The Scottish Government

« Previous | Contents | Next »

Listen

13 THE COSTS OF LONG-TERM CARE IN SCOTLAND 2006-2031

210. We examine below the evolution of long-term care costs in Scotland to 2031; based on the estimate of long-term care costs for 2005-06 discussed in Chapter 12 above. The estimates of future costs use the 2006-based demographic projections produced by the Registrar General for Scotland, which are the latest available.

211. The paper also considers how the cost estimates would vary if the demographic projections were those produced in 1998, rather than those of 2006. The 1998 data was what was available to the Care Development Group when it first produced estimates of the costs of free personal care in Scotland in 2001.

Uncertainty

212. Our modelling also examines the role of uncertainty, and its implications for the "funnel of doubt" associated with future care costs, which featured in the reports of both the Royal Commission and the Care Development Group. There are major uncertainties associated with the cost of care even if demographic projections are accurate. As well as the uncertainty associated with predicting the size and age structure of Scotland's future population, other sources of uncertainty include:

• Healthy life expectancy. This is an important determinant of the demand for long-term care. At present there is no consensus regarding the direction of healthy life expectancy in the UK. In some senses, the older population is more healthy than it has been in the past, as evidenced by its increased longevity. But on the other hand, the prevalence of increasing numbers of chronic diseases such as obesity, diabetes and liver disease may have an adverse affect on the state of health of the older population in the future;
• The provision of informal care. Informal carers play a vital role in providing care for frail older people. There are around 400,000 informal carers in Scotland, out of 5 million in the UK as a whole. ²² Ten % of Scottish adults provide informal care to a disabled child or adult. There has long been concern that informal care provision might fall if formal care was extended. There is no evidence that this has happened in Scotland, though there is some suggestion that informal carers have switched to other forms of care as a result of the provision of free personal care at home; ²³
• The balance of care. There has been a reduction in hospitalisation of frail older people and an increase in home care provision throughout the developed world in recent years. This not only has a positive effect on the well-being of clients: it also has implications for the costs of care as the balance moves towards home care, which is generally cheaper than hospitalisation. However, this is not always the case and occasionally some home care packages do cost more than provision in care homes. Hence, how older people as clients are supported through home care, care homes and long stay hospitals has implications for costs. The future balance between these forms of provision is a source of further uncertainty regarding long-term care costs;
• The costs of care. Long-term care costs are driven by the costs of labour, capital and by the efficiency with which these resources are applied to the provision of care. Capital costs are important in both the hospital and care home sectors. This is because there is often a clear opportunity cost associated with using land and buildings to provide long-term care. For hospitals, this may include additional provision of acute care; in care homes situated in areas where property costs are high, alternative usages may include housing. Past experience has shown that the costs of housing are volatile. To the extent that the costs of long-term care are affected by housing costs, the uncertainty associated with housing costs will be transferred to the costs of long-term care.

Other important aspects of capital costs include the costs of aids and adaptations and technological aids. The latter are increasingly being viewed as a way of reducing labour costs. While relatively cheap, their acceptance is not yet widespread. However, this may change, because the devices improve, due to policy interventions or because future cohorts of frail older people are more accustomed to dealing with technological aids. But again, this is another source of uncertainty.

Labour costs are important in all care settings, though this is particularly true in home care. Home care workers tend to be among the lowest paid groups in society. Given recent trends in the wages of the low paid, it might be anticipated that their wages, in real terms, would grow less rapidly than the economy as a whole. However, if higher qualification standards are imposed on the care sector, these workers will require a return for their investment in skills. The returns to different levels of skills in Scotland and the UK are reasonably well-known. ²⁴ More difficult to establish is the impact on wages of the increasing demography-driven demand for care workers. There is not an infinite supply of such workers, either from local sources or migrants. This is particularly true in parts of Scotland where there is no slack in the labour market and in rural areas where local provision is limited and transport problems further increase costs. These influences will tend to drive up the wages of care workers. But the responsiveness of care workers wage rates to increase demand is not well known. Hence, this is yet another cause of uncertainty in long-term care costs; and

• Care Policy. The final source of uncertainty in long term care costs is that associated with the policy environment. Care costs can rise due to the unintended consequences of policies or due to a lack of clarity in policies. Because the interactions between long-term care and other parts of the social and economic environment are so extensive, the range of policies which might affect the costs of long-term care is extremely broad. They include housing, health, social protection, industry regulation and employment.

