ABSTRACT
In this chapter, we tell a twofold history of field experiments in economics. The first strand comes from the history of laboratory experiments. In this strand, field experiments in economics developed as a way to get closer to the “real world”. The second strand originates from the broader history of field experiments in social sciences. Its aim is to use field experiments as a tool to assess public policies. Through the promotion of randomized field experiments, this tool is now considered to be able to produce evidence for policy. We argue that the two strands focus on different aspects of experimental control. The first strand interprets control broadly, including the control of factors deemed relevant to the inferences through the creation of artefactual environments, whereas the second strand focuses mainly on the randomization of assignments of treatments as a way of controlling for the distribution of unknown factors. We call the former direct and the latter indirect control.
