Annex 3 - Potential Cost Consequences Methodology & Literature Review
One of the four key research aims and objectives of the project is to consider if the Intensive Family Support Projects offer value for money.
This Appendix starts by describes the cost consequences methodology adopted in the economic evaluation and the reasons for taking this approach. It then provides a brief overview of the evaluations of a number of other Intensive Family Support Projects ( IFSP), focusing mainly on the costs associated with them and on some of the methodological problems encountered when undertaking this work. These include the Dundee Families' Project ( DFP), the Shelter Inclusion Project ( SIP), several other projects providing intensive support to families in northern England, and an Edinburgh-based project providing an early intervention case management service to families with young children. Although each project had its own specific acceptance criteria, most worked with vulnerable families exhibiting anti-social behaviour at risk of losing their tenancy and becoming homeless. Most families also had significant health-related issues (e.g. mental health problems; addictions to drugs and/or alcohol) and child protection was also often a major concern.
One of the biggest difficulties associated with evaluating projects with multiple objectives where the benefits may occur over many years is trying to determine what might have happened to the family in the absence of the intervention. Although family-specific estimates cannot be made of the services that they may have otherwise required (and of the costs associated with these services), it is possible to identify a number of possible short-term and longer-term consequences and their associated costs. This Appendix includes information on the costs of a variety of such consequences, drawn from Scottish data sources (where available) and supplemented by data from research in other parts of the UK. This element does not provide a comprehensive review of this literature, but it does identify its main implications for the economic evaluation element of this study.
Economists have traditionally used a variety of approaches when undertaking economic evaluations:
- Cost-minimisation analysis
- Cost-effectiveness analysis
- Cost-utility analysis
- Cost-benefit analysis.
The health service was one of first major public sector services to adopt such approaches and health economists have developed a variety of ways of evaluating health care interventions. Economic analysis is now an integral part much research in this area 25. Cost-minimisation analysis considers the intervention that achieves a specific outcome at the least cost (in terms of its use of scarce resources). Thus if home measurement of blood pressure is shown to be as effective as measurement by a GP or by a Practice Nurse, the cheapest approach should be adopted. Cost-effectiveness analysis values costs of interventions in monetary terms and compares these with a single primary outcome (e.g. the cost per life year gained for hospital dialysis versus kidney transplantation after renal failure). Cost-utility analysis is a specific form of cost-effectiveness analysis in which outcomes are measures in terms of Quality Adjusted Life Years ( QALYs). This approach recognises that people may prefer a few additional years in good health to a greater number of additional years in poor health. Cost-benefit analysis values all of the outcomes from an intervention in monetary terms as well as the costs. It has been used by planners to evaluate large public programmes (e.g. the Channel Tunnel) where the impacts may be felt across many different sectors of the economy.
Given the range of outcomes associated with the projects, the evaluation has adopted a form of cost consequences approach (as described by Coast (2004) 26) for the economic evaluation. This methodological approach identifies and tabulates all of the relevant costs (which reflect the value of the resources used) and the possible consequences associated with a particular intervention. It clearly shows decision makers what is included and excluded from the analysis and where information is quantitative and qualitative. No attempt is made to combine the costs and benefits formally into a single measure (as done in traditional economic evaluations), but decision makers can use the assembled information to draw their own conclusions. Therefore the overall objective of this economic evaluation is to consider the values of the resources used and possibly prevented from being used that are associated with the Scottish Intensive Family Support Projects to determine the extent to which the approach offers good value for money.
The economic analysis falls into several stages:
- Identification of project-related activity;
- Identification of resource use and associated capital and revenue costs by the projects and by other agencies and organisations;
- Identification of relevant unit costs of delivering the projects;
- Identification of short-term cost savings and possible long-term cost savings (i.e. the avoided cost consequences).
i) Identification of Project-related activity
Much of the information collected during the other elements of the evaluation (e.g. project aims and objectives; selection criteria; success and case closure criteria; facilities; staffing levels; anticipated 'steady-state' caseloads; family composition; referral sources; information about services used by family members at referral) provide important contextual information for the economic evaluation. However, the projects have also provided specific information for all referrals about dates of referral, dates of acceptance or rejection, and case closure dates.
This information provides important details about the length of time for which families are in contact with the projects. There are two main reasons why families do not proceed to work with a project:
- They do not meet the project's acceptance criteria and are therefore rejected by the project;
- They meet the acceptance criteria, but do not want to work with the project.
For some families, the time between receipt of a referral and the decision to reject a family will be relatively short (e.g. if the family clearly does not meet a project's specific referral criteria). In some cases the project will need to gather additional information before it can decide whether or not a family meets its acceptance criteria, which can take several weeks. However, even when families meet the referral criteria and are offered support, they will often work with the project for a while before deciding whether or not to 'sign up' to participate. It is important that this time is captured in the evaluation, due to its opportunity cost (i.e. the required project resources cannot be used for any other activities).
The date of closure for those families that agree to work with a project determines the amount of time for which a family is 'on the books' of a project. Cases tend to close for one of two reasons:
- By mutual agreement with a successful outcome, where the objectives have been achieved;
- The family ceases to comply with the project's requirements (e.g. they refuse to attend key meetings) - in such cases some benefits may have been achieved.
Data on the number of contact months, and their dates, are key elements of the economic evaluation. Although the amount of contact families have with the projects is likely to vary from week to week and month to month, these variations are not included in the analysis. This is partly because such data is very time-consuming for staff to gather and collate for all families, but also because it is the macro-level average total costs of the families that are relevant for decision making about projects' cost-effectiveness. However, local analysis of family-specific micro-level data on aspects such as contact levels with project staff can be used for project audits to help managers determine how resources are being used and whether they could be used more efficiently. Such audits can show, for example, whether families tend to be relatively time-intensive initially, with a gradual reduction over time to case closure, or if families' needs for staff support are more erratic and unpredictable. This information can be helpful when planning staffing levels and staff workloads.
ii) Identification of resource use and associated capital and revenue costs
- Costs incurred by the Projects
The primary source for identifying the capital and revenue costs of the resources used by each project was their annual (audited) income and expenditure accounts and their budgets, financial projections and financial monitoring for the current financial year. Each project was asked to provide copies of audited accounts for completed financial years (e.g. for 2005/06 if relevant, 2006/07 and (when available) 2007/08) and regular (e.g. quarterly) updates of expenditure against budget during 2007/08. Copies of any other relevant finance-related documents (e.g. Business Cases; internal financial reports) were also requested for each project on an ongoing basis.
The three Breaking the Cycle demonstration schemes only have revenue costs, as they only provide outreach support. Ideally, for comparative purposes, separate capital and revenue costs would be available for the two Families Projects (i.e. Dundee and Aberdeen) with core units, along with separate revenue costs for their outreach/dispersed services and core accommodation. However, in reality the revenue costs of providing core and outreach services are hard to separate, as staff work across both services and buildings are used to provide support to all families using the service.
Two other significant aspects relating to the use of resources also need to be taken into account to identify the true costs (as opposed to the accounting costs) of running the projects:
- Central overhead/support costs;
- Payments in kind.
To identify the true resource-related costs of running the projects, it is important that the costs of senior manager time provided by staff not specifically funded by the projects and the costs of centrally-provided activities such as staff recruitment, financial management, staff training and central administration are included. Projects may also benefit from staff seconded from (and paid by) other departments/organisations (e.g. social workers; youth workers) or from the provision of facilities such as rent-free office space. These resources generally have an 'opportunity cost' (i.e. they could have been used for other activities), especially the staff. It is therefore important to identify (e.g. through discussions with Project Managers) the extent to which projects benefit from such 'free' resources so that their true cost can be included in the financial cost calculations for each project. This is relatively straightforward for staff, whose salaries (plus relevant on-costs) can be used, but is usually much harder with facilities, as it may not be possible to identify a 'market rent' for such buildings (as they may otherwise have been unused or used to reduce pressures on space elsewhere). It is also be important to try to identify the value of the resources contributed by any other organisation which works closely (e.g. in partnership) with a project and is crucial to its operation.
Costs incurred by other agencies pre- and post-intervention
Many of the families will also generate additional Exchequer-related costs due to their participation in these projects. For example, a mother may be encouraged to receive help for previously untreated depression and a child may be placed in a pupil referral unit to start to address their schooling problems. The evaluation of six intensive family support projects in northern England (Nixon et al, 2006) found that families were generally making less use of most services when their cases were closed than they had been doing when their cases opened, although more use was made of youth workers and mental health workers when the intervention ceased. Many of the families are very vulnerable and continue to be so after their immediate tenancy and anti-social behaviour problems have been addressed.
Case notes for closed cases were studied during the second part of this evaluation to try to determine the extent to which families' use of other services had changed before and after the intervention. Where possible, the impact on other resources was estimated. However, it must be remembered that service use may alter for reasons beyond the scope of the projects' influence. It will also be influenced by local availability of (and possibly waiting lists for) many services (e.g. child and adolescent psychiatry). It can also be argued that, due to their needs, the families should already have been receiving these services. Therefore, although the interventions may result in additional short-term costs for other agencies and services, it is expected that these costs will be far outweighed by the longer-term benefits of project participation.
One of the problems associated with evaluating new initiatives is the lack of information about their longer-term effects. The researchers who undertook the evaluation of the six intensive family support projects in northern England were able to follow-up some of the families several months after their cases had closed (Nixon et al, 2007). Given their fundamental vulnerabilities and the fact that families' needs inevitably change over time, some of the benefits achieved at closure had been lost, resulting in them placing new demands on various services and agencies. However, it is important that any increased use of other services to address problems and difficulties is not seen as a sign of failure by the intervention. Indeed, getting families to recognise their needs and to seek help may be a sign of success.
(iii) Unit costs of project delivery
Two main unit costs are calculated:
- The average cost per family contact month;
- The average total cost per closed case.
Both of these have been used in other evaluations of similar initiatives and provide a sound basis for monitoring the development and progress of the projects. Values are calculated for each financial year, which enables comparisons to be made over time (although many evaluations tend to focus on relatively short timescales, often before projects have achieved 'steady state').
The average cost per family contact month is determined by dividing the (true) annual cost of delivering the project by the number of family contact months provided over the year. For example, a project working with an average of 10 families each month over a year delivers 120 family contact months. If the project's annual cost is £240,000, then the average cost per family contact month would be £2,000. The activity data provided by the projects enables the actual months of contact with each family to be used in the calculation.
The average cost per family contact month in a specific year can be compared with the estimated average cost per family contact month if the project was working at full capacity. For example, if the above project was resourced to work with 20 families per month (i.e. to provide 240 family contact months per year), then the minimum average cost per family contact month would be £1,000 if the project was always working at full capacity. Evaluations of other similar interventions suggest that it is very difficult for projects to run at full capacity all of the time, especially whilst they are getting established, recruiting staff, seeking suitable referrals, and building up their caseloads. It is therefore important to use real activity data on caseloads and the actual costs of delivering the projects to calculate the true average cost per family contact month for each project. This unit cost can then be monitored as the project develops and matures.
The second key unit cost is the average total cost per closed case, which will depend on the average amount of time for which families are in contact with the project. This cost is built up for a specific year by averaging the total cost for each family whose case was closed in a particular year. For example, if the average cost per family contact month is £1,000 and the family worked with the project for nine months during that year, then the total cost for that family is £9,000. If the family works with the project across two financial years, then the relevant average costs per family contact month should be used for the number of months falling in each financial year. The average total cost per closed case in a specific year is calculated by averaging the totals for each family whose case was closed during that financial year. However, it is also useful to identify and specify the range of times for which families are in contact with a project, as these may differ considerably, resulting in a wide range of total costs for closed cases. For example, two families with contact times of 3 months and 15 months and total costs of £3,000 and £15,000, respectively 27, would have an average contact time of nine months and an average total cost of £9,000 per closed case. The average values alone may mask considerable variations.