There is also ultimately a feedback mechanism, whereby the costs of public support for long-term care will not deviate hugely from what is seen as socially equitable, given all other demands on scarce public resources. If provision is seen as too generous or too meagre, political change is likely to take place, restoring spending towards what is perceived to be an acceptable level.

What is then important is what determines public perceptions of fairness in respect of provision for long-term care. Although this is beyond the scope of our remit, the issue of intergenerational equity - fairness between generations - is worth mentioning. Since long-term care funding in the UK and Scotland is based on a pay-as-you-go approach rather than being funded, provision of care to older people must be seen as fair by the working age population.

213. There is therefore a wide variety of sources of uncertainty associated with the future costs of long-term care. Rather than attempt to model these individually, we have used in our projections a generalised form of uncertainty associated with our central predictions of the demand for care and care costs.

Changes in Demographic Projections

214. Our central projections are based on current levels of healthy life expectancy, no significant change in the provision of informal care and no change in the balance of care. We do not assume improvements in efficiency or changes in the balance of care. We assume that the costs of care increase at 2%, which is around the same rate as the long-term growth rate of the Scottish economy. This is a common assumption made in such exercises ²⁵ and allows for some improvement in the quality of care, given the real increase in resources allocated to each client over the period. For many of the reasons listed above in relation to the evolution of costs, this assumption is also subject to considerable uncertainty. Accordingly, we incorporate uncertainty associated with the evolution of costs in our model.

215. We begin by considering the latest projections of the older population in Scotland, The most recent estimates of the size of the Scottish population by 5 year age bands through to 2031, based on 2006 data, are shown in Figure 1 below.

Figure 1: 2006-based Projections of Population of Scottish Population Aged 65+

Source: General Register Office Scotland

216. There is clearly a rapid overall growth in the number of pensioners, from 837,000 in 2006 to 1.36 million by 2031. The growth is particularly concentrated among the older old - those most likely to require long-term care. Table 16 shows the overall and annual growth rates between 2006 and 2031 implicit in the GROs projections.

Table 16: Population Projections

Source: General Register Office Scotland

217. The table also demonstrates the very rapid rate of increase in the number of older people in Scotland in the period to 2031. The number aged 90 and over will more than double over this period - equivalent to an annual rate of 4.7%. The number aged 75 and over, who are the most likely group to require social care, are expected to grow at an annual rate of 2.4% until 2031. This is an important number, to which we return. These 2006 projections illustrate the very rapid increases in the older population of Scotland that are expected in the next 2 decades. Maintaining and enhancing services to older people will present a particular challenge over this period and will require radical adjustment of both institutions and policies.

218. Demographic projections are subject to uncertainty. This is well illustrated in the changes to the GROs projection of the older Scottish population aged 65 plus between 1998 and 2006. Figure 2 below illustrates the change between the 1998 projection and the 2006 projection. These changes clearly affect both the size and age structure of the older population over this period.

Figure 2: Percentage Change in Population Projections for those Aged 65+, 1998 to 2006

Source: General Register Office Scotland

219. Between 1998 and 2006 projections of the number of Scots who would be aged 65 to 69 between 2006 and 2031 hardly changed. However, with older age groups, the projections changed dramatically. A principal cause of this change was due to the 2001 census, which revealed that life expectancy had been increasing in Scotland much faster than had been anticipated. Accordingly, the population projections for the "oldest old" age groups were revised upwards. The upward revisions were largest for those aged 90 plus: they also increased with time. The net result was that for those aged 90 and above in 2031, the 2006 projection was almost 70% higher than the 1998 projection. Given that it is predominantly the "oldest old" that require long-term care, these revisions have had a very substantial effect on estimates of the future long-term costs of care in Scotland; this effect is calibrated below.

Projections of the Cost of Long-Term Care

220. In this section we examine the evolution of the cost of long-term care in Scotland. Our central projection is based on our estimates of the total costs of long-term care in 2005-06. We estimated these (see Chapter 12) to be £2.3 billion. This covers contributions from local authorities, the Department for Work and Pensions ( DWP), the National Health Service, and from individuals themselves. We examine the likely trend in costs for each of these 4 groups separately.

