As simple as “three words”? Why caution is needed with a ‘Pay for Success’ approach to ending homelessness.

Market-based solutions to complex social problems can appear tantalising in their simplicity. In this post Lanie Stockman, Outcomes and Evaluation Specialist at Good Shepherd Australia New Zealand, questions how outcomes would be identified and measured if social services were funded based on ‘success’.

 From the heartless calls to ‘clean up’[1] our streets, to the campaign to make visible the real extent of homelessness in Australia by writing ‘none’ in the address section of this week’s census, homelessness is in the public discourse on an almost daily basis. Again.

 So it was interesting to come across a solution proposed on ABC News online that’s different to the usual “short term cash splashes”: Pay for Success. The idea is elegant in its simplicity: services that deliver programs to address homelessness should have their government funding tied to achieving “measurable life outcomes”. With the Productivity Commission’s current inquiry into the marketisation of human services, this idea is perhaps more relevant than ever.

 Innovations such as these are not new ideas and they seem to have had modest success. One example of a very small pilot social impact bond project in the UK to reduce recidivism found an 11% drop in reoffending – from 159 to 141 reconviction events annually for 100 people released from prison. This pilot was discontinued in favour of “wider reforms”. Still in the UK, there seems to be interest in learning from outcome based funding models. Research looking at how lessons of Australian incentivised employment services can be applied to the British context pointed out that while the model was operating successfully there was “concern about the quality of some service delivery and that many participants [who require intensive assistance] were receiving only limited help”.

 In developing countries, initiatives to improve access to sanitation, for instance, have trialled approaches where organisations receive payment from funders for installing quality latrines in rural areas – but sustained use of these latrines is yet to be comprehensively realised. In other words, this is an approach that clearly supports provision of outputs, but to what extent are sustainable positive changes in people’s lives achieved?

 The critical question here is – if social service delivery depends on results to receive funding, how are meaningful and sustained changes to people’s lives defined and evidenced?

 As discussed in other blogs on this website, evidencing outcomes is highly political. How do we decide what a meaningful outcome is? Who makes that decision? Research into measuring outcomes in the homelessness sector notes that because of the “multiple and complex difficulties facing homeless people, what may appear to be very small changes can, for some, still be highly significant". At Good Shepherd’s Youth Homelessness Service, it is not realistic for us to place every young person who comes to see us into sustainable housing – simply because there is not enough of it to go around. We do aim to support every young person who participates in the program to acquire improved living skills. This is so they are able access a broader range of housing options, such as the highly competitive private rental market, and are resilient to changes in personal circumstances that may negatively impact their living arrangements. Would supporting improvements to ‘small’ things in people’s lives such as life skills equate to success?

 To take an example from another (but related) sector, at Good Shepherd we’re currently evaluating the Waranara Centre – a program which aims to re-engage young people in their education so they can achieve their high school qualifications. Our preliminary results suggest that achievement of desirable outcomes is highly dependent on a young person’s age, previous experiences of school and personal situations – not solely their experience at the service that we provide. This is because most young people at the Centre experience multiple, complex challenges that affect their attendance and success at school. To illustrate a fairly common journey, one of our students made a promising start to Year 10 last year and was attending school regularly. However, by the second semester, his attendance declined very quickly and almost completely, coinciding with his family breaking apart and a subsequent negative impact on his mental health. With support from our service, this young person is now re-attempting Year 10 and progressing well. If we were funded to deliver this program according to a ‘Pay for Success’ model would we be paid on evidence of the student’s initial regular attendance? Or only when he achieves a high school qualification, which may be years in the offing to account for his personal journey? In other words, how does a ‘Pay for Success’ model operate when progress isn’t linear and results fluctuate?

 Proponents of ‘Pay for Success’ approaches would argue that accountability of service providers is critical – especially to those they seek to service. They may also assume that service providers do not evaluate their work to continuously improve services to best address the needs of service users. Yet social services have been monitoring and evaluating their practice to various extents for decades – often on shoestrings.

 Among the many potential pitfalls of ‘Pay for Success’ is that funding pressures would entrench skewed accountability to funders rather than service users. It would also put an end to efforts to genuinely and overtly learn from practice, especially when things don’t go to plan. Perhaps this is an approach with promise – and we should certainly be open to whatever it takes to end such a significant social issue as homelessness – but it is not as simple as “three words”. Let’s tread with caution.

[1] There are many recent examples including http://www.heraldsun.com.au/news/homeless-return-to-melbournes-elizabeth-street-despite-police-crackdown/news-story/33396870c760168dc1ac6a09f6f8f6c2