ࡱ> ]^\ܥhc e=~tش  ,<: QvzX;C NP v N4L Free Care for the Elderly Jim Cuthbert Margaret Cuthbert Introduction 1. The policy of free personal care for the elderly in Scotland was implemented in July 2002. Work on the costings of the exercise had been undertaken by the Care Development Group (CDG), a group established by the Minister for Health and Community Care, and chaired by Malcolm Chisholm. Research to assist in the costings was specially commissioned and the research papers were published by the Scottish Executive Central Research Unit: (CRU, 2001). The CDGs main conclusion with regard to costs was that the initial cost of the policy would be 125 million, rising to 142 million by year 3. (CDG, 2001) 2. Even before the policy was implemented in Scotland, the question of cost had caused considerable debate. Doubts, for example, were expressed by David Lipsey and Joel Joffe, two of the members of the Royal Commission on Long Term Care, who had produced a Note of Dissent (1999). In Lord Lipseys view, "If it was affordable and if it was going to people who most needed it, yes great, free care would be a splendid policy. Unfortunately it is not affordable." One contribution to the debate on the cost of the policy in Scotland was a paper by Cuthbert and Cuthbert (2002), in which it was argued that the costs had been substantially underestimated. 3 Recently, some members of the CDG have been quoted as saying that implementation of the policy could be too expensive: (Scotsman 12th October 2004). Nevertheless, the current state of the debate on costings is still unsatisfactory. The discrepancy between the cost estimates which have been produced by the CDG and ourselves is so large that it needs to be resolved. Moreover, the need for this resolution is heightened by two additional factors. First, the latest population projections by the Government Actuarys Department indicate that there will be more old people in Scotland than was previously projected, which means that the effects of any underestimation of costs will be multiplied. Second, the Liberal Democrats have adopted free personal care for the elderly for the whole of the UK as a major part of their pre-manifesto. The LibDem costings of 1.4billion for this policy for the UK appear to be broadly in line with the CDG costings for Scotland. So if the CDG has indeed underestimated its costings for Scotland there could be profound implications for the LibDem strategy. 4 The main new information in this paper is derived from examining, in greater detail than in our previous paper, that part of the CDGs work concerning the estimated costs of delivering free personal care in the community, as opposed to in residential and nursing homes. In the course of this critique we establish that there were flaws in the basic research commissioned by CDG which meant that the numbers of elderly disabled people in private households in Scotland were seriously underestimated by the CDG, by a factor of more than 2. This mistake appears to account for a major part of the discrepancy between the CDG and Cuthbert and Cuthbert costings. In addition, in other parts of our critique we point to other areas where the methodology is suspect. Background 5. The overall CDG costing for the introduction of free personal care in Scotland was built up from the following components: a. residential care: 22 million. b. nursing homes: 35 million for personal care, 15 million for nursing care. c. care in the community: 70 million building up over three years. In this article, we are primarily concerned with care in the community. This is not to say that we are content with the CDGs approach on (a) and (b): however, our critique of their treatment of those components is set out in our previous article (2002). 6. Prior to the introduction of free personal care, the main mechanisms for the provision of personal care for the elderly in the community were informal carers family and friends; formal care provided through local authorities, either paid for by the local authority or by means tested charges levied by the local authority on the individual; formal care arranged privately. The CDG starting point in costing community based free personal care for the elderly was an analysis of the home care provision provided through local authorities. On the basis of the most recent data returns from local authorities for 2000, and based on their assumption that 45% of this care could be regarded as personal care, they determined that the local authorities would forego around 10 million in charges if free personal care was introduced and if the existing level of service was maintained. 7. To this basic cost was added the amount currently paid for private personal care provision outwith the local authority system: a survey of private providers by the UK Care Homes Association (UKCHA) commissioned by CDG indicated that this was of the order of 10 million per annum. 