Microsimulation in a Cold Climate

1
Microsimulation in a Cold Climate

• David Bell
• University of Stirling

2
OPERA

• Older PEoples Resource Allocation model
• Addresses issues of population ageing
• Design has been reactive rather than proactive
• Partly due to funding issues
• Consequence coherence? but closely linked to
  policy process

3
Structure of Talk

• Structure of OPERA
• Applications
• Local Income Tax Burt Commission
• Indexation of Social Security Benefits Finance
  Committee
• Modelling Home Care Costs Audit Scotland
• Dementia Satellite Model Alzheimers Scotland

4
Structure of OPERA

• Population
• UK Households (Private/Non-Private?)
• Subsets region local authority?
• Main datasets
• Family Resources Survey
• (Boosted sample in Scotland)
• Survey of Personal Incomes
• What to do about communal dwellings?

5
Structure of OPERA

• Accounting Relationships
• Taxes and Benefits
• Non-behavioural
• Home Care
• Parameterised from external dataset
• Care Homes
• Calibrated from key statistics
• More data soon available

6
Structure of OPERA

• Software - Stata/Mata
• Statistics
• Distributions
• Panel
• Survival
• Directly integrate estimation results
• Graphics
• Wide range of flexible routines
• Choropleth maps
• Not currently in Stata Corp release

7
(No Transcript)
8
Example 1 Local Income Tax

• Proposal to replace council tax with local income
  tax
• Variant 1 uniform rate of local income tax
  throughout Scotland
• Variant 2 each local authority able to set its
  own local income tax
• What would be the distributional and spatial
  consequences?

9
Distributional Effects
10
Winners and Losers with Local Income Tax
11
Costs ( of Disposable Income) of Various Local
Tax Structures
12
Example 2 Projecting Consequences of Indexing
Social Security Benefits to Prices

• Rather than transit from Period 1 to Period n in
  unit time period increments, ignore the dynamics
  and reweight data based on externally projected
  control totals.
• Disaggregate geographically

13
Effect of Continued Price Indexation on Household
Poverty
14
Example 3 - Home Care CostsBased on Welsh Local
Authority Survey

• Distribution of Costs Highly Skewed
• 40 of clients account for 10 of costs
• 10 of clients account for 40 of costs
• Poses real problems for estimation, simulation

15
Modelling Home Care Costs

• Common specification of cost functions
• Log cost - log(y) Xß e
• But E(ln(y)) ? ln(E(y))
• Unbiased estimates of y difficult if e is
  heteroscedastic in x
• If ln(y) Normal (µxß, s2f(x)), then E(yx)
  exp(xß 0.5 f(x))

16
Determinants of Personal Care Costs

• Costs
• increase with disability
• decrease with age
• decrease with presence of informal carer
• unaffected by gender and ethnicity
• vary by local authority

17
Costs by Age and Gender UK as a Policy Laboratory
18
Costs by Index of Disability (Resource Need)
19
Charges by Disability
20
Model Calibration

• Estimate determinants of costs of care using
  Welsh dataset
• Estimate determinants of needing care and of
  being in receipt of local authority care using
  FRS data
• Match FRS disability classification with that
  used in Welsh survey
• Select most disabled of those receiving LA care
  in FRS sample to receive personal care match
  with proportions receiving LA personal care in
  Scotland (thus model mimics Scottish policy
  setting)
• Stochastic simulation of model to maintain
  distribution of costs rather than focus on point
  estimate
• Results weighted using FRS weights to represent
  UK/Scottish population

21
Example 4 - Simulating Dementia Costs

• Satellite Model
• Not restricted to private households
• Uses information on life expectancy, dementia
  prevalence, duration and costs
• Competing risks model
• Time aggregation to generate annual estimates

22
Simulating Dementia Costs

• Weibull hazard used to model months of life
  expectancy after age 65
• Scale and shape parameters set to approximate
  life expectancy patterns

23
Life Expectancy by Multiple Deprivation Decile
24
Prevalence of Dementia
25
Dementia Costs

• Types of what if questions?
• What if prevalence rises/falls?
• What if onset could be delayed by better medical
  interventions?
• What if cost structure changes?
• What if dementia sufferers cared for at home
  rather than in care homes?

26
Dementia by Deprivation Decile
27
Changed Individual Dementia Prevalence
28
Change in Costs Associated with Changed Prevalence
29
Modelling Dementia Care

• Model not specific to dementia could run a
  different/wider range of competing risks
• How to integrate calibrated satellite model with
  main dataset?
• How to deal with care home residents on whom
  liuttle socio-economic information available

30
Conclusions

• Calibration helps explain the obvious
• But is no more powerful than the data on which it
  is based
• Dont overplay the results
• Careful work with policy makers important,
  especially when calibration weak

31
Care Costs

• Most recent SE estimate of cost of providing FPC
  at home to pensioners in 2003-04 120m
• Model estimate 170m
• Consistent with LAs spending approx 50m prior to
  introduction of policy
• What about the personal care costs of those aged
  under 65 requiring PC?
• Model estimate 130m
• Fewer clients, higher cost per client

32
Model Results

• Weekly Costs by Age and Gender
• Household Net Income by Costs of Care
• Personal Care Costs by Disability
• Aggregate Annual Costs by Age and Gender
• Influence of Informal Carers On Costs
• UK Costs of Applying Scottish Personal Care
  Policy to Domiciliary Clients

33
Costs by Age and Gender
34
Costs by Household Net Income
35
Costs by Disability
36
Charges by Disability
37
Aggregate Costs by Gender and Age Group
38
How Does Presence of Informal Carer Influence
Local Authority Costs?

• Someone with an informal carer less likely to
  receive LA care
• Someone receiving LA care will receive less
  costly support if informal carer present
• This does not account for effects of informal
  care provision on labour market participation
• When local authority care available, informal
  carers may act as gatekeepers and/or provide
  other services