Recent emphasis in applied development economics has been on evaluating complex financial market interventions, such as microfinance programs, savings mechanisms, and innovative insurance products. However, these programs are often plagued by low participation, making the task of impact evaluation more difficult due to imprecision of estimates. Furthermore, the program impacts identified by different designs will vary when the effects of participation are heterogeneous. These two factors can combine to leave researchers with difficult choices.

This paper is a methodological exploration of research design in the presence of low participation and program effect heterogeneity. To put this issue in context, I use the example of index insurance, an innovative financial tool characterized by low participation rates. I focus on the choice between a research design based on randomized eligibility and a randomized encouragement design. Randomized encouragement designs offer a stronger incentive for program participation to a randomly chosen subpopulation, and then use the incentive as an instrumental variable in econometric impact evaluation.

When impacts are heterogeneous, the effect estimated by a randomized encouragement design will be biased relative to the impact of a program on participants. However, the possibility of greatly enhanced precision in estimation means that randomized encouragement designs may yield estimates closer to the truth than a randomization of eligibility in a given sample. In addition, greater unobserved heterogeneity will not necessarily increase this bias, contrary to intuition. These conclusions depend on the nature of the program and outcomes being studies, and ought to be considered carefully by researchers weighing alternative research designs.