Social security agency in Scotland: outline business case

Section A: Context and Introduction

1. Introduction

1. The devolution of social security to Scotland is an integral part of the Scottish Government's wider constitutional reform agenda following the Scotland Act 2016.

1.1. In March 2016 the Scottish Government announced that: "The social security system in Scotland can be seen to have a number of levels of delivery. This ranges from the governance of the entire system, the back room delivery functions which will process applications and arrange for payments to be made etc. to the user interface where customers will interact with the system. This system is in the process of being appraised over two stages."

1.2. A report was published ^[2] on our initial high level appraisal around the governance of social security in Scotland and the strategic case for change. It found that the governance body should have close links to Scottish Ministers and be flexible enough to respond as the social security landscape in Scotland unfolds.

1.3. Flexibility means having the capacity to expand and take on new work as well as being able to continually improve. A central agency with access to the wider resources of the Scottish Government family was seen as being able to deliver this flexibility. Minsters therefore decided that social security in Scotland should still sit within the Scottish Government family in order that it might be able to draw upon the strengths and resources of the parent organisation, when needed.

1.4. The work leading to this point, although not formally written up as such, was the process by which a case for change was initially made. It forms, along with the wider political debate and narrative around further devolution following the Independence Referendum, the Strategic Outline Business ( SOBC) case for social security delivery in Scotland.

1.5. Subsequently, work was undertaken to formally examine the potential options for delivery of social security in Scotland within the context of the decision to do so in the form of an agency. Work undertaken to extend the initial options appraisal into this second stage forms the bulk of the Socio-Economic Case and Financial case of this OBC.

1.1 The nature of this document

1.6. The nature of social security delivery is complex and is very different to a standard government investment project, particularly those of a capital nature. Whilst this OBC follows HMT guidance on the construction of business cases ^[3] it does so in a way that is consistent with the wider approach taken across Scottish Government that extends the narrow focus of the economic case under HMT guidance to consider the wider socio-economic impacts.

1.7. The form and structure of this OBC is driven by these factors. An initial section (Section A) discusses the context of the work and the approach undertaken.

1.8. Social security delivery within an agency could take many forms. Section B of this document details the process that was followed to develop the six options that are initially examined in this OBC. These options are primarily about how delivery takes place and who or what undertakes different components. The options are defined from the "bottom-up" in terms of the activities that a social security delivery system needs to undertake.

1.9. Section C applies the five cases approach. It examines the Socio-Economic and Financial Cases for the six options. An overview of the recognised Programme and Project Management ( PPM) methodology being used and governance structures is covered in the Management Case. It also revisits the Strategic case as discussed above and updates the position to take account of recent developments. Not least amongst this is the consultation on Social Security undertaken in 2016, the results of which are published as "A New Future for Social Security in Scotland".

1.10. The six options, crucially, represent a wide spectrum of possible delivery options. Each option would require very different contracting and procurement process and indeed would require significantly different organisational structures to deal with commercial aspects and interactions with other organisations. This means that the Commercial case developed in Section C is more about exploring these issues than providing detailed information on the specific way in which delivery would be taken forward.

1.11. Drawing together the evidence across the five cases leads to the creation of three hybrid options in Section D that represent the best combination of components of the options in terms of their ability to deliver against outcomes and their financial implications. These hybrids are assessed and this leads to the choice of a preferred option that forms the recommendation of this OBC.

1.12. In terms of the Financial case, it is important to note that the costs of administering these benefits refers to the staff, non-staff resources and infrastructure costs which will be needed under each option on an on-going basis, it does not include the amount paid out to those in receipt of benefits. The Fiscal Framework (detailed in the Strategic case in Section 3.12) between the UKG and SG makes provision for this money (based on equivalent DWP expenditure) to be transferred to Scotland. There are financial implications if the SG alter the nature of the benefits being transferred, top-up existing benefits or create new ones but these are future policy decisions that are not considered in this OBC.

1.13. Nor does this OBC examine changes in delivery policy from the current system. It is very clear that changes to this, specifically around the nature and frequency of assessments will have large impacts on the on-going costs and the expenditure on benefits. These decisions have not yet been made and are thus not considered at this point. However, within the Financial case, sensitivity analysis is undertaken on likely policy variation to ensure that such variation would not lead to different choices between the delivery options.

1.2 The scale of the delivery challenge

1.14. Expenditure on benefits in Scotland in 2015/16 was £18.3bn, 9.1% of the GB total. Of this 15.3% or slightly less than £3bn is to be devolved. See Figure 1 below.