221. As implied above, our central projection is purely driven by demography. We do not make any assumption about changes in healthy life expectancy, informal caring or the balance of care. Consistently with other forecasts of long-term care costs in the UK as a whole, we assume a 2% annual increase in the real costs of care. The exception is social security benefits: for some time these have been indexed on prices, and therefore have not increased in real terms. We assume that this practice will continue and therefore that there are no real increases through time in the value of Attendance Allowance or Disability Living Allowance to individual clients. This means that we would expect that the Department for Work and Pension's contribution will gradually fall relative to that of other public sources of support for long-term care.

222. Our projections also assume that the age distribution of care provision does not change. We know the age distribution in 5-year age bands for those 65 and over receiving different forms of care and social security benefits. Therefore, for example, as mentioned earlier, those receiving Disability Living Allowance (Care Component) are generally among the "young" old, while care home residents are more likely to be drawn from the "oldest" old. We assume that these age distributions do not change through time. This is consistent with the assumption that healthy life expectancy is not changing.

223. All our projections assume that the cost of long-term care for older people from all sources in 2006 was £2.3 billion. Over £1 billion was provided by local authorities, with the remainder drawn from the DWP (£503 million), the NHS (£343 million) and private contributions (£374 million). This supported 9,100 care home residents receiving free personal care, 5,900 care home residents receiving free nursing care and 57,000 local authority social care clients of whom 41,000 received free personal care at home. It also supported 139,000 clients receiving Attendance Allowance and 245,000 receiving Disability Living Allowance. Some individuals receive more than one of these services or benefits. Eligibility for different services and benefits is generally governed by complex rules and assessments.

224. Based on these assumptions, our central projection based on the 2006-based population projection shows care costs rising from £2.3 billion in 2006 to £3.3 billion in 2016 and £4.1 billion by 2021. This suggests that costs will increase at 4.4% per annum. Around 2% of this increase is due to increases in the unit costs of care, but population ageing drives increases in the volume of care. As mentioned previously, the annual increase in the population aged 75+ (those most likely to require social care) is expected to be 2.4% between 2006 and 2031. Thus, it is not surprising that our central projection is that the real increase in the overall costs of care will be 4.4%.

225. To further illustrate the effect of ageing, we compare cost estimates based on:

(1) No population ageing,

(2) 1998 population projections ( those available to the Care Development Group), and

(3) 2006 population projections.

The results are shown in Table 17.

Table 17: Costs of Long-Term Care in Scotland, under Different Demographic Assumptions

226. The first column shows how the costs of long-term care might evolve between 2006 and 2031 if there was no increase in the number of older people above 2006 levels. In this case, it is the increase in the real unit costs of care which drive overall cost increases. Costs increase by 50% over the entire period, an annual rate of 1.6%. This is below the 2% rate of increase set for most forms of care. But because the real growth rate of benefits is set to zero since they are indexed to price inflation, the average increase in total care costs is below 2%.

227. Introducing population ageing as it was forecast in 1998 adds £1.7 billion to the projected future costs of care in 2031 and more than doubles the real annual growth rate in costs to 3.3%. However, as mentioned previously, the 1998 population projections underestimated the number of older people in the Scottish population, particularly the "oldest" old. The current population projections suggest that this group will be much larger than envisaged in 1998. As a result, the forecast costs of care increase further, with the annual rate of increase rising to 4.3% and the difference between the 1998-based projections and the 2006-based projections rising to £1.3 billion by 2031. The Care Development Group, which deliberated during the early part of 2001, based its cost estimates on the 1998 GROs projections. Hence it did not anticipate the cost implications of subsequent projected increases in the number of the "oldest" old in Scotland.

228. The final column of Table 18 expresses the overall costs of social care as a share of estimated Scottish GDP in 2006 . It shows that the 2006 ²⁶ central projection implies that social care will account for a significantly larger share of GDP in 2031 than it did in 2006, rising from 2.6% to 4.6% over this period. The Wanless Review of Social Care (2006) proposes 3 scenarios for the provision and funding of social care in England. The first is a continuation of the current service model; the second involves changing the system of social care to achieve higher levels of provision of personal care and safety; the third involves improved social inclusion and well-being outcomes. All 3 options have broadly the same financial outcomes, resulting in increases of spending of between 139 and 142% between 2002 and 2026. The Wanless Review ²⁷ assumes 2% annual increases in the unit real cost of care. Our approach to unit care costs thus follows the same assumptions as the Wanless Review. However, it has proposed new models of care, which involve some reallocation of resources rather than simply an expansion of the existing structure of service, imply annual growth rates in care costs of around 3.7%, somewhat lower than the 4.3% increase projected for Scotland under our central estimate. By 2026, the Wanless models are projected to cost around 2% of GDP. Our estimates suggest that Scotland has already passed this level of support and that by 2026 the costs will have risen to 4.0% of GDP.