The experience of other similar projects indicates that the average total cost per closed case is likely to increase over time for two main reasons:
- Some families have more complex needs than others, requiring longer periods of intervention and contact than the relatively straightforward families;
- With experience, projects improve their abilities to identify which referred families are most likely to be suitable (and/or to improve the suitability of the families referred to them), thus reducing the numbers of families with very short contact periods with the project (e.g. due to a failure to engage; being too close to eviction for the project to be able to intervene).
It is therefore unlikely that the steady state unit costs of delivering the projects will be established until the projects have been running for at least 18 - 24 months, and possibly longer (depending on the lead-in time required to set up the project). One of the reasons for including the two Families Projects in Aberdeen and Dundee was to enable a longer perspective to be taken when identifying the unit costs.
These two unit costs are relatively simple to calculate from routinely-recorded data and to monitor. There are a number of other possibilities that can also be calculated, such as the cost per successfully closed case and the cost per family member. However, these tend to have a variety of shortcomings. For example, focusing only on the families whose cases have been closed due to the successful achievement of their objectives overlooks the many benefits that may also have been achieved even if a family prematurely terminates its involvement with a project. In addition, although the average cost per family member reflects the fact that families differ in size, it may be misleading because family compositions are often fluid and tend to change over time (e.g. older teenagers may move out; some children may be living temporarily with foster carers). Furthermore, a project may work mainly with only one or two members of a large family if the behaviour of other members is not a cause for concern.
Finally, unit costs can also be used to undertake some comparisons between projects, although this should only be done with caution. Projects in different places may work with families of differing sizes and with varying levels of complexity (e.g. due to local provision of other services relating to ASB). They may also have different underlying objectives. Such comparisons can be especially problematic if there are small numbers of service users. For example, if a core unit is designed and staffed to work with two resident families but only has one family in it for a period of time, the costs for this family will be higher than they would have been if two families had been in residence. Unit costs will generally be lower if projects are working close to capacity, but it is also important that the projects work with appropriate families to obtain the desired outcomes (and thus to deliver genuine value for money).
(iv) Avoided short-term and longer-term cost consequences
The final element of the economic evaluation is to consider the costs that may have been incurred by other service providers in the absence of the Intensive Family Support Projects. Given the uniqueness and complexity of families and their situations 28, it is not possible to compare the services used by those families experiencing the intervention with a matched sample of families not working with the projects. Instead, this element focuses on identifying potentially avoided costs, using professional judgement where feasible 29, to determine which may have been incurred by the families. It focuses on the costs of the Exchequer-funded services rather than on any costs that may have had to be met by the families themselves. Potential costs that would have fallen on the Exchequer include short-term costs associated with activities such as:
- Evicting (and possibly rehousing) families;
- Placing one or more children in foster or residential care;
- Youth justice services (e.g. Young Offenders Institute; legal and Court costs);
- Special education or training provision for those children and young people not regularly attending school or college.
In addition, there are also likely to be longer-term cost consequences of anti-social behaviour due to social exclusion and poor educational attainment impacting on employment and lifestyle opportunities (e.g. resulting in lifelong benefit dependency). Many of these fall on society and individuals as well as on Exchequer-funded services. Although much less certainty is associated with these costs (and their present value will be relatively low if they would occur may years into the future), it is nevertheless important to identify them so they can be considered as possible cost consequences.
Many potential quantitative cost consequences can be identified from the published literature, and these are described in the following two sections of this Appendix. Local values are also used in the evaluation with local values where these can be identified.
Finally, it should be noted that these types of projects are also likely to have a number of other, more qualitative outcomes. For example, residents in a neighbourhood may feel that it has become a safer, quieter and more pleasant place to live. Although it is not possible to attach financial values to these benefits, it is important not to overlook them. The cost consequences approach used in this evaluation allows these qualitative benefits to be captured and described. Policy makers can then use their discretion to interpret the relevance of the findings to their specific circumstances.
A3.3 Costs of similar projects
i) Dundee Families Project
An evaluation of the Dundee Families Project ( DFP) was undertaken by a team of researchers from the University of Glasgow from May 1999 to May 2001 (Dillane et al, 2001), drawing on data from November 1996. DFP was established to assist families who were homeless or at severe risk of homelessness as a result of anti-social behaviour and was run by NCH Action for Children Scotland (as it was then) in partnership with Dundee Council Housing and Social Work Departments. It works with families deemed to have exhibited a range of anti-social behaviour with the aim of enabling them to avoid eviction or be restored to satisfactory tenancy arrangements. This would help prevent the breakdown of vulnerable families or re-unite separated families. Almost all of the referred families were on low incomes and many were headed by a lone parent. Professionals often also had concerns about parenting and care of the children. The evaluation found that many families improved their behaviour and stabilised their tenancies. These improvements were usually sustained after contact with DFP ceased, although many families still had serious problems, especially relating to childcare.
Although a formal economic analysis was not undertaken, the cost analysis and information from key stakeholders included in the evaluation of the DFP suggested that if the DFP were not there, the immediate annual costs to housing (management and legal) and to social work (mainly from looking after children in public care) would outweigh the annual costs of the DFP itself. There were felt to be immediate savings in staff time and resources for a few agencies and professionals, especially housing. For social workers, the overall level of input did not necessarily reduce in the short run, though the nature of their fieldwork input would change. The researchers found that major long-term gains would accrue whenever the probable entry of children to foster or residential care is avoided, or when looked after children are able to return to their families.
The study also applied cost information to the alternative actions that would have been required had the DFP not been available to its recipient families. These calculations suggested that the DFP saved the local Council more money than it required to operate. The authors recognise that their analysis has many limitations - for example, it is based on family problems only persisting for one year and excludes a set of broader social costs "which were impossible to guestimate" - but nevertheless they conclude:
"In summary … the DFP could be said to have generated approximately £117,600 of savings per annum. At worst, therefore, the [ DFP] can be assumed to cost no more than the conventional way of dealing with these families. However, it is more likely that the [ DFP] actually generates real cost savings, particularly when long-term costs are taken into account. In addition, it has the potential to deal with families in a more effective way".
Nearly all of the stakeholders interviewed during the evaluation of the DFP agreed that it delivered three main types of long-term benefits to the recipients who engaged positively:
- Avoided high-cost options (e.g. children becoming looked after, supported accommodation for the family);
- Reduced behaviours (including crime) with potential long-term cost implications for society;
- Promoted the quality of life of family members.
ii) Shelter Inclusion Project
The Centre for Housing Policy at the University of York undertook an evaluation of the Shelter Inclusion Project in Rochdale (Jones et al, 2006). This project was launched in 2002 in partnership with Rochdale Metropolitan Borough Council. It provided a specialist floating support service to tackle anti-social behaviour and social exclusion amongst four types of clients - lone adults, adult couples, lone parents and family groups. Its income came through Supporting People and the Neighbourhood Renewal Fund.
It supported 74 households containing a total of 230 individuals over its three-year duration. Two-thirds of the households included children and just over two-fifths (43%) were headed by lone parents. However, in 70% of the households the anti-social behaviour was being committed only by an adult in the household. Most households were economically inactive, with only 3% of service users in work. Almost a quarter of the children (24%) were either temporarily or permanently excluded from school of missing significant amounts of schooling. Many households also had high levels of reported health and support needs. More than half of the households contained an adult with depression or other mental health problems. Just under a third included someone with a limiting illness or disability. Drug and alcohol dependency among adults was a problem in almost a quarter (23%) of the households). Households had also been subject to a wide variety of actions due to their anti-social behaviour, including eviction, injunctions and Anti-Social Behaviour Orders ( ASBOs).
Anti-social behaviour ended or improved in 70% of households that completed their time with the project. Among the 45 closed case households, 60% no longer exhibited any anti-social behaviour and a further 11% showed improvements. With regard to tenancy sustainment, over four-fifths (84%) of closed cases were assessed as being no longer at risk of homelessness. Of the 34 children and young people who received direct support with their education, 91% showed improvements in school attendance. Many service users reported improved self-esteem and feeling more in control of their lives. Table A3.1 summarises the project activity and the unit costs associated with this project.
Table A3.1: Shelter Inclusion Project - Summary of Project activity and costs: August 2005 - December 2006
Cases Closed in 2003/04
Cases Closed in 2004/05
Number of closed cases
Number of individuals in closed case households
Cases closed during period:
Average duration (months)
2 - 15
4 - 27
Average cost per client month
Average cost per family member month
Total cost per closed case:
£1,358 - £10.185
£3,276 - £19,873
Total cost per family member
The project was staffed to work with a maximum of 33 households per month. It was estimated that the project would cost £771 per household month in 2005/06 if the project worked at full capacity throughout the year. If the average contact time could be reduced to 12 months per household (e.g. by closing cases more efficiently), this would result in an average total cost of £9,254 per closed case household.
iii) Anti-Social Behaviour Intensive Family Support Projects
The findings of a study of six anti-social behaviour rehabilitation projects was undertaken by a team of researchers from Sheffield Hallam University for the Department for Local Government and Communities (Nixon et al, 2006). The six projects, which worked intensively with vulnerable families, were located in northern England. Five were managed by Action for Children (formerly known as NCH) and one by a City Council. They offered multi-disciplinary support for families exhibiting problem behaviour at risk of family breakdown and homelessness either in their own homes or by moving them to managed accommodation. The study focussed on 256 families, nearly all of which had high multiple support needs. Two-thirds of the families had three or more children four-fifths of the referred adults had poor mental health (especially depression) and/or problems due to substance misuse. Almost half of the families reported either intimate partner or intergenerational domestic violence. In 85% of cases, the intensive support resulted in families changing their behaviour so that complaints of anti-social behaviour either ceased completely or dropped dramatically.
The evaluation included an in-depth discussion of the costs and cost consequences of these projects. The majority of the funding came from Supporting People, though some projects also received income from other sources, such as the Children's Fund their local Council. These funders tended to have different requirements, which influenced the range and scope of some of the projects.
Two unit costs were calculated for 2003/04 and 2004/05 - the average cost per client month and the average total cost per closed case. Three of the six projects had a core unit for part or all of the evaluation period, but it was not possible to separate the costs of the core units from the overall accounts for these projects. Activity and unit cost data for the six projects are shown in Table A3.2.
These costs show clearly that the average cost per client month and the average total cost per closed case tended to increase over the two years. For example, for the projects without a core unit, the average cost per client month increased from a range of £813 - £1,305 in 2003/04 to £1,060 - £1,496 in 2004/05. During this same period the average total cost per closed case increased from a range of £3,954 - £5,991 to a range of £4,913 - £12,940. The latter increase was because the families leaving the projects in 2004/05 had generally had more complex needs and been in contact with the projects for longer periods. The table also show that, as expected, costs are much higher for projects with core (residential) units than for those only providing outreach services.
Table A3.2 - Intensive Family Support Projects - Activity and cost data: 2003/04 and 2004/05
Client months provided**
Cases closed during year
Average duration of closed cases: Months
1 - 8
2 - 8
2 - 13
2 - 11
8 - 12
Client months provided**
Cases closed during year
Average duration of closed cases: Months
1 - 20
6 - 39
1 - 24
1 - 21
1 - 16
1 - 25
Average cost per client month
Average total cost per closed case
Average cost per client month
Average total cost per closed case
*: denotes projects with a core unit throughout
º: denotes the project with a core for part of 2004/05
**: all clients, including those in core units (where applicable)
Although it was not possible to determine the Exchequer costs that would have been incurred by these families in the absence of the intervention, the evaluation includes an illustrative case study showing how a family with four children can easily generate public costs of more than £330,000 over a 12-month period (e.g. due to eviction, foster care, residential care and criminal justice provision) if their anti-social behaviour is not addressed. As shown in Sections A3.4 and A3.5 below, the costs associated with anti-social behaviour and non-sustainable tenancies can be high. Furthermore, the costs of providing foster care and more specialist residential care to children who cannot remain with their families can be very high. These types of projects can be very cost-effective, not only in the short-term but over recipients' lifetimes.
iv) Edinburgh Early Intervention Families Project
The Case Management Early Intervention Project (aka the Early Intervention Families Project) was established in Edinburgh in August 2005 with initial funding for two years. Although the project has been evaluated independently over two years by Brodies 30, its economic aspects were included in an evaluation undertaken by Sheffield Hallam and Herriot Watt Universities of the impact of local anti-social behaviour strategies at the neighbourhood level (Flint et al, 2007). The project works with vulnerable families in specific parts of the city where there is at least one child aged 12 years or under and the behaviour of one or more children is giving cause for concern. Each family is allocated a Case Manager who works holistically with them and with all of the other agencies in contact with the family to co-ordinate these services and to identify and fill any gaps. It is less intensive than the other models described above and has a stronger focus on prevention through early intervention and on working with families with at least one young child. It also draws on a New Zealand model known as Strengthening Families, which encourages families to develop and draw on their own resources to improve the functioning of the family as a unit. This is expected to reduce the subsequent development of antisocial behaviour.
