Jonathan Davis, Jonathan Guryan, Kelly Hallberg, and Jens Ludwig
When researchers conduct randomized controlled trials (RCTs) of social programs, the hope is that programs that appear promising in small trials can then be implemented at a larger scale. But how do we know whether a social program will be successful at scale without testing it at scale? Guryan and his colleagues propose a model to measure the economics of scaling up by focusing on the challenge of finding the necessary inputs to the program as the program grows. As programs are scaled, if any inputs are limited in supply, either the average cost of the input must go up, or the quality of the program will decline holding average costs constant. While acknowledging that exact costs of scale-up cannot necessarily be known, by using random sampling of inputs, the authors show that it is possible to create and test a program at a smaller scale while still learning about the effectiveness and cost-effectiveness of the program at a much larger scale.