Transcript Local Financial Benefits Modelling SERIG

Supporting People Local
Financial Benefits Model
South East London
Tom Ashton
17 December 2009
Objectives / expectations
The objectives of this session are to:
• Explain the Supporting People National Benefits Model and the work
we’ve carried out to
•
update it
•
build a local model
•
pilot that local benefits model
• Explain the context for the local model
• Demonstrate the local model
• Explain how the user guide works
• Answer any questions
Are there any other expectations we need to meet
in this session?
Agenda / Contents
• Introduction to / re-cap of the Supporting People National
Benefits Model
<15 mins>
• Detail of the recent project
• Updating the work
• Adding new client groups
• Developing a local model
<10 mins>
• The local model
• Context & where to find it
• Demonstration and user guide
<10 mins>
• Expectation review & questions
<10 mins>
• Our contact details
Introduction to / recap of
the Supporting People
Financial Benefits Model
We start by explaining how the
national model works...
The national model, for each of a set of client groups,
compares the cost of:
• existing arrangements (involving Supporting People)
• an alternative provision scenario.
The alternative provision scenario for each client group
was built on the assumption that, in a world without SP,
current clients would get a mix of:
• existing arrangements, with the SP-funded element
removed
• alternative residential care arrangements (e.g.
nursing care, psychiatric care, rehab etc.)
...it includes the use of two kinds
of costing
Costs of existing arrangements and the alternative
scenario include
‘Package costs’: These are the costs associated with
providing support (e.g. SP services, Housing costs,
Living costs, Social services costs, Benefits
administration)
‘Event costs’: These are the costs associated with
events that happen to clients (either positive, planned
interventions or adverse events e.g. Health
interventions, Costs associated with committing or
become a victim of crime, Costs associated with
becoming homeless (rough sleeping, tenancy failure
etc.)).
National Model Example:
Learning Disabilities (1)
For the 31,238 SP supported households containing people with learning disabilities,
the two scenarios were:
The existing arrangement:
100% (31,238): existing arrangement
Package cost (per household unit):
£39,010
Event cost (per household unit):
£2,327
Total cost (per household unit):
£41,337 x 31,238 = £1.29bn
An alternative arrangement:
35% (10,933): existing
without SP
65% (20,305): residential care
Package cost (per household unit):
£79,133
£27,185
Event cost (per household unit):
£2,327
£4,697
Total cost (per household unit):
£81,460
£31,882
x 20,305 = £1.65bn
+
x 10,933 = £349m
=£2.00bn
National Model Example:
Learning Disabilities (2)
The net financial benefit is £711m:
£2.00bn - £1.29bn = £711m.
100% (31,238): existing arrangement
Package cost (per household unit):
£39,010
Event cost (per household unit):
£2,327
Total cost (per household unit):
£41,337 x 31,238 = £1.29bn
An alternative arrangement:
35% (10,933): existing
without SP
65% (20,305): residential care
Package cost (per household unit):
£79,133
£27,185
Event cost (per household unit):
£2,327
£4,697
Total cost (per household unit):
£81,460
£31,882
x 20,305 = £1.65bn
+
x 10,933 = £349m
=£2.00bn
National Model Example:
Learning Disabilities (3)
Event costs: the point to note is that event costs are much higher for people who
don’t receive SP support or residential care. Some key events drive this
situation…
Event type
Cost per annum
(with SP or
residential care)
Being admitted to hospital due to general health issue
Cost per annum
(without SP or
residential care)
Additional cost
(without SP or
residential care)
£1,537
£3,236
£1,699
£351
£369
£18
£23
£47
£25
£237
£249
£12
£82
£87
£5
£8
£8
£0
Being a victim of burglary
£31
£32
£2
Being a victim of street crime (violent crime or mugging)
£59
£62
£3
Becoming a victim of homelessness
£0
£178
£178
Receiving home care provision
£0
£428
£428
£2,327
£4,697
£2,370
Visiting an A&E department
Being admitted to an acute mental health ward
Visiting a GP due to general health issue
Visiting a community health service (not mental)
Being visited by a community mental health nurse
Total
Similar work was done for other client
groups, producing a total benefit
Client group
Cost (£m)
Net financial benefit (£m)
People with Alcohol Problems
(20.7)
92.0
Women at Risk of Domestic Violence
(68.8)
186.9
People with Drug Problems
(30.1)
157.8
Homeless Families with support needs - Settled Accommodation
(32.5)
(0.5)
Homeless Families with support needs - Temporary Accommodation
(17.5)
28.5
Single Homeless with support needs - Settled Accommodation
(130.1)
30.7
Single Homeless with support needs - Temporary Accommodation
(106.7)
97.0
People with Learning Disabilities
(369.4)
711.3
People with Mental Health Problems
(254.4)
559.7
(55.4)
40.3
(198.2)
646.9
Older People in Very sheltered Accommodation
(32.4)
123.4
Older People receiving Floating Support and Other older people
(97.3)
628.0
People with a Physical or Sensory Disability
(28.4)
73.3
Teenage Parents
(24.9)
(18.3)
Young People at Risk - Settled Accommodation
(94.9)
26.6
Young People at Risk - Temporary Accommodation
(38.1)
26.7
