The Scottish Government

« Previous | Contents | Next »

Listen

5. The Analysis of Costs and Benefits

Introduction

5.1 As has been discussed above, the major problems in constructing a cost - benefit analysis of the SP programme are in the measurement and valuation of benefit. Additional difficulties arise in that SP is a highly diverse programme - in some respects a funding allocation which supports a diverse array of projects with widely variable specific objectives. Finally, as discussed above, the recording of key data makes the assessment of costs problematic both in terms of unit costs and by client group.

5.2 In the Matrix reports benefits were measured as reduced need for public spending. This measure is of interest and is relevant to an assessment but it is not a complete measure of benefit.

5.3 Nevertheless, in the light of the literature review, consultations and the case study work, we consider that there are unifying principles in terms of the aims and objectives of SP programmes which provide a structure for the cost - benefit analysis.

5.4 As discussed above, the fundamental aim of Supporting People is to enable vulnerable people to live independently in the community, in all types of accommodation and tenure. The primary benefits of SP relate to whether and with what quality of life it enables people to live independently. Thus one may say that SP aims to enable people to live independently with an acceptable quality of life. For some groups an acceptable or satisfactory quality of living in the community takes account of their impacts on others - thus for people at risk of offending or at risk of engaging in anti-social behaviour there are benefits to others from altering or managing their behaviour.

Supporting People - Impact Mechanisms and Benefits

5.5 Examination of the structure of the SP programmes and client groups and the consultations and case studies suggest that SP projects can be divided into three main groups or "families" defined in terms of their fundamental aims. These are:

• Projects which enable people to "move on" from a position of vulnerability to long term sustainable independent living
• Projects which give people long term support which enables beneficiaries to live "permanently" in the community with continuing support
• Projects which arrest decline in independence and enable people to maintain independent living for a longer time.

5.6 Broadly, one would expect the first type of support to be of limited duration while the other two forms of support are likely to be permanent. This has implications, in principle at least, for the way in which benefits may be considered - which will be discussed below. The distinction between the second and third types of effect is not exact but we consider the distinction useful. In broad terms, support under the second category tends to be more intensive than under the third, and it is likely that a very high proportion of individuals in this category will be unable to live independently without support. The third category corresponds in many ways to an "insurance" model - it is support provided to reduce the risks of adverse events experienced in independent living to an acceptable degree so as to increases the probability that an individual can stay in the community.

5.7 The manner in which and extent to which the benefits of these three types of SP may be assessed are considered below.

5.8 "Moving on" impacts. Certain types of support - focussed on particular client groups - have the aim of enabling the recipient to move to a position in which support is no longer needed and the person or household is able to live independently without unacceptable consequences for themselves or for others. These programmes do not or should not offer permanent support but their benefits should extend into the future. .

5.9 The most notable examples of this type of project are support to people who are homeless or threatened by homelessness, support to women escaping domestic violence, support to some people with mental health problems and, arguably, support to offenders or people with substance abuse problems. It may be, notably in the case of drug problems or offending, that SP alone cannot achieve the "change" required in behaviour - the role of SP is to enable people to maintain their ability to live in the community while receiving other support with the underlying behaviour.

5.10 The impact of this type of intervention may be quantifiable in terms of numbers of persons enabled to "move on" to a position of sustainable independence. Edinburgh Council uses an interesting risk scale for in relation to homelessness - ranging from actual lack of accommodation through various levels of insecurity to permanent housing.

5.11 It is a feature of this type of support that the effects of the programmes will be on the incidence of adverse outcomes in the assisted groups. For example, not every person at risk of homelessness who is supported by a programme will successfully maintain independent living and not every assisted person who does maintain independent living would have failed to do so without support. The measurement of impact could, therefore, be in terms of increases in the proportion of the assisted groups achieving positive outcomes or the reduction in the level of adverse events experienced by the client group as a result of assistance.

5.12 The impacts of "moving on" may be valued in terms of the costs of the condition from which people are moved can be measured ( e.g. as with domestic violence or, possibly, homelessness). What is perhaps most interesting about this group is that the benefits can be conceived of as an annual stream of benefits resulting from an intervention which is of limited duration. To take an example, support which enables or helps a person homeless person to sustain a tenancy may save the costs of repeated broken tenancies and episodes of homelessness in future. The same principle might apply to support to offenders, people with drug problems or people with alcohol problems. At the same time, it is necessary to recognise that in such instances the change in behaviour or capacity which enables people to "move on" may be due to more than the SP support.

5.13 Permanent Support. For some households a completely independent life in the community will never be possible but life in the community can be enjoyed with the level of support provided by SP either alone or in combination with other support. This situation is characteristic of people with permanent disabilities, people with learning disabilities and some people with mental health problems. The essential point is that the condition which gives rise to the need for support is permanent and changes slowly if at all.

5.14 In these cases the fundamental measure of benefit is the maintenance of community based living with an acceptable quality of life. It is reasonable to conclude that a very high proportion if not of the assisted persons would have an unacceptable quality of life or an inability to live in the community without assistance. No doubt some SP clients in this group would be able to survive independently but with a high (possibly unacceptable) level of adverse events such as increased illness or accidents. Nevertheless, it may be concluded that all recipients of assistance enjoy at least some benefits from the assistance.