During the period August 2005 to December 2006, the project worked with 37 families. Two-thirds were headed by a lone adult parent and each family had an average of 3.9 children (range: 1 - 9). As well as including one or more children whose behaviour was causing concern, 60% of accepted families were affected by parental mental health problems and substance misuse, over half by significant poverty and debt, and 40% were affected by poor parental physical health or learning disabilities. Slightly over three-quarters (77%) of accepted families had been involved in, or had generated complaints about their involvement in, antisocial behaviour. Four-fifths of families working with the project had had problems relating to the attendance or behaviour of their children at school and two-thirds of families were involved with the police and social workers at the time of referral or previously. Two-thirds of accepted families were living in City of Edinburgh Council tenancies. At least 40% of these were seriously overcrowded. The remaining third of families were either living in the private rented sector or in homeless accommodation. Over a third of the families accepted by the project were homeless or threatened with homelessness at the point of referral.
Table A3.3 - Case Management Early Intervention Project - Summary of Project activity and costs: August 2005 - December 2006
August 2005 - March 2006
April 2006 - December 2006
Number of new clients/cases:
Number of client months
Average number of clients (families) per month
Number of family members
Average number of family members per month
Number of closed cases
Cases closed during period:
Average duration (months)
3 - 8
4 - 12
Average cost per client month
Average cost per family member month
Total cost per closed case:
£2,664 - £7,104
£2,584 - £9,688
Table A3.3 summarises the project's activity and costs since its inception in August 2005 to December 2006. Two families had worked with the project for 17 months (i.e. since it started in August 2005) and almost half (10/21) of the families working with the project in December 2006 had done so for 12 months or longer. The table shows how the average duration for closed cases has increased over time, from 4.8 months (range: 3 - 8 months) during August 2005 to March 2006 to 7.7 months (range: 4 - 12 months) in April to December 2006. It also shows that the average cost per client (i.e. family) month has fallen from £888 to £646 and the average cost per family member month has fallen from £179 to £123 over these periods. Overall, the average cost per closed case increased slightly from £4,292 (range: £2,664 - £7,104) to £4,670 (range: £2,584 - £9,688).
Although these costs are of interest, they were not felt to represent the likely steady-state costs associated with the project after mainstreaming. This was partly no overhead charges were levied on the project during its pilot phase. Based on the assumption that the project has an average caseload of 20 families each month (i.e. it provides 240 client contact months each year) and costs £250,000 over a full year, the average total cost per closed case would be £12,500 if the average duration of contact with the project is 12 months. However, it would only be £9,375 if the average duration of contact was nine months 31.
Although this project tends to intervene sooner and works in a less direct manner with its families, these costs are similar to those associated with evaluations of other projects providing support to families with problems relating to their behaviour and their housing (Jones et al, 2006; Nixon et al, 2006).
The above discussion shows that although there is considerable variation across the projects in terms of the lengths of the period of participation and the costs per client month, they generally cost in the region of £10,000 - £13,000 per closed case. This cost tends to increase over time as projects mature and work with families with more complex needs, requiring longer periods of contact.
The economic analysis included within the Dundee Families' Project evaluation was relatively embryonic. The analysis included in the evaluations of the SIP, the IFSPs and the Early Intervention Case Management Project in Edinburgh tried to address some of these limitations, for example by including calculations of the unit costs for each family and by gathering generic information on some of the costs that may have been prevented due to the interventions. However, it was not possible in these evaluations to include the costs of changes in the use of (Exchequer-funded) services by the families because of their involvement (although there are many caveats associated with such work). The economic element of the evaluation of the Intensive Family Support Projects in Scotland provides an opportunity to build on existing experience and to try to address some of the shortcomings of the previous economic evaluations.
A3.4 Short-term cost savings
A considerable amount of published literature is available on the potential cost consequences for publicly-funded services of not tackling children's and families' antisocial behaviour 32. Some of these are short-term, such as the costs of tenancy failure, foster and residential care, and youth justice. As shown below, these costs can be very high, particularly if specialist provision is required. Long-term cost consequences are discussed in A3.5
The following categories of costs are considered:
- Looked after children:
- Costs of foster care;
- Costs of residential care and secure accommodation;
- Estimated total costs of services for children in care with varying support needs;
- Housing and Homelessness;
- ASB and Crime:
- Domestic violence.
Although it is not known what costs would have been incurred by the families receiving intensive support to help them address their problems, the costs presented below provide a range of useful values for estimating the potential cost consequences that may have been incurred in the absence of the families and Breaking the Cycle projects.
(i) Looked after children
The following information provides an overview of the numbers of looked after children in Scotland on 31 March 2006 (Scottish Executive, 2006):
- 12,966 children were 'looked after' by local authorities in Scotland - an increase of 6% from 2005 (though young people aged 18 and over are included for the first time - if the 216 of them are excluded the increase since 2005 is 5%;
- Of these, 7,220 were boys (56%) and 5,746 were girls (44%);
- 53% (6,855) were aged 11 years or younger; 47% (6,111) were over 12;
- The proportion of children who are looked after in Scotland has been rising steadily since 2000 and reached 1.16% in 2005/06 (the highest since 1982);
- The main statutory reason for being looked after is a supervision requirement;
- Statutory reasons for supervision requirements include:
- Commission of an offence;
- The child being out of control;
- Not attending school;
- Lack of parental care;
- The child "falling into bad associations or exposed to moral danger";
- A risk of offences being committed against the child;
- If a parent cannot look after the child;
- If the child needs to be looked after and accommodated to safeguard his or her welfare;
- With the majority of supervision requirements, children stay at home or are fostered;
- 56% of looked after children are at home with their parents (43%; 5,506) or with friends/relatives (13%; 1,726);
- 29% (3,731) of 'looked after' children are in foster care;
- Nationally, 13% (1,638) of looked after children were looked after in residential accommodation (ranging from 6% in Clackmannanshire to 31% in Orkney);
- 737 were in local authority homes (45% of those in residential accommodation and 6% of the total);
- 663 in residential school (39% of those in residential accommodation and 5% of the total)
- 84 were in voluntary homes (5% of those in residential accommodation and 0.6% of the total);
- 78 were in secure accommodation (5% of those in residential accommodation and 0.6% of the total);
- 76 in other residential accommodation (5% of those in residential accommodation and 0.6% of the total
- During 2005/06, 4,718 children started to be looked after and 3,882 ceased to be looked after;
- 74% of children are looked after away from home for longer than one year and 29% have 3 or more placements.
(ii) Foster care
- Costs of foster care in Scotland
Foster carers are usually paid a combination of allowances and fees. The allowances are payments made to cover the cost of looking after a child in foster care, including everything from food and clothes to pocket money. In 2007/08 current expenditure per week by Scottish local authorities on fostering allowances is about £460,000 and annual expenditure is almost £24 million. The fee payments are the remuneration for the foster carers' work, skills and experience.
A report on the costs of foster care in the UK (Tapsfield and Collier, 2005) was published by British Agencies for Adoption and Fostering ( BAAF) and the Fostering Network 33. It concluded that "a massive injection of funding from central government will be required for fostering services to meet the complex needs of children in public care". The estimated shortfall in funding in Scotland for 2005/06 was estimated as £65.5 million (against annual expenditure on foster care in 2003/04 of £47.4 million - £55.5 million 34).
The researchers found that the average weekly unit cost to deliver effective foster care services in Scotland in 2004/05 would have been £605 (i.e. £31,460 per year). This equates to about £680 per week (or about £35,400 per year) in 2007/08. This value includes allowances, fee payments, training costs, respite care costs and the costs of management and support. The report also estimated that total required expenditure on foster care in Scotland would need to be £121.8 million in 2007/08 to ensure sufficient numbers of foster care households.
The Fostering Network conducts a comprehensive annual survey of the fostering allowances and fee payments made to foster carers in every local authority in the UK. For many years it has published recommended minimum weekly allowances, based on a comprehensive assessment of the minimum needs of looking after a fostered child. In addition, it recommends an additional four weeks allowance each year to cover the costs of birthdays, holidays and a religious festival. The recommended weekly allowances for 2007/08 for children of different ages and the annual equivalents based on 56 weeks per year are shown in Table A3.4. It should be noted that these costs exclude any managerial and administrative costs associated with foster care incurred by the local authority.
Table A3.4 - The Fostering Network's recommended minimum weekly allowances: 2007/08
0 - 4
5 - 10
11 - 15
Annual Rate (56 weeks)
In terms of current expenditure on allowances, the survey for 2006/07 showed that 34% of Scottish local authorities paid at, or above, this rate. This had increased to 53% (i.e. 17 of the 32) in 2007/08, although seven authorities were paying £40 - £60 per week less than this recommended minimum weekly allowance. Extra investment of £4.26 million would be required in 2007/08 to ensure that all Scottish local authorities were able to pay the Fostering Network's minimum recommended rate. This represents an increase of almost 18% on current expenditure.
Scottish local authorities operate myriad different systems for the fee payment schemes that provide remuneration for the work, skills and experience of the foster carers. These are sometimes tailored to meet local needs or budgets. The Fostering Network's survey for 2007/08 showed that 16 local authorities (i.e. 50%) operate a fee payment system that is open to all approved foster carers. These schemes may or may not include higher fees for more skilled or experienced foster carers, but do provide entry level payment. Eight authorities (i.e. 25%) operate systems that do not pay all of their foster carers but that could very broadly be described a following a 'payment for skills' model (though sometimes it is couched in terms of being for foster carers who care for more difficult children). One authority operates a fee payment scheme that is only available to 'specialist foster carers, while seven authorities (22%) reported that they had no fee payment scheme whatsoever. The percentage of authorities operating no form of payment scheme is considerably higher in Scotland than in the other nations of the UK.
The Fostering Network recently undertook a survey of fee payments to foster carers in the UK (Swain, 2007). Within Scotland, the survey found that:
- 37% of foster carers receive no fee payments;
- 67% earn less than the national minimum wage from fostering;
- Only 11% of foster carers receive a fee comparable to that of care staff in a children's home (who receive £17,368 - £21,341 per year, at 2005 rates);
- 21% of foster carers claim state benefits for additional income and for 4% their only income is from benefits.
These factors make it very difficult to recruit and retain sufficient foster carers with the right skills and experience. Such foster carers are essential if children in public care are to be given the same opportunities to have a successful future as other children in society. The Fostering Network argue that, given the regulatory responsibilities and the time needed to devote to fostering, it is a job of work, not just a way of life, and that foster carers should not be treated as volunteers.
The final evaluation of the Early Intervention Case Management Project in Edinburgh (Brodies LLP, 2007) includes a number of potential savings associated with the Project provided by the Council's Children and Families Finance Section. These include annual costs for specialist carers of £35,500 (£683 per week) and annual costs for specialist respite of £14,859 (£286 per week), based on 2006/07 figures. An annual cost of £12,381 (£238 per week) is quoted for adoption and of £2,062 (£40 per week) for befriending.
The above discussion has shown that current expenditure on foster care in Scotland seems to be considerably lower than needed to provide sufficient numbers of foster carers who are adequately remunerated for their work. It suggests that this would cost an average of about £680 per week (i.e. about £35,400 per year) per foster placement, which includes all managerial overheads. However, current payments of allowances and fees to foster carers tend to be below the levels recommended by the Fostering Network and BAAF. Furthermore, they vary considerably across the Scottish local authorities.
