'Maximize results while managing the risks': Lessons from USAID's open innovation program

SAN FRANCISCO — Development Innovation Ventures, which provides grants to test and scale innovations, is celebrating its 10 year anniversary. The U.S. Agency for International Development’s DIV borrows from the venture capital model, allowing the agency to invest in high risk and high return projects that do not fit within the agency’s traditional structures. “Just like a VC firm, we don’t expect every investment will be successful.” --— Michael Kremer, co-founder, Development Innovation Ventures At this week’s virtual Social Capital Markets conference, funders and implementers shared lessons from DIV over the course of the past decade, including how rigorous evaluation can support the scaling of innovation. While for-profit startups can raise equity financing until they can build a market for their product or service, nonprofit innovators have a hard time raising money to develop product-market fit. “We struck out with a lot of funders early days because we didn’t really have a clear product,” said Gayatri Datar, co-founder and CEO of EarthEnable, which is working in Rwanda to replace dirt floors with earthen floors made of locally sourced materials and a proprietary oil. “We have literally iterated six different business models.” Funding tied to a particular business model could have led EarthEnable to go bankrupt, but DIV allowed the organization to experiment and iterate. When DIV launched, it was a unique initiative among bilateral donors, donors which are accountable to taxpayers and tend to avoid investing in projects with a high probability of failure. DIV’s future within USAID has not always been certain. The program was put on pause in July 2017, amid uncertainty around foreign aid funding. But the agency relaunched DIV applications in September 2018, announced 16 new awards last month, and it is currently accepting applications. “DIV remains open for business,” said Michael Kremer, who co-founded DIV and currently serves as its scientific advisor. A number of approaches set DIV apart from other funders of international development innovations, Kremer said. He emphasized DIV’s open approach to innovation, which supports evidence-based interventions across sectors, countries where USAID works, and pathways to scale. “We try to complement that openness with some principles about how we fund things,” Kremer said. “We try to use best practices from investors and academia to try to maximize results while managing the risks.” DIV’s tiered funding approach helps to manage those risks, he explained. Innovators can apply for funding across the stages of piloting, testing and positioning for scale, scaling, and generating evidence. DIV uses peer review of proposals, and engages with the development economics research community, in the application process. While innovators can apply for DIV at any stage, larger-scale funding requires evidence of impact, Kremer said. He won the Nobel Memorial Prize in Economic Sciences last year for his work on randomized control trials as a way of testing poverty alleviation programs, but not all grantees need to run RCTs, he explained. In later stages, Kremer said DIV wants to see rigorous testing of whether and how an intervention works to demonstrate a causal relationship between innovation and outcome. “The experimental method supports the process of continual improvement,” Kremer said. “Sometimes it gets pigeonholed as evaluation after the fact. I really think of it as a critical part of the innovation process, like beta testing.” Innovators who applied for DIV grants with previous rigorous evidence of impact, or partnered with development economics researchers in order to generate that evidence, were more likely to scale, he added. “That might be counter-intuitive to people who are used to thinking of that as a barrier, but I think it suggests impact evaluation can play an important role in innovation development,” Kremer said. DIV provided Village Enterprise, a nonprofit that provides entrepreneurship training in Kenya and Uganda, with funding for its poverty alleviation development impact bond but the COVID-19 pandemic has disrupted that effort. “We saw countries globally shut down right when our first outcome evaluation was scheduled to take place,” Dianne Calvi, CEO and president of Village Enterprise, said in an email to Devex. She said DIV, which is an outcome funder of the development impact bond, has shown flexibility in negotiating how to finish the project despite the delay. Calvi added that she sees more and more funders recognizing the importance of rigorous evidence in evaluating interventions. For example, the Global Innovation Fund, a nonprofit innovation fund headquartered in London, uses similar criteria to evaluate interventions. Since DIV launched 10 years ago, a growing number of funders have started initiatives to invest in innovation for development. But little work has been done to assess their returns on investment, write Kremer and his coauthors in a working paper evaluating the first 2.5 years of DIV’s portfolio. They suggest that development innovation funders build on lessons from DIV by considering open innovation funds, taking an expanded view of routes to scale, and understanding that not every innovation will yield dramatic social impact returns. DIV committed itself to a 15% target for the social rate of return for the portfolio as a whole, Kremer said in the SOCAP session. “Just like a VC firm, we don’t expect every investment will be successful,” he said. “We want to take risks, but we expect enough will have enough impact to cover the costs of the whole portfolio.” Investors in development innovations should expect some investments to fail, but compare the cost of an entire portfolio against the benefits that successful innovations generate, Kremer said. Maura O’Neill, former chief innovation officer of USAID who co-founded DIV with Kremer, said she hopes to see DIV source more proposals from low resource countries and scale development outcomes through people who are willing to pay for it, whether consumers, businesses, or governments. To realize its potential, DIV needs to figure out “how to develop more capability to successfully scale and sustain the outcomes” without long-term donor financing, she told Devex. USAID might support these efforts through its diplomatic or development discussions with host countries or companies that have the ability to produce and distribute at scale, O’Neill said.