Shortly after the new millennium began, before she became an associate professor of political economy at Stanford Graduate School of Business, Katherine Casey discovered a difficult global reality firsthand.
The foreign-aid donations she observed while working for the World Bank in Sierra Leone didn’t always have the positive impacts on the ground that donors claimed or hoped for.
Specifically, while large-scale efforts often improved infrastructure, they typically fell short on a key dimension: bringing greater democracy and inclusiveness to decision-making in rural areas. Here, development efforts encountered a deeply embedded cultural obstacle in the form of village chiefs, who typically held lifelong positions of power with few institutional checks on their authority.
In addition to ruling largely without democratic process or accountability, village chiefs tend to be much older and less educated than younger community members, especially those benefiting from recent investments in education in emerging regions.
“With the increased efforts to provide education,” Casey says, “young, smart, well-educated people were going back to their communities. Would the chiefs leverage this new talent [to improve development] or sideline them because they’re not part of the ruling elite?”
Finding effective development routes is crucial in countries like Afghanistan, Liberia, and Sierra Leone, where most rural communities fall beyond the reach of central state organizations and must provide a variety of public goods and services for themselves.
Sierra Leone, for example, endured 12 years of civil war ending in 2002, after 30 years of autocratic rule. Worse, the “chieftaincy” social structure works against progress. Related research shows that the most powerful chiefs often oversee communities with the poorest development outcomes.
Skill Versus Voice
Casey and collaborators set out to study two different approaches to development in such areas. The resulting paper, “Skill Versus Voice in Local Development,” coauthored with Rachel Glennerster (UK Department for International Development), Edward Miguel (University of California, Berkeley), and Maarten Voors (Wageningen University), compared typical large-scale community development efforts to a new, more focused strategy that identifies and encourages the use of talented “technocrats” who are not part of the ruling elite and thus might otherwise be ignored by chiefs.
Casey and collaborators systematically tested both approaches in a set of 236 villages in Sierra Leone.
The first strategy, which has become very popular with foreign aid donors, is commonly known as “community-driven development.” Essentially a long-run effort to democratize local institutions, community-driven development provides communities with financial grants to build public infrastructure and requires residents to make decisions in a more democratic and participatory way.
The second, more customized intervention identified high-skill community members, or “technocrats,” and encouraged chiefs to entrust them with development-related work. To find untapped managerial talent outside traditional hierarchies, the researchers asked community members to nominate capable individuals, then administered a skills test assessing individuals’ ability to create and manage local development projects. The research team nudged communities to place high-potential managers in lead roles writing real grant proposals for submission to an infrastructure grants competition run by the government.
“The team made a public display of saying this person won the leadership ‘lottery’ and had the highest score on our test,” Casey says. “The chiefs weren’t obligated to pick them, though.”
The research team evaluated both approaches and compared them to the status quo of traditional rule, via randomized control trials. One clear finding was that under the status quo, chiefs typically refused to cede decision-making authority. As the researchers write, “Left to their own devices, chiefs fail to delegate complex project tasks to high-skill community members, even when it appears to be in the community’s interest to do so.”
That changed when objective skill-related information entered the picture. Many chiefs, the study found, delegated grant-writing to high-skill members when publicly encouraged to do so.
Importantly, those who scored well on the management capabilities test produced much better grant proposals — as judged by independent experts — and were more likely to win a grant from the government. Providing brief management-skills training improved performance even further.
The community-driven development approach, in contrast, created fewer detectable improvements, even though it is an order of magnitude costlier than the skill intervention. Use of more democratic processes in public deliberations, for example, resulted in no improvement on key measures. And while chiefs in those villages were marginally more likely to delegate to high-skill residents, the strategy didn’t affect grant-competition performance.
“Community-driven development was supposed to be transformational,” Casey says. “The idea of empowering the poorest people to take more control over local governance was inspiring. But research showed it wasn’t having as much effect as everyone hoped. Not surprisingly, it turns out that restructuring political hierarchies that have been in place for a very long time isn’t something outsiders can do easily.”
Ultimately, the skill approach proved a high-impact, low-cost way to catalyze development in underserved regions. That’s not to say community-driven development is without value, Casey says: “It delivers sustainable value in terms of public goods, but less in terms of improving democratic processes. Giving more voice to poor people is still an important goal; we just need to work harder and be creative in finding other ways to achieve it.”