- Other estimates of the costs of foster care
Several studies have estimated costs of foster care outwith Scotland. The Personal Social Services Research Unit's ( PSSRU) Unit Costs of Health and Social Care 2006 (Curtis & Netten, 2006) estimates that the unit cost per child per week of foster care in 2005/06 was £513 (comprising £311 per child per week for the boarding out allowance and administration; £152 per child per week for social services, including the costs of a social worker and support; and £50 per child per week for other services, including education). This gives a cost of £26,676 over a full year (including a total of £16,172 for the boarding out allowance and administration).
Data for 2004/05 provided by a District Council in north-west England used in the Shelter Inclusion Project (Jones et al, 2006) showed an average weekly cost for foster care for a child of £392 (£269 for local provision and £641 for provision by others). This gives an annual cost for foster care of almost £20,500. In the study of several Intensive Family Support Projects in northern England (Nixon et al, 2006), the Social Services Department associated with one of the projects provided estimates of £700 - £900 for weekly fostering costs, depending on the need and level of care. This gives an annual cost of £36,400 - £46,800. Foster placements in another project's area were calculated to be a minimum of £40,000 per year.
- Summary of costs of foster care
Table A3.5 summarises the relevant findings from the above studies. It should be noted that these costs may be underestimates for some children, especially those with emotional and behavioural difficulties. These costs were considered in a study by Ward et al (2004) and are discussed after the discussion of the costs of residential care and secure accommodation.
Table A3.5 - Summary of estimates of costs of foster care for children
Cost required to attract sufficient numbers of foster carers in Scotland
£680 per week
£35,400 per year
Tapsfield and Collier, 2005
BAAF and the Fostering Network, uplifted to 2007/08 costs
Cost of specialist carers
£683 per week
£35,500 per year
LLP Brodies, 2007
City of Edinburgh Council, Children and Families Finance Section, 2006/07 values
Foster care - unit cost per child
£513 per week
£26,676 per year
Curtis and Netten, 2006
Personal Social Services Research Unit ( PSSRU) estimates for 2005/06
Metropolitan Borough Council in north-west England
Average of £392 per week
(£269 for local provision and £641 for provision by others)
Annual average cost of £20,384
Jones et al, 2006
2 other Councils in northern England
£700 - £900 per week
£36,400 - £46,800 per year
Nixon et al, 2006
Cost depends on need and level of care required
Minimum of £40,000 per year
(iii) Residential care and secure accommodation for children
- Scottish data
Nationally, 1,638 of looked after children were looked after in residential accommodation in 2005/06. Such accommodation includes local authority homes, residential schools and secure accommodation. Scottish data on the costs of such accommodation are considered in a recent study for the Scottish Executive Education Department on secure accommodation in Scotland (Walker et al, 2006).
Between 200 and 250 young people are admitted to secure care in Scotland each year, with about 90 placements at any one time 35. In recent years there have been a number of changes and developments in the provision and use of secure accommodation in Scotland. These include the growth of a range of community-based 'alternatives' for young people who can safely be accommodated within an open setting, such as schemes offering enhanced or intensive community-based support (which may include electronic tagging) and specialist foster care. Partly because of the nature of the children's hearing system, secure accommodation in Scotland is quite different from similar provision in other parts of the UK. One of the key differences is that it is located within residential child care provision. A high proportion of young people are admitted primarily on welfare grounds. However, policy and practice issues inevitably have resonance with those in other parts of the UK.
The recent SEED study (Walker et al, 2006) compared the experiences of two groups - one comprising young people admitted to secure accommodation and the other consisting of young people receiving 'alternative' services. Girls predominated in the secure sample, and were often seen to be putting themselves at risk through drug use, running away and risky sexual behaviour. As well as comprising mainly boys, the alternative sample included more young people aged less than 14 years and more young people for whom offending was a primary concern.
During the three-year research period it quickly became apparent that residential or community-based services seldom offered a 'direct' alternative to secure accommodation. One of its key messages is that the decision to admit certain young people to secure accommodation is shaped as much by the service provision context around them as by the needs and behaviour of the young people themselves.
The study calculated average weekly costs for a range of placements types for looked after children and these are shown in Table A3.6. Although these figures mask the wide variation in the ways in which such services are provided to children and young people, they do provide indicative costs of the various types of placement.
Table A3.6 - Average weekly cost of placements for looked after children
Type of placement
Cost per week
Cost per year
LA residential unit
6 local authorities
11 residential schools
Close Support Unit
4 close support units
5 secure units
Fast Track costings
The researchers estimated pathway costs over a two-year period. Full information on the pattern and duration of placements in the year prior to secure authorisation and the year after were available for 42 of the secure care sample and 18 of the sample receiving alternative provision. The average (i.e. mean) total cost of services over the two years for the 42 young people in the secure care sample was £185,650 per person, compared with an average cost of £117,000 for those receiving their 'secure' care in alternative settings. One of the significant costs in relation to the secure sample is the cost of their secure provision.
However, the researchers also found significant differences in the average costs of services for the young people in the two groups during the year before they were placed in or considered for secure accommodation. The average total cost of services received during this period was just over £61,200 per young person in the secure group compared with just under £40,000 per young person in the alternative provision sample. In the year after placement commenced, the average cost was £124,000 per young person in the secure sample and £77,100 for the alternative provision sample.
Three main routes into secure or alternative care were identified - from children's homes, from residential school, and from home. Table A3.7 shows the mean and median costs and the range of costs associated with each pathway into secure or alternative care. Although there is a tendency for the alternative provision to be less expensive than secure provision, there is wide variation in cost within the two samples and overlap between them. It should also be noted that the costs for secure accommodation services and residential schools include education costs. The researchers were unable to attribute costs for education provided to those children and young people in other placements. Although education in a secondary school costs approximately £5,000 per pupil per year (based on Scottish Executive data for 2005), few of the children and young people in the sample were in mainstream education without additional educational support. Audit Scotland (2003) found that it cost approximately £7,800 per year to support a pupil with special educational needs. However, there was wide variation - from £3,000 per pupil to £17,500 per pupil - across local authorities.
Table A3.7 - Pathway costs over two years for children and young people receiving care in secure placements and alternative settings
Sample n = 60
(n = 42)
Children's unit (26)
£66,800 - £354,400
Residential school (11)
£144,100 - £271,700
£112,400 - £166,400
£66,800 - £354,400
(n = 18)
Children's unit (8)
£58,200 - £217,000
Residential school (3)
£121,000 - £205,200
£20,800 - £148,200
£20,800 - £217,100
Because of the differences in the two study groups, it was not possible for the researchers to compare their outcomes in a meaningful way. However, one of the key distinctions between secure placements and community-based intensive support was that the latter worked closely with parents and other family members, whereas admission to secure accommodation could potentially isolate the young person from their family. Many of the young people came from uncertain or unstable family situations. Moving into secure accommodation locates their problems within themselves, whereas community-based approaches enable family-based problems to be addressed. Furthermore, community-based intensive support could last for several years. However, a notable feature of secure accommodation was that it facilitated reintroduction to education and other services from which the young people had become disengaged.
The final evaluation of the Early Intervention Case Management Project in Edinburgh (Brodies LLP, 2007) includes a number of potential savings associated with the Project provided by the Council's Children and Families Finance Section. These suggest that the residential care costs for children can vary from £81,432 to £204,516 per annum (2006/07 figures based on charges to other local authorities, including internal costs but excluding external costs such as external management and other external support services).
- Other estimates of the costs of residential care
The Personal Social Services Research Unit (Curtis and Netten, 2006) identifies costs in 2005/06 of £2,285 per resident week for the establishment costs of a local authority community home for children (of which £2,192 is due to salary and other revenue costs), rising to £2,459 per resident week if the costs of other external services (including health, education, social services and youth justice) are included. These result in annual costs of almost £128,000 per child. Similar costs are also presented for community homes for children provided by the non-statutory sector, with establishment costs of £2,330 per resident week, rising to £2,405 per resident week (i.e. about £125,000) if the costs of external services are included.
Local data used in the Shelter Inclusion Project (Jones et al, 2006) showed that the average weekly cost of a place in a Children's Home is £2,710 (with a cost of £2,239 for local provision and £3,266 for provision by other Councils) for a Metropolitan Borough Council in north-west England in 2004/05. Therefore the average annual cost of a place in a Children's Home was about £141,000. These values are similar to those provided by two of the Councils involved in the DCLG study of Intensive Family Support Projects (Nixon et al, 2006). One Council in northern England estimated local costs of £2,000 - £2,500 per week (i.e. £104,000 - £130,000 per year) (dependent on need and Statement of Purpose), whilst another estimated an annual average cost in its care homes for looked after children of £125,000 - £200,000.
- Summary of costs of residential care (including secure accommodation)
The above studies provide a range of estimates for the weekly and annual costs of residential provision for looked after children. These are summarised in Table A3.8.
Table A3.8a - Summary of estimates of costs of residential provision for children
LA residential unit in Scotland
£1,400 per week
£72,800 per year
Walker et al, 2006
Cost of residential care
£ 81,432 - £204,516 per year
LLP Brodies, 2007
City of Edinburgh Council, Children and Families Finance Section, 2006/07 values
Local Authority provision
£2,285 per resident week
£118,820 per year
Curtis and Netten, 2005
Establishment costs only
Non-statutory sector provision
£2,330 per resident week
£121,160 per year
Metropolitan Borough Council in north-west England
Average of £2,710 per week (£2,239 for local provision and £3,266 for provision by other Councils)
Annual average cost of £141,000
Jones et al, 2006
2 other Councils in northern England
£2,000 - £2,500 per week
£104,000 - £130,000 per year
Nixon et al, 2006
£125,000 - £200,000 per year
Table A3.8b - Summary of estimates of costs for residential education provision
LA residential school in Scotland
£2,100 per week
£109,000 per year
Walker et al, 2006
Close Support Unit in Scotland
£2,775 per week
£193,700 per year
Secure care in Scotland
£3,725 per week
£193,700 per year
For children who are out-of-school in a Local Authority in north-west England
£2,000 - £3,000 per week
£150,000 - £200,000 per year
Nixon et al, 2006
Local Authority Secure Children's Home (10 - 14 year olds and vulnerable 15 year olds)
£185,780 per place
House of Commons Committee of Public Accounts, 2004
A Secure Training Centre (run by private contractors for young people aged up to 17 years)
£164,750 per place
House of Commons Committee of Public Accounts, 2004
Part of the variation is due to the different ways such care can be provided. Furthermore, the costs per child or young person will depend upon their pathway through such provision, especially if the need to be place in secure accommodation or to receive intensive support in community settings. The average (i.e. mean) total cost of services over the two years for the 42 young people in the secure care sample was £185,650 per person, compared with an average cost of £117,000 for those receiving their 'secure' care in alternative settings. In the year after placement commenced, the average cost was £124,000 per young person in the secure sample and £77,100 for the alternative provision sample.
(iv) Estimated total costs of services for children in care with varying support needs
The costs identified above may, however, underestimate those incurred by some children, as children with multiple difficulties, complex needs and offending behaviour can cost considerably more than the costs presented above. A study by Ward et al (2004) (reported by the PSSRU in Curtis and Netten, 2006) looked at the different combinations of additional support that may be needed by children in care. In their sample of 478 children, they found that 27% showed no evidence of additional support needs, 45% displayed one additional need, 26% displayed two, and 2% of children displayed three or more. Several examples are given, showing the costs associated with a particular child in each category over an 87 week period. They are presented below as they illustrate the types of costs that may be incurred and their potential magnitude.
"Child A" is an example of a "low cost" teenage boy in care who is due to move on from his long-standing foster family but who otherwise has no additional support needs. The costs (2005/06 values) to the local authority over the 87-week period were £36,653 (i.e. an average of £421 per week), with an additional cost of £11,328 associated with other services (though it should be noted that the majority of these costs were associated with mainstream schooling and attendance at a Further Education College, which is desirable rather than preventable expenditure, and for physiotherapy for an existing neck injury, which is likely to be specific to this child).
