This working paper describes the revolution in the United States in support for the use of randomized controlled experiments to evaluate social programs. Focusing on the welfare reform studies conducted between 1970 and the early 2000s, it presents the major challenges to winning this support and how they were overcome.