Figure 1 - Total Benefit expenditure in Scotland 2015/16

1.15. Expenditure on benefits to help disabled people with the additional costs that they face, namely Disability Living Allowance ( DLA) and Personal Independence Payment ( PIP) and Attendance Allowance represent around 2/3rds of the expenditure. See Figure 2 below.

Figure 2 - Breakdown of benefit expenditure in Scotland 2015/16 on benefits to be devolved

1.16. The variation in size of the benefits to be devolved is significant as is their complexity. Most are not primarily dependent on income but on an individual's circumstances and health.

1.17. Administration of these social security benefits in Scotland is currently provided by the Department for Work and Pensions ( DWP). Section B of the stage 2 options appraisal looks to estimate the costs of six options for providing this function out with DWP.

1.18. These are:

• Option 1 - The agency centrally delivers social security in Scotland;
• Option 2 - The agency delivers social security in Scotland through local offices;
• Option 3 - The agency delivers non-discretionary benefits and local authorities deliver discretionary benefits;
• Option 4 - The agency delivers cash and benefits 'in kind' as goods, services or concessions;
• Option 5 - The agency provides governance but delivery of social security is done by others e.g. via procurement or a Service Level Agreement ( SLA) and
• Option 6 - The agency provides governance but delivery is embedded in a range of existing public services.

1.19. These options can be thought of as representing somewhat stylised views of the possible delivery landscape. See section 1.4 for an initial discussion and section 2 for a full elaboration. The following parts of this section provide wider context on the environment under which this OBC has been undertaken.

1.3 Link between Outline Business Cases and Options appraisal

1.20. This document is written as a formal OBC but it also provides a record of the Stage 2 appraisal of options for social security delivery in Scotland. The approach is based on HMT's Green Book.

1.21. The appropriate form for an options appraisal under the approach common in Scotland is a neutral document that provides the evidence base for decision making but does not itself make the decision. For other examples, relating to appraisal see Scottish Transport Appraisal Guidance ^[4] and more widely, Housing Need and Demand Guidance ^[5] . This is detailed in Section C of this document under the headings of the five case OBC model but Section C specifically and deliberately does not impose value judgements or reach a decision. As such, Section C of this OBC is similar to separate options appraisals conducted elsewhere in Scottish Government. Section D carries this process forward in order to arrive at a preferred option.

The completion of a study, as documented in a final report, allows the rationale behind a potential transport options to be presented in a clear, evidence-led manner and provides the information required by a decision maker to make an informed choice of the most appropriate options for design development. STAG Technical Database Section 1.1

The HNDA should be factual in scope and bring together the evidence upon which subsequent housing policy and planning policy decisions and interpretations should be based. HNDA Managers Guide Section 8.1

1.22. The standard approach within options appraisal is to compare options for an intervention against a base case when no action is taken. This is referred to as a "Do-nothing" or a "Do-minimum" option. For social security this is assumed to be DWP continuing to deliver all benefits in Scotland. Convention dictates that this is referred to as option 0 but this should not be taken to imply that it is actively being considered as an option. This option is the basis of the cost model detailed in Section C. However, this option is not practically viable as it both counters the point of devolution and DWP would be unable to deliver it (and allow for any changes from the current system). As such its prime purpose is as a comparator on the costs (and the basis by which the costs of the options are assessed).

1.23. The nature of social security and specifically the determination of the form of the delivery of social security means that a standard cost-benefit approach is not sufficient to provide the information that is required to choose between options. This is demonstrated at the start of the Socio-Economic case - a standard Cost-Benefit analysis of the options is undertaken but is not helpful in assessing the characteristics of the systems (other than cost) that vary between the options. Instead, a multi-criteria approach ( MCA) has been adopted to provide a broad spread of evidence to inform the socio-economic case. This includes drawing on submissions to the social security in Scotland consultation that ran from August to October 2016 ^[6] as well as a number of workshops with internal and external experts. The criteria used in the multi criteria analysis are shown in Figure 3 below and fully discussed in Section C.

Figure 3 - Criteria used in MCA

1.24. Note that each criteria has a number of sub-criteria. There is no weighting given to the criteria. Instead, the multi criteria analysis provides information on the benefits of the options against the range of criteria and, in a similar manner to Transport or Housing, it is down to the person making the decision to weigh the importance of that evidence against the associated costs.