"Child B" represents a "median cost" child with some emotional or behavioural difficulties who is living with foster carers. The costs to the local authority over the 87-week period associated with the foster placement were £33,217 (i.e. an average of £382 per week), with an additional £17,467 falling on other agencies (although a significant proportion of these costs relate to mainstream schooling, considerable amounts are also due to weekly appointments with a clinical psychologist and with a speech therapist and also for time spent with a personal teaching assistant for part of the period).
"Child C" is an example of a "high cost" child with emotional or behavioural difficulties and offending behaviour. The costs associated with this child cover a 74-week period due to his decision to disengage from local authority placements during the study period. Prior to the start of the study period he was placed in secure accommodation on five separate occasions. He had also had placements in various residential homes, schools and foster placements, many of which had broken down. During the 74-week study period he experienced ten different placements, mainly out-of-authority. The total cost to the local authority was £226,620 (i.e. an average of £3,062 per week). He also generated additional costs of £70,560 over the study period (mainly due to £66,731 for Youth Offending Team ( YOT)involvement and criminal costs, as he refused to attend school or engage with health services).
The final example is "Child D", who represents a "very high cost" child in care with disabilities, emotional or behavioural difficulties and offending behaviour. He experienced nine placements during the 87-week study period, mainly in agency residential and foster home located out of the local authority area. The monthly charges for these ranged from £3,302 to £10,504. High levels of social work time were needed to support the placements due to the distances involved. Because he was so difficult to place, increasing amounts of social work time were also required to find placements that would accept him. The costs of changing placements were calculated at over £1,000 per move. The total costs falling on the local authority over the 87-week period were calculated to be £428,667 (i.e. an average of £4,927 per week). Costs incurred by other agencies were estimated to total £12,927 over the period. These arose mainly due to expenditure of over £6,700 on home tuition costs (at £35 per hour) after he was permanently excluded from school.
The above findings are summarised in Table A3.9.
Table A3.9 - Summary of average weekly costs to local authority of children in care with varying support needs
Child A: low cost - no evidence of additional support needs
Child B: median cost - child with emotional or behavioural difficulties
Child C: high cost - child with emotional or behavioural difficulties and offending behaviour
Child D: very high cost - child with disabilities, emotional or behavioural difficulties plus offending behaviour
These figures show not only how much time spent in care generally costs, but also how much extra it can cost when children have multiple difficulties and needs. Thus the potential savings to other agencies responsible for care placements and youth offending resulting from the projects, even in the short-term, could be considerable.
(v) Costs relating to housing and homelessness
- Costs of tenancy failure
A research study on the use of possession actions and evictions by social landlords (Pawson et al, 2005) shows that although the vast majority of such evictions are triggered by rent arrears, a small proportion is to counter anti-social behaviour. The study found that practice varied widely across social landlords as to when and to what extent they intervened to try to address the problems and to prevent the need for legal action and eviction. The researchers also found that few social landlords recorded or collated extensive data on the costs of possession actions. Such costs include court costs and legal fees, rental losses, relet costs (which may include repairs), security costs for voids, and staff time. The study estimated that the landlord costs to evict a tenant for rent arrears are £2,000 - £3,000, rising to £6,500 - £9,500 when the eviction is due to anti-social behaviour. However, the authors feel that these figures are underestimates and that landlords' cost-accounting methods show many weaknesses, with many relevant costs being subsumed under generic budget headings.
These values for the costs of evictions can be compared with estimates of the costs of tenancy failure from other sources. For example, the Audit Commission (1998) calculated that the cost of tenancy failure to a housing authority for vulnerable tenants living in the community was £2,100 per failed tenancy. Shelter (2003) estimated costs of £1,913 for "standard" cases and £3,190 for "complex" cases, while Crisis (2003) estimated the costs as ranging from £1,610 - £4,210.
A Housing Association in the northern of England (Nixon et al, 2006) estimated that preventing the eviction of a family saved them an average of £4,115 per household (comprising Court costs/legal fees of £500; rental loss of £390 based on an average void turn around of 39 days and a rent of £65 per week; average relet costs of £2,500 where the tenant is evicted; security costs of £120; and a saving of £605 in staff time through avoiding Court preparation time). The Housing Association pointed out that these figures were only estimates and that they do not take into account the cost savings of complainants not terminating their tenancies as a result of the reduction in anti-social behaviour. Local costs in another Council area were estimated as £5,000 per eviction (including Court costs, rental loss and officer time) and £23,400 for temporary accommodation for a homeless family (based on an average length of stay of 6 months). Although staff working for the Council where the Shelter Inclusion Project was located did not know the local costs associated with terminating a tenancy, they estimated that the average cost per legal case was about £5,000, regardless of whether or not the tenants are evicted (Jones et al, 2006).
These figures are summarised in Table A3.10. Although they encompass a wide range and have several methodological weaknesses, taken together, they show that considerable costs of about £5,000 - £9,000 can be avoided for housing departments, housing associations and social landlords by the prevention of tenancy failure. Furthermore, there will be no need to provide temporary homelessness accommodation for the evicted family and other tenants living in the neighbourhood are less likely to terminate their leases if the ASB reduces or ceases, resulting in considerable additional prevented expenditure.
Table A3.10 - Summary of costs of tenancy failure
Costs to a housing authority for vulnerable tenants
Audit Commission, 1998
Costs to housing authority:
Costs to housing authority
£1,610 - £4,210
Landlord costs to evict a tenant
£2,000 - £3,000
The authors believe that these figures are likely to be underestimates due to accounting weaknesses
Pawson et al, 2005
Landlord costs to evict a tenant due to ASB
£6,500 - £9,000
Average cost per legal case associated with terminating a tenancy in a Metropolitan Borough Council in north-west England
Jones et al, 2006
Costs of eviction incurred by a Housing Association in northern England
Includes Court costs, legal fees, rental loss, re-let costs and officer time
Nixon et al, 2006
Estimated Exchequer costs in a Local Authority in northern England
£5,000 for eviction plus £23,400 for temporary accommodation (6 months) for homeless family
The £5,000 includes Court costs, rental loss and officer time
- Benefits of preventing family homelessness
The Office of the Deputy Prime Minister 36 commissioned some research on the Benefits Realisation of the Supporting People Programme (Matrix Research and Consultancy, 2004). It considers eight key Programmes, including homeless families. This model found that the largest potential benefits related to the maintenance of tenancies in permanent accommodation and securing and maintaining permanent tenancies for families in temporary accommodation. These benefits were valued at £45.92 million per annum, compared with total Supporting People expenditure on the Programme of £52 million per annum (for around 7,300 household units receiving temporary accommodation and support). Benefits in terms of reduced use of health services were valued at £33.91 million and crime reduction at £1.73 million (i.e. delivering tangible benefits with a total estimated value of £81.56 million from the Supporting People expenditure on homeless families). These data show that preventing family homelessness should result in considerable financial benefits to the Exchequer.
However, the authors also cite that there is evidence of many other benefits from working with homeless or potentially homeless families that cannot readily be assigned monetary values. These include an improved quality of life (e.g. due to improved health, greater independence, and decreased vulnerability) and greater family stability from allowing families to stay together and to deal with other issues in their lives such as education, unemployment, mental health problems and behavioural problems. Children's health and educational achievement may improve, with long-term benefits and children may also be less socially isolated and at less risk of bullying if they have stable tenancies and do not have to change schools frequently.
- Summary of costs relating to housing and homelessness
The above studies show that are considerable financial and non-financial benefits associated with preventing tenancy failure and family homelessness. Costs are saved not only for accommodation providers, such as local authorities and housing associations, but also for providers of other services, including those relating to health and crime. The intangible benefits for individuals, families and neighbourhoods are also important.
(vi) Cost consequences relating to anti-social behaviour and crime
A wide range of estimates have been made of the costs relating to anti-social behaviour and to crime, although it is not always easy to distinguish between the two activities. Furthermore, some studies focus on generic anti-social behaviour, whereas others consider the costs associated with specific types of anti-social behaviour. These are summarised below.
- Anti-Social Behaviour
A significant review of the economic and social costs of anti-social behaviour was undertaken by Whitehead et al (2003). This study considers the costs falling on a wide range of agencies due to anti-social behaviour, not just on housing authorities. In many cases these costs will in part be determined by local policies on when and how to intervene when problems arise. The practical problems of defining, measuring and interpreting the economic and social costs of anti-social behaviour are also discussed in detail. Material is drawn from many sources (some of which are now rather dated), and covers the costs associated with responding to activities such as noise, rowdy behaviour, nuisance behaviour, abandoned vehicles, intimidation/harassment and criminal damage/vandalism. The authors state that:
"At the lower end, [unit] costs are of the order of £20 - £50. At the upper end, there are examples of over £1m. For the vast majority of incidents where action is taken however the costs vary between £100 and £10,000. These are very general estimates based on estimates using widely varying approaches. They do not directly reflect costs of non-alleviation such as increased vacancies. Most importantly they exclude any net costs to victims".
Their findings for specific types of anti-social behaviour are summarised in Table A3.11.
Table A3.11 - Costs associated with specific types of anti-social behaviour
Category of ASB
Nature of Action
Unit Cost Estimates
'Lower' end ASB
£20 - £50
Whitehead et al, 2003 (costs are borne by a variety of agencies)
Vast majority of incidents
£100 - £10,000
'Upper' end ASB
£1 million +
Direct costs to environmental services:
£3 - £70
Dignan et al (1996)
- including prosecution
- including imputed staff time.
Cost of Housing Department informal intervention
Cost of transfer of tenancy
Cost of legal action/possession
£365 - £3,900
Treated as per noise
a) Nuisance neighbours
Cost of legal action to LA
Cost of legal action to LA
£142 - £305,000
Hunter et al (2000)
Cost of legal action to HA
£500 - £80,000
Hunter et al (2000)
b) Neighbourhood disputes
Costs of possession action
£1,000 - £5,000
Atkinson et al (2000)
Costs of eviction
Hunter et al (2000)
including staff time
£245 - £1,000
Cost of possession order
£1,500 - £5,000
Dignan et al (1996)
Cost of injunction
Cost of collection and disposal
Reading Borough Council (2003)
Cost of disposal
£60 - £100
Jill Dando Institute (2003)
Cost of common assault offence (including social costs)
Brand & Price (2000)
Cost of individual incident (inc social costs)
Brand & Price (2000)
Cost of incident against commercial/public sector (inc social costs)
Brand & Price (2000)
Average cost per LA dwelling in Bradford
Cost per incident assuming 5% incidence
Source: Whitehead et al (2003)
Several other studies have also considered the costs at an aggregate level associated with anti-social behaviour. A day count of anti-social behaviour undertaken by the Home Office's Anti-Social Behaviour Unit in September 2003 estimated that anti-social behaviour cost agencies in England and Wales at least £13.5 million, which equates to an annual cost of around £3.4 billion. At a more local level, a study evaluating the costs of responding to and preventing anti-social behaviour in Rotherham MBC by Crowther and Formby (2004) estimated that the annual costs were at least £3.3 million and probably closer to £4.0 million. These figures are very similar to the estimate from Leeds Local Authority of annual expenditure of £3 million - £5 million on anti-social behaviour reported by the Social Exclusion Unit (2000). It should be noted that none of these estimates takes full account of the costs to victims (both financial and psychological) of anti-social behaviour. These findings are summarised in Table A3.12.