8. The CDG then considered the issues of unmet demand and the potential for substituting formal care for informal care. They considered a range of information sources: in particular, they commissioned Stearns and Butterworth (2001) to explore the anticipated demand for personal care by the disabled elderly in private households and the likely substitution effects arising from the introduction of free personal care. The CDG estimated that meeting unmet need for personal care services would cost between 15 and 25 million per annum. Meeting the extra demand arising from a potential shift from informal to formal care would cost an estimated 20 to 25 million. It was assumed that both of these costs would build up over the first three years of implementation of the free personal care policy. 9. In the following sections, we present a critique of important aspects of the background research and of the CDGs costings. Critique of Stearns and Butterworth Approach to Estimating Numbers of Disabled in Private Households 10. The following table, abstracted from Stearns and Butterworth table 3.5, shows their estimates of the total disabled population in Scotland aged 65 and over by various categories of domicile. The estimates in the first row of the table, (numbers in private households), were derived by them from the Department of Works and Pensions (DWP) Disability follow up study to the Family Resources Survey of 1996/97, and the 1985 OPCS survey on disability. To these survey results were added estimates of the numbers in each of the other categories shown, to give their estimate of the numbers disabled in the population. Table 1: Disabled Aged 65 and Over in Scotland: Stearns and Butterworth estimates. Disabled in1985 estimates1996/7 estimatesPrivate Households310,950142,224Special Needs Housing 33,261131,198Residential Care 14,185 15,037Nursing Home 15,000 15,730Long Stay Hospital 8,735 6,098Psychiatric Hospital 3,777 5,045Total Disabled Population385,908315,332 In Stearns and Butterworths analysis, it is the group labelled private households in the above table which they believed constituted the population from which the potential demand for personal care in the community would arise. Note that the table shows a decline of over 50% in the numbers disabled aged 65 and over in private households between 1985 and 1996/97. Also note that they assumed that the population in special needs housing is distinct from the population in private households. This is an important assumption and, it turns out, a mistaken one. 11. In terms of the standard descriptions used by government departments, residents of most special needs housing are categorised as private households: (sources - General Register Office for Scotland, DWP, and Office of National Statistics website). Note also that in the 1985 and 1997 surveys of disability, which Stearns and Butterworth are using, special needs housing is an integral part of the overall sampling frame of private households, used both in the selection of the sample, and grossing up the sample results: (as confirmed by personal communications from the DWP and from the authors of the Disability Follow Up Survey report.) So, when Stearns and Butterworth state, in their footnote 17, To the best of our understanding, both surveys do not include special needs housing as private households. they are quite mistaken. 12. Thus, in terms of the definition of private households as used by the government, in terms of the definition used in the source surveys, and, most importantly, in terms of the practical question of whether residents of special needs housing constitute a potential demand for personal care in the community, Stearns and Butterworth are wrong to exclude special needs housing from the definition of private households. 13. Recognition of this mistake then leads on to another extremely puzzling feature of their figures. Given the way that the Disability survey was conducted, elderly disabled residents of special needs dwellings are necessarily already included in the private households category in Stearns and Butterworth table. So, from their figures as shown in Table 1 above, if there are around 142,000 elderly disabled residents of private households, (and this de facto includes special needs housing), (line 1), and, around 131,000 elderly disabled residents of special needs houses, (line 2), then the logic of this implies that there are only around 11,000 elderly disabled residents in all other non-special needs housing private households. This would clearly be a nonsensical conclusion to draw. A much more likely explanation, which we now examine in more detail, is that there has been a further mistake in Stearns and Butterworths calculations, in the process of grossing up the disabled survey results. 