In the past decade, several initiatives have been established to invest in innovative goods and services designed to addressed development problems, including by ADB. Emerging evidence suggests that the social returns on such innovation portfolios are large, and driven by a handful of high-reach, high-impact innovations. We characterize the distribution of innovation scale using data from development programs supported by USAID between 2010 and 2014 through its Development Innovation Ventures portfolio. In investigating the correlates of innovation scale, we find that a given innovation was more likely to reach one million users if it leveraged existing distribution platforms, had low unit costs, and was built on rigorous evaluations. We hypothesize that these characteristics are negatively associated with barriers to entry, suggesting a path for public and philanthropic funders to generate large social returns by investing in innovations that are unlikely to attract financially motivated investors because of limited profitability. We build a model consistent with these findings and discuss implications for the design of social innovation funds as a complement to profit-oriented investing.

JEL codes: O30