5.15 Even if the foregoing argument is accepted, the problem of valuing the benefit remains. One possible approach to valuation would be based on the principle that the benefit of SP support can be measured as the minimum cost of securing an acceptable alternative outcome - which may be residential care or simply the provision of support similar to that provided by SP under a different funding heading.

5.16 In contrast to the "moving on" case, one can argue that for this type of support benefits and costs occur over the same time period. In particular, the support has to be maintained indefinitely if the "alternative" outcome and its costs are to be avoided.

5.17 Arrested Decline. For some groups, support enables people to maintain a certain quality or way of life, or independence, longer than would otherwise be the case. The difference from the previous group is possibly a matter of degree but we still consider the distinction to be meaningful. This is characteristic of support for older people where support can enable people to delay - possibly for their whole lives - the need for more intensive support. In some cases it may even extend their lives.

5.18 In the "permanent support" cases the most reasonable assumption is that a very high proportion of those in receipt of that support could not sustain their quality of life without that, or alternative support. For the groups where the objective is arrested decline it is considered that for any individual it is the probability of maintaining an independent life that is affected by support. Thus one cannot conclude that if support was withdrawn then all the former clients would immediately become unable to sustain an acceptable quality of life: rather, the withdrawal of support would increase the proportion of households who would experience an inability to sustain independence or a deterioration in their quality of life/an increase in the number of adverse events.

5.19 A distinctive feature of these clients in this category is, moreover, that their circumstances can change fairly quickly and generally in one direction. Thus as people age, the probability that they will be unable to maintain an independent life of an acceptable quality rises. Of course, this can be dealt with in various ways and is still frequently addressed by arrangements made within families. However, the general principle can be illustrated in Figure 5.1. With unit costs of support steady, the net benefit of providing support rises with age so that targeting can increase total benefit.

5.20 The benefits of a service which maintains a certain acceptable quality of life for a person will vary between individuals. For some beneficiaries the effect will be to prevent or delay a total loss of independence while for others the effect will be to sustain a quality of life and-or mitigate the effects of age or ill health as reflected in adverse events. This suggests that benefits - and by implication their value - will vary across the client group.

5.21 The Matrix studies recognised this variability of benefits as a general phenomenon in its analysis. With regard to older people the studies argued for the existence of a wide range of effects including items such as reduced need for in-patient care, reduced need for various types of treatment and reduced experience of crime. However, as discussed above, that study counted as benefits of SP a number of impacts for which there was no evidence and, arguably, no logical foundation - though the actual financial significance of many of those events proved very minor.

5.22 The overall social benefit of a service which "arrests decline" is best conceived of as an annual flow of benefits which requires a sustained level of expenditure.

5.23 It will be recalled that the Matrix approach was to measure benefit as the (mainly) public expenditure cost of the outcome prevented by SP - examples being the cost of hospital treatment avoided or the avoided additional cost of residential care. However, the cost of an alternative "outcome" is likely to be, at best, a minimum measure of benefit. It is unlikely that most people would regard the benefit of avoiding a painful injury and medical procedure as being measured by the financial cost of the required treatment. However, we do lack information which would enable the "consumer benefit" of SP services to be measured.

Framework for Cost Benefit Assessment

5.24 Within the limitations which have been discussed above, an assessment of the benefits of the Scottish SP programme is set out below. The approach is focussed on client groups and on the three mechanism of impact described above and is based on the (limited) evidence from the literature (including the other studies) and on the consultations and case studies. The assessment is both qualitative and, as far as possible, quantitative/monetary.

5.25 Cost benefit analysis requires that data on cost and benefits can be compared for the same groups of individuals. As discussed in Section 2, the cost data for SP cannot be allocated to client groups at a highly detailed level but can only be only be analysed at or above the level of the revised "super groups" identified above. However, in assessing benefits alone we can examine the evidence relating to more disaggregated client groups

5.26 Turning to benefits, the estimation or calculation of benefits of SP programmes requires three stages:

• The identification of beneficial changes in the "experience" of adverse events or in the "circumstances" of the client group which can be brought about by SP projects
• The quantification and measurement of those changes
• The monetary valuation of those changes ( i.e. establishing a unit value for each change).

5.27 As has been discussed, each of these stages is problematic and the extent to which each can be achieved varies between client groups and activities. In general, the first step is the one most likely to be successfully implemented. The intended effects or benefits of programmes can usually be described and, in a number of cases, assessed in qualitative terms. However, it is the case that the Matrix studies included a number of expected or supposed benefits which were open to question.

5.28 The degree to which the second stage is capable of being successfully implemented is variable but generally limited. Evidence on the quantitative impact of interventions is relatively rare and the reliability of the evidence which is available is open to question. It was in response to this situation that the Matrix study adopted a largely "what if" approach.

5.29 The extent to which a monetary valuation of benefits can be produced is also variable but generally very limited. The approach adopted in the Matrix studies was to measure benefit of SP in terms of social costs avoided and that reports did deploy this approach in a very thorough manner. However, avoided costs are not the only benefit of programmes - individuals experience improved well-being as a result of programmes and this has a value.

