It seems as though all that's in the news these days are stories of how the big, bad banks are ruining our lives.

To be fair, that impression would be largely accurate. JPMorgan Chase (NYSE: JPM) recently settled charges that it rigged municipal-bond bids over a 13-year period, and the Justice Department just charged a former Bank of America (NYSE: BAC) executive with similar activities. And who could forget Wachovia -- now part of Wells Fargo (NYSE: WFC)--– and its 2010 settlement of charges that it had been complicit in laundering vast quantities of Mexican drug money? The list goes on. You almost have to laugh to keep from crying your eyes out.

But there are some points of light out there in all the muck -- examples of how banks could do better, how they could possibly serve society instead of taking advantage of it. Goldman Sachs (NYSE: GS) has recently, quietly, offered an example of how that might look.

Now, this is the same Goldman Sachs that, just a few months ago, was under the particularly uncomfortable glare of the media spotlight for one of its investment holdings. The company reportedly owned a 16% stake in a website that was "[t]he biggest forum for sex trafficking of underage girls in the United States." It would be hard to overemphasize the sheer horror behind that statement. In fairness, Goldman Sachs' higher-ups probably had no idea what it owned. But wait a minute: Isn't that crazy in and of itself? The fact that giant financial institutions sit on piles of investments, to some degree indifferent to what their money is being used for, indicates a bigger problem in the system.

But Goldman's latest venture could turn out to be incredibly socially beneficial. It's called the social-impact bond, and it's a brand-new shiny thing.

Profiting from prisons: vice or virtue?
My colleague Alyce Lomax has written about the moral bankruptcy of for-profit prisons and the perversity of making money from deteriorating prison conditions. I couldn't agree with her more. But tweak the notion of the investment a bit, and suddenly it becomes a completely different picture. Goldman Sachs has just launched the United States' first social-impact bond, and it has everything to do with investing in and profiting from prisons. The difference, though, is in the alignment of incentives. And it could provide a sea change for some businesses.

In its simplest form, a government entity contracts with a private investor to provide a defined set of social services. The government pays the investor entirely based on achievement targets. In Goldman Sachs' case, those targets are recidivism rates. The targets are tiered: On one end, we have particularly high recidivism on the part of former inmates, and Goldman Sachs gets absolutely nothing. On the other end, if recidivism is reduced to zero, Goldman Sachs makes a mint. Moderation exists in between.

Strengths and weaknesses
On the upside, this structure directs investors' energies squarely toward clear objectives. On the downside, this structure directs investors' energies squarely toward clear objectives. Both statements are true.

Readers may be familiar with an old mantra that you don't manage what you don't measure. Whatever elements of the prison experience are not defined in the terms of the bond may not be addressed.

Now, if Goldman Sachs determines that a positive and humane prison environment contributes to declining recidivism rates, so much the better. But at least in theory, the company could conclude that torture and food deprivation are the best strategies and -- barring pesky international humanitarian law -- could pursue such methods with impunity. Still, this structure is worlds better than the model that simply pays out a fixed price for prison maintenance, regardless of outcomes.

Social-impact bonds have been tried before, albeit on a limited basis. There has been an experimental attempt in Peterborough, England, also targeting prison recidivism rates. It is too early yet to glean any lessons from measurable results, but the effort has attracted attention as a de facto proof of concept. If these initial forays are successful, we should expect to see expanded use of social-impact bonds in the future.

It's the incentives, stupid.
Whether this proves to be a social bane or benefit has everything to do with how incentives are aligned. Recidivism is a brilliant goal to target and is likely to result in major improvements in the prison experience. But if we are to expand this concept further out to other programs and social realms, we will have to be awfully careful about our targets. The desired outcomes should be clear and measurable, and there should be mechanisms for detecting undesirable, unintended consequences to experimental programs.

With those caveats, the social-impact bond could offer an elegant means to harness the power of economic incentives for societal benefit. This is capitalism at its best, in that it rewards creative thinking and innovation in solving social problems. Imagine if we could use such a structure to reduce hospital readmission rates and medical errors, or to improve high school graduation rates in neglected communities. The possibilities are endless. Good for Goldman Sachs for providing a rare positive example in an otherwise craven space. Here's hoping it's just the beginning.

If you're interested in a deeper look at one of the big banks, consider our premium report on Bank of America. It comes with a year of updates and analyzes the key risks and opportunities for investors. We've done the leg work for you, so all you have to do is read.