So how did the session at ASE conference go?

On e last thought: Another place to go for lessons learned is international development. In that world there is increasing interest in using randomization as part of the evaluation of program effectiveness. See, for example the chapter by Dufflo and Kremer in Reinventing Foreign Aid (MIT Press) http://bit.ly/ATTHg. Adam Jaffe long ago floated a similar idea for technology policy, but never got much traction. http://bit.ly/YcmJP

Now this is actually a good idea--not for the totality of a fund, but for a sizable fraction. Looking to the long term, there is real value in setting up a structure that will allow for high quality information to be gather over time on the characteristics of program effectiveness. Not easy, but again there is substantial learning out there from which to draw.

Otherwise, I expect we're all in agreement that evaluation needs to be as minimally budensome as possible. See e.g. essay by Brian Trelstad of Acumen in Innovations journal http://bit.ly/xQMcS.

Sorry I'll miss the investment panel as I'll be on the education panel. Looking forward to catching up afterward.

The exit strategy approach you suggest makes sense. The analogy then is to corporate venture capital, particularly where the corporation is a potential customer for the funded venture, or its products.

If you're talking about a venture that would sell goods or services to the government at scale, I think this model would potentially work well. Numerous precedents exists. One is the CIA's In-Q-Tel. (The Chesapeake Crescent Initiative, with whom I have been working, has a proposal in the works for similarly structured funds organized around national needs in health [VA as buyer], defense [DoD as buyer], and energy [GSA as buyer]).

There is more likely to be a problem when the new venture is providing a good or service that is competitive with existing government programs. The corporate track record in sustaining activities that canabalize existing activities isn't very good. There is not much reason to expect government--any government--would be substantially better.

My other concern about the direction proposed--emphasizing later stage--is that, while it is certainly low risk, it may also be low return on the marginal dollar spent. The yield on SBIR funds is very high, in part because the dollar amounts are low and the focus is early stage. It is important to nurture the supply side of social ventures. That means seed funding to validate concepts that have promise.

Always easy to identify winners at the finish line... not so easy at the starting gate, or even at the bend.

Doesn't have to be an either/or, though. Particularly for such a new initiative, some structured experimentation makes sense.

The correct structure of the fund will follow from its function. Are funds primarily intended to achieve scale? To prove a concept? Or to subsidize provision of a good or service that creates social value, but for which markets do not exist or are undeveloped?

If to achieve scale, the growth of the venture may be most effectively achieved by funding the venture's most promising projects, as in the case of Commerce Department's Advanced Technology Program (now the Technology Innovation Program). ATP awards were highly competitive and substantial—typically in the range $1-$5 million—and were critical to the development of a number of high-impact technology companies.

If to prove a concept, then, again, it may be better to help the company grow by providing staged grant support at the project level. Here the model is the Small Business Innovation Research (SBIR) program. (Ref. study completed last year by the National Academy of Sciences of the SBIR program: <a href="http://www7.nationalacademies.org/sbir/" target="_blank">http://www7.nationalacademies.org/sbir/ ). SBIR Phase I awards are $100K. Successful completion of Phase I permits firms to compete for Phase II awards of $750K.

If to provide a good or service that creates social value, but for which markets do not exist or are undeveloped, then it is possible that the SBIC model makes sense. The history and documented impact of that program are worth a careful look. What looks good on paper doesn't always work out so well in practice.

"Venture funding" without both an equity stake and mentoring/support is a grant. From a marketing standpoint it may make sense to call an entity that provides grants a "venture fund." But calling it so doesn’t make it so. It is critical to keep in mind that such “venture funding” without an equity stake and real expertise or networks to offer support to entrepreneurs is not in any meaningful way analogous to "venture capital" or angel investments.

One venture fund focused on social innovation that seems will merit that name is the one launched at the Skoll World Forum last month by the Aspen Network of Development Entrepreneurs (ANDE). Like Lemelson, Acumen, and Endeavor Global (who are members) ANDE will focus on mid-sized firms with potential for high growth. Ref. <a href="http://bit.ly/Qu03" target="_blank">http://bit.ly/Qu03

Can the government develop and implement an ANDE-like fund? Or should the Social Innovation Fund, more like the SBIR program, support promising projects in companies to get them to the stage where they can be supported in the next stage by private social investment funds (an approach to public-private that focuses on staged complementarity rather than simultaneous co-investment).

Looking forward to seeing you again and to joining this discussion at the ASE conference.

[The literature on entrepreneurial finance is, obviously, vast. For anyone interested, two surveys of mine:
* Handbook on "Financing Entrepreneurship": <a href="http://www.e-elgar.co.uk/Bookentry_contents.lasso..." target="_blank">http://www.e-elgar.co.uk/Bookentry_contents.lasso...
* Paper I presented at Kauffman/OECD conference last summer on policies to support high growth firms: "Entrepreneurship, Opportunity, and Growth" <a href="http://ssrn.com/abstract=1376427" target="_blank">http://ssrn.com/abstract=1376427 ]