5.30 Table 5.1 sets a summary assessment of the present study concerning the evidence on impacts in relation to the groups for which some reasonable evidence is considered to exist. The table follows the three step process for benefit assessment which was described above in that it is based on the available evidence the nature of the effect of SP projects, the quantification of that impact and the valuation of the impact. A merit of this framework is that it can, in principle, be updated as better information comes to light and the information "gaps" can be filled. It should be seen, therefore, as a framework of analysis rather than a fixed statement.

5.31 For each client group the table details:

• The form(s) of impact ( i.e. moving on/permanent support/arrested decline) which are relevant
• A qualitative description of the nature of impact and benefits created programmes for this groups
• An assessment of the degree to which impacts can be quantified
• An assessment of the available monetary measures of benefit

5.32 The table also includes a subjective assessment of the quality of the information on quantification and valuation - enabling the main strengths and weaknesses of the present information base to be highlighted by group.

5.33 As may be seen in Table 5.1, the data on quantified impacts and values of impacts are generally weak across all groups. It is important to emphasise that the assessments here are essentially comparative - even where data are said to be "good" they do have limitations. In order to conduct a reasonably rigorous analysis of benefits one would wish to see the data for both quantified impacts and values to be classified as good: the only client group which appears to meet this requirement is women fleeing domestic violence..

5.34 There is no clear pattern in terms where weaknesses lie - data on impact are not, on average, better or worse than data on valuation.

5.35 The benefit assessment as reviewed in this table does, however, lead to the view that the beneficiaries of SP can be divided into two broad groups.

5.36 The first group is those who are receiving fairly specific support services (which often come close to the boundary with "care") and, in principle, are receiving support without which they would be unable to maintain their existing quality of life or even live in the community at all. For this group there is little doubt about the existence of impact but much greater uncertainty over the scale and value of the benefit.

5.37 This does assume that no one is receiving the service who could, in fact, manage without it. While it may be unrealistic to assume that there is no case of "unnecessary" service provision, we have no evidence to suggest that this occurs on any scale.

5.38 The groups which we consider to fall into this category are:

• People with learning disabilities
• People with acquired brain injury
• People with dementia
• People with mental health problems (in most cases)
• People suffering psychological trauma (in most cases)
• People with physical disabilities
• Older people (in some cases)

5.39 Insofar as direct comparisons can be made, the Matrix Welsh study - which may be regarded as setting out the "up to date" Matrix position - assumes that without SP a high proportion of clients with Learning Disabilities would otherwise require care or "other" interventions, that dependency levels would rise among people with mental health problems and that a fairly high proportion of older people would require other interventions.

5.40 The second group consists of services - mainly advisory in nature or focussed on support to behavioural changes where the effects of SP are to improve the chances of a "good" outcome or to reduce the risk of a poor outcome. For women escaping domestic violence, SP projects reduce the risk of further violence (though the key question is the extent to which SP project support is needed to enable women to escape abuse). Similarly, advisory services to homeless people reduce the risk of homelessness. However, not everyone in these groups experiences a direct benefit since some (or many) users may not value the support. For this group the evidence on both impacts of SP and value of impacts is patchy and limited.

5.41 The client groups which we consider to fall in this category are :

• Homeless people
• Users of alcohol and drugs
• People with poor social skills
• People at risk of offending
• People vulnerable due to young age
• Women escaping domestic violence (in some cases).

5.42 The Matrix assumptions are somewhat mixed - assumed effects on "adverse events" for people with drug and alcohol problems are modest (the 5% default) while effects on offending are also assumed to be modest. However, effects on the experience and costs of homelessness assumed by Matrix are substantial.

5.43 It must be stated that no very strong evidence base is brought forward by either Matrix to justify the assumed effects of SP on behaviour.

5.44 The assumptions and estimates made in relation to benefits for client groups in the present study are considered below. The "client numbers" referred to in this analysis are those recorded in the Scottish Executive statistics. As discussed above, we recognise that these estimates are subject to some degree of imprecision since individuals have multiple characteristics and the client group in which they are classified is, to some extent, arbitrary. Nevertheless, there are no alternative data on the distribution of SP clients between groups.

Table 5.1 Client Group Benefit Measures - Summary

ANALYSIS OF BENEFITS BY GROUP

5.45 People with learning disabilities This group comprises just under 7,000 people. In England the Matrix study found that this group accounted for a very substantial element of SP expenditure and a high unit cost but that the benefits of this expenditure in terms of "avoided cost" were relatively low (barely 10% of costs). Most of these savings came from assumed reductions in spending on health care based on arbitrary assumptions. Given that the major financial benefit "driver" in the Matrix methodology involved the avoidance of residential care costs, that study evidently concluded that most members of this group would be able to remain in the community without SP support.

5.46 The Matrix study in Wales concluded that SP had a powerful effect in helping people to live independently, with the alternative for many households being residential care. This study indicated that the households which did continue to live in the community without SP would experience many more adverse effects. These assessments were based on the results of consultations with SP lead officers and other stakeholders.

5.47 The work of the present study has concluded that this client group in Scotland shares key cost characteristics with the English client group. In Scotland as in England the group has by far the highest unit costs of all client groups (approximately £16,000 per client in Scotland and £12,000 in England) with costs many times the programme wide average.