229. Within the overall increase in spending on long-term care, local authorities net costs would more than double, based on the central projection. Table 18 shows estimates of net NHS costs, local authority costs ²⁸ and the costs of provision of free personal care at home between 2006 and 2031. Local authority net costs rise from £948 million to £2,746 million. NHS costs rise by a similar proportion.

Table 18: Long-Term Care Costs to Scottish Government

230. Personal care costs, which are paid by local authorities increase from £256 million to £813 million between 2006 and 2031. They comprise around 21% of all long-term care costs. It is incorrect to expect that changing or removing this policy would cause a very substantial reduction in the overall costs of long-term care. This is because a significant proportion of those that receive free personal care at home, who comprise around 80% of all those receiving long-term care, would be unable to pay for the cost of this care. Current rules governing charging for personal care outside Scotland are complex, and tend to recover a relatively small proportion of the total costs. For example, a recent survey in Wales ²⁹ estimated that only around 14% of the costs of free personal care are recovered in charges.

231. Costs mainly payable by the Scottish Government (some contribution would come from local government taxation) increase from 1.4% of Scottish GDP in 2006 to 2.7% in 2031. Unless there is a marked increase in the share of taxation in the Scottish economy, quite significant cuts will have to be found in other programmes to continue to fund the present structure of care. Our present estimate ³⁰ is that a 1p increase in the basic rate of income tax in Scotland, which is within the devolved powers of the Scottish Government, would raise £370 million ³¹ or 0.4 % of GDP.

Uncertainty

232. So far, we have described the central projection of our model of the costs of long-term care. As mentioned earlier, there are good reasons to avoid reliance on a point forecast for policy development. Policy should take account of the inherent uncertainty in projecting the costs of long-term care. These include future trends in healthy life expectancy, informal caring behaviour, the balance of care and costs.

233. Diagram 1 below - Funnel of Doubt in Care Cost Projections - gives a flavour of this uncertainty. It is based on a repeated simulation of the model, each with different levels of uncertainty associated with the volume and the cost of care. Each of the components of care and their costs are equally likely to increase or decrease. The chance that the volume of care varies by more than 5% either positively or negatively is around 30%. The chance that the costs of care vary by more than 1% in either direction is also around 30%. The variability associated with the simulated projections also increases with time. We take the averages of the simulations and also calculate how much they vary. The diagram then shows the central projection for the costs of long-term care to the NHS, local authorities and the Scottish Government. Alongside the central projections are statistics based on the simulations showing values either 1 or 2 standard deviations above or below the central projection.

234. While no great significance should be attached to the precise values associated with the variability of the simulation, the figures do illustrate that relatively small changes in both the volume and costs can have quite significant effects on the evolution of the costs of care. But at present, it is very difficult to predict which factors might influence changes in the volume and costs of care and in which direction. This is particularly true, the further projections are taken into the future.

235. And while projections can indicate a general direction of travel, it is important to recognise that those illustrated here are based on a continuation of the present policy configuration. The evolution of costs and policy are not independent. Increasing costs are frequently a spur to the reconfiguration of services or of making them more efficient.

Conclusion on projecting the future costs of long-term care

236. We have attempted above to project the future costs of long-term care in Scotland. If there is no change in the demand for care in each age group and costs increase at around 2% per annum, then the overall costs of care would increase by around 4.3% in real terms between 2006 and 2031. This is somewhat higher than the 3.7% growth rate implicit in the predictions of the increase in the costs of long-term care contained in the Wanless Review of Social Care (2006). Free personal care is a relatively small component of the total costs of care. Future costs of long-term care are still likely to increase significantly, even were the free personal care policy to be revised or abandoned.

237. However, there is significant uncertainty associated with projections of the costs of long-term care which could go in either direction. Actual spend is also responsive to changes in policies which different administrations follow. Finally, it is clear that early intervention to influence the efficiency of delivery may have significant long-term benefits in terms of reduced costs, without any necessary diminution in the quality of service.

Funnel of Doubt in Care Cost Projections

Department of Economics, University of Stirling

« Previous | Contents | Next »