| With Respect to Old Age: Long Term Care - Rights and Responsibilities | ||||
|
Chapter 2 The costs of Long-Term Care Now and in the Future This chapter looks at the current expenditure on long-term care services for elderly people, the numbers using such services, their sources of income and their assets. A baseline is established for the Commission's projections on the future demand and costs of long-term care. The chapter concludes by looking at what might happen within the range of doubt around these projections.
2 In considering how care can be afforded in the future, we first need to examine the current demand for and the costs of long-term care for older people in the United Kingdom. This can mean either long-term care services which individuals currently pay for themselves privately, contributions towards the cost of state services or the money which the state pays on their behalf. We do not include in this calculation anything for the costs of unpaid care provided by family, relatives and friends. We explain why later on in this chapter. Having set out the costs of long term care now we go on to:
2.1 It is important to be clear that we are making projections to use as yardsticks for comparisons. They are not predictions or forecasts about the future. This would require forecasting both the future policies of current and future Governments and changes over the coming decade in public expectations and preferences. A projection takes the best information known about a subject, makes plausible assumptions, often based on past experience, about how different the future might be from the present and then takes the figures forward to show what the future might hold. Because the future is unpredictable one single projection is not enough. A number of projections are made on different assumptions to give a range in which the future might be found. This gives us a baseline from which we can cost the effect of policy changes. 2.2 We have used a model for future costings developed for the Department of Health by the Personal Social Services Research Unit of the London School of Economics at the University of Kent.(2) This uses the best information available for the United Kingdom in respect of the underlying factors which will influence future demand and costs. 2.3 Our terms of reference asked us to look at the numbers of people requiring various forms of long-term care through the first half of the next century up to 2051. However, any such projections become ever more uncertain the further away we move from the present. In looking at pensions, for example, the Government used a model that projected pensioner's income to the early part of the next century. The differences in the possible effects of small changes each year soon mount up because of compounding. Projections made for 10 years ahead suffer less from this than projections made for 20 or 30 years ahead; projections for 50 years' time are therefore very uncertain indeed. They suffer from a lack of reliable and consistent data which has dogged our work from the very beginning. While in this chapter we illustrate some projections to 2051 as best we can, we consider that they are not reliable beyond about 20 years. Our recommendations as to how long term care should be paid for are intended to apply within this timescale. 2.4 Later in this Report we make a specific recommendation about a National Care Commission for older people. We want to say straightaway that one of the most important functions of this body must be to keep the projections up to date, where possible improving them, and recommending changes to Government if these seem necessary in the light of future developments. This is important to the future dignity and financial security of older people in generations to come. The National Care Commission we propose should be made responsible for making and publishing projections about the overall cost of long-term care at least every five years [Recommendation 2.1].
CURRENT NUMBERS OF PEOPLE AND EXPENDITURE ON 2.5 Table 2.1 summarises information about the numbers people receiving long-term care services by type of service. About 600,000 people over the age of 65 are getting home care from a local authority. About 480,000 older people are in care homes - that is about 1 in 20 of all elderly people. More detail is available in Research Volume 1. Table 2.1: Number of people in the UK receiving long term care services by type of service and funding.