5.48 Consultations conducted in the course of the study with providers of services and local authority lead officers confirmed that these costs reflect intensive levels of support - staff costs per hour and hours of support per person per period being the main determinants of unit cost. While levels of learning disability vary, the consultations also supported the conclusion that many if most people in this group would not be able to maintain independent living without intensive support.

5.49 It is concluded that virtually all of the people in this client group in Scotland receive significant benefits from the programmes in that the support provided is, for most, critical to their ability to live in the community. That said, it is uncertain how many of the people receiving SP would have to move into some form of care if support was withdrawn, and how many would remain in the community albeit with a greatly increased burden on family members.

5.50 The English evidence is that about 60% of people with learning disabilities live with family members (parents) ³⁷ and that most people with learning disabilities who do not live with family members are in care. The Matrix analysis included consideration of an alternative benefits estimate in which the expected impact of withdrawal of SP would be a large increase in the requirement for residential care.

5.51 It is the conclusion of this study, based on the consultation work and the data review and taking account of the Welsh Matrix study, that without SP (or equivalent support) a high proportion of clients would require residential care. The Matrix study ³⁸ reports that the average cost of a residential care place for this client group would be at least £16,000 per annum. If we assume that 50% of the client group would require such care without SP then we have an "avoided" cost of £56 million. If the proportion is 75% then the cost saving is £84 million.

5.52 This is by no means the only benefit of SP. It is generally held that the quality of life for someone living in the community with support is, all other things being equal, superior to that for someone living in an institution so that SP can enable people to enjoy a better quality of life. SP can also provide benefits to carers (generally family members) in terms of a reduced burden or care. This may amount to 20 hours per week in some cases. This "released time" might enable people to take up employment and could, in principle, be valued in terms of earnings which would be "lost" without SP. However, there is a lack of information on the situation of carers and we cannot say how many are able to work because of the availability of SP. Nevertheless, it is not unreasonable to assume that the value of this time "released" would be substantial and could represent a significant offset to the corresponding economic cost of the service.

5.53 As argued in the English and Welsh studies, it is likely that if persons with learning disabilities did continue to live in the community without SP support then those persons would experience higher levels of adverse events such as illness or injury but it must also be acknowledged that estimates of such effects are highly speculative. There may also be other effects on quality of life from SP support which cannot be readily quantified.

Individual Case Study - Person with a Learning Disability

Jackie is in her fifties and was living in a group home (registered as a care home at the time) with 4 others. She has a learning disability and has severe eczema which regularly led to her being admitted into hospital for treatment. Her eczema was so severe that she has experienced problems with her sight. Jackie is someone who has always found it difficult to live with others. She has been described as having 'challenging behaviour' around relationships with others. Her eczema and her living arrangements in turn were causing Jackie to become depressed.

Through a process of person centred planning, Jackie, her family, social worker and support staff who knew her well, agreed that Jackie should be supported to live in a flat of her own.

Jackie now has her own flat. She receives support from staff during the day and she can access an over night worker if required.

Due to the change in her support arrangements, Jackie's stress levels have significantly reduced and this has resulted in her eczema clearing. Jackie has not been admitted to hospital for over 2 years. She looks better and as such she is more confident and able to take part in many aspects of life.

Jackie has developed many new skills, such as looking after her home, paying her bills, managing her money, making meals, organising property repairs and accessing many different community facilities and resources. She has also started a part time job.

Jackie still finds it difficult to maintain relationships but continues to work on this area with staff support.

5.54 People withdementia The group consists of people whose condition will not improve and who require support to enable them to maintain life in the community. People with dementia and their carers can access a range of support through local authorities and voluntary organisations. These other sources of support provide the care that SP cannot provide and are may be more critical in maintaining independent living. According to the official data, about 1,300 people with dementia receive SP support. However, as noted in section 2, there may be many more people with dementia than this receiving SP support - the problem being that they have been classified under some other client group (probably older people). There is, however, no alternative estimate of the total number of people with dementia receiving SP support in Scotland.

5.55 It is estimated by Alzheimers Scotland that there are 63,000 people with dementia in Scotland so even if the 1,300 figure is an underestimate it appears that SP is not, in itself, a major source of support for this group.

5.56 The withdrawal of SP would lead to an increase in the numbers of persons with dementia requiring residential care and would affect the quality of life of clients and their carers in the cases where people continued to live in the community. There is no firm basis for monetary valuation of these latter benefits. Costs and benefits for this specific group were not analysed in the English or Welsh studies.

5.57 People withacquired brain injury This group is very small - just over 300 people. Brain injuries vary greatly in nature and severity and most do not lead to severe permanent impairment. However, it is reasonable to assume that those sufferers receiving SP support are at the more severe end of the spectrum (though obviously not completely incapacitated).

5.58 The Scottish Head Injury Forum estimates that 16,000 Scots each year suffer a brain injury and that in 5,000 cases this leads to some form of disability. It is evident that SP is a minor factor in support for this group.

5.59 There limited published evidence on the benefits to this group though it is likely that the assistance is being targeted on people who do require support to enable them to maintain a good quality of life. The costs and benefits for this group were not analysed in the English or Welsh studies.

