Community Driven Development in Sierra Leone

Researchers evaluated the impact of two separate interventions by providing them to a randomly selected subset of villages from a set of 236 villages in the Bombali and Bonthe districts. The first was a community driven development project (known as the “GoBifo” program) that ran from 2005 to 2010. The second was an intervention to delegate managerial responsibilities of development projects to high-skilled individuals, which was conducted several years after the CDD project was implemented. 

1. Community Driven Development (the “GoBifo” program): In 2005 and 2010, researchers provided block grants to a randomly selected subset of villages to fund public infrastructure.  The first phase ran from 2005 to 2009 and included roughly US$5,000 dollars in block grants per community in the 118 intervention villages. The grant was coupled with six months of intensive social facilitation to promote collective action and more inclusive and accountable local governance. The participation and inclusion of marginalized groups, such as women and youth, was central to the program. A second less intensive phase of GoBifo began in 2010 with an additional grant of US$1300 to support youth empowerment activities in 60 of the 118 treatment communities. These grants were managed by a village development committee (VDC), which was mandated to include representatives of marginalized groups, and provided training. The committee was encouraged to make the process of implementation participatory and democratic. 
2. Technocratic selection: In 2016, the local elected governments in the study districts ran a competition to award grants for small-scale infrastructure construction. Entering the competition required a detailed project proposal and budget, submitted to the district government office. A committee of elected Local Councillors evaluated and ranked all proposals and awarded implementation grants each worth US$2,500 to the top twenty proposals. Researchers leveraged this competition to evaluate the effect of an intervention to technocratically select the managers of the community development project. 

Researchers chose high skilled individuals based on community input and a standardized test to measure basic literacy, past experience, etc. All communities were asked to nominate potential technocrats during a community meeting, who then took the standardized test. The highest scorer in this test was named the local technocrat.

Next, researchers randomized all 236 programs and comparison villages from the GoBifo program into three groups. In the first program group, local leaders were nudged to put the local technocrat in charge of managing the proposal. In the second program group, a management training was provided to the local technocrats alongside the nudge. The third group received no intervention and served as the comparison group.  

Researchers collected data in 2009, shortly after the first intensive phase of GoBifo activities had ended, to capture any impacts on the stock and quality of public goods, as well as measures of institutional change and collective action. Researchers collected data from the original 236 villages in the treatment and comparison groups in 2017, to assess long-term impacts on institutional changes. 

Researchers also collected information on the prior beliefs of academics and policy experts in Sierra Leone and OECD countries on the long-run effects of CDD, by asking them to predict the magnitude of treatment effects.

Finally, researchers measured the quality of proposals submitted in the 2016 government grant program through review by different sets of local development professionals. They used the quality of proposals to compare the long-term impact of the CDD program with the technocratic selection intervention, and contrast these with the status quo of the traditional chiefs’ authority.