Table A3.12 - Aggregate costs associated with generic anti-social behaviour
Daily cost of ASB to agencies in England and Wales
Antisocial Behaviour Unit, Home Office, 2003
Annual costs of responding to and preventing ASB by Rotherham MBC
£3.3 - £4.0 million
Crowther and Formby, 2004
Annual expenditure on ASB by Leeds Local Authority
£3 - £5 million
Social Exclusion Unit, 2000
Anti-social behaviour can lead to criminal behaviour, which has higher associated costs. For example, a report by the Audit Commission (2004) on the reformed youth justice system includes an illustrative case study of a 15-year old male teenager who has had been involved in criminal behaviour. The case study includes some estimated costs associated with his behaviour, including approximately £13,000 associated with police time, Youth Offending Team ( YOT) involvement and Court appearances relating to theft and taking a car and approximately £51,500 for a six-month custodial sentence in a secure unit. Local crime and disorder costs (e.g. for a youth committing car crime) in one of the areas in the study of the Intensive Family Support Projects (Nixon et al, 2006) were estimated as £36,575 per year (about £95 per day) for provision in prison or Young Offenders Institutes and to have an overall annual cost (i.e. including aspects such as tagging and probation) of at least £50,000.
A more comprehensive approach to identifying the costs associated with criminal activity against individuals and households in England and Wales was adopted in a study of the economic and social costs of crime undertaken for the Home Office in 2000 (Brand and Price, 2000) and subsequently updated to 2003/04 values in 2005 (Dubourg & Hamed, 2005). These studies consider the costs as falling into three categories:
- Costs in anticipation of crime;
- Costs as a consequence of crime;
- Costs in response to crime.
The costs for 2003/04 are reproduced in Table A3.13a. They include both the Exchequer consequences of crime (e.g. the costs associated with the criminal justice system and the NHS) and the costs borne by the individual (e.g. in terms of the physical and emotional impact of the crime) and by society (e.g. in terms of lost output). The disaggregated costs associated with the criminal justice system are shown in Table A3.13b.
Table A3.13a - Estimated average costs of crimes against individuals and households in 2003/04 by crime type and cost category
Costs in Anticipation of Crime (£)
Costs as a Consequence of Crime (£)
Costs in Response to Crime (£)
Average Cost (£)
Physical & Emotional Impact on Direct Victims
Value of Property Stolen
Property Damaged / Destroyed
Criminal Justice System
Violence against the person:
Burglary in a dwelling
Theft - not vehicle
Theft of vehicle
Theft from vehicle
Attempted vehicle theft
Source: Dubourg and Hamed, 2005
Table A3.13b - Estimated average Criminal Justice System costs of crimes against individuals and households in 2003/04 by crime type and CJS cost category
Costs in Response to Crime (£)
Average Total CJS Cost (£)
Non-legal Aid Defence
Other CJS Costs
Criminal Injuries Compensation
Violence against the person:
Burglary in a dwelling
Theft - not vehicle
Theft of vehicle
Theft from vehicle
Attempted vehicle theft
Source: Dubourg and Hamed, 2005
The original study by Brand and Price (2000) indicated that the benefit of preventing the 'average' offence in 2000 was valued at about £2,000. Recent research on the cost of exclusion (Prince's Trust, 2007) used the data for 2003/04 to estimate that the average cost per crime committed by young people aged 10-21 years (weighted by the prevalence of the types of crime among young people) was almost £4,600 in 2004. These values, which include the impact on individuals and society as well as on Exchequer-funded services, show that considerable cost savings are associated with preventing criminal activity. Some of the costs associated with criminal activity by young people are summarised in Table A3.14.
Table A3.14 - Costs of criminal activity by young people
Costs associated with a teenager involved in criminal behaviour
£13,000 for police time, Youth Offending Team involvement and Court appearances; £51,000 for 6-month custodial sentence at Young Offender Institution
Audit Commission, 2004
HMP and YOI provision in a Local Authority in north-west England
About £95 per day and £36,575 per year. Overall annual costs of at least £50,000. Overall annual cost includes aspects such as tagging and probation
Nixon et al, 2006
Average cost per crime committed by a young person aged 10 - 21 years
Prince's Trust, 2007
(vii) Cost consequences for education
Some of the costs associated with education provision have been referred to in the above discussion (e.g. under the costs of looked after children) or are considered as part of the discussion of the long-term cost consequences of anti-social behaviour. Expenditure on school education in Scotland in 2005/06 was £4,138 per pupil in primary education and £5,771 per pupil in secondary education (Scottish Executive, 2007).
The final evaluation of the Early Intervention Case Management Project in Edinburgh (Brodies LLP, 2007) includes a number of potential savings associated with the Project provided by the Council's Children and Families Finance Section. These suggest an annual cost for education support of £8,557 (£165 per week) in 2006/07.
New Philanthropy Capital published a report in June 2007 on the costs of truancy and exclusion from school (Brookes et al, 2007), which is considered in more detail as part of the discussion of longer-term cost consequences. The report calculates that it cost £831 (in 2005 prices) to manage a permanent school exclusion. Most excluded children enter a Pupil Referral Unit ( PRU), which provides a narrower curriculum in a specialist setting with a higher teacher:pupil ratio. The annual costs of alternative provision for excluded pupils are shown in Table A3.15.
Table A3.15 - The destination and costs of alternative provision for excluded pupils
Alternative Education Provision
Cost per annum
% of Excluded Children
Pupil Referral Unit
Average weighted total
Source: Brookes et al, 2007
Based on an average cost (in 2005 prices) of £4,355 for a place in a mainstream school, the report calculates that the average extra cost of educating a permanently excluded pupil is £7,181 per year.
(viii) Cost consequences for health
Many members of families involved with intensive family support projects have a variety of health-related problems at referral associated with poverty and low socio-economic status, such as alcohol and drug misuse, obesity, depression, and self-harm. They may be taking prescription drugs for depression and may be attending their local A&E department frequently (e.g. due to domestic violence). Whilst it is recognised that some additional NHS health-related expenditure may be incurred because of the projects (e.g. leading to referrals for help with mental health problems or drug/alcohol misuse), these additional costs will be very small compared with the likely subsequent costs for the NHS (and other agencies) of not identifying and addressing some of these problems (e.g. the costs resulting from a mother having a serious mental breakdown).
Research on the impact on health is scarce, but limited estimates are included in some of the studies considered in the discussion of long-term consequences.
(ix) Costs of domestic violence
Members of some of the families involved with intensive projects are likely to have experienced domestic violence, either from a partner (current or former) or from a child (e.g. a teenage son on his mother). For example, domestic violence in some form was a significant issue for almost half of the families working with the intensive family support projects (Nixon et al, 2006). A major study of the costs of domestic violence was published in 2004 (Walby, 2004). Based on the Home Office framework for costing crime, it estimates the cost of domestic violence for the state, employers, and the men and women who are subjected to it. The total annual cost of domestic violence to services (criminal justice system, health, social services, housing, and civil legal) amounted to £3.1 billion (plus a loss to the economy of £2.7 billion). The aggregate component costs (costs are not often presented at a unit level) are shown below:
- Criminal Justice System ( CJS): Domestic violence costs the CJS about £1 billion per year, which is about one quarter of the CJS budget for violent crime. The largest single component is that of the police. Others include prosecution, courts, probation, prison, and legal aid.
- Health Care: The cost to the NHS for physical injuries is around £1.2 billion a year. This includes GPs and hospitals. Although physical injuries account for most of the NHS costs, mental health care is estimated to cost an additional £176 million per year.
- Social Services: The annual cost is about £0.25 billion and is overwhelmingly for children rather than adults, especially those caught up in the co-occurrence of domestic violence and child abuse.
- Housing: Expenditure on emergency housing includes costs to Housing Authorities and Housing Associations for housing those who are homeless due to domestic violence, housing benefit for such emergency housing, and the costs of refuges. These elements amount to a total of £0.16 billion a year.
- Civil Legal: These services cost over £0.3 billion per year, about half of which is borne by legal aid and half by the individual. This includes both specialist legal actions such as injunctions to restrain or expel a violent partner, as well as actions consequent on the disentanglement of marriages and relationships such as divorce and child custody.
These costs, which are summarised in Table A3.16, exclude any human and emotional costs. Domestic violence leads to pain and suffering amounting to an estimated additional £17 billion per year, which is not included in the figures. Even if these non-Exchequer-related costs are excluded, a considerable amount of the expenditure of some statutory services arises due to domestic violence, and some of these costs will be avoided due to the work undertaken by intensive support projects.
Table A3.17 - Costs of domestic violence
Total cost to services
£3.1 billion per year
Services include criminal justice system, health, social services, housing, and civil legal
Loss to the economy
£2.7 billion per year
Human and emotional costs
£17 billion per year
(ix) Costs of addressing conduct disorders in pre-school children
The following discussion is included for comparative purposes because it provides estimates of the costs of providing a parenting programme to the parents of young children displaying conduct disorder based on a randomised controlled trial. Conduct disorder is a major health and social problem and the most common psychiatric disorder in childhood. The diagnosis is given to children who display persistent severe anti-social behaviour such as tantrums, verbal and physical aggression, lying, stealing, and violations of other people's rights (Scott, 2007). It is estimated to affect at least 5% of children aged 5 - 15 years in the UK and the USA. For those children with early onset in pre-school years, conduct disorder often persists into adulthood and predicts poor employment prospects, marriage breakdown, and self-harming or anti-social criminal behaviour. A study of the financial costs of social exclusion (Scott et al, 2001, see also A3.5 below) showed that the economic implications of severe behavioural problems in childhood are serious and enduring.
Parenting is a key determinant in childhood behaviour and a recent study of a parenting programme for parents of children at risk of developing conduct disorder (Hutchings et al, 2007) included analysis of its cost effectiveness (Edwards et al, 2007). The findings of this programme are discussed below as it indicates the costs associated with such a programme 37. These can be compared with the costs associated with delivering intensive family support projects, which provide a much wider range of interventions.
The study comprised a pragmatic randomised controlled trial of a group parenting programme - the Webster-Stratton Incredible Years basic parenting programme - delivered in the community through Sure Start in north and mid Wales. The programme is delivered by two trained leaders in 12 weekly sessions. Parents of 153 children aged three and four years at risk of developing conduct disorders (identified by health visitors) were recruited to the study and randomised to the intervention group or a six-month waiting list control. Full economic data were available for 116 of these families. The main outcome measure was the incremental cost per unit of improvement on the intensity score of the Eyberg child behaviour inventory.
The economic analysis estimated a cost of £73 per one point improvement on the intensity score (95% confidence interval: £42 - £140). It would cost £5,486 to bring the child with the highest intensity score to below the clinical cut-off point and £1,344 to bring the average child in the intervention group within the non-clinical limits on the intensity score. The mean cost per child attending the parenting group was £1,934 for eight children and £1,289 for 12 children. These costs include the initial costs and the costs of materials for training group leaders and are comparable with the costs of most psychological treatments. The programme seemed to be most cost effective in those with the highest risk of developing conduct disorder. The clinical study also found important benefits to the parents' mental health and the behaviour of siblings. The authors concluded that the parenting programme involved modest costs and demonstrated strong clinical effect.
Although this programme is not directly comparable with the intensive family support projects considered in this present study, the costs are nevertheless interesting. Although the intensive family support projects discussed in Section A3.3 generally cost about £10,000 - £13,000 per family, the projects work intensively with all family members (and families often including three or more children under the age of 18 years) over a prolonged period of time in situations where poor behaviour and other problems may already be entrenched. By contrast, the parenting programme considered here focuses mainly of preschool age at risk of developing conduct disorder. Such programmes (and, indeed, programmes delivering parenting skills to parents with children of school age) may be a useful element of intensive family support projects. The costs discussed above suggest that intensive family support projects may offer good value for money, given their wider impacts.
(x) A case study illustrating potential short-term cost consequences
An illustrative case study cited in the evaluation of intensive family support projects in England (Nixon et al, 2006) shows the estimated costs over a year for a vulnerable family with four children that did not work with an Intensive Family Support Project in Table A3.18. It shows that the short-term costs associated with non-participation can be considerable (£334,000 over 12 months for this family), suggesting that the costs associated with delivering an IFSP can be covered by achieving 'successful' outcomes in the short term for a relatively small number of families.