Individual Case Study - Acquired Brain Injury

Claire is 48 and lives on her own in her own tenancy. In November 2005 she was referred to the Brain Injury Team with brain infection (encephalitis). Initial visits involved the Brain Injury support worker and social worker with the Team's assistant psychologist in order to gain multidisciplinary perspectives. Initial problems presented were a severe memory deficit and low mood.

Claire presented with global memory problems and was confused as to what bills she had and what needed paid. Her home was in disarray and she kept no food in the house.

Claire has been using weekly support since November. The support delivered has included assessment of daily living skills, coordinating and supporting attendance for appointments (including the housing department), help with security and upkeep of her home, reading correspondence and helping support responses, support to carry out weekly shopping and meal preparation as well as medication prompts and help to keep in contact with family and friends.

The support has helped to prevent the loss of her home due to debt arising from unpaid bills as well as prevent hospitalisation due to poor eating habits and further health deterioration. In addition Claire has used the support to help conquer the social isolation she has been at risk of and reduce the risk of exploitation she may be at risk of due to her vulnerability.

Claire's support network is being increased as her health continues to deteriorate. Further work is anticipated to ensure she can remain in her community. This work is multidisciplinary, with the Brain Injury Support Worker, Supporting People funded, being essential to the provision of support and key to the co-ordination of all required services.

5.60 People with psychological trauma This group comprises just 73 people. Little is known about the impacts of SP on this group. They may best be considered a sub-set of the group of clients with mental health problems.

5.61 People with mental health problems This group comprises about 6,500 persons and is one of the larger client groups. Mental health problems cover a wide spectrum and are often transitory. The literature and the consultations suggest that the main benefits of support to this group lie in avoiding the need for residential care, preventing episodes of homelessness and quality of life improvements. There may also be benefits to carers.

5.62 The Matrix study for England indicates that most SP support to this group is "low level" ³⁹ but the evidence on unit costs shows that this is a relatively high cost area of support.

5.63 The Matrix study made fairly arbitrary assumptions concerning the degree to which SP prevented people being taken into care (about 10% of cases) and the impacts on homelessness. Although there is doubt over the scale of these effects, the effects themselves are plausible. The Matrix study also made assumptions concerning impacts on physical health and crime which are less obviously well -founded.

5.64 The assumptions in the Welsh study were effectively the same as those in the Matrix study in England.

5.65 The view that people with mental health problems are at high risk of homelessness is well founded. Various research studies have shown high levels of mental illness among homeless people and Matrix quote an Audit Commission study which shows that almost half the people identified as living in the community with mental health problems face eviction problems. If SP reduces this risk substantially, we might expect that it prevents a substantial number of episodes of homelessness each year

5.66 Other benefits to this group include quality of life improvements and possible benefits to carers.

Individual Case Study - Person with a Mental Health Problem

Alice is 49. Before receiving a visiting housing support services from the voluntary sector, Alice was drinking heavily. She had a chaotic lifestyle, poor physical and mental health, resulting in memory problems and confusion. Her teenage daughter was looked after by the local authority because the situation at home was too difficult.

A year later, Alice has been alcohol free for 8 months. Her home environment has improved greatly and she has been supported to deal with her debts. Her health and her memory have got much better. Most days, Alice gets out of the house, with support, and is now looking at local college courses. Alice's daughter is back at home with her mum.

5.67 People with physical disabilities This is the second largest group with just over 9,000 clients. Disability problems cover a wide spectrum and vary in intensity and severity. The literature and the consultations suggest that the main benefits of support to this group lie in avoiding the need for residential care, preventing episodes of homelessness and quality of life improvements. There may also be benefits to carers.

5.68 There is, however, little evidence on the impact of SP on the life experiences of this group, though it is believed to overlap with the older people group. The group was not included in the English or Welsh studies.

5.69 A high proportion of people with physical disabilities who receive care at home are elderly and it is likely that for many of these the absence of SP support would compromise their ability to live independently. That said, almost all admissions to care homes by people with physical disabilities are for respite not long term care and we would expect impacts to be on respite care.

5.70 Older People. This is by far the largest SP client group. The data from the Matrix studies provide useful insights in relation to this group. We consider that the principal benefit of support to these groups rests in reduced, or more accurately delayed, admissions to more intensively supported accommodation. That is to say, SP helps people living in the community to avoid or delay a move to sheltered housing or to residential care while SP support to people in sheltered or very sheltered housing enables some of those people to avoid or delay a move into residential care. This can be expressed as a reduced probability of an individual requiring to move into more intensively supported accommodation in a year.

5.71 SP support is provided to 83,000 older people in Scotland each year. The population over 65 in Scotland is about 900,000 so that, as in England, just under 10% of people of retirement age and above are in receipt of SP support.

5.72 Annual admissions to Scottish care homes in 2002 for people admitted as "older people" were 11,887 for short stay/respite care and 4,600 long stay admissions. It is to be noted that long stay residents - who are almost all over 75 - are rarely discharged for reasons other than through death, hospitalisation and transfer.

5.73 Lettings of special needs housing (which are not only to older people) in the Scottish social sector amount to around 2,500 per annum.

