Social Impact Bonds - Approach With Caution (Part Two)
In my previous post, we looked at the financing cost of social impact bonds (SIBs), their potential for fostering innovation and the transaction costs. In this second section we will take a closer look at whether SIB programs could crowd out or overshadow other social programs, and the challenges involved in measurement and proving that real change has taken place.
There is a risk that SIB-funded programs will distract service providers and governments from equally important, but less immediately impactful, programs. This is similar to a trend that occurred in microfinance, where investor hype around micro-credit loans distracted from the equally important but less lucrative micro-savings, micro-insurance and financial literacy services that were also necessary to deliver a positive social impact.
While SIBs are a new and exciting tool, they should not distract from the complex inter-related set of services that are often necessary to deliver positive social impacts.
Measurement and Causation
This brings us to one of the greatest challenges of SIBs: demonstrating true impact, or change, in people’s lives. Third-party verified evidence is needed to show that behavioural change is due to an SIB program and not to random environmental factors, other programs or normal changes in people’s lives. Without this evidence, investors won’t be getting paid and the SIB model is compromised.
Things gets particularly complicated when a population is part of the SIB program but also impacted by several other government programs. If these other programs change during the life of the SIB program, there could be an effect on the recipients.
Proof of causation requires comparing the population receiving the treatment with another group that is on average equal (aka the placebo). This is difficult in the real world and requires thorough measurement, starting with a baseline and progressing over time, controlling for external variables. It also greatly increases the cost of a program, although improving the measurement of social programs is likely often worth the investment. Many NGOs currently do not have the capability to administer this level of measurement. Delivering statistical proof of causation also requires a sufficiently large sample size, requiring a large program.
Finally, what works in Toronto might not work in Chicago. This makes it difficult to predict the impact of a program, which increases risk and therefore cost, and potentially limits the scalability of SIBs.
In general, I am very excited about the trends of improved measurement and a focus on outcomes in the social sector. Continued innovation in measurement and IT will make it easier for smaller organizations to demonstrate impact, and hopefully over time this obstacle to proving change will be overcome.
SIBs are not a silver bullet, and a profitable one is only likely under particular conditions. A financially successful project must be large enough to cover its transaction costs and to statistically demonstrate impact. The impact itself must become evident within a relatively short time frame in order to justify a 7-15% discount rate. A program must not distract from, or result in a lack of funding for, other programs that may not be able to demonstrate impact on the same scale as an SIB program.
Finally, fairly sophisticated measurement techniques are required, implying a certain level of scale and complexity. If a program is already this sophisticated, it should be able to demonstrate its value to society and government should be funding it, especially if government will end up paying more to private investors down the road for the same program.
It remains to be seen whether or not the benefits of leveraging new sources of capital, reduced government risk, encouraging innovation and focusing on outcomes will outweigh the challenges listed above. My feeling is that SIBs will find a sweet spot in the range of financing options for social program — somewhere above philanthropic grants and below fully implemented government programs.
I think it is important to be cautious about the role of private sector finance in the social sector, in relation to SIBs. There is certainly a role for SIBs in social financing, but it is important to keep in mind it is just one tool to be used, in particular situations, if it is to be effective and efficient.
The most informative source for me while researching this post was the Young Foundation’s publication on SIBs, which is also the source for the diagram below outlining a series of factors that should be considered when implementing an SIB.
Note: SocialFinance.ca offers a list of resources relevant to social impact bonds as well.
Photo credit: http://www.flickr.com/photos/alexandredebie/8175955121/