Impact Investing

Debunking the Myths Behind Social Impact Bond Speculation

An examination of SIBs recently conducted in Maryland should give pause to governments and nonprofits looking to take the leap.

Anytime Goldman Sachs’ is speculating on investments linked to broader societal outcomes, we should probably take notice. As NPR noted recently, New York City has signed a Social Impact Bond (SIB) deal that would allow Goldman “to profit by helping troubled teens.” Ironically enough, other jurisdictions across the world are looking at replicating the financial creativity in New York City to solve the budget problems precipitated by pre-2008 financial innovation. A closer examination recently conducted in Maryland should give pause to other governments and nonprofits looking to take the leap.

Originally conceived as part of the UK’s “Big Society” agenda, SIBs are a new type of performance-based contract. Their innovation is to include third-party investors who provide money upfront to fund the operations of a social service program. If targeted outcomes are achieved, the government agency holding the contract pays for the program by reimbursing investors. If the program does not meet its targets, the investors lose their entire investment. This, in theory, shields service providers and governments from performance and outcome risks.

How a Social Impact Bond for a criminal justice reentry program works.

The fact that investors are providing upfront capital has led many to believe that the government’s upfront budget obligation will also decrease. Proponents note that in the SIB model, the government will only have to “pay for success.” Even better, success saves the government money because the model changes the focus of programming to outcomes and their measurement, stimulating innovation for programs directed at high-cost individuals.

After conducting research for the Maryland General Assembly, I found, however, that this last paragraph is built largely on myths:

  1. Myth #1: New capital for social programs. This myth ignores the fact that if successful, the government will have to pay for the program. In addition to the fiscal imprudence of accruing debt for social services, states are generally restricted from creating liabilities in their operating budgets without providing matching funds. Thus, it should be no surprise that Massachusetts, like the UK national government, is pre-funding its pilot SIB program. In New York City, Bloomberg’s personal foundation is providing a grant on behalf of the government to cover the payments.

    Given the costs of attorneys, consultants, program evaluators, the potential for a return on investment to third-parties, and a second tier of program managers, using an SIB relative to direct financing will therefore increase pressure on the budget, as the government must set aside more funds than even the investors provide to the program.

  2. Myth #2: The programs will save the government money. An independent evaluation by RAND Europe of the first SIB pilot program in Peterborough, UK, found that the prison reentry program “is too small to deliver substantial ‘cashable’ savings” for the government. My analysis found that even for a relatively large reentry pilot program in Maryland, a 10 percent reduction of reimprisonment for treated individuals would at best produce a 6 percent discount to the cost of operating the pilot program—using highly optimistic assumptions. If the costs do not stack up in criminal justice, it seems unlikely they would materialize in other types of programs.

  3. Myth #3: The government pays only for success. This assumes that governments and nonprofits will actually be able to enforce the byzantine contracts in the event the program does not meet its targets. As with every attempted SIB implementation, which have uniformly noted the complexity and difficulty of designing a contract, RAND Europe found that in Peterborough, “complexity in some instances meant that the actual transfer of risk is not clear.” 

    This complexity is inherent to the model. Attempting to manage social service through contract attorneys, consultants, program evaluators, and an all-or-nothing payment model will inevitably produce a contract that is extremely complex, and therefore more likely weak. As the United Kingdom’s prior fiasco with the London Underground demonstrated, financial creativity does not prevent the government from holding the bag on failed projects, even when the risk appears to be on the private sector.

  4. Myth # 4: A focus on outcomes will encourage innovation in programs. Given the high-stakes nature of the outcome payment, investors will be more likely to select programs with a proven record and with evaluation techniques that maximize the chances of demonstrating a positive outcome. In the case of Peterborough, the project skimmed providers and the prison location based on unique locational advantages and a long-standing record of provider accomplishment. Even more problematic, the evaluation technique cannot control for (or estimate) these advantages. Thus, in Peterborough, the government is likely to pay a risk premium to investors for a program that is already known to work but may not be replicable or scalable.

Although the benefits commonly associated with SIBs are undoubtedly appealing to cash-strapped governments and nonprofits still recovering from the Great Recession, without an understanding of how social impact bonds actually work, these endeavors in financial creativity may become expensive experiments that leave governments with the ultimate risk and providers with broken or contested contracts. Before leaping, nonprofits and governments should evaluate SIBs as a public private partnership, with an eye toward their own financial and operational risks relative to the efficiencies of direct government financing.

Tracker Pixel for Entry


  •'s avatar

    BY, SEL Alliance for Massachusetts (SAM)

    ON April 9, 2013 10:12 AM

    Dear Mr. McKay:

    Your analysis of SIBs, based on your research, calls into question some basic presumptions about the place of government in the lives of its citizens. 

