In this chapter, we report a field experiment in Burkina Faso, which aimed to isolate the effects of information content from other channels of influence through which an information intervention could affect voting behavior. The experiment was carried out in 38 rural municipalities, prior to the 2016 municipal elections. These 38 municipalities had been controlled by the same party for the past two electoral cycles, since the first nationwide municipal elections in 2006. In our experiment, we presented 741 randomly selected study participants with detailed information about their previous municipal government’s performance along nine indicators of municipal service quality in the areas of health, primary education, water access, and civil services. These indicators reflected national standards for municipal services, i.e. widely accepted service delivery targets. Simultaneously, a control group of 752 study participants was presented merely with information about the indicators of municipal government performance, without any information on the actual performance of their previous municipal government. Thus, our experiment varied study participants’ access to information about municipal government performance, but it held their knowledge of service delivery targets, as well as the method of information delivery, constant across the treatment and control conditions.

Malte LierlMarcus Holmlund
Publication type: 
Working Paper
October 04, 2017
Program area: