Last time out we learned how social impact bonds were initiated in the UK. We saw that they are not simply bonds (debt instruments with fixed interest payments and a maturity date). They are a set of complex contracts where payments are contingent on achieving measurable social outcomes.
The particular SIB we want to look at here is the fourteenth one set up in the UK. It was created in 2012 to fund a homelessness program at St Mungo’s, a well-established English charity.
Setting up the St Mungo’s SIB
The St Mungo’s SIB was a joint project of several government departments, led by the Cabinet Office. They came up with £5 million for the project after lobbying efforts by Social Finance Ltd., a consulting company that seeks, as its name implies, to bring together the social and finance sectors. Social Finance and another social innovation organization, the Young Foundation, were commissioned by the government to develop the business case for a SIB on homelessness.
The business case they wrote candidly stated, “We have been unable to identify a robust evidence base linking interventions with realistic expectations of outcome improvements” (page 15). Undaunted they laid out three potential models of intervention. The government eventually selected the most expensive of the three. The winning model involved assigning each homeless person to a personal “navigator.”
The Navigator Model
The navigator model was designed so that a key social worker – the navigator – would support each homeless client with an individualized assessment, help him or her find a way through the network of provision, and sustain this support over time. The navigator would be given a budget to support a personalized intervention in the homeless client’s life. Navigators would try to achieve the outcomes specified in the SIB.
In our interviews, people at St Mungo’s said the navigator model turned out to be impossible to operationalize, for financial and practical reasons. The homeless people selected for the program (more on this below) slept on the streets of central London. Most of the hostels or shelters they could move to were at least an hour away, since property prices make building shelters in central London prohibitively expensive. Navigators would be unable to work with people on the streets of central London and also visit the people who had been moved to accommodation on the outskirts of the city. The navigators would simply have spent too much time travelling.
Despite this problem, the navigator model became a key feature of the business case because it created a highly entrepreneurial approach to service delivery. Each navigator was considered a small enterprise with flexibility (within the constraints of their personal budget) on how to achieve the specified outcomes.
The implicit assumption of the navigator model is that “solving” social problems is best done at the level of the individual. By design, the navigator model cannot address any systemic causes of homelessness. It assumes that a person is homeless because of the choices he or she has made, and attempts to help them make better choices.
This is entirely consistent with the economic theories underlying neoliberal government policies. These theories assume, for instance, that unemployment is a rational choice on the part of workers to work fewer hours during a recession in response to offers of lower wages. This insight probably comes as a surprise to anyone who has actually been laid off. If you would like to learn more about the economics behind SIBs, please read The Profit Doctrine (Chernomas & Hudson, 2017).
Choosing the Targets
Once the decision was made to use individual entrepreneurs to deliver the services, the next step was to assign them specific homeless people to target. The city government had a fairly sophisticated database of the homeless population in central London. This was the “Combined Homelessness and Information Network,” or CHAIN.
The city used CHAIN to identify specific homeless individuals and establish baseline measurements on their condition and behaviour. Without this baseline, the SIB contract would essentially be unworkable, because payments to the service provider would depend on changes in these measurements.
There are three basic categories of homelessness in the CHAIN database:
those who have just become homeless, many of whom quickly return to stable housing
those who are chronically homeless, who face multiple problems such as mental illness or drug dependency, and
a middle group, referred to as “Inbetweeners” in reference to a British TV show, who have had multiple separate episodes of homelessness but who have not yet been labelled as chronic.
The SIB would focus on the Inbetweeners. Every one of the 830 Inbetweeners in CHAIN was included. This prevented the delivery agency from gaming the system by selecting the “best” candidates. It still biased the experiment, however, because out of all the homeless people, this cohort was the most amenable to intervention and would therefore provide the best results for the SIB. Remember, this project was not just about homelessness. It was about whether SIBs work.
Together, the CHAIN database and the navigator model became the key elements of the initial business case for the London Homeless SIB. Detailed analysis was then performed on three interrelated economic questions:
What outcomes could be achieved with the Inbetweeners, measured in terms of particular aspects of their homelessness, such as housing and health?
Given these potential outcomes, what types of payments could be offered to the delivery agency as incentives to reach desired targets?
Given these payments, what cost savings, if any, could accrue to government, compared to traditional ways of funding homelessness programs?
The initial business case took a rationalist, deductive approach based on implicit ideals about healthy, productive individuals. The St Mungo’s staff members we interviewed said the initial business case contained the “kind of ideas that a non-expert on homelessness would come up with.”
In response, St Mungo’s staff took an inductive approach based on their street-level experience with homelessness. The staff basically worked out what they could achieve with the money that might eventually be paid at the end of the SIB, and attempted to arrive at performance metrics that were achievable with the money available.
Only after negotiation did the target outcomes become realistic and achievable. In fact, the final performance metrics seem to have been chosen to ensure that the SIB would be profitable.
In the next part of this series, we will look at these metrics.
All photos taken off the south coast of Turkey in 2012. No navigators were harmed in the taking of these pictures.
Chernomas, R., & Hudson, I. (2017). The profit doctrine: Economists of the neoliberal era. London: Pluto Press.