How much time is enough… but not too much?
“(…) the definition of outcome metrics are extremely important for a contract such as a Social Impact Bond. For the service provider, the outcome metrics are the motivation to be socially efficient (…), for the commissioner, they guarantee a social issue is taken care of in an efficient way, for the investor, these outcome metrics have to be set in a way to optimize both the financial return and the social impact of their investment.”
European recidivism rates vary between 60-70% (UNODC Handbook, 2012), which means that in every 10 convicts, 6 to 7 commit another crime after the end of their previous sentence. Although the exact reoffending rate is unknown in Portugal, by considering Europe as a benchmark, it is fair to assume that the Portuguese rate is at least higher than 50%.
The Portuguese prisons are overcrowded, with an occupancy rate of 120% on average (Aebi, M.F. & Delgrande, N., 2015), and with a daily state’s cost per inmate of approximately 40€ (Dores, A.P. & al., 2013).
In general, while the current system does not seem to be able to make offenders desist from crime, the costs for both the government and society are considerably high, ranging from the financial costs, the insecurity and the psychological harm, to name just a few. Recidivism is a social issue recently studied within the Social Impact Bond (SIB) Research Programme, which incubated a feasibility study whose challenges will be deepened in the present article.
The tough reinsertion in the community upon release is seen as one of the main causes for recidivism (Griffiths, C. T. & al., 2007). The stigmatisation of prisoners makes it challenging for them to build a new life upon release and reach stability by, for example, finding a job or an apartment. However, the lack of offenders’ means to face the competition on the job market or more generally in the community is just as demanding. Ex-offenders may face difficulties regarding their level of education, their professional experience, their past’s aftereffects or current self-image. APAC Portugal, the intervention the feasibility study focused on, is intended to correct this issue and help inmates prepare for what is waiting for them at the end of their sentence, when they get their freedom back. The programme is divided in different phases, each with the goal to tackle this issue, through education, workshops, creativity and work. The first kind of programme of this type was launched in Brazil in the eighties and inspired dozens of countries since then. The recidivism rate of the Brazilian counterpart approximates 8.4% (IBJ, 2009).
In the public sector’s perspective, Social Impact Bonds make it possible to pay only for successful social interventions and enable innovation while shifting financial risk to a private investor. Moreover, if these social services result in successful enterprises in the long run, not only will some public sector’s social costs be alleviated through the intervention (less prisoners, less crime, successful reinsertion which again brings fewer social assistance costs) but also could the social issue be partly resolved – or, at least, some people may enjoy a better life (less crime, more safety).
For the aforementioned reasons, the definition of outcome metrics are extremely important for a contract such as a Social Impact Bond. For the service provider, the outcome metrics are the motivation to be socially efficient in order to further get funding for the project. For the commissioner, they guarantee a social issue is taken care of in an efficient way. As for the investor, these outcome metrics have to be set in a way to optimize both the financial return and the social impact of their investment.
Challenges when constructing a SIB, especially in the area of recidivism, is to time the outcome metrics and payment right. Recidivism is, per se, easily measurable, by looking at the recidivism rate of the project. But would that be a fair outcome metric for a programme intended to help offenders reinsert into the community? If a person is financially independent, she can take care of herself and her family and is presumably partially socially active. For this reason, employment was chosen to be the second relevant outcome metric. For both metrics, several points in time were defined in order to reduce the initial investment capital. Indeed, the sooner the prospective payments flow, the smaller is the initial capital requirement to launch the project.
The timing to measure the outcome metrics is key, however difficult to set. As mentioned before, the outcome metrics should be a motivation for the service provider to reach social enhancement. The sooner the better, since the payment is needed to continue the operations and repay investors. The state wants the outcome metrics to be relevant and proportionally representative of the work done and the social impact of the programme. Measuring recidivism after two weeks clearly does not make sense, but measuring it after ten years, even if relevant and interesting, is too long for the service provider to take. The investors get a better internal rate of return (IRR) with money flowing sooner but are also investing in an SIB to have a social impact. Investors in SIBs are ready to give away a little bit of their financial return if it allows the project to have more social impact. This trade-off is complicated to solve and the ruling can be financially and/or socially tough for the various parties.
Finally, the following outcome metrics were chosen, with the following timing: recidivism is suggested to be 18 months upon release. Indeed, many studies say that most of the recidivism cases occur during the first year upon release (National Institute of Justice). The service provider would get a predefined cohort payment from the outcome payer if only 20% or fewer offenders were to reoffend. As for employment, the service provider would get an individual payment per ex-offender entering the professional market and another payment if the ex-offender maintains his job for at least 6 months. The metrics of employment include the non-reoffending clause, since an active person cannot be in prison.
The main reason for the aforementioned outcome metrics and timing is the belief that, after 18 months of freedom, a working person is decently stable and has enough to lose to commit a crime again. The arbitrary decision was driven by the 3-sided trade-off mentioned above.
The meticulousness with which the metrics and their timing had to be reflected on was strongly influenced by the effect and the reach the end decision would have. First, the project could be seen as viable although it is not, which would consume its social relevance and could bias further projects. However, if the service provider does not receive deserved funds, the potential of the project might be wiped out.
This project could be scalable to any other region. If its pilot is successful, it implies that an effective and efficient mean to give a better life to communities was found. The reach of such a statement is as big as the riskiness to influence its success.
Milena integrated the second edition of the SIB Research Programme and developed a feasibility study for a Social Impact Bond for the methodology of APAC.