1.25. The focus of the Financial case is to provide a steady state estimate of these costs. This is the typical annual cost for the social security system to operate normally. This normal operation would be characterised by each benefit having relatively small and smooth changes in numbers of people accessing them and administration costs year to year, although some degree of on-going change will be a part of the steady state social security system. Currently, two of the main benefits being devolved (Disability Living Allowance ( DLA) and Personal Independence Payment ( PIP)) are not in steady state, owing to the current transition of working age DLA claimants to PIP. Due to limitations on caseload forecast data, it is not possible to accurately predict long-term costs in, say, 2030. Therefore a simplification is made; administration costs are provided for 2020/21, on the basis that this is close to being in steady state. This relies on the assumption that all cases will have migrated from adult DLA to PIP.

1.26. In line with standard practice, the costs of each option are fixed in 2014-15 prices (based on the latest data available). This assumes that costs are constant in real terms. The Financial case then makes realistic and sensible assumptions about how caseload might vary in the future, based on demographics and projects the steady state costs forward. It also makes some basic assumptions about implementation costs, specifically regarding Information Technology. These are combined in a standard way to provide a measure of the long-term discounted cost of the system that would be found in any standard OBC. This is followed by a section that examines appropriate levels of optimism bias ( OB) in terms of implementation costs.

1.4 Scale, approach to options and challenge

1.27. The scale of this options appraisal is unprecedented in Scotland in recent history. By way of comparison, an options appraisal was undertaken for the Forth Replacement Crossing ( FRC) as part of a wider Transport Strategic Projects Review in 2006-7 ^[7] . The final three reports are broadly equivalent to the material in this OBC and run to several thousand pages. Like the FRC work this OBC provides full detail of the analysis and thinking that will inform any decision.

1.28. The total cost of the FRC project, which was estimated to be between £1.325 billion to £1.350 billion as of winter 2015 is slightly less than half what the annual expenditure on social security will be in Scotland even if the administration and (initial) implementation costs are excluded. Whilst there are large differences between the operation of a social security system and the construction of an estuary crossing, the scale and importance of the decision over the form of the social security system in Scotland is clear.

1.29. Another way of thinking of the choice of options is that it provides the information necessary to answer a series of questions, each of which have two possible answers and some of which have follow-ups. Once these questions have been answered the form of the agency will become clear. These questions are:

• Should the social security agency administer / govern the administration of all devolved social security benefits in Scotland?
• Should social security in Scotland make some provision for face-to-face contact (through local offices) in addition to that which would be provided for assessment and how is it best to provide this?
  If yes, should it be part of the agency or be in partnership with local authorities?
• Should we, as much as possible, aim to deliver social security through already available public sector services and organisations?
  If so, should it be local authorities or other public sector organisations?
  Should any aspect of social security be delivered by others such as the third or private sector?
• Should the agency deliver social security assessments for disability related benefits itself or work with others?
  If others, should it be private/third sector or public?.
  If public should it be local authorities or other existing public services?
• Should the social security agency in Scotland be responsible for providing benefits in cash only or offer a choice of cash and non-cash options ( e.g., discount cards, energy discounts, physical goods)?

1.30. And also a question in more general terms:

How best can we harness digital services for social security delivery in Scotland?

1.31. There are therefore five questions plus five sub-questions plus the consideration of digital which the options in this OBC aim to answer. To answer each question in a traditional manner by constructing all possible options would require a matrix of 2 ^[10] or 1024 options plus consideration to digital services. This would clearly be impractical. The approach taken of using the six options that are detailed in Section B allows these questions to be considered in a framework that is manageable. Note that these answers require an assessment of the benefits as well as the costs.

1.5 The five cases model and the choice of the preferred option for social security

1.32. The standard approach to government business case development follows the Five Case Model:

• Make a robust case for change - the 'strategic case';
• Optimise Value For Money in terms of economic, social and environmental benefit - the 'socio-economic case';
• Be commercially viable - the 'commercial case';
• Be financially viable - the 'financial case' and
• Be achievable - the 'management case'.

1.33. Section C of this document presents the current five cases for the social security agency. The complexity of the options means that the choice of the preferred option is a more complicated process than for many other projects or programmes.

1.34. Following Section C, Section D considers the merits of each of the six options in line with their respective costs. To account for the complexity of the system, Section D then constructs hybrid options based on the Socio-Economic and Financial cases that can best deliver the outcomes discussed in the Strategic case.

1.35. These hybrids are subjected to further analysis and a preferred option established as the conclusion of this outline business case.