|
Abstract:
|
"This paper presents an impact assessment of a peacebuilding program in Colombia. The program was started in one region of the country and, through development aid, transmitted to 15 other regions. A common criticism of impact evaluation is that evidence of program impact is localized to the study population. Little evidence exists to confirm if successful programs achieve the same impacts when 'scaled-up' to the national level or when transmitted to different regions-fundamentally an issue of ex-ternal validity. We evaluate program transmission success by comparing outcomes in towns located in the original region to outcomes in towns from five different regions of Colombia where the program has been transmitted. We assess outcomes based on the program goals of increasing cooperation. To measure cooperation we conducted a series of experimental games (a trust game, a public goods game, and a voluntary contribution to local public services) with almost 2000 Colombian citizens from these regions. We show that the metric by which one measures cooperation is important to determining the effects of the program's expansion. We find that beneficiaries of the peacebuilding program are more likely to exhibit cooperative outcomes in these games. We then investigate the conditional average treatment effects (CAFE) of the program to determine what types of subjects are most likely to benefit from the program in the expansion regions and find that male subjects were the most likely to benefit from the program."
|