    With respect, when you begin to debunk a myth with the parenthetical phrase, “In addition to the fiscal imprudence of accruing debt for social services….”  I must object before I finish reading your sentence. 

    I think there are many of us, seniors like me, those who are disabled, even in children who need a good education or a meal to fill their empty stomachs, don’t find it imprudent accruing debt to serve those without the means to care for themselves. 

    This idea, that government should care less about its citizens and more about its finances has been argued since the time of the Founders.  You have taken the Jefforsonian view (although when President he was as much a Federalist as Alexander Hamilton) and I hold to the Washingtonian view, as did Franklin Roosevelt, that government’s responsibilites lie in providing for the underprivileged.

    It’s a fair discourse, but your clear desires to NOT fund any social program as “imnprudent” on their face makes the rest of the article rather flimsy on evidence.  I feel that it would require an article of equal length to refute many of your statements.

    But for those of you who are interested in financial side of government, I funded a study through my nonprofit called the “Economic Burden of Crime and Substance Abuse in Massachusetts” by Dr. Clive Belfield of Columbia who found that 22% of the state budget was spent on the consequences of crime and substance abuse with less than 1 % of prevention.  These social issues are not going away.  Out of the 2008 budget, $6.6 billion was spent on these two issues alone.  Shall we make no attempt to lessen this expenditure or just say “no” to new ideas, yet proven and go on as before, seeing the costs rise each year?  Isn’t that “imprudent?”

    Does Mr. McKay want to simply banish these individuals who eat up so much of our state funds?  Shall we house them in prisons as we do now, ad infinitum, which only costs us more money?  Should we “just say no” and be done with it? Shall we allow the criminals and the addicts walk around the streets without help or hope or rehabilitation?

    The SEL Alliance for Massachusetts ( is engaged in implementing a long term educational plan that will reduce the cost to government AND also relieve the misery of these social and public health maladies . 

    SIBs offer that opportunity of proving that we can, in the long term actually reduce these twin evils that so effect our society.  Foundations and social investors can view programs before they invest.  They can talk to the children effected.  They can talk to the teachers who believe that when we teach children how to think for their own mental health.  In other words, they can due their due dillegence and then pick out specific data collection points to alter through their program.  Of course, it’s messy.  You have to work for success.

    Mr. McKay, rather than debunk, do an analysis that does what my father taught me to do - you don’t have to go to school to do it.  Draw a line down the center of the paper and label once side PRO and other side CON.  In this case, the PROS outweight the CONs.  Doing nothing is never the answer.  Not spending short money for long term gain is an imprudent fiscal policy.

  • Dear Mr. McKay,
    I don’t agree with you. There are some questions about SIB that you don’t point out in your article and they are important.

    First thing: SIB are usually focus on “prevention” of specific social problems. These social problems already have a cost for the government and normally the action from the Government to tackle the problem is focused on mitigation of the consequences, leaving the prevention work normally to the NGO’s.

    Second: SIB are a new form of looking for funds for the NGOs that work in social services in a context where governments (local, national…) have serious cash-flow difficulties. Therefore, is better for anyone (even gov) to get more time to pay for something than not paying at all. (Usually those NGO arrive where the social services don`t, and that’s why those “social providers” are given grants although the money not always arrives). So the solution is allowing private funds to participate and giving more time to gov. to plan their future cash-flow and debt.

    Third: the difficulties you said in contracts… Well show me just one contract with at least 3 parties and involving money that is not complex or clear 100% (that’s why lawyers and courts have so much work!) 

    Fourth: there is really a saving in the public finances, if the SIB is well designed. Meaning that it has to be a real saving in the total cost of the SIB, taking into account how is the actual cost for the gov to tackle that problem and how much is the SIB program saving to the gov for avoiding this problem. (For example, in Peterborough the cost for the gov of keeping and mantaining people in jail versus the cost of the SIB program (avoiding reoccurrence) + interests;  has to be smaller).

    Fifth: the percentage of return for private investor depends on the risks they take. The more risk you take (with new and innovative programs) the more benefit you get. In this case, being conservative is not the best business.

    Sixth: SIB are mid term programs, and they can tackle problems that are not in the agenda of politics or don’t have short-term results.