14. Further evidence for this can be found by comparing Stearns and Butterworths estimate of the GB elderly disabled population in private households, (which they also obtained by grossing up results from the disability survey), with the DWP published estimate. The DWP report on the Disability survey, (Grundy 1999), gives the number of disabled persons aged 65 and over and resident in private households as 4,082,000. The Stearns and Butterworth report,based on exactly the same survey, gives the much smaller estimate of 2,944,114 as the size of the UK disabled total for those aged 65 and over. (Their figure should not be UK as the survey was GB). 15. A number of other points add to our concerns over the accuracy of the Stearns and Butterworth figures: in particular, the population figure which they quote for private households in Scotland in their Table 3.5 was derived by subtracting all other estimates from total estimated population. The figure given for the total population in 1985 is wrongly recorded as 723,516: it should have been 742,588. More importantly the figures in the different categories do not add up to the total population, with a missing group of 27,755 in 1996/97. 16. We therefore have strong evidence that Stearns and Butterworths procedures for grossing up the disability survey results are seriously flawed. The implication is that we cannot obtain a satisfactory estimate of the elderly disabled population from their figures. We therefore now go back to the available survey evidence to derive a revised estimate. 17. The first step is to estimate the number of elderly people in private households in Scotland by age band. We do this by subtracting from mid year 1996 estimates of the total population those in particular communal establishment categories (see table 1 above). The overall estimate of the elderly in Scotland in private households derived in this way for 1996/97 is Total population =779,462, from which we subtract Residential Homes = 15,037 Nursing Homes = 15,730 Long term Geriatric = 6,098 Long term psychiatric = 5,045 giving a private household population for the elderly in Scotland of 737,552. [This estimate is consistent with the 2001 census findings of around 95% of the elderly population living in private households.] 18. If we then apply appropriate disability prevalence rates to the different age bands in this population we can obtain estimates of the numbers of elderly disabled in private households in Scotland. We use GB prevalence rates from the Disability Survey. The primary reason for using GB rates is that the disability survey team did not believe that the sample sizes for the survey were large enough to provide useable estimates of disability rates by age band in Scotland: (Stearns and Butterworth, para. 2.7). 19. It is important to note, however, that our use of GB prevalence rates is, if anything, a conservative assumption. The DWP survey team have calculated age standardised rates of disability for Scotland and GB, (Grundy 1999, Table 2.6): Scotlands rate is 200 per 1,000, compared with the GB rate of 198 per 1,000. This means that disability rates for individual age bands are slightly higher on average in Scotland than for the corresponding age bands in GB. Our use of GB prevalence rates to estimate disability in Scotland will therefore lead to a slight underestimate, unless the slightly higher average rate of disability in Scotland came about as the result of Scotland having much higher rates of disability than GB for those aged under 65, and much lower rates than GB as a whole for those aged over 65. Two pieces of evidence mean that this possibility can be discounted. a) First, in the DWP survey, 8.5% of all disabled in GB live in Scotland, with 9.6% of the most severely disabled living in Scotland. As severity tends to be associated with age, it is likely that disability rates for the elderly in Scotland are relatively higher than GB, rather than lower. b) Secondly, 5% of the Scottish population under 65 receive Disability Living Allowance, (DLA), compared with 4% of the GB population: while 25.2% of the Scottish population aged over 65 receive Attendance Allowance (AA) or DLA, compared with 21.4% of the GB population: (DWP (2003), and private correspondence from DWP). While the criteria for receipt of AA and DLA are more restrictive than the measure of disability used in the Disability follow up survey, these figures again suggest that disability rates are relatively higher in Scotland compared with GB for those over 65. Both of these pieces of evidence support our view that, in using GB prevalence rates, our estimate of the elderly disabled population in Scotland is likely to be an underestimate. 20. Table 2 shows the derivation of the estimates of the population in private households for the relevant age bands. The age distribution of those in communal establishments was derived from the age distribution of those in residential homes given in Scottish Community Care Statistics, 1999. Table 2: Elderly Population in Private Households in Scotland, 1996 AgeMid Year Population Population in Communal DwellingsPopulation in Private Households65-69238379241023596970-74207494241020508475-79150501743914306280-8410439074399695185plus786982221256486GROS total77946241910737552 Table 3 shows the derivation of our estimate of the number of elderly disabled in private households in Scotland. Table 3: Prevalence and Estimated Numbers of Disabled by Age: Scotland 1996/97, Using GB Disability Prevalence Rates AgeGB Prevalence Rates of Disability per 1000Elderly in Private households in ScotlandEstimates of number elderly Disabled: Scotland% elderly disabled65-692962359696984729.670-743382050846931833.875-79650143062929906580-84709969516873870.985plus838564864733583.8Total73755234822947.2 The figures imply a total disabled population of 348,000 for Scotland. A cross check can be derived by starting from the DWP published estimate of 733,000 for the total number of disabled of all ages in private households in Scotland. We do not know the proportion of disabled in Scotland aged over 65 but the corresponding GB percentage is 48%: applying this to the figure of 733,000, this would imply an estimate of 352,000 for the disabled elderly population in Scotland. 21. Finally, it is worth noting that there is another, technical, reason connected with the DWP survey, which means that estimates based on the 1996/97 survey are likely to be slight underestimates. This arises because of the sifting technique used in the survey, which means that the incidence of disability in the specific age range from 65 to 74 is likely to have been underestimated: (see Grundy (1999)). 22. Given the above, we feel confident in taking 350,000 as our central estimate of the elderly disabled population in private households in Scotland. This contrasts with the Stearns and Butterworth estimate of 142,000. As we shall see, this difference has profound implications for the CDG estimates of the initial costs of implementing free personal care in the community. 23. There are also important implications for the future cost profiles of implementing free personal care. Based on their estimate of a decline in the disabled elderly population in the UK, (though they should have said GB), from 3.4 million in 1988, to 2.9 million in 1996/97, (that is, 0.6 percentage points per annum, as a percentage of the total elderly population), Stearns and Butterworth postulate that the elderly disabled population as a whole in the UK might have been declining by 0.2 to 0.3 percentage points a year from 1988 to 1996/97, (as a percentage of the total elderly population). In their paragraphs 3.14 and 3.16 they then state: It is conceivable that the decline in disability in Scotland is greater than the decline in the UK as a whole, but for a variety of reasons the estimate for the UK is methodologically a stronger estimate. These more moderate rates of decline (e.g., 0.2 to 0.3) are also potentially more plausible for conservative estimates of the anticipated future rate of decline in the Scottish population. Therefore, an assumption of constant health expectancy for predictions of the cost of free personal care seems unnecessarily conservative. An assumption of at least a modest continuing decline among the population in private households is probably justified. In fact, both of these statements now look highly questionable. On the basis of DWPs published figures, the disabled elderly population in private households in GB increased substantially from 1988 to 1996, (from 3.4 million to 4.2 million): and the corresponding Scottish disabled elderly population, on the basis of our estimate, increased from 343,000 to 350,000. Critique of the CDGs Costings. A: The 45% Claim 24. In paragraph 7 above, we have seen how the CDGs costings of community based personal care relied, in part, on the assumption that 45% of the home care provision provided through local authorities was personal care. As indicated in the Stearns and Butterworth report, this assumption was based on an estimate from a study carried out by West Lothian Council in its strategic services review: (West Lothian, 2000). In fact, the derivation of the 45% figure from the West Lothian study is based on a misreading of the West Lothian evidence. As West Lothian staff and the relevant table in the West Lothian report (Table 2) confirm, of the total amount of personal and domestic care delivered by West Lothian Council, 55% was personal care. This puts the West Lothian data more in line with other available sources of information. For example, available to the CDG at the same time was an estimate from the Resource Use Measure Group of the NHS for Scotland which in its pilot survey of nine areas in Scotland estimated that 61.5% of home care was personal care: (NHS, (2001)). Further, the UKCHA survey of private providers shows that, of the care they supply through local authorities, only 28% of all hours is for practical domestic care, the remainder covers personal, night time and live in care, with personal care being 60% of all care. In the light of the above, the 45% claim