2.6 We have had to build from scratch a picture of the current total expenditure on long term care as there is no one central source of for the figures. Government figures on health expenditure for people with long-term care needs are included among the figures for all health service expenditure. We have also had to find a way of estimating what people spend on long-term care from their own resources. Our starting point has been the use of a range of long-term care services for older people with disabilities that prevent them from looking after themselves or from doing household tasks like cooking or shopping. These services include home helps, day care, meals on wheels, NHS community nursing services, NHS long-stay beds, residential care and nursing homes. This information can be found from Government statistics and sample surveys of the population like the General Household Surveys for Great Britain. Information on the costs of these services comes from work done by the Personal Social Services Unit at the University of Kent.(3) Costs and use when multiplied together give the total costs shown in 2.7 In Table 2.2 we set out our estimate of long-term care costs in the United Kingdom in 1995 (the latest year for which all figures are available) as £11.1bn. This includes expenditure by the NHS and social services, but not GP services, nor housing or leisure services relevant to community care or unpaid care. It includes the "accommodation" costs of hospitals and care homes but not the living expenditure or rent of people living in ordinary or sheltered housing. It does not include social security expenditure attributable to people with long-term care needs - this is discussed below. About £7.1bn of this £11.1bn figure is paid for by the state directly via the NHS and social services. Older people themselves pay about £4bn. This element of the estimate is subject to considerable uncertainty because there are no reliable statistics collected about private expenditure on long-term care. Table 2.2: Current expenditure on long-term care services 2.8 About £8.3bn of the £11bn total is spent in residential care and nursing homes and the remaining £2.7bn on home care. These figures are broadly in line with those provided independently by industry experts.(4) 2.9 The Department of Social Security and the Department of Health and Social Services for Northern Ireland(5) have provided us with figures which show how much social security expenditure in the UK is attributable to people with long term care needs. In 1995/96 (our base year) this was about £6bn. Some of this expenditure - such as the continuing expenditure on higher levels of Income Support for people with preserved rights in residential care and nursing homes is taken account of directly in the £7.1bn of state expenditure referred to above. The remaining expenditure is taken into account in our figures as private expenditure and appears to the extent that it represents spending from Social Security benefits or other income by individuals on items, such as charges for local authority home helps, or the purchase of private domestic help or payment of fees to care homes. THE INCOME AND ASSETS OF OLDER PEOPLE 2.10 In our terms of reference we were asked to have regard to the income and assets of older people. These indicate the extent to which older people are capable of making contributions now and in the future to the costs of their long-term care from their income and assets 2.11 An analysis of inequality in the UK(6) places the income of older people in perspective with other age groups. It found that nearly two thirds of those over 70 are among the poorest 40% of the population, and only half as likely as the average of other age groups to be among the richest 40% of the population. In its analysis of older people's income and wealth, Swiss Re(7) concentrated on looking at income patterns in those over 55 years of age and acknowledged the difficulties faced by many in funding long-term care from income and savings. It reported average weekly income of those between the ages of 55 - 69 to be £92.37, with only 3% of that age group having a net weekly income of £275.00 or over. Excluding the value of their home, 57% of those aged 55 - 69 have savings/investments less than £3000.00 with only 8% having savings/investments over £30,000. 2.12 For the Commission, the DSS analysed income and asset data of single older people aged 75 and over from the 1995 - 96 Family Resources Survey(8) linking it with council tax banding as a proxy for housing equity. Table 2.3 is illustrative of the proportion of the levels of assets held by the population of single people aged 75 and over in private households and their incomes. It also gives a broad indication of the proportion of this age group that may, depending on income, benefit from the current state financed residential or nursing home care, and possible changes in the upper asset limit of present means test rules (see Chapters 4and 6). It should be stressed, these figures are estimates. Table 2.3: Assets and Incomes of older people: Distribution of assets, and approximate mean and median incomes of single people aged 75 or over in private households by asset band 2.13 The figures in the Government's recent pension review(9) emphasise these disparities and indicate that since 1981 the gap between the richest and poorest pensioners has grown; the poorest 20% of single pensioners having an average weekly income of £68.00 with the richest 20% having £205. Over the next 25 years the review expects that, on average, the value of pensions paid on retirement will increase in real terms, and for the upper quartile by as much as the rise in earnings. However, it notes that this growth in pensioners' income will not be shared equally. 2.14 Our first projection - the "base case" - contains a set of reasonable and plausible assumptions about what might happen in the future to factors for example, the number of older people, over which there is little or no influence or control - even if the Government could make policy changes to influence the future. After that we change those assumptions to test the range of possibilities with a "sensitivity analysis". The future will lie somewhere in the range of our projections. This range of uncertainty is known as the "funnel of doubt". 