  • BY Joseph Bute

    ON April 11, 2013 10:40 AM

    I think Mr. McKay misses a key part of the SIB concept - namely that government incurs all sorts of costs associated with a social problem like recidivism.  Let’s take another example, gang violence reduction.  If an NGO operates a program that can effectively reduce the number of incidents in a neighborhood the current model of funding will almost inevitably lead to a reduction in funding support with what is already regarding as a non-core program.  By that I mean, we take the cost for police, fire, the judicial system as “core” where intervention programs are usually discretionary and time-sensitive.  But that apporach under-funds the very programs that could reduce the cost to the government for these services.  A typical shooting incident in a city involves tens of thousands of public dollars - from police response, to crime scene investigators to emergency rooms to coroner’s office to the state’s attorney plus the indirect costs of the role violence plays in lowering property values, discouraging investments, reducing social activity in a community.  The list goes on.  So instead of a one to one relation with the intervention, the multiple of public expense in all forms is more like 10x or 20x for each successful intervention.  Here is the genius in the SIB concept.  If we started adding in ALL of our costs associated with a social problem then we should acknowledge that programs that change that dynamic - even one incident at a time or one block at a time - prsent the possibility of freeing resources that are otherwise misdirected to these incidents.  McKay’s assessment is far too one to one, cause effect.  I think the SIB is pointing directionally in the right way - much more to learn through this model but the potential is enormous.

  • Kyle McKay's avatar

    BY Kyle McKay

    ON April 11, 2013 12:49 PM

    Joseph, Andrea, and SAM,

    I agree with all of you that there are in fact social problems that would save the government money if the problems were properly addressed. However, there are key questions to ask:

    1) Should we focus programs only on interventions that produce savings to the government?

    2) Are social impact bonds the best way to achieve cost savings?

    3) What role should larger policy changes have in achieving the desired outcome?

    In answering these questions, a comparison needs to be made to direct government financing through direct governmental service provision or contracting.

    Looking at criminal justice, for example, a privately administered reentry program delivered post-exit represents a very narrow way of reducing recidivism and addressing social problems associated with criminal justice systems. As I discuss in the full report to the Maryland General Assembly (, other options include policy changes in sentencing and employment records.

    Creating an investment market for service programs may hinder these larger policy changes, as public agencies will spend enormous amounts of time trying to design contracts, collect data, and implement programs with private service providers (i.e. there is an opportunity cost for programs). Additionally, in order to prove that a specific program caused an outcome, research designs must be carefully done under controlled circumstances. Making large policy changes in the middle of an intervention under research will make it very difficult, if not impossible, to assess the impact of the program under study. 

    If we know that a type of program works, then the cheapest way to fund it will be through direct government financing (otherwise you would unnecessarily end up paying a premium to investors). If we aren’t sure if a program works, then we need to conduct research. Including investors who are trying to make a return during the research process, as I note above, is likely to distort the research process.

    Independent of the research value that a pilot program can yield with a high stakes payment scheme, the scale of pilot programs are seldom large enough to produce cost savings, as RAND Europe has shown with Peterborough and we demonstrated in Maryland using detailed fiscal analysis. This is obvious if you skip the headlines and look at marginal cost savings, which are much smaller than average costs.

    In short, the discussion on whether or not to pursue a given program should be a separate one from how to finance and deliver the service. Because governments should responsibly budget for contingent liabilities, social impact bonds are more of a contracting tool then a panacea to underfunded programs—especially when you consider the need to build contract design, evaluation techniques, and a return on investment into the program budget.

    Hopefully that clarifies a few points for everyone.

    ~Kyle McKay

  • Bravo Mr. McKay. You presented very valid points and commend you on your clear thinking on this matter. We sometimes don’t like the facts if they are inconvenient but they are none the less true.

  • BY Emma Tomkinson

    ON April 27, 2013 01:36 AM

    Hi Kyle - great post and I echo Doug in commending you on your clarity of thought. ! I think the important message you help convey is that social impact bonds are used by different jurisdictions to achieve different objectives, but those objectives are really hard to realise. What I love about social impact bonds is not that they’re a panacea for delivery of innovative preventative services in the face of declining budgets, but that they force governments, service providers and investors to sit down and rethink how to best deliver change for those most at need. They’re a tool for change, and I hope that all those who embark on the long and difficult process of developing a SIB that they come out with better organisations, better services and better measurement systems on the other side - so far this seems to have happened. But you’re right, as they seek to make this tool suit the mix of stakeholders around the table, they may come up with something that doesn’t deliver the potential of the abstract objectives they began with (or used to sell the idea to the electorate).

  • BY O.T.C.

    ON May 3, 2013 04:52 AM

    Thank you Kyle for this thoughtful and well-researched article.
    I am wondering if you could say more about the private capital side. As this is an investment with expected returns, what are the sources of these returns? Do they solely consist in capital transfers from municipal budgets? Or do philanthropic contributions (such as Bloomberg Philanthropies in the New York City example) also compose part of the private investment returns?

  • Thank you Kyle for presenting a reasonably thoughtful piece, challenging the notion of SIBs as a panacea to stretched municipal budgets. But you miss a critical element in your SSIR article, that you subsequently address in your clarification post, and which needs to be shouted from the rooftops: SIB/PFS is not a funding mechanism, it is a financing mechanism. The decision as to whether or not it is a good idea to layer additional costs on strained municipal budgets for proven effective programs is one discussion. How to finance those programs is another. To me, the best use of the structure is to offer costs savings to contractually obligated state expenses, not to “prove” that some new program will work… with the result that more state spending becomes mandatory.