appears untenable. A more reasonable assumption would be that 60% of the domestic care provided by local authorities related to personal care. This has implications for a number of different aspects of the CDG costings. a) For one thing, it implies that the cost of personal care which local authorities provided free, before the introduction of the policy of universal free personal care, was underestimated by about 18 million: (since this was already being provided free by local authorities, this does not add to the additional cost of free personal care: but it does increase the size of the base cost of the service.) b) More directly, the 60% figure implies that the cost of providing free that element of personal care which was previously purchased from local authorities should have been estimated at 13.3 million rather than 10 million. c) There are also implications, as we shall see below, for the unit cost of personal care, and hence for the estimated cost of unmet need. B: The CDGs Estimate of the Cost of Unmet Need 25. Paragraphs 5.40 to 5.42 of the CDG report deal with the topic of unmet need, as follows: the Group therefore commissioned external work to assist consideration on these and other matters which could have a considerable effect on demand but which are more difficult to quantify accurately. 5.41 We asked Aberdeen University to do some further work for us on estimating the levels of current unmet need for community based personal care services. To do this they used information from the Family Resources Disability Follow Up Survey, although the sample size was not ideal. They found that levels of reported unmet need for personal care did not exceed 10%. Moreover, cost was not the most commonly reported cause of unmet need. Instead, not knowing help was available, not knowing where to find help, or wanting to help oneself, were reported more often. 5.42 This level of reported unmet need was also confirmed broadly by the Scottish Household Survey and from the British Household Panel Survey, though neither relate specifically to personal care. We also made use of information from the CareNapE (elderly) assessment tool which had been piloted in Glasgow. The work culminated in an estimate of unmet need for personal care services in the community in the range between 15 to 25 million. 26. Note how brief the CDGs description is of their approach to this important topic: and also how limited their information sources are. For example, the pilot of the CareNapE assessment tool in Glasgow related to 460 people in Govan. It is not clear how individuals were selected for this pilot: but what does seem clear is that it would be difficult to base an estimate of unmet need at the national average level on information which relates to one specific area and which will therefore be conditional upon local perceptions, and also upon the particular level of provision for personal care which the relevant local authority has seen fit to provide. 27. The primary information upon which the CDG is founding its estimate relates to the estimate of unmet need of at most 10% derived from the Aberdeen, that is, Stearns and Butterworth, analysis of the disability survey: (the CDG states that the Scottish Household Survey and from the British Household Panel Survey broadly confirms this estimate). How does this 10% estimate relate to the CDGs money estimate of unmet need as being in the 15 to 25 million range? 28. This assessment is rendered difficult because the CDG gives no detail (other than the limited information in the quotation given above), on how they actually calculated their money estimate. The CDG do not even define what the denominator of their 10% unmet demand percentage actually is. However, it is clear from study of the Stearns and Butterworth report, (specifically table 4.13), that the 10% unmet demand figure relates to the percentage of the elderly disabled population in private households reporting current unmet need for formal in-house services. 29. This is where the significance of the Stearns and Butterworths underestimate of the private household elderly disabled population comes in. Given their estimate of 142,000, this would imply on the basis of 10% unmet need, that 14,000 elderly disabled people were experiencing unmet need for personal care in the home. However, as noted above, a reasonable estimate of the size of the private household elderly disabled population in Scotland in 1996 would be 350,000, implying that a more realistic estimate would be that 35,000 were experiencing unmet demand. 