For a World Bank conference in 2016, this year’s Nobel Prize winners Abhijit Banerjee, Esther Duflo, and Michael Kremer wrote a paper titled “The State of Economics, The State of the World.” Perhaps unsurprisingly, the paper was about the role of randomized controlled trials in development economics and policy. But it highlighted a unit in the US government: Development Innovation Ventures (DIV), a USAID program.
As explained in the paper, “DIV holds a year-round grant competition for innovative solutions to a range of development challenges, pilots and tests them using analytical methods, and scales solutions that demonstrate widespread impact and cost-effectiveness.” The fund provides small grants (up to $200,000) to test whether promising new ideas can work (proof of concept), more significant grants to test the effectiveness of those ideas—frequently using randomized trials (up to $1.5 million), and then larger grants (up to $5 million) to scale successful models.
One early example of a DIV success built on a pilot that had been successfully tested outside of DIV. Researchers distributed stickers for commuter minibus drivers in Kenya to post inside their buses, encouraging passengers to complain if the driver was speeding. They recruited 2,300 minibus drivers and offered half of them small cash prizes to keep the stickers in place. The pilot led to a 60 percent reduction in insurance claims involving injury or death compared to buses without stickers—at a cost of $7 per year of life saved. But could this work at scale? DIV provided support to replicate the experiment with five times the sample. The stickers led to 140 fewer road accidents in a year.
Kremer was a driving force behind the creation of DIV, so perhaps it isn’t surprising that he and his fellow randomistas were positive about it. But they aren’t the only ones. (CGD researchers have long championed the unit.)
The results speak for themselves. Just looking at the first two years of DIV’s portfolio, Banerjee, Duflo, and Kremer suggested that the social rate of return on projects exceeded the target for success of 15 percent. More recent analysis—presented earlier this year—suggests a striking social rate of return above 75 percent, with those road safety stickers in Kenya reaching 40,000 minibuses, software for community health workers reaching 20 million workers in India and beyond, election monitoring technology reaching more than 6 million people in Afghanistan and several African countries, and more. A rate of return like that comes not just from reaching lots of people with effective interventions, but also by doing so at a reasonable price.
DIV, which was founded in 2010, suspended program applications in July 2017 because of budget uncertainties. But with support from the US Congress, DIV was able to resume project funding in September 2018 on the back of a $23 million budget. And USAID Administrator Mark Green has warmly praised the fund—and Kremer’s role in creating it. “As a founder of DIV,” Green said, “Michael revolutionized innovative approaches to solving humanity's greatest challenges by lowering bureaucratic barriers to encourage anyone to submit innovative ideas to USAID and subjecting those innovations to rigorous analysis of impact, cost-effectiveness, and ability to scale.”
It’s rare that good ideas in economics get such strong bipartisan political support in Washington. Hopefully that’s a sign of a secure future for DIV.
CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise.
CGD is a nonpartisan, independent organization and does not take institutional positions.
