Excellent. Thank you.
Just to provide a quick introduction, I'm an associate professor at the Sauder School of Business and am responsible for running an initiative called ISIS, which focuses on social innovation and social finance in British Columbia in Canada. We act as an incubator and research centre focused on applied research in the domain of social innovation in general.
I was involved in the British Columbia committee looking at social entrepreneurship and social innovation, and social impact bonds were one of the recommendations at least for exploration to that committee. We recently authored papers looking at green financing bonds as well, which is a similar kind of mechanism, as well as the field of impact investing in Canada and North America.
I would like to start by echoing some of my colleague Andy Broderick's not so much concerns, but caveats about the potential for social impact bonds. In the first instance I'd say the focus in much of the early discussion, and rhetoric and enthusiasm for impact bonds I think is on the wrong aspect of the program, whether they are pay for performance or social impact bonds. The real enthusiasm in the early days was about mobilizing private sector capital and philanthropic capital on a performance basis to address intractable social problems.
My view is that in the first instance, we need to focus primarily on the effectiveness of the programs rather than on the ability to mobilize new capital. With that in mind, I'd like to provide a few comments about pay for performance or social impact bonds in the context of the criminal justice system.
As I'm sure as a committee you have already learned, the flagship example, the point of reference is the Peterborough program in the U.K., which focused on 3,000 short-term prisoners serving sentences of less than a year. The very positive result from that program was that the intervention of the St Giles Trust reduced reoffending rates by over 65% and saved something in the region of £4.7 million.
There's demonstration there of very high levels of effectiveness and a quality of service in terms of the rehabilitation of prisoners that went far beyond what was being delivered by conventional social services. There were very positive results in the context of that initiative.
I'd like to highlight some of the reasons that I think the criminal justice system, in theory at least, is well suited to this form of intervention, whether that's pay for performance or some form of bond.
One of them is this is genuinely an intractable problem in almost every developed country. The failure rate of the conventional system is very high, and if the goal is rehabilitation, then the room for success in this particular area is also very high. If the reoffending rate is 75%, even a 25% reduction in reoffending is a dramatic success compared to the status quo.
The second reason I think programs in this area are a good starting point is that the core metrics are relatively easy to track and are unambiguous at the level of that prisoner population. You know pretty clearly when someone has been intercepted again by the criminal justice system, and it's relatively easy to measure that.
The third reason is that unlike health interventions, which is another area where people have spoken about pay for performance because you're investing in prevention, the results can appear relatively quickly since the key period for reoffending is within the first one to two years, so the savings and the proof of concept for government could be relatively immediate.
The fourth reason is there are and historically have been very well-established models of small-scale interventions by non-governmental organizations and community organizations that have proved highly effective at the small scale.
Those are the reasons that I think exploring pay for performance and impact bonds in this area is a positive idea.
I think some of the risks worth noting are that since, by definition, the interventions involve working with convicted criminals, there are reputation risks and other forms of risk associated with the success and failure of the program.
The second reason is that criminal justice in general as a public policy issue is highly polarized, with very strong positions on either side of the debate about rehabilitation versus punishment. What this means is that any initiative launched even on a pilot basis will be closely scrutinized by very vocal organizations.
The third reason is there are in Canada and particularly in the U.S. very strong private sector interests associated with the current system, and in Canada and the U.K., there are public sector interests and union interests associated with maintaining the status quo that may challenge the success of this.
The final two points really relate to the public policy context in general. What I believe is necessary to test social impact bonds or pay for performance is some form of prototyping and bounded experimentation. Historically, government programs have not been tolerant of the kind of experimentation that we allow for and see in the community sector and also in the private sector through the activity of entrepreneurs and social entrepreneurs. I think any attempt to explore this further will have to involve a higher degree of tolerance for experimentation and failure than is typical of government programs.
Finally, since within the criminal justice system there is a disproportionately high number of aboriginal citizens who are intercepted by that system, I think we need to be profoundly sensitive to the historic treatment of first nations by criminal justice and the residential schools programs.
My comments about the likely sources of capital really focus on a reality check about what private investors in particular might look for from this kind of program. In the early stages, I don't believe the private sector would mobilize quickly in response as an intermediary to providing capital to scale this up. My view is that if there is a desire to test this and prototype and pilot these kinds of projects, it will require government funding to underwrite it, with stakeholders from the philanthropic sector and potentially from local governments for it to work.
The reason the private sector, I think, will hold back is that typically for institutional capital investments, they would require investments of $25 million or more to address the transaction costs involved. I don't see an intervention on the scale of $25 million being realistic initially for a prototype or for a pilot study. In contrast, public sector and philanthropic capital is likely to be more appropriate in the early stages. Community organizations and philanthropic institutions having a stake in the investment I think is important because it ensures a level of accountability and engagement with the success of the program.
With those caveats, my view is that to proceed with this on any scale would require a commitment to a design approach and a prototyping and piloting effort. Any program in this area would genuinely have to provide autonomy to the community organizations involved. The expectations with respect to financial return would have to be reduced, with an emphasis instead on the effectiveness of the programs, at least during the early stages. You would have to evaluate partner organizations, such as the St Giles institute, to demonstrate that they have the capacity to scale sufficiently so that it isn't just a one-off example with no capacity to go to scale.
There would have to be a commitment to iterative learning by all partners from the experience, and a recognition that for many community organizations it's a real challenge to shift from a conventional community approach to the delivery of programs, an NGO approach, to the more business style of delivery, one that's metrics driven, that's required for these initiatives. I would argue that it's non-trivial for most community organizations to shift to those kinds of management metrics. There's a risk that the shift to a more managerial approach could undermine the effectiveness of the program.
My view is that one should avoid overstating the potential for these projects to attract private financial capital. In the short run the emphasis should be on effectiveness rather than on financial efficiency. If the federal or provincial governments sought to pursue this, they should recognize it's going to cost more money in the short term because reductions in reoffending rates don't translate into savings unless you're able to reduce the prison infrastructure. It's going to cost more money initially to try this with the goal of reducing the fiscal burden in the longer term. It would require real independence and independent investment to ensure there's sufficient room to try a pilot and experiment with these kinds of initiatives.
Those are my main comments.