30. There is then the problem of converting these numbers into money equivalents. The CDG do not explicitly state what the average cost of personal care delivered in the home would be: and in fact, there appear to be two inconsistent estimates implicit in their figuring, as follows. a. First, based on the amount of home care provided through local authorities and the CDGs assumption that 45% of this is personal care, this implies a unit cost of 1,100 in 2001 for recipients of personal care through local authorities. b However, when the CDG were estimating the cost of providing the personal care in the home which is currently purchased privately, the CDG estimated that 10 million would be needed to fund the care purchased by the 3,000 purchasers of private care. This implies a unit cost of 3,300 per person. 31. If we take the lower of the above estimates, of 1,100 pounds per recipient, then the cost of unmet demand would be 15.4 million on the basis of the mistaken Stearns and Butterworth estimate of the disabled population: this figure appears consistent with the lower end of the CDG estimated range for the cost of unmet demand, (which, it will be recalled, was 15 to 25 million.) However, given our estimate for the size of the disabled elderly population in private households, a unit cost of 1,100 would imply that the cost of unmet demand would be around 38.5 million: while the higher 3,300 unit cost would imply around 115 million. 32. What can we say about these figures? a. First of all, it might be objected that the 10% figure for unmet demand is itself too high, since, as the CDG note, cost was not the most commonly reported cause of unmet demand. However, two of the other reported reasons (not knowing help was available, and not knowing where to find help), will be largely removed by the operation of the free personal care policy. Indeed, if significant numbers of people continue to be unaware that free care is potentially available, then this would be an acid test for the policy having failed in at least one important respect. It is also likely that the introduction of the policy could initiate a significant cultural shift, so that accessing free personal care comes to be seen as very much a norm and an entitlement. For all of these reasons, it seems unlikely that the effective factor for unmet demand will be significantly less than 10%. b. On this basis, the estimate based on the mistaken SB figure for the elderly population has to be rejected: as has correspondingly, the CDG estimate. c. Thirdly, we have illustrated above the effects of unit costs of 1,100 and 3,300 - both of which have some basis in different aspects of the CDG report. The figure of 3,300, (corresponding to approximately 8 hours per week at 8 per hour), is the average cost of care consumed by those who, prior to the introduction of the free care policy, had the resources and the incentive to purchase private domestic care: this figure could well therefore be somewhat higher than the average for those who currently experience unmet demand. On the other hand, the 1,100 figure depends on the suspect 45% personal care assumption: increasing the 45% to 60%, as we have argued above is more reasonable, would increase the 1,100 to 1,500. This would still only equate to around 30 minutes of care per day, (at a cost of 8 per hour), which still looks somewhat low: it would be even lower than 30 minutes per day if the CDG had used the higher cost per hour as shown in the NHS survey commissioned by the Chief Nursing Officer. 33. On this basis, a conservative estimate might be a unit cost of 2,000 for each individual with unmet demand, corresponding to around 70 million per annum to meet unmet demand. This contrasts with the CDGs estimate of from 15 million to 25 million, suggesting that the CDG underestimated the cost of unmet demand by 50 million on a conservative estimate. C: The CDGs Estimate of the Cost of Private Provision 34. In paragraph 8 above, we noted that the CDGs estimate of the cost of domestic personal care currently purchased privately was 10 million. This estimate is a function of the method used to scale up the sample results from the UKCHA survey. UKCHA state clearly in their report that they have a number of concerns, not least that they did not have a full data base of private providers of care in Scotland. The researchers were thus faced with the problem of grossing up from the sample to obtain population estimates. They did this by assuming that those firms which had provided information to the UKCHA survey would be responsible for the same percentage of the total amount of personal care bought privately, as they were of the total amount of care purchased by local authorities from the private sector. This appears a very questionable assumption: indeed there is some evidence to suggest that some firms are likely to specialise in providing care for local authorities- which would tend to invalidate the assumption used in grossing up, implying that an estimate of total private care produced by grossing up on this basis would probably be an underestimate. As UKCHA themselves readily agree, the method of grossing up actually used was only one of a number of ways of grossing up. Had another indicator been used, e.g., based on the number of firms giving useable responses, (79), out of the known 146 suppliers, then a larger estimate of 15 million would have resulted. This again would not have covered all the personal care provided by companies unknown to the UKCHA, nor the care provided on a paying basis by neighbours etc. On the basis of the available evidence it does not appear possible to reach any firm conclusion, other than that the CDG estimate of 10 million is probably an underestimate, and quite possible an underestimate by 5 million or more. D: The CDGs Assumption of Improved Health Expectancy. 