget
  changes
• Use SimLeeds to model the oppositions proposals
  for pensioners
• Use SimLeeds to formulate and evaluate new
  policies

29
Source The Guardian, 22 March 2000
30
Source The Guardian, 22 March 2000
31
Estimated spatial distribution of pensioner
couples in Leeds
32
Estimated spatial distribution of pension
increases under the Conservatives proposals.
33
Human systems modelling -RS and spatial
microsimulation for the generation of population
microdata

• S (ed,h,g, ed,h,x1,x2,,xn) (1)
• from
• M (ed,h, x1,x2,,xn) (2)
• R (ed,h,g) (3)
• where
• S estimated spatially disaggregated population
  microdata set at the house level.
• M microsimulation output - spatially
  disaggregated population microdata set
• R remotely sensed data set
• ed the Enumeration District location of each
  household
• h the housing type
• g the exact geographical co-ordinates of each
  house
• x1xn socio-economic and demographic attributes

34
(No Transcript)
35
Remotely Sensed data
Microsimulation model output
36
Microsimulation research agenda

• Combining census data and remotely census data
  for the generation of population microdata
• Analysis of electoral behaviour
• Models such as SimLeeds can be linked to Virtual
  Decision Making Environments
• Adaptive rule-based agents approaches
• Dynamic microsimulation / event modelling
• SimIreland
• SimYork, SimBritain, SimWorld!