A fifth area has to do with SIBs and program evaluation. The thing around program evaluation with SIBs is that it tends to be quite complex. This involves a lot of costs if it's going to be done right. In addition, there's often no guarantee that the results will be accurate.
Providing a meaningful evidence-based approach to results requires social experimentation methods in which at least 200 participants per year are randomly selected and are matched with an identical control group in order to try to determine what your actual results are and that they're accurate. SIBs required detailed population data on subject groups and control groups. Again, this is quite costly.
It's also difficult to know often whether the results that are attributed to the SIB approach itself are a result of what the SIB has done, whether it's the unique combination of the service interventions that are unique to the SIB, whether it's about more resources that are targeted to that particular service area, or whether it's due to another policy change that has occurred randomly in conjunction with the project.
It's interesting that, in the case of Peterborough Prison in the U.K., it turned out that the sample chosen there was not a random sample; it was actually volunteers. That had the effect of biasing the sample, so that one would expect more positive results from the way the sample was selected. In fact in the mid-term report, in terms of outcomes, what they found was that there had been very modest positive results, but this may have been due to sample bias, in large measure. Interestingly, in another case in the U.K., with a similar program around recidivism in the prison system, they actually came up with negative results. But that was also probably the result of other sorts of factors, because there was a tough-on-crime policy that was implemented around the same time in the U.K., which probably influenced the results of those outcomes. It's hard to sort of decipher what's going on with the results, which makes interpretation of the evaluation often very difficult.
The other thing I think that's worth noting here is that in the case of Peterborough, U.K., which is the first SIB project, as Meghan mentioned, it was cancelled and they moved towards a full sort of privatization model. There was never a full result in the case of that original SIB project, which is unfortunate.
A sixth area of concern is that SIB projects are risky for participating non-service providers to some degree as well. Non-profits do not have exclusive control over results, and a poor outcome could cause them the loss of their reputation. The reputation is the prime asset that a non-profit organization has, and that could cause them considerable problems if the project goes wrong for reasons out of their control. It could actually result in their going out of business and the loss of jobs and the loss of essential services they're providing to the communities they serve.
Non-profits are also at a disadvantage compared to private-sector competitors when bidding for an SIB contract, because they have limited access to capital and lack of financial track records of these kinds, because of some of the rules around their financing. The SIB tool is also quite biased towards the largest non-profits, as well as private social service providers. So probably, in the case of the non-profit sector, only the biggest non-profits could be involved with these. This, in part, is problematic because small non-profits are often the most innovative service providers because of their particularly close relationship to client groups, and especially vulnerable client groups in particular.
SIBs further distance the relationship between government and non-profit organizations because the relationship is now with the SIB intermediary. That limits the opportunity for non-profits to inform public policy based on their own grounded expertise. Some of that disconnect that might operate between SIBs and their connection to government and how it may inform policy and programming is another consideration one would want to take a look at.
In terms of concluding thoughts, SIBs are no panacea and should not be viewed as simply a replacement for public service provisions and grant-based program funding to the non-profit sector. Also, while there are potential advantages to SIBs, including a new emphasis on supporting non-profits design, quality of life outcome measures with their clients, as well as longer-term payments for non-profit service provisions that reduce some of the burden of the reporting requirements, these types of changes could also proceed without the transition to a SIB model. Some of the recommendations that were made by the blue ribbon commission on grants and contributions in 2006 point the way to some reforms here for the non-profit sector that don't involve a SIB model.
The non-profit sector has traditionally been run on principles and values of care that have served Canadians very well overall, and SIBs risk placing further pressure on the non-profit sector to market this type of provision, which as we've suggested has some dangers to it that one would want to be rather cautious with.
As a final comment, this is definitely a case of proceed with caution and let the evidence come in and see what the real costs and benefits of SIBs will be before jumping into it wholesale.
Thank you very much.