One of the best news stories to come in the wake of the financial crisis was L’Affaire du Chaton, in which the accusation was lobbed that Goldman Sachs literally abandoned a group of stray kittens. Goldman, apparently recognizing that there are limits to the amount of profit-seeking the public is willing to tolerate, set not one, but two spokespeople to quell the looming media disaster.

Which is what I’m reminded of when I read this story about Goldman Sachs’s investment in social impact bonds sold by Utah to fund its preschool program.

As I understand it, Utah’s Granite School District needed money to finance its preschool program – which, it believed, prevented at-risk students from needing expensive special education later.

So Utah’s United Way of Salt Lake sold Goldman bonds.  The money was used to finance the preschool program, and Goldman was to be paid by the United Way and Salt Lake County to the extent that the program did result in cost-savings by reducing the need for special education.

The problem was that Utah itself set a rather specious standard for determining whether the pre-school program avoided the need for special education, by inflating the numbers of at-risk students.  Because the at-risk student metric was inflated, the program appeared to be wildly successful – providing Goldman with a hefty profit.

This is fascinating on a couple of different levels.  First – if I’m understanding the situation correctly – it highlights a problem with social impact bonds.  I’m guessing that it’s typical in politics to exaggerate the benefits of expenditures.  Here, the founders of Utah’s preschool program wanted to “sell” the program to the state, and made inflated claims about how much money could be saved later.  Their goals may well have been benign – they wanted funding for a good program that would help people – but in a world of political calculation, limited resources, and lobbying, they felt that they needed to overstate the benefits of their program in order to get anyone’s attention.

I assume that this  is business-as-usual in politics – not a great thing, but not exactly shocking.  And it’s mostly fine until Utah decides that it will pay actual cash money to an actual outside investor based on these inflated savings projections.  Which, I assume, is a problem that plagues all social impact bonds.

But the other level on which this situation fascinates me is Goldman’s (apparently) kitten-abandoning level of venality.  It is difficult to believe that Goldman was unaware of the flaws in the metrics when it made the investment; yet, it had no compunction about draining Utah’s school districts of funds that were intended for preschool and special education.

In this case, however, there probably aren’t many people agitating for Goldman to make reparations.  Utah, the United Way, and the Granite School District aren’t going to want to admit the flaws in their own metrics, so we don’t have a kitten-defending constituency.  Just lots and lots of payouts to Goldman.

Maybe Goldman would have done better to have purchased social impact bonds for the kittens.