Peace, Recovery, and Reconciliation: Randomized Evaluations in Post-Conflict Environments

May 12, 2016

Twenty-five years ago, Sierra Leone plunged into a decade-long civil war, during which neighbor turned on neighbor, 50,000 people were killed, and over half the population had to flee their homes. Sierra Leone is far from unique: roughly a quarter of countries worldwide were experiencing civil war when the Sierra Leone conflict started. How can individuals and groups recover from these violent conflicts?

A new randomized evaluation by Jacobus Cilliers (Georgetown University), Oeindrila Dube (NYU), and Bilal Siddiqi (World Bank), implemented by Innovations for Poverty Action and funded in part by J-PAL’s Governance Initiative, is both encouraging and concerning. The results, released in Science, suggest that bringing communities together to discuss their experiences and having perpetrators ask for forgiveness can increase trust and encourage people to contribute more to their community. At the same time, reliving the horrors of war left individuals feeling more anxious and depressed.

In a Perspectives piece accompanying the article, J-PAL affiliates Katherine Casey (Stanford) and Rachel Glennerster (J-PAL) place this research in the context of other evaluations of post-conflict interventions, including community-driven reconstruction and various support programs for ex-combatants.

Community-Driven Reconstruction: These programs provide communities with grants to invest in a local infrastructure project, which they must choose in an open and participatory way. Randomized evaluations have generally found that the programs are effective in improving infrastructure, but not in promoting trust, raising contributions to public goods, or changing the way communities interact:

  • In Sierra Leone, a community-driven development program successfully established village-level organizations and tools to manage development projects, but there was no evidence that the program led to fundamental changes in local institutions or decision-making.
  • In Liberia, a program improved measures of cooperation, but only under specific circumstances.
  • In the Democratic Republic of the Congo, there was no evidence that a program had any impact on participation in community meetings, demand for accountability, or other changes in local governance structures.

Support for Ex-Combatants: These programs aim to prevent further violence among ex-combatants and other youth involved in violent or illicit activities by providing support such as cash grants, employment opportunities, and cognitive behavioral therapy (CBT). Randomized evaluations have found that these relatively short interventions can lead to meaningful behavior change:

  • In Liberia, an intensive agricultural training program increased participants’ employment in agriculture and average wealth and decreased the amount of time they spent in illicit activities.
  • Another program in Liberia, which combined CBT with a cash grant of US$200, reduced aggressive and criminal behavior among participants by up to 50 percent for a year.

This existing research, along with the new study on community reconciliation, highlights some promising approaches to social and economic recovery in post-conflict environments. However, the question of how to address the mental health consequences of war remains unanswered.

See more news coverage:
The Washington Post
Daily Mail
Pacific Standard