SIBs don’t work for complex problems because they’re unaccountable to service users

By Stephen Sinclair, Neil McHugh and Michael J.Roy

Supporting change through social policy is an iterative process which develops participatory relationships with users. Social Impact Bonds fail on this front, rendering them unfit for purpose.

SIBs are intrinsically unsuitable for programmes that seek to transform people’s lives. That’s because such transformative interventions typically succeed only when they involve partnership with the people who are in need of help – and that’s not one of SIBs’ strengths.

SIBs are generally unaccountable to service users; they usually answer to funders. That’s their core failure and limitation. An inadequate relationship with service users means that SIBs typically lack sufficient understanding of – and sensitivity to – the causes of social problems that is required to achieve transformative change.

We should face up to these issues because, although there are only 32 SIBs operating in the UK, there is a considerable amount of enthusiasm for them. They have, and continue to receive, a lot of vocal and financial support from government.

Difference between technical and transformative interventions
In social policy, it’s helpful to distinguish between technical and transformative interventions. The former tend to be simple responses that encourage relatively uncomplicated changes and behaviour. Often they are about increasing service use, such as, for example, raising the uptake of vaccinations. SIBs could be suitable for such programmes and randomised controlled trials could be one way to evaluate them.

In contrast, transformative interventions try to generate profound changes in the dynamic between conditions and actions. So they must engage the agency of service users or the recipients of those interventions. These interventions address complex problems, often with multiple stakeholders, and try to ameliorate poor conditions. They’re very much contextual. Crucially, they’re about doing things with people rather than doing things to them. So, they might involve trying to reduce or tackle obesity and the problems around that very complex issue. Or they might seek to reduce health inequalities which are a major problem within the UK.

Recipe for social transformation
For social interventions to be considered transformational, they must engage with the means and the identities of those they seek to help. They should cultivate capability for independent living. There shouldn’t simply be a focus on producing outcomes. A good way of thinking about this is to consider education policy: it should be about cultivating independent and critical thinking rather than focusing simply on outcomes and achieving certain rates for test passes.

This approach should encourage us to rethink best practice. Where transformation is concerned, it shouldn’t be about uniform standardisation. Policy should respond to varying conditions and individual circumstances. What works in one area may not work elsewhere. We should resist the urge to scale up interventions indiscriminately.

Different approaches to evaluation
Evaluations should, in turn, provide insights that help us understand these very complicated human interactions and social conditions. Good policy development and evaluation should be about understanding interactions between agents and the subjective experiences of service users. It should be about getting to the bottom of why things work, which is a difficult and challenging issue.

There tend to be two approaches to evaluation. One tradition – let’s call it “positivist” – is exemplified by much of the evidence-based policy movement. It tends to favour randomised control trials in trying to attribute causality.

An alternative tradition – let’s call it “phenomenological” or “interpretivist” – regards policy in relation to complex social problems as less linear and more of an iterative process. It’s based more on reasoning from incomplete or imperfect data. It highlights, to some extent, the limitations of the evidence-based policy movement.

A great deal can be learned from the field of public health. It has a long history of engaging in, developing and evaluating complex social interventions. Indeed, the Medical Research Council (MRC) has issued guidance on how to develop and evaluate complex interventions. The MRC says it’s an iterative process and that the first step should be to understand subjects’ views and how they understand their own condition and treatment.

The political nature of social policy
So social policy, when trying to do anything complex or transformational, shouldn’t try to reduce issues to mechanistic terms such as providing inputs and then producing outputs. We should also recognise that the design of social policies is a political process. Indeed, simplifying complex conditions and problems into depoliticised inputs and outputs, far from being apolitical, is actually a form of ideology itself.

Nevertheless, SIBs tend to be represented as a technocratic and depoliticised approach to policy. They involve shifting responsibility for welfare policy away from government. In this vision, the role of government is diminished, becoming about purchasing or brokering public services.

SIBs are highly political
We would, instead, argue that SIBs are highly political in two key respects. First, they turn public services into commodities, rather than being services to citizens that require a relationship with those citizens. So SIBs “commodify” public services. Second, they give financial value to those fictitious commodities. So SIBs change what were non-monetary social relationships into financialised commodities.

This process diminishes and dehumanises service users who become reclassified as customers. This diminution is not immediately obvious because it is masked by language. Indeed the language of consumer culture – in which the key agent is the customer who is always said to be “right” – might seem to raise rather than diminish the status of the service user. But this is actually a fraudulent representation.

A real customer has control of the cash, providing him or her with consumer sovereignty. In reality, with SIBs, there is very limited consumer sovereignty because citizens’ welfare entitlements become transformed into contracts between organisations and stakeholders.

In this change, the quality of these services is altered. The previous public sector ethos is lost, replaced by a profit incentive. As a result, the service user, instead of being at the heart of everything, becomes an externality, a troublesome afterthought to the main activity which is a series of transactions between principal stakeholders.

What next for SIBs?
Where does all this leave us? SIBs encourage a focus on what is measurable. They also prioritise outcomes rather than trying to understand how these effects are actually achieved. We would argue that complex multi-dimensional social problems require a different approach that involves collaborative policymaking. This policy should involve trying to alter conditions for service users’ rather than imposing strict theories of change upon them.

Moving forward from SIBs
Recognising these deficiencies, we propose a different approach which prioritises service users’ perspectives and which is a first step towards co-determination of policy. A good way forward is to use Q methodology, a research tool used in psychology and in social sciences to study “subjectivity” (namely views, opinions, values and beliefs). It was developed in the last century by the psychologist and physicist, William Stephenson.

This approach is a well-established way to identify and describe viewpoints. First, a card sort (of statements) is used to generate data. Then, there is a form of factor analysis which identifies patterns and similarities between individual card sorts. These methods reveal the structure and form of subjectivity on a particular topic, allowing subjective perspectives to be described.

Q methodology has potential uses in policy development and evaluation because it could help to ascertain the relative importance of problems, the causes of problems, as well as potential solutions and policies. However, it’s just one practical, albeit important, way to actualise service users’ perspectives and so improve policy design and evaluation.

Crucially, this approach is aligned with genuine social innovation whose starting point is that it should empower service users. It should enhance their capacities. That’s missing in SIBs. If social policy is really going to make a difference, then we need to replace SIBs with more participatory approaches to policy development.

Dr Stephen Sinclair is a Reader in Social Policy in the Yunus Centre for Social Business and Health, Glasgow Caledonian University
Dr Neil McHugh is a Researcher at the Yunus Centre for Social Business and Health, Glasgow Caledonian University.
Dr Michael J. Roy is a Senior Lecturer in the Yunus Centre for Social Business and Health, Glasgow Caledonian University