Scaling Social ImpactNovember 11, 2009 by Kimberly Dasher Tripp
I’m Kimberly Dasher Tripp, a program officer here at the Skoll Foundation and a first-time contributor to the Skoll Foundation blog. I look forward to the dialogue!
I am at the Fuqua School of Business at Duke this week attending the Scaling Social Impact conference. The event is hosted by the Center for Strategic Philanthropy & Civil Society at Duke University’s Sanford School of Public Policy, the Center for the Advancement of Social Entrepreneurship (CASE) at the Fuqua School, and the Bridgespan Group.
Last night two academics (Marc Epstein, Rice University and Funda Sezgi, IESE Business School) presented research they have conducted regarding scaling. Jeff Bradach of the Bridgespan Group did a remarkable job summarizing crosscutting themes from the two pieces of research. Bradach framed it using four questions about scale:
1. What are we trying to scale? Meaning, what is the essential “it”? We are trying to figure out how we know a model works. So, if we are scaling the wrong thing, we aren’t measuring or creating measures for the right piece of the puzzle. Epstein cited an example in the area of microfinance: should we simply be scaling access to capital or should we be scaling human capital along with the financial capital in order to support these loans? If every loan officer has an average of 500 clients, do you realistically have the structure to measure impact beyond repayment rates?
2. What are the pathways for scale? Or, what are the processes by which a concept scales? Organizations need to think about the power of running different scale models, identified by Greg Dees at CASE as fitting three broad categories: dissemination, affiliation and branching. The interaction and tension among the three models can be very instructive. Do underlying fee/income structures allow for better scale? Does government or policy support help account for scale?
3. How do you ensure fidelity to the model? Sezgi’s evaluation of the Aravind Eye Hospital noted that the model required a deep discipline of execution with respect to scaling values, not just processes. So, Bradach posed, if you do not exert specific control over the scaling of your values, are some versions of your replicated model not, in fact, your model? In evaluative terms, how can you be sure that the model does or doesn’t work if some branches or versions of your model are not actually your model?
4. What about scaling impact versus scaling organizations? We must take into consideration what scale really means. Bradach used Teach for America as an example. Isn’t some of the most potent impact of Teach for America the work of some of its alumni after their Teach for America experience is over? Is talk about scaling the organization perhaps missing something about impact?
The conference is clearly off to an interesting and thought-provoking start!