Paper short abstract:

Researchers conducting randomized controlled field experiments convene “public randomization ceremonies” to make randomized resource allocation appear fair to retain research participants. I draw on theories of ceremony to examine this practice as both an aesthetic and an epistemic object.

Paper long abstract:

What does it mean for rationality to rest on ceremony? Development economists use randomized controlled field experiments to evaluate the impact of policy interventions in low-income countries. Because it is difficult to manufacture placebos of social interventions, researchers need randomized resource allocation to appear fair to enroll and retain research participants. To demonstrate the fairness of randomization researchers organize "public randomization ceremonies" (PRCs). In PRCs research participants gather to take turns drawing their treatment assignment and displaying it to the group.

While STS analysts have discussed ceremonial aspects of science such as award ceremonies and public meetings as critical to the work of science, I consider the PRC as an element of the experimental apparatus that make experimental knowledge production possible. Drawing on theories of ritual and ceremony, technologies of transparency, and public engagement with science, I examine PRCs as both an aesthetic and an epistemic object.

I suggest that PRCs formal structure presents the conflation of fairness with transparency as unquestionable and precludes discourse about alternative meanings of fairness. PRCs also function as rites de passage in which individuals "officially" become consenting research participants: the control group gives up rights to the resource being distributed and to contestation of the distribution process and the treatment group agrees to follow the study protocol. But because public randomization ceremonies are "junction ceremonies" produced for individuals among whom personal ties are fragile, the consensus they build is tenuous. Control groups do resist forcing researchers to do repair work to sustain their experiment.