2.15 We know that a population with more older people would be likely to need a greater volume of long-term care. In the light of increased demand we have projected what the overall health and social care costs would be, the cost to the state and to individuals. Projected costs are shown as a percentage of national income (gross domestic product - GDP). In Research Volume 1 we set out the projections in much greater detail. 2.16 The PSSRU model which forms the basis of our work looks at the receipt of long-term care and incorporates measures of all the factors affecting the demand for, use of and costs of long-term care that we describe below. Details of the model have been published by PSSRU.(2) All the results we describe here are the Royal Commission's own projections using underlying data from published sources which are incorporated in the PSSRU model. This model differs from those developed by the Institute of Actuaries and London Economics (10)(11) in that it does not attach a financial cost to unpaid care. WHAT AFFECTS THE DEMAND FOR AND COST OF LONG-TERM CARE? 2.17 This is a complex area discussed in detail in Research Volume 1. Here we outline the five most important factors which affect the future demand for and costs of long-term care. 2.18 Figures 1and 2 show what has happened and what is projected to happen to the population of the United Kingdom as a whole since 1901. The population has been growing since the turn of the century. There has been a long-term upward trend in the number of people aged 65 and over as life expectancy has improved considerably. From the turn of the century the number of older people has increased by 400%. Since 1931 the number of older people has doubled. The overall trend established is projected to continue until about 2030 when the population stops growing, as a result of past falls in birthrates. In a sense the UK has already lived through its demographic "time bomb" earlier in this century. The future is much more manageable. 2.19 By 2050 the biggest relative increase in older people is expected in the number of those aged 85 and over - "the oldest old". They will be three times more numerous in 2050 than now. The United Kingdom is not alone among the developed countries in expecting such increases in numbers, but is among the group of countries facing a smaller increase in numbers. Research Volume 1 Chapter 6 gives more information on the international perspective. 2.20 In addition to the birth rate, the future size and composition of the population is the result of great strides made in the UK in improving the health of individuals, public health generally and material standards of life. After the post-war and 1960s "baby-booms" the birth rate fell, so the proportion of working to retired people will change substantially after the first quarter of the next century. However, higher productivity from the workforce would make it possible to support a higher ratio of non-working to working people and to do this even if older people's need for help increases. 2.21 The health of older people is a key determinant of the need for long-term care. If the larger number of older people are living longer but:
2.22 Health expectancy measures the numbers of years of life that will be free of chronic illness (which limits a person's ability to carry out tasks like normal housework or shopping) or severe disability (which prevents a person from looking after their personal care needs such as feeding, going to the toilet, bathing or which inhibits their mobility). Any model which projects the future need or costs of long-term care must therefore incorporate a measure of "health expectancy". Our model reflects this by looking at measures of dependency - by using information on activities of daily living; including the ability to carry out domestic tasks and well as personal care tasks. 2.23 A worldwide debate has been taking place about health expectancy trends among older people. The best evidence we can find about the United Kingdom suggests that the factors which are causing us to live longer are also resulting in the extra years of life being free from severe disability. Research from the USA tends to support the view of a general improvement in health expectancy. There is reason for optimism here but we sound a note of caution. 2.24 Our caution concerns the data on which the optimism is based. The data came from the General Household Survey which asks people to report health problems or disabilities. This gives a snapshot of people's health at any one time. While this data can be used to monitor trends in health expectancy, a better way would be by following directly a representative sample of people over time and asking them about their health and how it has changed. But no such longitudinal survey exists for the UK. Had it existed it would have increased our confidence about the trends we assume in health expectancy. Such a survey could also produce other related data, which would be useful at national and local Government levels and in the financial services industry. Getting such data involves committing effort and resources. We have already recommended that our projections should be updated every five years. We recommend also that to improve the information available to the National Care Commission the Government gives urgent consideration to setting up a national longitudinal survey to provide data to reliably monitor trends in health expectancy [Recommendation 2.2]. The Supply of Unpaid Care from Families, Relatives and Friends 2.25 Any changes to the supply of unpaid care could have important effects on the demand for paid care and thus the cost to individuals or the state. By unpaid care we mean care and support from family, relatives and friends which in terms of hours (and perhaps in terms of monetary value) far exceeds what the state and individuals provide by way of paid for care. It is sometimes assumed that Government can influence the supply of unpaid care by taking certain measures to ensure that families and relatives care for their older people. However, in a free society a Government can do little, short of compulsion or other draconian measures, to influence the way in which families or relatives decide whether to care for their older members. Later on we say what we think will happen to the supply of unpaid care based on changes in society and on past trends.
