Self-funders and relative needs

Self-funders and relative needs
Project Advisory Panel
17 April 2013
Principles
• The Relative Needs Formula (RNF) for adult social care
is a mechanism that determines each local authorities’
share of total social care funding from central
Government.
• Fundamental principle: to ensure equal opportunity of
access to services and support for equal need
– Conventional interpretation: that each council should have
sufficient net funding so that they can provide:
– … an equivalent level of support (services or otherwise) to
all people in their local population
– … who would satisfy a national standard eligibility
conditions
– Net public expenditure requirement (NPER)
Net public expenditure requirement
(NPER)
• Determined by:
– Access and support test, which determines how
much support a person receives (including
nothing) depending the severity of their assessed
need.
– Funding system test: whether a person is eligible
for any public support on the basis of relevant
non-need criteria, usually concerning the person’s
financial circumstances.
External factors
• NPER will vary from LA to LA according to factors
beyond the control of the LA:
– Needs characteristics of population (Access and
support test)
– Wealth-related characteristics (funding test)
• Also,
– Supply conditions (at least in the short-term)
• RNF aims to adjust funding share to LAs to
account for NPER consequences of different need
and wealth patterns between LAs
How is the RNF determined?
• Aim: to estimate how NPER varies between areas
according to the needs and wealth characteristics of
those areas
• Estimate a formula on this basis:
– NPER = f(needs, wealth)
• Conventional method:
– Use historic LA net expenditure as measure of NPER
– Measure NPER for small area e.g. wards, or Census areas
(LSOAs)
– Measure needs and wealth characteristics for each small
area e.g. age structure, benefits uptake rates…
– Use statistical methods (multivariate regression) to
determine the relationship between NPER and external
factors
Social care reforms
• What will funding reforms mean for the RNF?
• Dilnot reforms
– Capped risk protection
• Mostly affects self-payers in care homes
– Deferred payments
• Loans from councils save people from (initially) selling their homes
– Minimum eligibility?
• Will mean a change to future access and funding
tests…
– … historic expenditure will be a poor guide
• So use different methods for currently-supported
population and newly-eligible population
– Likely to imply separate funding formulae: current, capped
risk and deferred payment
Currently-supported NPER
• For currently-supported population:
– Use conventional small area method
• Small area used: LSOA
• Councils to download (current supported) service use data
by their pre-care LSOA
• Use unit costs to determine ‘current NPER’
• Use need and wealth characteristics of pre-care address
LSOA
– Need e.g. age structure, Pensioners living alone rate, AA uptake
rate
– Wealth e.g. Pen Cred uptake rate, average house price
• Fit multivariate need and wealth models
Newly supported people following
Dilnot reforms
• Need to estimate the NPER of people that
would have been self-payers currently (prereform) but will become state supported in
the future (post reform)..
– … Dilnot NPER
• A number of methods:
– Use data on the current number of self-payers
(pre-reform)
– Simulation modelling
(1) Using current self-payers
• Dilnot NPER =
– (a) the number of self-pay recipients
– (b) the unit costs of care that councils would have to pay
– (c) and the proportion of the self-pay population that would become eligible
under Dilnot.
• Bulk of Dilnot recipients will be in residential care
• So we need a way to predict numbers of self-pay residents (SPRs) by small
area based on need and wealth
–
–
–
–
SRP in LSOA = Total beds x occupancy – supported residents
Current address of self-payers is their ‘pre-care’ address
Needs: as above
Wealth: as above, but house price not applicable because most residents will
have sold their homes.
• Instead use average price of care home: high priced care homes have higher proportion
of self-payers
• Alternatively use population surveys such as ELSA to link self-pay care
home admission with wealth and need characteristics in the survey
– But sample sizes are small.
Data sources
• CQC has care home beds by LSOA
• LA downloads = supported residents by LSOA
• Need data on:
– Occupancy rates
– Prices
– Proportions self-funded
• Use survey and Census data
– Our own survey of care homes
– Existing surveys
Issues
• For deferred payments, pre-care home address is the
responsible LA
• So we may need to collect some data on pre-care
home address for self-payers
– Need a bespoke survey data collection
• Practically no data on self-pay non-residential
– Must assume same relative need pattern in this case
• Still need a estimate of the proportion of SPR who hit
the Dilnot cap and/or seek a deferred payment
– Calculated using simulation (also see below)
• Dilnot NPER will be an approximation of actual
expenditure implications of Dilnot
(2) Simulation and profiling
• Simulation uses data on the underlying distribution of need
in the population..
• … estimate numbers and types of service users directly on
that basis
• … uses both estimated relationships (e.g. demand for care
services) and deterministic rules (e.g. means-testing rules)
– rules reflect prevailing service model and funding system
• These ‘rules’ can be altered in light of reforms
• Can be used to predict numbers of people receiving
support on the basis of the Dilnot reforms directly using
their needs and wealth characteristics
• RNF for Dilnot funding requirements synthesised directly
• Or, the proportion of SPR who hit the Dilnot cap and/or
seek a deferred payment can be simulated
Discussion points
• New RNF(s) mostly needed to allocate ‘Dilnot funding’ to
councils
– Likely to embody a different pattern of need/wealth than
current council spend
• But…
– No actual experience to go on when estimating Dilnot funding
requirements
– Generally poor data on self-payers
• Some LAs collect data on SPs but far from systematic
• So are we using a reasonable method given these data
problems?
• How could the method be improved?
• Will it be credible (enough)?
This research has been commissioned and funded by the Policy
Research Programme in the Department of Health. The views
expressed in this presentation are not necessarily those of the
department.