35. In calculating their forward projections of the cost of care for the elderly, the CDG built in an assumption of a 0.25 percent per year reduction in the proportion of the elderly population requiring services, reflecting increasing health expectancy: (CDG, para 5.21). This assumption is explicitly linked to Stearns and Butterworths estimates of increasing health expectancy: (CDG, para 5.13). However, we noted above that Stearns and Butterworths findings of improved health expectancy for the elderly, in both Scotland and the UK, are dependent on their mistaken grossing up of the results from the disability survey. Once this has been corrected for, there is little or no evidence from the disability survey results of any consistent improvement in health expectancy. The effect of this on the CDGs future costings is very significant. If the assumption of improving health expectancy is removed from the CDGs model, then the effect is to increase the cost of long term care services for the elderly by about 130 million per annum by 2022: (note, however, that only part of this increase will relate to expenditure on personal care.) Conclusion. 36. In summary, the various points we have identified above have a very significant impact on the CDGs cost estimates. The effect is to a) increase the base cost of providing personal care in the community by some 18 million, due to the doubtful 45% personal care provision assumption: as already noted, this is provision which local authorities already made, so this is not new money. (see paragraph 24). b) increase the initial cost of the policy by approaching 60 million per annum, as set out in the following table:- millionCDG EstimateOur Revised EstimatePersonal care previously purchased from local authorities (para 24)1013.3Unmet need (para 33)15-2570Private provision (para 34)1015 The principal revision is due to the effect on the assessment of unmet need of the mistake in calculating the size of the elderly disabled population in private households, but with the doubtful 45% assumption, and the doubtful method of grossing up the UKCHA survey also contributing. The above two points relate to the base and initial costs of free personal care for the elderly in the community. In addition, as regards projections there is the effect of: c) increasing the longer term cost of care for the elderly by about 130 million by 2022, due to the dubious assumption made by the CDG about improving health expectancy. (see paragraph 35). It must also be made clear that the latest Government Actuary Department projections, which have further increased the numbers of elderly in the population would imply additional costs to (c) above. 37. In the context of the debate about the affordability of free personal care, the cost increases identified in (a) to (c) above are significant amounts. It is also relevant to set the cost estimates at (b) above against the information in Cuthbert and Cuthbert, (2002). While in that paper we did not give a single point estimate of the cost of introducing free personal care in the community, we did set out a grid, (Table 2.2 in that paper), which gave different cost possibilities in what we regarded as the feasible range of what the new policy might be expected to deliver. We had also suggested that the CDG costings lay very much at the unlikely lower extremity of the costings considered in our grid. The effect of the additional 60 million identified in this paper is to move the adjusted CDG costings very much more towards the centre of our grid. In that sense, we have accomplished one of the tasks we set at the start of this paper- to reconcile the CDG costings with our own. 38. In our earlier paper we argued that the effect of the significant underestimation of the costs of implementing the new policy would probably manifest itself as much in the dimunition of the quality of service as in cost over runs. Recent anecdotal evidence from some local authorities suggests that part of this quality reduction is already manifesting itself by a switch of home care services away from domestic care towards more targeted personal care. In other words, free personal care may be being bought partly at the expense of an increasing number of the poorest elderly people either having to do without domestic care which they previously received free, or having to pay for such care. 39. It is hoped that the findings in this paper will have a significant effect on the debate