The following post originally appeared on Markets for Good.
Markets for Good is a welcome push for the development and use of data to inform better decisions and, ultimately, better philanthropy. I could not be more supportive: indeed, “Better Data, Better Decisions, Better Philanthropy” is the tagline of the organization I lead.
But I have two worries about this effort. My first worry is that there is data… and then there is data.
In the current moment, in which “big data” is almost fetishized, I fear the chasing—and use—of data in a less than discerning way. We should all be careful not to promote the use of potentially misleading metrics, such as administrative cost ratios or “lives touched.” In the words of one nonprofit board chair denied funding on the basis of that metric, “I could give a lollipop to every kid in Boston and do really well.”
We should be interested, instead, in philanthropy supporting the development of data and performance management systems of the kind promoted by Mario Morino and David Hunter—ones that speak to the efficacy of the work and drive continuous learning and improvement. We should also support the development of systems that serve the needs of multiple organizations pursuing shared goals as demonstrated by the Stuart Foundation in its child welfare work. This kind of work ain’t easy. Or cheap.
This relates to my second concern, which is that, when it comes to nonprofit performance assessment, funders talk the talk but too many don’t even crawl the crawl, much less walk the walk. At the Center for Effective Philanthropy, we have surveyed CEOs of large foundations and seen that they believe in holding nonprofits to higher standards of evidence. Yet we know from our surveys of grantees that funders in general do precious little to support nonprofits in doing the difficult work of performance assessment.
Assessing the performance of a human services organization is far, far more challenging than assessing the performance of a business—requiring different analytic approaches and techniques.
In our surveys of 41,000 grantees of 284 foundations, we see that at the typical foundation, less than 10 percent of grantees report getting assistance beyond the grant for the development of performance measures. And in a recent, separate survey of nonprofit leaders who comprise our Grantee Voice panel, we see that 71 percent of foundation grantees say they do not get any support—financial or non-financial—from any of their foundation funders.
It’s not that nonprofits aren’t trying to do this work. On the contrary: our survey and a recent survey by New Philanthropy Capital in the UK shows that nonprofits care about assessment and are working on it. But they’re not getting the support they need and want. There are exceptions—well-known ones such as Edna McConnell Clark Foundation and lesser known ones such as St. Luke’s Foundation in Cleveland. But they are clearly exceptions.
Where will nonprofits get the support for this work if not from foundations? The median organizational budget of the foundation grantees in our dataset is just $1.3 million. They are stretched thin already.
It’s not just that funders don’t generally step up to support nonprofits with needed financial and non-financial support to build performance management systems. It’s that many continue to focus on administrative cost ratios and impose arbitrary limits on “overhead,” which is often ill-defined, in ways that actually create a disincentive for nonprofits to do this kind of work. As authors such as Caroline Fiennes in the UK have demonstrated, in her excellent book It Ain’t What You Give, It’s the Way You Give It, that is sheer madness—and runs counter to a focus on results.
For Markets For Good to result in meaningful change, a big part of the emphasis must be foundations stepping up and supporting the development of good, credible data and robust nonprofit performance management systems.
It won’t be easy, but it’s what is needed.
Phil Buchanan is President of the Center for Effective Philanthropy. You can find him on Twitter @PhilCEP.