2.26 There are about 5.7 million people providing some hours of informal care,(12) most of whom will be caring for older people. Most carers spend no more than about 2.27 Table 2.4 shows the characteristics of people providing informal care. More women than men provide informal care. People aged between 45 and 64 comprise the single largest group of unpaid carers. The largest group providing unpaid care are those providing help to their parents or their parents-in-law. The General Household Survey shows that of all those providing unpaid care, spouses are the most important source of help with domestic tasks and personal care. Table 2.4: Characteristics of people providing informal care, GB 1995 2.28 Whether there will be a reduction in the supply of unpaid care in the future is one of the most difficult questions we have been asked to consider. This involves looking very carefully at the different generations and how they have experienced family and working life, but also looking beyond the supply of long-term care to looking at evidence about what happens when older people ask their families to help. There is genuine concern about the effects on the supply of unpaid care because of changes in family structure brought about by falls in birth rates, higher divorce rates, re-marriage, greater family mobility and less living together of families across generations. We have considered evidence from researchers and in particular from the Carers National Association which contained new research from London Economics. 2.29 Older women now are less likely to have spouses. The immediate next generation of older people are more likely to have been married and have children who could care for them and are also likely to be married when the need for long-term care arises. However, for subsequent generations a greater number of older people may have no spouse or partner to care for them because of changes in marriage patterns, including cohabitation. This has been taken account of in the model. 2.30 An analysis of the 1994 General Household Survey (13) shows a very strong link between the availability of unpaid care from families, relatives and friends and the actual receipt of unpaid care. This suggests that the demand for unpaid care cannot be looked at in isolation from the supply of it and that where older people need care and have spouses or children in the same household or nearby, they generally receive care from them. 2.31 We can see no evidence of working women being less willing to provide unpaid care [see Research Volumes 1and 3], although we are acutely aware of the personal costs to them in both time and money of providing care in this way. We address the needs of carers in Chapter 8. 2.32 Research from London Economics for Carers National Association(14) using a new source of information - the British Household Panel Survey, provides some interesting, but - as the authors themselves acknowledge - far from complete insights into patterns of unpaid care and the tendency of people to care for their elders. The data shows that while about 14 - 15% of the population are providing unpaid care at any one time, there is a remarkable turnover of people providing such care, but the overall supply of care has not gone down. 2.33 The research looked at the preparedness and likelihood of various groups of people to provide unpaid care, including people who had been unemployed in the previous year but were now working, and also at carers not living with an older person as the carers's wealth and mobility increased. It found that unemployed people were less likely to provide unpaid care, and that carers not living with an older person were more likely to provide care if they were better off. Overall London Economics concluded that the supply of informal care would decrease over time but adduced no evidence to support this conclusion. However, we would expect that as fewer people live with their older relatives, care by "extra-resident" carers will be more important. Of course, the type of care given might change - for example more care at a distance, and less hands on personal care. 2.34 Table 2.5 shows the relationship between levels of dependency and receipt of informal care drawn from an analysis of the 1994 General Household Survey. At first sight it might seem strange that nearly half (46%) of all older people without any level of dependency receive support from an unpaid carer. This is because unpaid carers of older people are more likely to be involved in informal helping than in heavy duty caring, and also because of the way the General Household Survey records help given to all older people whether they are disabled in any way or not. Such help is generally the product of a two way relationship within the family but it is a very different kind of help from intensive support given on account of disability. Table 2.5: Dependency and receipt of informal care: England 1994 2.35 There is a measure of uncertainty which surrounds any projections about the availability of informal care. The model takes account of a decline in the number of older people who are likely to be married (except that for very elderly men who are more likely to be married). Apart from these changes (the effect of which is minimal - see Research Volume 1), changes in household type and basic demographic changes make it difficult to make assumptions about the future availability of informal care from other members of the family apart from partners. Therefore, in our base case we conclude that after allowing for the factors described in this paragraph there will be no real change in the future availability of informal care. 2.36 The use of services is clearly important in our model. Our projections are largely based on the volume of care services currently used, and assume that the volume of services provided to each older person will be broadly the same in the future as in the past for any level of disability. However, use of services is and has been determined by political decisions and is influenced every day by personal preferences, and we simply do not know how services will respond to demand in 20, 30 or 50 years time. We have however, modelled changes which assume that fewer older people go into care homes or that paid for care services will be provided by local authorities to older people who presently rely on help from unpaid carers. 