Actuaries were naturally excited by the final report of the Dilnot Commission on Funding of Care and Support. It is rare that major public policy proposals are so clearly built around observations on the nature of risk and uncertainty.
Many actuaries quickly recognised the Commission’s ‘capped cost model’ as a variant of the ‘limited liability model’, which has featured in debate on long-term care funding for many years.
The limited liability model in context
The core idea of the limited liability model is that the state should pick up the ‘tail-risk’ costs associated with needing care. This is because private insurers find it so difficult to offer pre-funded insurance for care in the face of inevitable uncertainty around whether individuals will need care and how long they will need it for.
Problems in pricing such risk stem from the presence of so many unknowable factors that could determine care-related liabilities — for example, the potential for new drugs to radically extend life expectancy for individuals with Alzheimer’s disease.
Previously, companies offering pre-funded care insurance sought to deal with these issues in a range of — ultimately unsatisfactory — ways. Many products in the US market impose limits on total pay-outs, capping the liability of insurers but thereby leaving customers exposed.
Is it true that the private sector can’t handle the tail-risk for care costs?
There can be no doubt that pricing disability and longevity risk years into the future is enormously difficult for insurers. But two points deserve to be made.
First, providers of immediate needs annuities have long pointed out that, at the point-of-need, it is much easier to price care-related longevity risk. The nascent market in immediate needs annuities for residential care costs is testament to this.
Second, novel proposals have been developed for offering pre-funded insurance to the population, underwritten by the private sector. In 2008, the current author proposed a National Care Fund for long-term care, updating the model in 2011 by exploring how it could be delivered through a public-private partnership.
In this more recent version, individuals would make pre-funded contributions to a National Care Fund, which would use the profits from an investment fund to purchase annuities on the bulk annuity market for individuals with care needs claiming on the Fund. In this way, households would effectively have access to pre-funded care insurance, but the private sector would only take on those care-related risks it is equipped to handle: annuities priced at the point of need.
The Dilnot Commission’s interpretation of the limited liability model
The Dilnot Commission’s ‘capped cost model’ proposes that individuals should have their costs capped at £35,000, and that the means-test threshold for ‘assessable wealth’ for those in residential care should be increased to £100,000.
The increase in the wealth threshold for residential care to £100,000 has caught the eye of many stakeholders as a quick fix for giving households certainty that, even under extreme scenarios, their family would be guaranteed to inherit a fairly sizable chunk of wealth. It is widely tipped as the funding reform most likely to appear in the 2012 social care White Paper.
But what is the prognosis for the £35,000 cap on individual liabilities?
In the detail of the ‘capped cost’ model, there is an important feature that many commentators and stakeholders have missed — the individual liability capped at £35,000 is determined using local authority needs assessments.
In recent years, as local councils have been pressed to offer care users personal budgets and cash-based direct payments, they have been compelled to develop new resource allocation systems (RASs) to enable them to allot a financial value to the support they will provide to someone, whether as a direct payment or a directly provided service. In this way, councils increasingly — and this is still a work in progress — allocate all individuals entitled to support a weekly sum, whatever the form that support takes.
The principal intellectual innovation of the Dilnot Commission was to be first to spot that RAS mechanisms could also be used as a way of determining individual liability for those not receiving public support — and in this way, produce a new version of the limited liability model.
Previously, proposals based on the limited liability model have typically been built around narrow assessments of disability, with a financial value allotted to each unit of disability.
However, under the Commission’s ‘capped cost’ model, a council would assess someone’s needs and determine the financial value of support they should be given according to the council’s criteria. If someone falls below the threshold of the means-test, they will receive this amount as actual support. If someone falls above the threshold, the council will record how much they would have received but for them being too wealthy. When the accumulated value of these notional, recorded packages of support reaches £35,000, the council will then begin paying actual support, and the amounts in question will cease to be notional.
The ‘capped cost model’ could therefore more accurately be called the ‘capped exclusion from means-tested support’ model. It records how much individuals would have received from their council but for being too wealthy, and when the amount of support that individuals are excluded from totals £35,000, they are effectively reassessed on a ‘means-blind’ basis.
What will the ‘capped cost’ model mean for private insurance products?
The key aspect of the ‘capped cost’ model for private insurers considering products that would provide protection against the £35,000 liability is that it is local authority needs-assessments — not just a person’s disability and longevity — that would determine this liability. In addition to assessments of (predictable) disability, these assessments take account of the availability and receipt of care from family members, and also vary according to individual council decisions on levels of support for different types of need.
The implications of these features of council needs-assessments for insurers are absolutely crucial. Insurers pricing pre-funded long-term care insurance can price products for trends in expected disability and expected longevity. However, a person’s liability in the ‘capped cost model’ — ie. when they reach the £35,000 threshold — is determined not just by their disability and longevity, but by the availability and receipt of informal (family) care, and what their council happens to provide to individuals with different levels and type of need; for example, taking account of the unit cost of different services in a particular area.
There are sound reasons for why councils use these factors in allocating resources to individuals in the local population with care needs. Ignoring the availability and receipt of informal care would see individuals receiving support to purchase formal care despite being able to rely on family members — ultimately, an inefficient way for the state to spend on care and support. Local decisions on levels of support mean that councils can look at the availability and cost of different types of services in local care markets, and allocate public support more efficiently in light of these facts.
But the result is that the £35,000 liability in the ‘capped cost model’, determined as it is by local authority needs-assessments, is — strictly speaking — uninsurable. For example, it would be impossible for an insurance company to offer insurance that would make regular pay-outs until a person reached the £35,000 threshold, since insurers will have no way of modelling when that will occur.
This will have important implications for the protection against the £35,000 liability that insurers could offer under the ‘capped cost’ model, whether in the form of pre-funded insurance or disability-linked annuities.
The most likely complementary product that could be brought to market would be akin to critical illness insurance — a lump-sum £35,000 pay-out upon experiencing a defined level of disability, which may or may not be when a council would class a person as needing support. But there would nevertheless be potential for a number of peculiar outcomes that would leave the customer confused and frustrated.
For example, if someone claimed on such insurance and received their £35,000 cheque, in the eyes of their council, they may not yet have any ‘assessable need’ on account of the availability of informal care. It would then be up to individuals to choose whether and how quickly to spend down their £35,000 sum, given their meter of ‘notional support’ would not be increasing at the same rate.
So although the nature of the individual liability in the ‘capped cost’ model does not preclude pre-funded insurance products (or disability-linked annuities), the detail of the model will affect the form of those products and the experience of consumers. These awkward outcomes are perhaps inevitable when two different mechanisms for allocating resources — private insurance and local authority needs-assessments — come together.
How this would affect the take-up of private insurance under the ‘capped cost’ model remains to be seen. Take-up rates in the leading overseas markets — the USA and France — have only ever edged up to around 10% and 15% respectively. With the unique factors that inhibit the English market — not least the operation of ‘free’ personal care in Scotland — the pre-funded long-term care insurance market could well remain defunct even following any implementation of the ‘capped cost’ model.