5.74 We have set out these numbers to test the plausibility of the Matrix "5%" assumption (found in both the English and Welsh studies). If indeed 5% of SP older clients were to become more dependant were the service to be withdrawn then the implication is that 4,000 new admissions a year would be required - split between sheltered housing and long term care (it is assumed that the 5% impact is only need for long term care though it might be argued that part of the impact is on short term admissions). In the broadest terms this represents an increase of almost 60% in the combined total of special needs and long term residential care lettings.

5.75 To consider this from another perspective - at present an individual in the Scottish over 65 population has a 0.4% chance of becoming a long term care resident in any year or a 0.27% chance (at most) of moving into sheltered housing. If we assume that the effect of SP is to reduce these risks for SP clients to zero in any year (an extreme assumption) then the figures for the "non SP assisted" population become 0.6% and 0.3%. If we then assume that withdrawing SP would mean that those formerly assisted had the same risk factor as the rest of the older population then the effect would be to produce from 83,000 people about 500 long term care admissions and 250 moves into sheltered housing (moreover, some of the moves into residential care would be from sheltered housing). In total this is 0.7% of the client group moving into more intensively supported accommodation.

5.76 These figures are far lower than the Matrix numbers. Moreover, the Matrix figures envisage all moves to more supported accommodation being to residential care but about half of the older people in Scotland receiving SP are in mainstream (not sheltered) accommodation and withdrawal of SP support might, therefore, merely result in people in this group moving to sheltered accommodation rather than to care. The assumption is that these people could live in sheltered accommodation withoutSP support. Even if we allow for the possibility - indeed the probability - that SP is already targeted on the most vulnerable older people, the Matrix assumptions look high.

5.77 We consider that a more credible assessment would be that withdrawing SP would result in the probability of a former client entering residential care being 0.02 per annum ( i.e. a 2% chance) - which is not a conservative assumption given that the part of the Scottish older population not receiving SP has, at most, at 0.6% chance of entering long stay care in a year and a 1.8% chance of entering residential care of any kind. The cost implications of moves to more sheltered housing are not clear but we might assume that the chance of moving to sheltered housing from mainstream accommodation is increased by 0.005 ( i.e. 0.5%). The Matrix argument that some of the people who would stay put would require some form of continuing support appears credible. A complicating factor, which must be noted, is that the viability of much sheltered housing in the absence of SP is uncertain as SP is a main source of funding for the services associated with sheltered housing. It is possible, therefore, that some of those people assumed to move to sheltered housing would, in fact, have to move to residential care.

5.78 There is also evidence, notably from the Telecare study mentioned above, that SP support can substantially affect hospital discharges and "bed blocking". If we take the average of about one hospital admission for every two older people per annum and assume an average reduction in length of stay of 1 day then we have an estimate of about 40,000 days saved per annum. There will be other impacts of SP on adverse events and on quality of life - though these are hard to assess far less quantify.

Individual Case Study - Older Person

Peter is a 69 year old man with respiratory and alcohol problems. He was referred to the housing support service via Occupational Therapy, for assistance with financial planning, dealing with correspondence and resettlement issues around a move into sheltered housing.

Peter was supported to take control of his own money and formalising arrangements, such as the use of a bank account. The support worker also provided assistance to make sure that Peter was receiving the correct benefits. This resulted in Peter receiving some additional money.

The support worker also assisted Peter to contact the councils gardening service, as the over- grown garden had been causing some disputes with the neighbours. Domestic support was also provided, to enable Peter to take more responsibility for his home.

This level of support continued until he was allocated a flat in a sheltered housing development. Support was provided to enable Peter to move from his house and move into his new home.

Peter has not consumed alcohol for over 10 months.

When asked about his life and how he has changed, Peter said 'I didn't know I had it in me to cope with this side of my life again and it's down to the support I've received from my worker.'

5.79 Homeless People. Support is provided to 28,000 people who face the risk of being or have recently been homeless. The benefit of this service lies in reducing the risk of homelessness. The cost of an episode of homelessness has been estimated in various ways to include such items as the cost of emergency accommodation and the cost of broken tenancies. However, it is unclear how far these estimates measure net costs - e.g. allowing for the subsidy cost of providing social housing which is the usual route out of homelessness.

5.80 We also lack very firm evidence on the effectiveness of SP services to the homeless. Given the aim of the SP projects, we might expect one impact to be on repeat homelessness. While data on repeat homelessness in Scotland over the last few years show a reasonably sharp reduction - from 10% of cases of homelessness leading to a repeat episode down to 8% - this difference is almost entirely due to changes in Glasgow and may reflect data collection changes. The Matrix study in England assumed that repeat homelessness would halve as a result of SP support, rising to a reduction of 80% in the Welsh study - though not on the basis of any evidence.

5.81 Certainly SP is being directed to a high proportion of homeless people - support was provided to 29,000 people in 2004-05 which compares with 40,000 assessments of homelessness in that year so the programme is large enough to impact on the aggregate figures.

5.82 The appropriateness of the assumptions on repeat homelessness in the English and Welsh studies to Scotland is unclear. For example, if we take the recent reduction in repeat homelessness as the maximum impact of the SP programme then we might conclude that the impact of SP is to reduce by 20% the incidence of repeat homelessness for those supported.

