The Sloan Foundation and other science funders naturally care about making basic research more productive, or, in other words, increasing the value of scholarly outputs per grant dollar we spend. Among many factors influencing that ratio, funder policies could be potentially important. How we attract applications, what we require applicants to do, and how we select among them all help shape the priorities and progress of the scientific enterprise. While funders like Sloan have broad control over our procedures, measuring the specific effects of changing those policies has traditionally been considered difficult at best. When it comes to evaluating policy implementations of any kind, the Abdul Latif Jameel Poverty Action Lab (J-PAL) is arguably the world’s premier organization dedicated to the design, execution, and interpretation of randomized controlled trials (RCTs). Based on J-PAL’s work since 2003 on international development, three of its founders shared the 2019 Nobel Prize in Economics. With Sloan support, J-PAL’s North America branch was launched in 2013 by Amy Finkelstein, who was named a MacArthur Fellow five years later. What does J-PAL’s mission have to do with scientific productivity? The role of new discoveries and innovations in economic growth is an old story among economists, of course. J-PAL North America has therefore concluded that policies for accelerating the progress of science should be among those it studies to help alleviate poverty. But there are two main rate-limiters for conducting RCTs in that area: first, not everyone knows how to design, carry out, and interpret a rigorous RCT; and second, those who have that expertise also need to partner with institutions willing to host such rigorous evaluations. J-PAL’s new Science for Progress Initiative (SfPI) aims to fix that. Co-led by MacArthur Fellow Heidi Williams from Stanford and economist, entrepreneur, and philanthropist Paul Niehaus from the University of San Diego, SfPI will not only train and support a community of researchers, but also match them to organizations with important questions to answer about making science more productive. As a step towards accomplishing this, SfPI leader Heidi Williams has helped organize a workshop at the National Academy of Science on 'Experimentation in Federal Funders.' Along with presentations from NSF, NIH, and other science agencies about what they want to learn from running policy evaluations, the U.S. Patent and Trade Office described how, inspired by what they already learned through randomized evaluations, a whole new office has been created there to run more RCTs in cooperation with academic researchers like the ones affiliated with SfPI.