Brief 40: Development Assistance and Collective Action Capacity

In recent years, aid donors have embraced Community-Driven Reconstruction (CDR) projects that seek to rebuild trust through the collective provision of public services. After one such project in Liberia, a team of researchers conducted a follow-up experiment aimed at determining the impact of CDR projects on the capacity for collective action. The researchers found substantive, large effects of the CDR on how much people donated, but found that these effects only existed for cooperation between mixed gender groups; women only groups cooperated at high rates whether or not they were exposed to CDR programming.

Read Full Study

Category: Conflict and Violence, Public Service Provision

Tags: CDR, CDD, community, cohesion, post-conflict, behavior, Liberia

Date of Publication: Wednesday, March 2, 2016

EGAP Researcher: Macartan Humphreys, Jeremy Weinstein

Other Authors: James D. Fearon

Partners: Photo credit David Stanley, available via Creative Commons BY 2.0

Click to Download the Data

Geographical Region: Africa

Research Question: How do Community-Driven Reconstruction (CDR) projects affect trust and cooperation in a society?

Preparer: Seth Ariel Green



Liberia is a West African country of about 4.5 million people. About 85% of citizens live below the national poverty line. Liberia held democratic elections in 2005, formally resolving two devastating civil wars from 1989-2003. Though the wars ended after the elections, Liberia—like many post-conflict societies—deals with issues of mistrust between citizens.

Research Design:

In 2006, the UK government and the International Rescue Committee funded and implemented Community-Driven Reconstruction (CDR) projects across Liberia. Under these projects, a number of local councils throughout Liberia were given the chance to choose infrastructure projects for their villages and to implement their construction.

The research team then conducted a behavioral experiment four months after project completion, when both communities who implemented a CDR project and some who did not were promised additional funds for a new project. They then learned that the exact amount of funds provided was dependent on the behavior of randomly selected, anonymous local citizens. Each selected citizen was given about $5 and told that they could choose how much money to keep and how much to donate. For half of the participants, the implementing NGO doubled donations, and for the other half, they multiplied them by 5. In half of selected villages, all of the anonymous participants were female (“women only condition”), and in the other half, the participants were mixed gender (“mixed gender condition”). The intention was to look for evidence of a link between CDR programs and subsequent willingness to support community projects.
The researchers found an unexpected result. The CDR did substantially increase how much people donated, but this effect was concentrated entirely in the mixed gender condition. That is, mixed gender groups who had had experience with a CDR project gave substantially more than those who had not experienced a CDR, but there were no significant differences in the women-only condition between those who had experienced the CDR and those who did not. In fact, with or without the CDR, women who were in the single-gender condition gave substantially more than both men and women in the with/without CDR variants of the mixed gender condition.

The authors hypothesize that this was because in the women-only condition villages, information spread through traditional women's networks, whereas the CDR was helpful in facilitating cooperation in the mixed-gender conditions, where citizens were likely to have previously had fewer between-gender interactions about public affairs.

Policy Implications:

First and foremost, this experiment shows the challenge (or perhaps impossibility) of finding a “one-size-fits-all” development project. An innovation that spurs trust in one place might have no impact in another, depending on local conditions and local power structures.

Second, policymakers would do well to pay attention to local power structures, including informal information networks like those that linked women in the single-gender condition.

Third, Community-Driven Reconstruction can have a substantial impact on both material and interpersonal well-being, shown by the increased generosity of many people who lived in villages with CDR projects. This is one of the very few CDR experiments that found evidence for positive social effects. The importance of this set of benefits, though difficult to precisely measure, should not be underestimated.