5.83 The effects of SP are not, however, confined to repeat homelessness. The Matrix work argues that 30% of those persons who are temporary accommodation and who receive SP assistance are thus enabled to obtain permanent housing and that 30% of rough sleepers who receive SP support are enabled to move to temporary accommodation. East Ayrshire council provided the study team with data on the impact of housing support on the ability of people to sustain tenancies and also stated that without support many of the assisted households would not have been housed at all. In the light of the English and Welsh studies and the Scottish evidence we the study team concluded that it would be appropriate to base benefit assessments on the assumption that the effect of SP is such that 40% of those persons assisted would otherwise be homeless. This assumption has been applied to other client groups where a major part of the impact of SP is on the incidence of homelessness ( e.g. including people who abuse alcohol or drugs).

5.84 Estimates of the cost of an episode of homelessness vary. In the Matrix study ODPM research is cited to indicate that the social cost of an episode of homelessness is £630 per person or about £1,800 for a family of three. Further benefits of about £360 per person (about £1100 for a family of three) arise from supposed effects on health. At the same time, the literature review undertaken for this study has found examples of costs per year of homelessness for households ranging from £9,000 to around £40,000. However, the literature review also clearly showed that very high costs were associated with cases where people had complex and severe problems which extended beyond the simple lack of suitable accommodation the literature suggests that many, indeed most, instances of homelessness last much less than a year and that the majority of persons who are classified as homeless do not experience the complex problems exemplified in the "high cost" cases.

5.85 High costs do arise where homeless people have to be provided with temporary accommodation, especially if this is for a prolonged period. However, most persons classified as homeless are living with friends and family and for the majority the alternative to being housed is not to be placed in temporary accommodation. The study team examined data provided by Glasgow City Council for 2005/06 ⁴⁰. The data showed that of persons classified as being in priority need just 30% became social tenants while only 10% were moved into temporary accommodation - most found other solutions or became "lost contacts". Of course, it is reasonable to assume that people helped by SP have more complex needs than the "average" homeless person in the first place.

5.86 Homelessness is certainly known to be associated with increased levels of illness and even crime and substance abuse. The Matrix studies made estimates of these impacts - though the supporting evidence is very limited.

5.87 The view of the present study team is that a cost of an episode of homeless prevented by SP of £5,000 is a reasonable estimate. While we make separate allowance for certain health care costs this figure does allow for the possibility that SP clients are more vulnerable than average are more likely than average to require temporary accommodation or to experience personal problems as a consequence of homelessness.

Individual Case Study - Homelessness

David is 16. He has misused drugs and alcohol and has a history of violent behaviour. He was thrown out of the family home by his father after persistent arguments.

In the unit providing Housing Support Services to homeless young people, David has been assisted in developing his independent living skills. He has been supported to learn how to plan his shopping, prepare meals, carry out housework, learn to budget and has been supported to access the appropriate benefits to which he is entitled.

David's support worker has encouraged him to make contact with his family - and he is now in regular contact with them, although returning home to live is accepted as not being an option.

David expressed an interest in becoming a tradesman, such as a joiner or plumber. He has therefore been given support to access a construction training scheme.

David is developing skills and confidence and continues to work towards moving on to a tenancy of his own.

5.88 Users of Alcohol and Drugs. The benefit of SP support to substance abusers is presumably first in terms of increased stability of tenancies and reduced episodes of homelessness. This presumably has a further benefit in supporting the processes of behaviour modification which may mitigate other effects of drug and alcohol misuse such as criminal activity and its consequences. That said, we do not know what contribution SP makes to the maintenance of such tenancies and we do not know how much impact stable tenancies have on success in rehabilitating drug users.

5.89 The Matrix studies assume a fairly large impact on homelessness - a reduction of 25% in homelessness among SP supported people with drug problems. The Matrix studies assume that without SP drug users would live in the community but experience many more adverse effects such as hospital admissions and offending behaviour. The English Matrix report cites research which indicates that the annual social cost of a problem drug user is £11,400. The reliability of these estimates is uncertain but if we take the Matrix figure as a benchmark, we can say that the benefit of preventing a person in this group from losing a tenancy consists of the following avoided costs:

• The cost of a broken tenancy and of consequent homelessness
• The impact of a broken tenancy on the incidence of "adverse events"
• The cost of adverse events (including impacts on quality of life).

5.90 People with poor social skills. We consider that this group have certain parallels with substance abusers. The primary impact of SP will be to reduce homelessness. There will also be some effect on anti-social behaviour with consequent benefits but we cannot assess these effects. As above, it is expected that these vulnerable groups are more likely to experience repeat homelessness than are the general population.

5.91 People at Risk of Offending. We consider that the effect of SP on this group is to enable people to maintain tenancies and indirectly to contribute to the likelihood that the person will be helped to avoid re-offending. It is likely that this group will experience more repeat homelessness than are the general population. Reducing homelessness will also have some affect on offending. The Social Exclusion unit report on Reducing Re-offending by Ex-prisoners ( ODPM 2002) reported a high correlation between homelessness/insecure housing and re-offending. However, it provided no clear guidance evidence on the extent to which improvements in the housing position of potential offenders would reduce offending.

