ABSTRACT
In randomized controlled trials (RCTs), programs to improve the transparency and social accountability of public services sometimes have transformative effects, and sometimes have null effects. Null effects are often interpreted as no effects. This chapter reviews evidence from an RCT of a community scorecard program in Indonesia and Tanzania that had null average effects on maternal and newborn health care outcomes. These null effects did not indicate no effects. Participants generally reported increased perceptions of their capacities to improve their communities, and the program led to tangible, memorable improvements in a significant proportion of communities where the program was offered. The effects were highly varied, often subtle, and not unusual, weakening their impacts on aggregate health care outcomes across communities where the program was offered compared with communities in a control group. But this same variability and normalcy were also strengths, reflecting adaptability and appropriateness in solving local problems. The results do not that the program had no effects. Even where information and encouraged participation are insufficient on their own to cause significant improvements, they can be an empowering, locally appropriate complement to available resources and capacities for improving the responsiveness, accountability, and effectiveness of a valued public service.
