Social Impact Bonds And Their Limits

There are lots of things the public sector is involved with that are incredibly valuable if they actually work, but they're often difficult to do and the public is naturally reluctant to support policy interventions that they don't think are going to accomplish anything. David Leonhardt touts one potential solution to the problem, so-called "social impact bonds" that function as a kind of loan to service-providers:

The idea goes by one of two names: pay for success bonds or social impact bonds. Either way, nonprofit groups like foundations pay the initial money for a new program and also oversee it, with government approval. The government will reimburse them several years later, possibly with a bonus — but only if agreed-upon benchmarks show that the program is working.

If it falls short, taxpayers owe nothing.

What we tend to do right now is first look at a problem that it would be hugely valuable to solve, then we discount what we're willing to spend on solving the problem for our estimate of the low probability of succes. The result is that we appropriate less money than it would be worth spending to solve the problem, but with no guarantee that our intervention is helping at all, and with some surety that if our intervention is helping we're not doing enough of it. Then you get growing cynicism and passivity. The "impact bond" structure could solve that. It could allow a situation whereby successful policy interventions get lavish funding, failed ones get no funding, and the public grows more optimistic and generous about solving social problems. Obviously, though, there are about a thousand practical issues with this. Jeffrey Liebman has a report for CAP looking at many of them and how they can be solved.

I do, though, have a concern about this on another level. The underlying presumption of the proposal is that our programs lack efficacy because of a technical problem that needs a technical solution. I'd say we have something closer to a cultural or ethical problem. I think we had a version of that in DC when people were pretty openly saying that the city government should be seen as a jobs program, rather than a provider of high-quality social services. And we have it when Representative Roy Blunt (R-MO) openly talks about how it's more important to protect farm subsidies than to feed poor families. SNAP is a program that could be improved. But you can see from Rep Blunt's attitude that he doesn't want it to be improved. He suffers from an ethical disorder in which he's not interested in the well-being of poor Americans. And he's operating in a political culture where this fact isn't something he's trying to hide, he's discussing it openly in newspapers. Given Blunt's views about the merits of funding farm subsidies vs funding nutrition assistance, it's clear that given the opportunity to tweak SNAP program design Blunt would not be interested in increasing its efficacy as a nutrion program—he'd want to increase its efficacy as a hidden subsidy to agribusiness.

Which is all just to say that politicians can always screw things up if they want to.