5.92 The Matrix studies assume that most adverse events are reduced by 5% while tenancy failures and homelessness are reduced by 50% and 33% respectively.

5.93 The benefit of preventing crime can be held to be broadly equivalent to the cost of convicting and incarcerating/punishing an offender. This is essentially the approach taken by the Matrix study, relying mainly on the Social Exclusion Unit report cited above, and producing a figure of £81,000 per ex-offender prevented from offending over two years.

5.94 This calculation appears to assume that criminals only commit the crimes for which they are convicted - which is self-evidently untrue. Professor John Farrington of Cambridge University has argued that the effect of unsolved and unrecorded offences is such that a year's offending could amount to £148,000 while Home Office calculations suggest £100,000 per annum. We consider that the evidence in the literature and from our consultations- while qualitative rather than statistical - supports the view that prevention of homelessness has a substantial effect on reducing anti-social behaviour and related "adverse event" and this is built into our benefit assessment..

5.95 People Vulnerable due to Young Age. We consider that this group is essentially a sub-set of those at risk of homelessness. We consider that the effect of SP on this group is to enable people to develop skills to enable them to sustain their own tenancy/live independently.

5.96 Women at Risk of Domestic Violence. This client group is perhaps the most soundly researched in the Matrix study and we see no reason to depart greatly from the approach of that study. The main effect of SP is to enable women (and their children) to escape abusive partners. We are content to adopt the Matrix estimate of the cost savings related to health care and crime. We do have concerns over the assumption that SP reduces the risk of women suffering violence by 80%. That figure is derived from the evidence that in 80% of cases there is no further assault after a woman separates from an abusive partner and implies that this separation would not be possible without SP. This assumption is inherently questionable and seems especially inappropriate in Scotland, where SP is only part of the mix of funding for women's refuges. We suggest reducing the figure to 50% ( i.e.SP is given "credit" for that proportion of the impact of separations)

SUMMARY OF COST AND BENEFITS

5.97 We can now draw together the estimates of the annual benefits of SP for the groups discussed above. This is done below in Table 5.2. The table sets out those benefits which, it is considered, can be measured with reasonable confidence in monetary terms and lists the main benefits which are considered to exist but for which monetary measures cannot be established. As has been made clear above, the quality of the evidence in relation to the impact of SP limited in almost all cases and very poor in some. Table 5.2 details the assumptions which have been made. These are based on the literature - including the other studies - and on consultations. The general approach has been to confine monetary valuation to those impacts and benefits where it is considered reasonable evidence exists and not to include hypothetical or poorly supported impacts. This means that the monetary valuations of benefits are considered to be fairly robust and relatively conservative.

5.98 As explained above, we cannot conduct a cost benefit analysis at the level of these groups as the cost data cannot be presented at this level. We have, therefore, presented the cost and benefit assessment at the level of the "supergroups" in Table 5.3.

5.99 We have also considered the consequences of a "wholesale" acceptance of the assumptions and arguments in the Matrix studies. This is not straightforward as there are differences between England, Wales and Scotland in the client groups covered. Nevertheless, Table 5.4 sets out an estimate of the Scottish benefits based on the assumptions of the Matrix Wales studies.

5.100 The main results indicate that financial benefits equal to about 110% of costs can be identified for these groups (who represent most of the SP budget). This far from being the total measure of benefits - there are very substantial unquantifiable benefits and the results indicate that SP is good Value for Money.

5.101 However, it must also be acknowledged that we cannot be certain that SP projects are delivering services at the lowest possible cost or best possible Value for Money in every case- that assessment would require extensive project level analysis.

5.102 Benefits exceed costs for three of the four groups and are close to costs in the other case. However, we recognise that there are many uncertainties inherent in the estimates of impacts of SP and that evidence is fairly limited on the effects programmes which aim to affect the behaviour of client groups ( e.g. drug abusers, potential offenders).

5.103 It may be noted that there are also various projects aimed at a tiny proportion of the potential clients in Scotland. These projects may be worthwhile in their own right but it is difficult to make any assessment in the present study. It is certainly the case that such "one off" projects are hard to relate to any strategy for SP.

5.104 The areas in which we may have most confidence in the effects of net cost-benefit of SP are in relation to its core client group of older persons and persons with various types of disability. There is relatively strong quantitative and qualitative evidence of substantial benefits for this group. As with the other three studies, it is clear that the major driver of financial benefits from SP is in keeping people out of various forms of residential care.

5.105 As Table 5.4 shows, the adoption of the Matrix (Wales) assumptions produces a slightly higher overall figure for benefits but most of the difference is accounted for by the figures on women escaping domestic violence and people threatened with homelessness (both in the Homeless supergroup). We have explained above why we do not consider that SP should be credited with all the benefits which accrue from women being able to escape an abusive partner and the Matrix study also assumes that the prevention of homelessness has a big impact on a range of adverse events. Overall, however, the present study's results are within the same broad range as those of the Welsh study.

Table 5.2 Monetary Value of SP Benefits

Table 5.3 Monetary Costs and Benefits by Supergroup

« Previous | Contents | Next »