about the affordability of free personal care. We also hope that the paper will have another effect as well. If we are correct in our analysis, (and we are confident that we are), then our findings raise very serious issues about the quality assurance of publicly funded research upon which important policy decisions are based. Our view is that the ultimate responsibility for quality assuring publicly funded research should rest with government. This is partly because such research is often published under the imprimatur of a government department, (as it was in this case, by the Central Research Unit of the Scottish Executive): but more importantly because, if wrong decisions are made, then the effects will rebound on government, and on the public. We suggest, therefore, that serious consideration needs to be given to improving the methods of quality assuring research, before it influences important policy decisions. References: Attendance Allowance, DWP, November 2003 Central Research Unit, (2001), Providing Free Personal Care For Older People, Research commissioned to inform the work of the Care Development Group, Scottish Executive. Care Development Group, (2001), Fair Care for Older People, Care Development Group Report. Jim Cuthbert and Margaret Cuthbert, (2002), Free Personal Care for the Elderly: A Critique of Costings, Fraser of Allander Quarterly Economic Commentary, March 2002, Vol. 27/ No1. Disability Living Allowance, DWP November 2003 Grundy E, Ahlburg D, Ali M et al (1999), Disability in Great Britain, Department of Social Security Report 94. HMSO, London. NHS, (2001) , Development of Resources Measure: Progress Report, Scotland. Royal Commission on Long Term Care, (1999), With Respect to Old Age, CM 4192-1. The Stationery Office, London. Sally C. Stearns and Suzanne Butterworth, (2001),Demand for and utilisation of personal care services for the elderly, Health Economics Research Unit, University of Aberdeen West Lothian, (2000), Strategic Social Services Review.  Personal care is defined in CDG(2001): the definition was developed from the Royal Commission on Long Term Care (1999, page 68)  BBC News, 26th July 2000  Launch of pre-Manifesto, 14th September 2004, LibDem website.  Special needs housing falls into a number of categories defined as follows: Sheltered housing: (34,976 dwellings in 1996). This includes houses with amenities installed such as handrails, non-slip bathroom floors, and a warden service. Very sheltered housing: (719 dwellings). This also offers a greater level of care and support through the service of extra wardens, full-time carers or domiciliary assistance and the provision of meals. Amenity Housing: (18,535 dwellings). This is similar to sheltered housing but without the warden service Other Dwellings with alarms: (27,769 dwellings). Only in the small very sheltered category is it likely that the mere fact of living in a special needs house implies that the residents requirements for personal care are automatically taken care of by virtue of the facilities and services provided with the dwelling. For all other categories of special needs housing, (apart from a few cases where it is possible that a warden might provide services akin to personal care), residents in special needs housing will look to the same sources for their personal care needs as any other member of the private household population. PAGE 1 .Axhxxhxhxxhxhxx.........)()();Hklxy .e#O2L\^  X#%&&&&)()))++,,X.q.*////012667<A==j>>>s?@D EEE0GGGLL}LM NԿbU]]a Ucc]^]a a ]uDPa Ua U^a  P]bch P]bc]c uDPU^J NNNOUUU[sssz"{^|F.7LVrzlm;<=JLWXtv./56789<=8uPaP uDPh uDP ]a cuDPcVaacUa UU^a ]ba 8(:;H:;  .Y"BC>?\](  4 h8Vh!f#/>OPckst'12LT\]^ l   ( ##%%t)u),,,.-.//1:1X1r11122447 9d;<<@=A=========== 9f(====>> >>>> >!>(>->4>5>;><>C>H>N>O>V>W>]>c>i>j>u>|>>>>>>>s? 9f 9f$s?w????????@@@@@@$@%@+@,@1@2@8@<@=@D@E@K@L@Q@R@X@\@]@d@e@k@l@ / H #   #$l@o@p@v@z@{@@@@@@@@@@@@@@@@@@@@@@@@@@BBCDDDEEIHIJ7 / H #'JKLLLlL}LNdOQR\T?UUUUXV$WZY[[]]__aaccef+g,giiiTmmqqYsZs[ss&xz7+zzz#{|7Fǁԁ.167LRUVruxyzF̅ͅlmmnol B (op}Sg`ѓ<WP-.9:;<=h`K@Normala c@ Heading 1U@ Heading 2^$@$ Heading 3^a $@$ Heading 4Ua "@" Heading 5Uc"A@"Default Paragraph Font O fm-vol-iss-date1c"O" Normal (Web)dd&O& Body Text 2ha  B@" Body Texta (O2( Body Text 2 ha  @B Footer 9r )@Q Page Number"@b" Footnote Textc &@q Footnote Referenceh&O& Body Text 2h]Oarticle1]bc O Body Text 3U @ Header 9r kx=sv==  N8NO =s?l@Jzo=PQRSTUVWX= !@QTimes New Roman Symbol "ArialTimes-Roman "Verdana"h꣣&꣣&)x=CFree Care for the ElderlyMargaretJ & M Cuthbert  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[`hRoot Entry FN4L_WordDocument~CompObjjSummaryInformation(  FMicrosoft Word Document MSWordDocWord.Document.89qOh+'0  $0 X d p |Free Care for the Elderly<G MargaretH840NormalJ & M Cuthbert20.Microsoft Word for WindoDocumentSummaryInformation8   FMicrosoft Word Document MSWordDocWord.Document.89qxself= Free Care for the Elderlyws 95@G@L.~@T4L@T4L)x՜.+,0@HX` hp xself= Free Care for the Elderly