Our research is motivated by the emerging consensus that the accountability, competence and inclusiveness of governance institutions are important determinants of economic development, yet there is little agreement on the specific types of programs or reforms that will successfully engender better functioning institutions in less developed countries. Moreover, it is unclear whether it is possible, or even desirable, for external actors like foreign aid donors to attempt to restructure local power dynamics.
Among donors, non-governmental organizations (NGOs) and governments, arguably the most popular response to this challenge of reforming institutions lies in “community driven
development” (CDD). More specifically, CDD programs combine block grants for local public goods and small enterprise development with intensive, long term facilitation that aims to reduce the coordination costs that impede collective action and empower marginalized groups—like women and young men—in decision-making.
We evaluate the impacts of one particular CDD initiative, the “GoBifo” Project (which means “Move Forward” in the local lingua franca), which was funded by the World Bank and the
Government of Sierra Leone. Implemented over a four year period (2005 to 2009), the project provided communities with US$ 4,700 (approximately $100 per household) in block grants for local development and skill training projects. Program facilitators further worked with communities to establish representative village development committees, agree to medium term development plans, and set up community bank accounts. Throughout this process, the facilitators actively encouraged women and young adults (aged 18 to 35 years) to participate in project meetings, take on leadership roles and manage their own projects.
Using a rigorous randomized experiment framework, we find that the project led to large, positive gains in what we call the “hardware” of development. Specifically, we see that the
“treatment” communities that received the project have more local public goods—like latrines, community centers and seed banks—and that these goods are of better quality than in the control” communities that participated only in the research. Moreover, beneficiary communities are better off economically in that they have greater village-level market activity, as measured by the number of petty traders and goods on sale, as well as more household assets and amenities.
Yet when we turn to the “software” side of development, we do not find any program impacts on participation in decision-making, the capacity to engage in collective action outside the project sphere, the voice of women and youth, or political and social attitudes. As these institutional dynamics are difficult to measure, we used three novel “structured community activities” (SCAs) to directly observe communities: i) making a communal decision; ii) responding to a matching grant opportunity; and iii) managing a public asset. Implemented after the GoBifo Project ended, these activities reveal that women and young adults in treatment communities were no more likely to speak publicly during a community meeting to decide between two comparable assets; communities were no better able to raise the matching funds necessary to redeem vouchers that subsidized the cost of building materials; and local elites were no more or less likely to capture an asset that was given to the community for free.
Taking these results together, we find that this CDD program in Sierra Leone was a reasonable mechanism for delivering small scale public goods and improving material welfare, yet did not lead to any lasting changes in local collective action performance, participation in decisionmaking, or village institutions.