    As with so many things in finance and politics, what may be a powerful tool with a narrow application is being sold as a general solution to post-financial-crisis budget woes. And after what are likely to be a few high-visibility failures, the baby may be thrown out with the bath water.

  • BY Jeff Mowatt

    ON November 11, 2014 01:47 AM

    Kyle, Thank you for lifting the lid, which proves impossible in Ukr media due to the enormity of investment in media promotion.

    There’s an old joke in the which asks ‘what is a management consultant’ to which this reply is given. “That’s someone who walks into your office and mugs you, leaving a report recommending that you’re mugged by thre of his closest friends”. As you point out, this comes at a considerable  

    In our efforts to leverage investment for social innovation prior to the existence of SIBs, we had argued the case for a strategy for overall cost reduction to the state. Our primary focus being childcare reform in Ukraine, to remove childen from institutions and place them in loving family homes.

    Where a Social Impact Bond approach is applied to this cause and it is being trialled in several UK local authorities, it will obviously lead to cherry picking for the greatest reward and lowest risk projects in order to meet investor expectations.

    .  .

  • Of course, “cherry picking for the greatest reward and lowest risk” in the context of most SIB service delivery models would mean that the most at-risk cohort is selected, as doing so will offer the greatest delta between the control group and the population receiving services. Which isn’t all bad….?

  • Jeff Mowatt's avatar

    BY Jeff Mowatt

    ON November 11, 2014 07:49 AM

    If that was the case, “impact investors” would be falling over themsleves to support us.

    “There is no substitute for a loving family environment for growing children. Existing state care institutions do not and cannot possibly provide this – despite occasional, lingering claims that state care is the best care for children. This attitude is a holdover from Soviet times when the state was idealized as the best possible caretaker for all, including children. Stark reality does not support that notion.

    While this section has strong focus on financial aspects for reforming childcare in Ukraine, these are just financial numbers to demonstrate that this can be done for an overall, long-term cost reduction to state budget. That is to say, simply, this reform program is at the least financially feasible. The barrier between old and new is the cost of the transitional phase.

    However, it is essential to not get lost in financial numbers and budgets. These are only important to show how this will work and will end up costing less money as the new program is fleshed out and the old program is closed. Most important is the welfare of each of these children. There are at this time, for example, numerous institutions across Ukraine where children die on a daily basis from little more than lack of knowledge about how to help them. The actual cost of helping them immediately is nothing more than one-day workshops for existing staff, to demonstrate basic, simple medical interventions common in the West. These institutions are generally closed to the outside world, difficult to access due to imposed secrecy, and are mostly in very rural areas where even the closest neighbors have no idea of the reality of these facilities.”

  • BY John Rogers

    ON December 8, 2014 03:33 AM

    SIBs are based on market logic and the assumption that there is only one tool in the bag - the dollar or any other national currency - to do the job of mobilising resources.

    The global time banking movement has shown how the simple incentive of a reward for service to the community - a time credit - can have many of the prevention effects claimed for SIBs.

    If it takes a village to raise a child, it takes a community to prevent crime and other antisocial problems. Time banks have shown how to prevent reoffending by engaging the community. Check out the Fair Shares project in Gloucestershire, UK, that ran from 2004-2013:

    ...or Edgar Cahn’s Time Dollar Youth Court in Washington D.C.

    “Youth that participate in TDYC and successfully complete the program have a 5% recidivism rate after six months and a 9% recidivism rate after one year”

    “Since January 2003, all youth who were diverted to TDYC, whether successful or unsuccessful, only have an11% re-arrest rate within one year, as compared to the
    recidivism rate of the District of Columbia, which was 25% in 2007 “

    “According to a study done by the Urban Institute, affiliated with the U.S. Department of Justice, taxpayers save $9,200 for each case diverted to a youth court

    “TDYC operates at a cost of $511.00 per child, saving the District over $8,900,000
    “TDYC saved the District $5,700,000 between January 2009– June 2009 when over
    600 youths were referred to TDYC instead of the juvenile justice system while op
    erating with a staff of 8 and the support of volunteers”
    “On August 17, 2009, the District of Columbia’s Justice Grants Administration denied TDYC the core funding of $450,000 for fiscal year 2010”

    Not a Goldman Sachs dollar in sight and savings of nearly nine million dollars to the taxpayer and yet the local funders cut the core funding of 450,000 dollars. Excuse me, but what kind of math is that?

Leave a Comment


Please enter the word you see in the image below:


SSIR reserves the right to remove comments it deems offensive or inappropriate.