Table A3.18 - Case study: potential short-term cost consequences for the Exchequer
Rachel is a single mother with four children - Matt (14), Kaylee (13), Johnny (10) and Emma (6). The children have little or no contact with their father, who left their mother shortly before Emma was born and is currently in prison for drug dealing. They have lived in their present home for several years, but their tenancy is now at risk because of the behaviour of some of the children. Matt is described as being "out of control" and "unmanageable" by Rachel. He has been involved in car theft, house burglaries and shoplifting, is often drunk, noisy and abusive, and on occasions has hit his mother, once breaking her arm. He has been excluded from school because of bullying, but was never a regular attender prior to being excluded. Kaylee is part of a local group of older teenagers who live on the estate. They are often seen smoking and drinking on the streets and vandalising property. She rarely bothers to attend school, despite Rachel's best efforts, and is often rude and aggressive. Johnny is hyperactive and has recently been diagnosed as dyslexic. He enjoys school and attends regularly, but is not making much progress. Emma is treated as the baby of the family by her siblings, who tend to spoil her. She suffers badly from asthma, which can restrict her physical activity, and she is very overweight However, she is friendly and confident and is doing very well at school, which she loves. Rachel is morbidly obese and smokes heavily. She left school with no qualifications and has never been employed. She is prone to bouts of depression, when she tends to ignore the children and let them look after themselves. Consequently the children are often hungry and grubby and the house is a complete mess.
If their current behaviour continues, they will be evicted from their home and the following approximate costs could be incurred during the subsequent 12 months:
Six-month custodial sentence in a secure unit for Matt
Six months in a specialist (out of area) Children's Home for Matt
One year in a local Children's Home for Kaylee (who refuses foster care)
One year in local foster care for Johnny (@ about £600/week)
One year in local foster care for Emma (@ about £600/week)
It is assumed that the rent for Rachel's place in a women's hostel (or other form of temporary accommodation) costs about the same as the rent on the family's current home and that these are paid for by Housing Benefit.
Involvement with a project could prevent some (or possibly all) of these short-term costs, although some additional costs would be incurred. For example, Rachel needs help to address her weight problems, depression and smoking and improving her parenting and life skills; intensive involvement of a learning mentor or other educational specialist with Kaylee and Johnny. A dietician may be able to reduce Johnny's hyperactivity and improve everyone's diet and weight. The YOT needs to work closely with Matt and may be able to prevent him receiving a custodial sentence. He also needs to learn to control his aggression and his drinking and he should receive some form of education or training, possibly in a specialised unit. If the members of this dysfunctional family can recognise and tackle their problems, this should save considerable amounts of expenditure not only in the short-term but also in the future, as well as improving their overall quality of life and future prospects.
Source: Taken from Nixon et al (2006), Anti-Social Behaviour Intensive Family Support Projects: An evaluation of six pioneering projects, Communities and Local Government (p147)
A3.5 Longer-term cost savings
The cost consequences considered above relate to some of the potential short-term cost consequences of reducing tenancy-related problems and anti-social behaviour. However, it is also important to recognise that problems experienced during childhood are likely to have much longer-term cost consequences. A number of research studies have addressed these issues, which include:
- The costs of social exclusion;
- The costs of poor educational attainment;
- School attendance and educational attainment;
- Truancy and exclusion;
- Being ' NEET';
- Literacy difficulties;
- Youth disadvantage;
- The impact on earnings from employment:
- Impact on lifetime earnings;
- Fiscal benefits of increased employment.
There are considerable overlaps between many of these studies. Their cost-related aspects are described below.
(i) The costs of social exclusion
Members of families engaging in ASB are likely to experience social exclusion. A study by Scott et al (2001) of the financial cost of social exclusion published in the British Medical Journal ( BMJ) compared the cumulative costs of public services used through to adulthood by individuals with conduct problems and conduct during childhood. The study followed 142 children from an Inner London Borough from the ages of 10 to 28 years. They were divided into three groups - 'no problems', 'conduct problems', and 'conduct disorder' (i.e. a persistent and pervasive pattern of ASB in childhood or adolescence, where typical behaviours include disobedience, tantrums, fighting, destructiveness, lying and stealing) 38. Conduct disorder behaviour is strongly associated with social and educational disadvantage. The study, which is referred to and used in several subsequent pieces of research, found that crime incurred the greatest costs, followed by extra educational provision, foster and residential care, and state benefits. Health care costs were comparatively small.
By the age of 28, the mean individual total costs for each group were £70,019 for the 'conduct disorder' group, £24,324 for the 'conduct problem' group and £7,423 for the 'no problem' group 39. Thus the costs for individuals with conduct disorder were ten times higher than for those with no problems. The study concluded that:
"Antisocial behaviour in children is a major predictor of how much an individual will cost society. The cost is large and falls on many agencies, yet few agencies contribute to prevention, which could be cost-effective".
(ii) The Costs of poor educational attainment
- Links between school attendance and educational attainment
Research reported by the National Audit Office in a report on improving school attendance (National Audit Office, 2005) includes the following data from a survey of 30,000 16-year olds in England:
- 60% of non-truants achieved five or more GCSEs at grades A to C, compared with 40% of occasional truants and 13% of persistent truants;
- 2% of non-truants achieved no GCSEs, compared with 5% of occasional truants and 25% of persistent truants;
- 96% of non-truants were in education, employment or training, compared with 89% of occasional truants and 66% of persistent truants.
These findings illustrate clearly the importance of reducing truancy levels to improve educational attainment. Although these data relate to England rather than Scotland, there is nevertheless likely to be a similar correlation between school truancy and poor educational attainment in Scotland.
- Costs of Truancy and Exclusion
New Philanthropy Capital published a report in June 2007 on the costs of truancy and exclusion from school (Brookes et al, 2007). It considers the social costs and the returns from effective ways of tackling social problems and draws upon partial data and a series of indirect proxies "to piece together a picture of the lifetime costs associated with exclusions and truanting". 40
It includes the costs:
- To the education system;
- Of lower earnings;
- To the health service;
- Of higher crime;
- To social services.
All calculations use 2005 prices and all future costs are discounted at 3% per annum.
The report estimates (conservatively, it states) that the average excluded child costs £63,851 to society (in present value terms). This includes costs to the child in future lost earnings resulting from poor qualifications and also costs to society in terms of crime, health and social services, as shown in Table A3.19. More than three-quarter of the costs fall on society. Each year, there are over 10,000 new exclusions from school, producing a total cost of £650 million per annum.
Table A3.19 - The costs (present values) of a permanent school exclusion (2005 prices)
Cost to individual *
Cost to society
Total cost of an exclusion
* lower than the gross lost earnings gap of £21,175 due to income tax and NI contributions of £6,988
The cost for the education system is based on the assumptions that the average age of exclusion is 12_ years and that the average additional cost of educating a permanently excluded pupil is £7,181 per year (i.e. £11,536 - £4,355, see Table A3.19). The aggregate cost to the education system of an exclusion comprises administrative costs (£676 discounted value, as this will occur 8 years in the future in 2013) plus alternative education costs (£19,434 over 3_ years from exclusion). These total £20,110.
The authors cite research showing that excluded children are:
- Three times more likely than their peers to leave school with no qualifications;
- Half as likely to get a degree;
- 37% more likely to be unemployed.
The combined effect of lower qualifications leading to lower pay and higher unemployment means that there is a predicted gap in earnings for permanently excluded pupils of £21,175 over a lifetime. This comprises a cost of £14,187 to the individual and a burden to taxpayer of £6,988 (i.e. the present value of lost tax and NI receipts associated with the gap in earnings) - i.e. expected loss of income to the taxpayer every time a child is permanently excluded from school.
The most common cause of school exclusion is persistent disruptive behaviour. The costs for the NHS due to school exclusion are calculated from the study in the BMJ of the costs of social exclusion (Scott at al, 2001) described above. Relative to the group with no problems, those with conduct problems and disorders each incurred an average additional cost of £1,019 in health costs. This includes hospital inpatient and outpatient costs, psychiatric costs and costs arising from abortions or miscarriages. Due to a lack of data from other studies, it was not possible to estimate any additional costs falling on the NHS after the age of 28 (the upper age limit in the BMJ study).
The lifetime cost of crime committed by excluded child is estimated to be £15,527, based on data from the BMJ study between the ages of 10 and 28 and on data linked to standard rates of recidivism and the age profile of the prison population after the age of 30.
The additional costs to social services comprise two elements which total £6,021. The cost of social services' involvement with excluded children is based on previous research and calculated to be £1,169 per year in 2005 prices, or £3,1665 over the assumed average of 3_ years between exclusion and reaching 16 years of age. The BMJ study provides data on the costs of residential and foster care for children with conduct disorders and problems, and these data give a total (discounted) cost of £2,856 between the ages of 10 and 16 years.
A similar approach is used to show that the average cost of a persistent truant is £44,468, as shown in Table A3.20. This figure splits roughly fifty-fifty between costs to the individual and costs borne by the rest of society. An estimated 198,000 children in the UK are persistent truants, missing at least five weeks of school per year through unauthorised absences. They represent 2% of the total school population of the UK.
Table A3.20 - The costs (present values) of a persistent truant (2005 prices)
Cost to individual*
Cost to society
Total cost of a persistent truant
* lower than the gross lost earnings gap of £33,694 due to income tax and NI contributions of £11,113
In terms of the costs borne by the education system, work by the National Audit Office in England on improving school attendance estimated that educational welfare services in England cost £108 million in 2002/03. This gives a cost per truant of £706 per person per annum in 2005 prices. Based on the assumption that the costs to the education welfare system are only felt over two years, when the child is 11 and 12, gives a total cost of truancy of £1,200 to the education system in 2005 prices.
The impact of truancy on qualifications, and therefore on earnings, is higher than for excluded children. This is because the majority of those who are excluded receive alternative provision and therefore are better educated than those who are absent from school and disengaged with education. For example:
- 60% of persistent truants fail to get any A-C grade GCSEs;
- One in four gets no qualifications at all;
- Earnings are on average 13% lower for truants.
The combined effect of lower qualifications leading to lower pay and higher unemployment results in a predicted gap in earnings of £33,694 over a lifetime. As with exclusions, some of the lost earnings are income tax payments and NI contributions lost to the taxpayer. These come to £11,113. Adding in higher unemployment benefits totalling £518 means that a total of approximately £11,631 in 2005 prices is lost to the taxpayer.
The estimated extra lifetime cost to the NHS of a truant is estimated to be £832 in 2005 prices, based on data from the BMJ study (Scott at al, 2001). As this study only considers extra costs up to the age of 28 the estimate is very conservative. For example, persistent truants are more likely to smoke, drink, take drugs and be sexually active than their non-truanting peers. A recent study on pupil drug misuse in Edinburgh (McAra, 2004) found that 23% of the surveyed truants reported weekly alcohol consumption compared with 7% of non-truants in the same study. Twenty% of truants smoked, compared with 3% of non-truants. These behaviours carry long-term health risks, but as data to quantify the extra risk for truanting children was not available, the associated costs to the NHS could not be included.
Truants are both more likely to commit crime and to become the victims of crime. Sixty-five% of teenagers who truant once a week or more self-report committing offences, compared with 30% of their peers (Stevens and Gladstone, 2000). Based on the BMJ study (Scott et al, 2001) and taking the costs of the group with conduct problems as a proxy for truants, this study estimates that the lifetime extra cost of crime committed by a truanting child is £6,776.
To estimate the additional cost to social services, the group with conduct problems in the BMJ study (Scott et al, 2001) is used as a proxy for truants. Brookes et al (2007) estimate that an extra £1,967 is spent on residential and foster care for truants between the ages of 10 and 16.
- The costs of being ' NEET'
A study by Godfrey et al (2002) provides some preliminary estimates of the costs associated with young people being 'not in education, employment or training' ( NEET) when aged 16 - 18 years. It estimated the additional costs that would be incurred by a defined group of 157,000 young people who were NEET at the end of 1999 compared with the hypothetical situation that these young people had the same current and future experience as the rest of their contemporaries. Costs were interpreted broadly to include costs to individuals, their families and the rest of society. The study attempted to provide estimates across the lifespan of the defined group and considered current, medium- and long-term costs. Implications for public finance costs, including changes to benefit payments and taxes, were also investigated. The study authors found that it was not possible to quantify all costs and were only able to provide estimates for educational underachievement, unemployment, inactivity, crime, substance abuse, and health. Health and crime costs were generally found to be relatively low compared with the costs of educational underachievement, underemployment and unemployment. Two specific case studies based on a hypothetical male and female teenager illustrate how costs can accumulate to about £300,000 41 over a person's lifetime. Many children engaging in ASB have poor records of educational attendance and attainment, and this is likely to increase their probability of being NEET in their late teens.