2.37 For the foreseeable future long term care will be a "people industry". It needs people to wash, dress and provide for the intimate needs of older people who can no longer do these things for themselves. The costs of care in the future will therefore be largely influenced by the earnings of people providing personal care which make up about 70% of the total costs of such services(16). These will need to be enough to attract the right quality of worker to this kind of work, at the right level of productivity and, of course, to comply with what ever minimum wage level is in force at the time. 2.38 As purchasers of long-term care, Local Authorities have a clear interest in ensuring that there is sufficient such care is available now and in the future at prices they can afford. So they must have regard to the future economic viability of organisations who supply and provide care. On the other side care providers, whether in the public or private sector, have to control costs and purchasers have an interest in this as well. 2.39 Other factors at work include expectations about what care is provided, in what surroundings and the quality of that care - although improved quality may not necessarily cost more and there may be better ways of doing things at the same or less cost. As far as expectations are concerned the example cited most frequently is that people will increasingly expect better standards of accommodation, such as single rooms and en-suite facilities. About a quarter of all people living in nursing homes now share a room(16) and this change from routine sharing of rooms has occurred over a relatively short space of time. The effect on overall costs of changes like this has been negligible. The future is harder to predict. There is an argument that as people become wealthier - as older people will overall - they will consider treating elements of long-term care as a "luxury good" in the economic sense, of which they will consume more as their incomes rise. On the other hand there is a view that most people do not want to spend more than is absolutely necessary on long-term care in residential care or nursing homes, preferring instead to spend money in their homes on aids, adaptations and technologies that would reduce their reliance on help from other people with care tasks. We could make assumptions about what people will want to do with their extra wealth but we are not in a position to do that with any degree of certainty. And in any case while the real incomes of older people will rise overall, the rise will be unequally distributed. 2.40 Unit costs might rise faster than general inflation, perhaps more in line with average earnings increases. The pattern of the past has been that the indices of NHS and Personal Social Services prices (i.e. the price of what it takes to provide care) have risen in real terms by 1.5% and 1% each year since around 1980 respectively. Some commentators have suggested that in the future these costs will increase more in line with average wages (1.5% to 2% per annum on average in the long run) or the general increase in national income (GDP at 2.25% per annum on average in the long run). However, wage costs are only about two-thirds of the costs of long-term care so there is not a simple one-to-one link with earnings. 2.41 In considering our cost projections, the Government will no doubt wish to make a direct comparison with its own calculations. These are done on broadly the same basis, except that we have used a half percentage point higher figure for increases in unit health care costs than used by the Department of Health in its memorandum to the House of Commons Health Select Committee(17). The impact of the National Minimum Wage has also to be considered. This will be expected to make a step change in wage rates, which will mainly affect the independent care sector. It will add to overall costs only to the extent that it is not paid for by efficiency gains, but the long run rate of growth in social care costs is unlikely to be affected. 2.42 We have had it put to us that instead of using a model built from the factors we have described to project future costs we should simply draw a straight line into the future based on past trends in NHS and social services expenditure of about 3.5% each year (1% more than our projection). This line should also be adjusted upwards by a further percentage point to reflect the fact that incomes will rise and so will the use of long-term care services - the "luxury good" referred to in paragraph 2.39. While this approach has the attractions of simplicity we think it over-simplifies a complex area and could give a very distorted picture of what the future might be like. For example, past expenditure increases take account of an entirely different demographic environment to what is expected in the future, and also takes account of past policy changes, for example the increase in Income Support expenditure for residential care and nursing homes . THE BASIS OF THE COMMISSION'S COSTINGS 2.43 Our base case makes the following assumptions:
2.44 Using these assumptions we modelled for the period 1995 (when the model starts) to 2051 to line up with population projections. We have already said that after 2031 the projections are very uncertain. Table 2.6 sets out the current and projected costs of long-term care. These figures are at 1995/96 prices so they are directly comparable across the years. We have also estimated costs as a percentage of future national income (GDP). 2.45 The base case projects that the cost of long-term formal care for older people (paid for by both individuals and the state) could rise from £11.1bn (1.6% of GDP) in 1995 to:
2.46 This implies an annual rate of growth in long-term care expenditure of 2.5% each year. During this period national income will have more than trebled. If national income grows more slowly then we would expect slower growth in the future costs of long-term care, but the share of national income would not be affected. 2.47 The model also shows the share of expenditure between the public sector and private individuals assuming no change in the present funding system described in Chapter 4. The increase in contributions from private individuals reflects the use of housing assets (as a result of rising levels of owner-occupation amongst older people) to pay for residential care. To the extent that the funding system is changed, the private and public share will change correspondingly. Table 2.6: Base case projections of long term care costs for older people, UK at 1995/96 prices 2.48 Factors over which there is little or no effective control impact on the base case. We vary assumptions to explore upper and lower limits of uncertainty. The factors which we have varied are listed below. We changed more than one assumption because certain changes appeared to go together. Some scenarios are however less plausible than others (for example it is more likely that health expectancy will improve than decline), particularly those where a number of things go "wrong" all at the same time. The results are shown in Figure 3. The boundaries of the possible future costs of long-term care are defined by two of these sensitivity analyses:
2.49 As each year passes the difference between the lowest projection and the highest projection gets bigger. The difference is around £10bn by 2021. The uncertainties from then on are so great that little predictive reliance can be placed on these figures, but it is still useful to compare trends between different scenarios. 2.50 We now test six sensitivities designed to establish the range of uncertainty around the base case, without assuming any policy changes. These are: 2.51 The results are very sensitive to the real increase in the unit costs of care. This is, in fact, the most sensitive assumption of all in the model. Therefore, we have considered alternative assumptions in the sensitivity analysis. The consequences of increases of 1.5% increase per annum for social services and 2% for health services are shown. Faster growth in 85+ age group 2.52 Historically projections by the Government Actuary's Department (GAD) for the numbers of people aged 85 and over have proved to be underestimates of actual numbers. This assumption increases the number of elderly people over the age of 85 by 1.26 million in 2051. This is a 30% increase in that age group and an 8% increase in the population overall. Faster growth in 85+ age group and higher cost 2.53 The cost implications of both faster growth in the 85+ age group and higher price rises occuring together have also been considered. The assumptions for faster growth in the 85+ age group and the higher increase in input prices of pay and other goods are as above. 2.54 If dependency levels at each age grow faster in the future than current levels, the implication is that for each age group in the elderly population there will be an increase in the incidence of chronic ill health or severe disability. This is, in fact, a very pessimistic view. To test the sensitivity, age specific disability rates have been assumed to rise by 1% until 2031 and remain constant thereafter(ii). On the other hand a 1% decrease in age-specific dependency rates (implying a healthier older population) would almost halve the costs between 1995 and 2051 compared with our central assumption. 2.55 We have tested the implication of reducing by 1% the number of single people living with others. The model is not sensitive to this change probably because the number of single people living with others is a small proportion (just under 20%) of all single people. Faster growth in 85+ age group and higher costs and fewer carers 2.56 The assumptions for the 85+ and higher costs are as above. We have taken a more extreme assumption on informal care. We halved the number of single people living with others by 2051 (compared with the base case), and then halved the probability that single people living alone will receive informal help with domestic tasks. The increase in formal services is assumed to be in line with existing users at similar levels of dependency and no informal care support. 2.57 In this chapter we have tried to provide an answer to the question about the future demand for and costs of long term care by making projections based on the number of people receiving them - the future size and structure of the population of older people, their health, the availability of unpaid care, the current costs of these services and how those costs will increase in the future. We recognise that while Government cannot directly control many of these factors it must nonetheless be prepared to respond to them. We are pragmatic about the value of these projections. We cannot forecast the future. The best we can do is project forward a realistic "base case" and accept that the future may lie within it, subject to wide margins of uncertainty. Little reliance can be placed on our projections beyond the first 10 - 15 years which is why we are recommending that the projections should be revised every five years. If costs do rise more rapidly than projected, society itself will have to take decisions about whether and how they should be met. The need for flexibility to respond to changes in the future is clear. We now go on to look at what we want from a funding system, at the current system and what it delivers and possible alternatives to it. Footnotes (ii) If they had been assumed to continue to rise until 2051, more than 100% of all very elderly women would be dependent.
| ||||