Young People Leaving Care
(12.7)
(0.7)
(1,612.4)
3,409.4
Offenders or People at risk of Offending, & Mentally Disordered Offenders
Older People in Sheltered Accommodation
Total
Remember that the client groups are of different sizes
Note also that coverage doesn’t include all client groups
The recent work with CLG
(Feb – June 2009)
We carried out three tasks in our recent
work with CLG
•Updating the data in the national model
•Adding new client groups
•Converting to a local model
The National Updating was primarily a
matter of housekeeping
• Every data source that was cross-referenced in the original
model was checked, and was updated if the source had been
updated
• The cost indexation (which uses the RPI) was adjusted to bring
all costs forward to 2009 levels
• Other minor adjustments were made, including an adjustment to
our treatment of the groups of older people to group those
categorised as “other” (primarily people receiving community
alarms or Home Improvement Agency services)
• We adjusted our estimates of package costs, since discussion
with Hilary Bartle – and independent consultant who had
previously been deputy programme director – suggested that
they had been under-estimated in the original work
The updating of the national model drove modelled benefits up from
£2.77bn to £3.41bn. A number of factors drove this, although a key one was
inflation in the cost of escalated interventions (residential care etc)
We added the following client groups
•People with alcohol problems
•Teenage parents
•Young people leaving care
•People with physical or sensory disabilities
Together with the client groups already covered,
described in the previous section, this coverage to 93%
of Supporting People spend nationally.
Some client groups remain un-modelled – these are
the smaller groups: People with HIV/AIDS, Rough
sleepers, Refugees, Travellers & a generic client group
The Local Authority work had three
key strands
• The three strands were
1. Engaging through stakeholder meetings in London
2. Engaging through regional group meetings
3. Piloting
• Engaging through a stakeholder meeting in London (31
March) produced agreement on the best approach to a local
model, and follow up meetings were held in May
• Regional group meetings were also used to inform people of
progress
• The model was piloted in early May with Hampshire County
Council, Middlesbrough Council & Leicester City Council
We are now returning to Regional Groups to publicise the
model
These figures are based on 2004/5 SPLS data
We considered a number of options, and
selected one of them with local
authorities
Option 1
Option 2
Option 3
Option 4
Client numbers
Client numbers
Client numbers
Client numbers
Package costs
Package costs
Package costs
Alternative scenario
composition
Alternative scenario
composition
Event reductions from
SP
Data that can be
edited in local model
Using one option has the advantage that it produces
consistency of output. This is our preferred option; it allows for
local variation but avoids asking people to make judgements
that may be difficult to make at local level.
Option 2 is what has been implemented.
The Local Model
The Supporting People Financial Benefits
Model is published on CLG’s website
There are two links:
A report on the national model can be found at
http://www.communities.gov.uk/publications/housing/financialbenefitsresear
ch
The local modelling work – together with a user guide can be
found at
http://www.communities.gov.uk/publications/housing/financialbenefitsguide
The first point of context is that local modelling
has “offensive” and “defensive” purposes
•The “offensive” purpose is to allow Authorities to see
what they are gaining from using Supporting People
services
•The “defensive” purpose is to allow them to defend
these costs to other stakeholders
•To some extent there has to be an assumption that
benefits locally are similar to benefits for the national
picture
The exact level of benefits that Authorities gain from
Supporting People services will depend on their efficiency
& their quality of delivery. It is not in the scope of the
modelling to support judgements about that.
In terms of “offensive” modelling
the context is as follows
20
We will demonstrate the local model
•We will imagine ourselves to be representatives of
Birmingham City Council – and we will edit relevant
parts of the model
The documentation for the local
model is similar to the pre-existing
documentation
The model documentation breaks down into six key
sections:
5 pages that explain in outline how the
model works
1. Summary
2. Model purpose and logical structure
3. Quick start guide
4. The spreadsheet
10 pages that explain how to use the
5. Glossary
6. Technical appendix
Glossary that includes detail of how
the modelled client groups are
constructed
model
20 pages that explain how the model
works in full
National documentation was also
updated for CLG to use
Questions
Expectation review
AOB
Our contact information
Contact details
Hilary Bartle
Peter Becque
Tom Ashton
Independent Consultant,
Hilary Bartle Consultancy
[email protected]
Operational Research
Consultants, Capgemini
[email protected]
http://www.uk.capgemini.co
m/services/consulting/or
07843 412 528
0870 904 4007
[email protected]
0870 904 5147
Claire
Hempenstall
[email protected]
0870 195 1417