- The long-term costs of literacy difficulties
The KPMG Foundation 42 commissioned a study of the long-term consequences of literacy difficulties for individuals and for society ( KPMG Foundation, 2006). Around 6% - 7% of 11 year olds in England leave primary school with very poor literacy skills equivalent to those of the average seven or eight year old. Slightly less than one-in-ten (9.2%) of all boys leave primary school at this level of reading.
Reviewed research showed that children with early literacy difficulties often experience negative lifetime experiences. For example, literacy difficulties are linked to costly special educational needs provision, to truancy, exclusion from school, reduced employment opportunities, increased health risks and a greatly increased risk of involvement with the criminal justice system. The study attaches costs to these risks and summed them over the life course to the age of 37 (the last point at which reliable survey data were available). The resulting costs to the public purse to age 37 43 arising from failure to learn to read in the early years at primary school are estimated at between £44,797 and £53,098 (present values) for each individual. These estimates are conservative and do not include savings that could not readily be quantified, such as the economic effects of reduced spending power, social housing costs, the costs of generally poorer health, the costs of substance abuse over the age of 18, and the costs of intergenerational effects on literacy skills.
The estimation and full quantification of social costs depends on four critical pieces of data:
- Population numbers;
- Prevalence rates (i.e. what % of the population have this problem or incur this cost);
- Typical frequency and/or duration of problem (i.e. number of episodes and over what period of time);
- Unit cost information (actual or proxy) for each specific type of social cost.
Unit cost information is taken from other published cost-benefit studies and from national sources for health and social care services, criminal justice and benefit receipts. Other costs have been calculated from first principles using agencies' data. All costs are quoted at 2006 prices and a discount rate of 3.5% has been assumed for all present value calculations.
As with other similar studies, this study focuses on five different types of costs for children who have not learned to read by the age of 7:
- Education costs;
- Employment costs;
- Being ' NEET';
- Health costs;
- Costs of crime.
These costs are summarised in Table A3.21. Two total amounts are presented, primarily to capture different assumptions about crime-related costs (although the lower bound value also excludes the costs of maintaining statements of special education needs in secondary school). Costs of housing benefit and social housing are excluded due to a lack of data on the housing status of adults with low literacy skills.
Table A3.21 - Estimated cost consequences associated with unaddressed literacy difficulties
Annual or One-off Amount
Special needs support - literacy and behaviour (whole of KS2)
Special needs support - literacy and behaviour (whole of KS 3 & 4)
Costs of maintaining SEN statement - secondary education
Educational psychologist time (3 hours @ £85 per hour)
Truancy (12 or more half days of unauthorised absence in first term of 2005/06 school year)
Tax and NI revenue (male)
Tax and NI revenue (female)
Cost of unemployment benefits (male only)
Costs of being NEET:
Social costs (male)
Social costs (female)
Depression (male and female)
Obesity (male and female)
Prison costs (male)
Present value of total cost per person to age 37 without intervention
The cost of a permanent exclusion has been estimated to be £10,555. This comprises a cost of £1,000 per exclusion in administrative costs and an annual cost of £12,555 for a place in a Pupil Referral Unit less an assumed £3,000 for the age-weighted pupil unit that would otherwise have been spent on the pupil.
The direct cost associated with an episode of truancy (which is defined differently from the New Capital Foundation study) has been estimated as £1,675. This is based on costs for 2003 of £1,500 for one prosecution, plus 2 hours of Education Welfare Officer time (at £15 per hour) uplifted to 2006 values. This will be a conservative estimate, as multiple episodes of truancy are probable. However, the authors could not find any data on numbers of truancy episodes resulting in prosecution.
The costs of being NEET are drawn from the study by Godfrey et al (2002) discussed above, which provides estimates of the social costs incurred by this group as a result of underemployment, poor health, substance abuse, teenage pregnancy and involvement in the criminal justice system. These have been applied to the percentage of the very low literacy group that would have been NEETs from 16-18, having first taken out costs already covered elsewhere in the current study so as to avoid double counting (i.e. Job Seekers Allowance), costs of crime and (for males only) the costs of being teenage mothers. These produce annual costs of £5,802 (male) and £10,072 (female) at 2006 prices. Costs are higher for females because of the high rate of teenage pregnancies in the NEET group.
Health costs only cover depression and obesity, as the study authors were unable to quantify costs associated with generally poor health. Costs of substance abuse (including alcohol) are included in the costs of being NEET aged 16-18. Costs of smoking are not included at any age since the tax benefits accruing to the Exchequer were assumed to cancel out the costs to the health service associated with smoking.
Two data sources have been used as the bases of the estimated costs of crime that result from early failure to learn to read. The authors used the study of social exclusion by Scott at al (2001) to estimate a cost of crime by males of £46,550 spread over ages 11 to 27, which includes costs of court appearances, probation, youth justice and imprisonment. Data from HM Prison Service give mean prisoner costs of £35,862 (under 18) and £25,880 (adult: 19-37). The resultant estimated cost impacts depend crucially upon the assumptions made about the percentages of the prison population who would not be in prison had they had better literacy skills.
Therefore employment-related costs form the largest category of potential savings due to improving literacy levels, with costs to the education system and the costs of crime providing the next largest categories.
- The costs of youth disadvantage
The Centre for Economic Performance at the London School of Economics has recently undertaken some research for the Prince's Trust on the aggregate costs associated with youth disadvantage in the UK (Prince's Trust 2007). This work focuses on three main aspects:
- Youth unemployment;
- Youth crime;
- Educational underachievement.
It draws on data from a wide range of sources and explores the inter-relationships between the above aspects and their impact on social exclusion. It also considers some of the costs for individuals and for the economy (including the UK's productivity and international competitiveness) associated with young people being 'not in education, training or employment' (i.e. being ' NEET'). For example, there is a significant cost to the individual in terms of reducing their lifetime chances and opportunities (e.g. youth unemployment has been estimated as imposing a wage scar on individuals of between 8% and 15% over time). Youth unemployment and inactivity are estimated to cost the Exchequer about £20 million per week in Job Seeker's Allowance and to result in a productivity loss to the UK economy of over £70 million per week. The estimated cost of youth crime in Great Britain is in excess of £1 billion per year. Furthermore, underachievement at school (with its resultant impact on skill levels and the workforce) is also partly responsible for the relatively poor performance of the UK economy and the productivity gap between the UK and its competitors. The UK has between 10% - 25% lower output per hour than France, Germany and the US, much of which can be attributed to poorer levels of skills (and a shortfall of capital investment). Being NEET also has knock-on consequences on health status (including on mental health/depression), which in turn imposes long-term costs on individuals and society.
The following quotes are taken from the report. They illustrate the importance and cost-effectiveness (for individuals, neighbourhoods and society) of initiatives such as the IFSPs that reduce anti-social behaviour and youth offending and promote educational achievement:
"The cost of educational underachievement in the labour market in terms of unemployment and wage penalties is significant. And underachievement at school appears to increase the probability of turning to crime and negatively affects the health and emotional well-being of the individuals concerned".
"[This research] reveals that interventions helping young people get into work, stay on in education or avoid crime represent excellent value for money given the measurable costs of social exclusion".
(iii) Impact on earnings from employment
- Impact on lifetime earnings
Evans & Eyre (2004) modelled the lifetime analysis of current British social policy. They constructed models of lifetime income streams (from earnings, benefits and pensions) for people with various lifetime experiences. The analysis shows that, based on the policies in operation when they undertook their research, people with low earnings throughout their lives (e.g. due to being relatively unskilled) experience many restrictions in terms of access to financial benefits such as owner-occupation and pension schemes. They are "trapped out of opportunity". This situation is likely to be experienced by many people who are low educational achievers at school. The risk of this may be exacerbated by failed tenancies and anti-social behaviour during childhood.
- Fiscal benefits of increased employment
A recent report for the Department for Work and Pensions ( DWP) (Freud, 2007) has considered ways of reducing dependency and increasing opportunities in the context of exploring options for the future of welfare to work. Part of this research explores the fiscal benefits of increased employment. For example, it estimates that:
- The gross annual savings to the DWP of moving an average recipient of Incapacity Benefit into work is £5,900, with the wider exchequer gains (offsetting direct and indirect taxes paid with additional tax credits) raising this figure to £9,000 per year;
- The equivalent figures for a recipient of Job Seeker's Allowance are £4,100 and £8,100, respectively.
- The DWP's gross savings on lone parents are £4,400 (with no further Exchequer savings because of the weight of extra childcare elements of the tax credit system balancing other tax revenues).
The report also recognises that those on benefits often do not work for many years. For example, once a person has been on incapacity benefits for a year, they are on average on benefit for eight years. If the full annual Exchequer saving of getting a person on incapacity benefits into work is around £9,000, a genuine transformation into long-term work for such an individual is currently worth around £62,000 per person to the State 44. Although Projects providing intensive support to families with anti-social behaviour and at-risk tenancies do not have increasing employment as a stated objective, these figures provide a good indication of some of the potential longer-term cost consequences for the Exchequer of periods of unemployment. Improving family members' education, skills and training (e.g. through or as a consequence of the project's intervention) will reduce the likelihood of unemployment and therefore of such costs being incurred by the Exchequer.
The importance of acquiring education and qualifications is also recognised within the DWP research. Data are presented from the UK's Labour Force Survey showing that about half of working age adults with no qualifications are not in employment. The DWP report also emphasises some of the benefits for health and well-being associated with work, based on other research undertaken for DWP (Waddell and Burton 2006 45).
All of the above publications show that preventing failed tenancies, reducing anti-social behaviour and encouraging children to attend school can have significant financial benefits in terms of "saving" (i.e. preventing) costs that might otherwise have been incurred. Although the focus has mainly been on the tangible cost consequences for Exchequer-funded services, considerable benefits will also be enjoyed by individuals, families, local neighbourhoods and communities, and society. Many of the benefits associated with intensive family support projects will be enjoyed by adults as well as by children and young people. Some of the benefits are expected to extend over a person's lifetime (and, indeed, may also extend to future generations). Table A3.22, which is taken from the report by Nixon et al (2007) looking at the longer-term outcomes associated with some of the families who had worked with the intensive family support projects considered in their earlier study, summarises the main cost consequences associated with such projects.
Table A3.22 - Summary of main cost consequences associated with IFSPs
TO THE EXCHEQUER
Current Short-term Costs/Resource Savings
Short-term Costs Prevented
Longer-term Costs Prevented
Legal and other costs associated with eviction
Foster/residential care (including secure provision)
Criminal Justice (e.g. Young Offenders Institute)
Police and Criminal Justice
Additional revenue via taxes on wages and family expenditure and NI contributions
TO INDIVIDUALS AND FAMILIES
Short-term and Longer-term Benefits
Improved family functioning and cohesion - less risk of family breakdown
Improved financial management skills (including receiving all benefits to which entitled)
Improved access to support and other services (e.g. for treating existing problems relating to health and education)
Improved education and training, leading to acquisition of qualifications and skills which enhanced employment opportunities and life chances
Improved lifetime earning potential
Less reliance on benefits (and criminal activity) as source of income
Better health (including mental health)
TO NEIGHBOURHOODS, COMMUNITIES AND SOCIETY
Reduced ASB, crime and fear of crime
Reduced neighbour disputes and tensions
Lower stress levels (leading to better health and improved productivity at work)
Neighbourhoods do not become run-down (reducing crime and numbers of voids and knock-on impacts for police and housing providers)
Improved social capital within communities
Potential to contribute to improved workforce productivity and international competitiveness
Source: Nixon et al, 2007