This is the second part of a two-part post on social impact bonds. Please see Part 1 for a description of how they work as well as the benefits ascribed to them by supporters such as New York Times writer David Leonhardt and Harvard Professor Jeffrey Liebman.
Liebman, Leonhardt and others need to move beyond sloganeering to learn how to increase the odds that various market disciplines can combine with organizational, network and personal disciplines to promote effective performance instead of cancerous performance. Their all-too glossy, air brushed pin up pictures of ‘market discipline’ jeopardize instead of advance the odds of success for social impact bonds. The conventional wisdom spouted in their writings actually increase the odds of ‘wrong result/wrong way’ for social impact bonds.
For example, Liebman expresses concern that non-profits benefiting from social impact bond investment might unduly manipulate promised performance results by cleverly figuring out how to pick and choose among the people they are trying to help. Yes, it’s fair to worry about whether and how organizations might manipulate markets (again, though, let’s please note how regulatory and cognitive capture combined with ‘market discipline’ and ‘performance’ in the financial sector over the past decade. Now that’s world-class manipulation!)
But, it’s worrisome that an expert like Liebman doesn’t appear to grasp that the same ‘market discipline’ he’s praising includes the disciplines of customer and market segmentation. Private sector organizations pick and choose the customers they wish to serve versus those they won’t. Tiffany’s and Zales quite intentionally do not serve the same customers. And these choices are key to success as competitors subject to ‘market discipline’. Why, then, would we construct social impact bonds to prevent social service agencies from practicing this same market discipline of segmentation?
Liebman, though, worries about social service groups ‘creaming’ the easiest-to-serve customers. But this means his notion of ‘market discipline’ carries castor oil and spinach demands for social service groups that are not required of private sector players. Yes, if we are serious about doing our best to solve serious social problems, then we should embrace the ethical obligation to respond to the full spectrum of customers and beneficiaries. And, if there are government resources involved, then at least in the US there are (or at least used to be) legal requirements to seek ways of benefiting all and not just some. But no ‘market discipline’ of which I’m aware demands that the all competitors serve all segments all the time.
Liebman goes on to celebrate how ‘market discipline’ will permit private sector investors to choose social service agencies without any government interference. Well, not so fast. Most ‘market disciplines’ involving collaboration recognize the reality that players who collaborate inevitably have a say in choices about with whom they will collaborate. Even private equity investors in, say, car rental companies lack total control over the joint venture partners and suppliers of those companies. They might recommend to management that they collaborate with, say, some chain of body shops. But, if the body shops don’t want the business, then the investor’s choice is stymied.
Similarly, the social service agencies working to reduce recidivism in the UK prison are likely to need the active collaboration of government employees and managers of that prison. It is predictable, then, that the prison folks will get involved – whether before or after the fact – with choices about the agencies. Yes, the private investors can propose such groups to the prisons; but the prison folks will have a say. This is just common sense – and, by the way, critical to success because the real commitment of prison employees to work closely with the social service agencies is needed to meet the 10% goal. So, one would actually want them to have a say in this matter. But, it’s a sad corollary of the ‘markets good/governments bad’ ideology that this ‘say’ by the prisons is considered interference while the parallel phenomenon in the private sector is not so much meddling as yet another Panglossian paean to the best of all possible worlds of ‘market discipline’.
Liebman’s monochramtic use of ‘market discipline’ threatens the wrong result/wrong way outcome because it denies non-profit and governmental organizations the use of strategies open to private sector competitors. It’s as if Liebman wishes the non-profit and governmental groups to run the race with one arm tied behind the back.
But it doesn’t end there because Liebman goes on to define winning the race – performance itself – in ways that would astonish any private sector company in any market. He (with Leonhardt’s applause) points to studies of government programs, studies that demonstrate the failure of programs to demonstrate results according to what is called the ‘scientific gold standard of randomly assigning individuals to a program or control group’ followed by finding ‘rigorous evidence’ that the programs work. Link:
This is no longer about ‘market discipline’. It’s about a different standard of performance being demanded of non-profit and governmental organizations versus private sector companies.
Before the social scientists scream, please note: I’m neither saying that random assignments of individuals to program versus control nor social scientific evaluations grounded in correlation and causality are useless. They’re very useful. We learn from them. And the Brookings study referenced by Liebman and Leonhardt is excellent.
No, what I’m saying is that neither private sector companies nor markets impose this standard of performance – especially against market challenges involving the level of complexity related to reducing recidivism or the other sorts of challenges Liebman correctly notes could benefit from social impact bonds: kindergarten readiness, employment services for hard-to-employ groups, health and disability interventions, and college retention services.
Consider kindergarten readiness. Years of working with CAP Tulsa, one of the nation’s best non-profits at providing early childhood education to low and very low income people, have taught me just how many factors influence a child’s readiness (health of the mother, socio-economic stability of parents, characteristics of neighborhood, use or not of day care and/or early childhood setting, quality of that organization and more) – not to mention an even broader array of factors that influence what happens to a ‘ready kindergartener’ from the moment of entering school to, say, graduation (or not) from high school (readiness of public school teachers and principals to take the ‘hand off’ of the child in the best way and myriad educational, family, neighborhood, social, legal, medical, spiritual and other factors from then on.)
CAP Tulsa has thoroughly embraced a range of performance disciplines and its clear that many more children and families are better off for it. But, I seriously doubt that CAP Tulsa could fully satisfy the social scientific levels of proof suggested by Liebman and Leonhardt’s writings.
In the private sector, neither investors nor executives nor employees nor customers would hold companies to social scientifically provable linkages between their efforts and their claims. Yet, that’s what happens routinely when it comes to ‘performance’ and the social sector. People are just not appliances. And the toughest socio-economic challenges are ones where people, all the messy realities that affect people, influence the outcomes – whether performance happens at all let alone whether it is sustained over time. The UK social impact bond aimed at reducing recidivism, then, is wise to measure performance by recidivism itself over some limited time frame – not by whether the prisoners ever return to prison over the course of their whole lives – and also not by whether those who avoid returning to prison are fully, effective functioning human beings in society.
There is one last element about social impact bonds that ought to make us at least a bit nervous.
Again, I’m for trying them. Still, viewed cynically, these bonds become yet one more way for financiers to strip-mine future cash flows from government. Liebman notes that the success rate is likely to be spotty – and, consequently, quite high rates of return may be needed to attract capital, which means lots of future cash flows from governmental efforts will be used to repay financiers instead of being reinvested in more effective and efficient government. Again, fair enough if and to the extent this form of capital is well tailored to fill a need not addressable by other approaches.
But, one needs to at least wonder why governments faced with the prospect of real and attractive returns on investment through scaling up solutions that work would turn to the private financial sector for the money instead of using taxes, budgeting and other means? One answer could be: having endured more than thirty years of demonization as ‘bad’ – in part because it lacks the ‘market discipline’ that is ‘good’ – government reaches a point where political and cultural ‘norms’ deny it access to capital from any source other than the private sector. The beauty of this, of course, is that if social impact bonds work, free market fundamentalists can and will sing the song “I told you so”. But, as they sing, we’ll have forfeited yet one more way that the legitimate functions of government to privatization in ways that end up serving investors more than people. If this were to happen, then social impact bonds would be much more of a trick than a treat.