Brief 34: Violent Conflict and Behavior in Burundi

This study investigates the relationship between exposure to violence during the Burundian civil war & current preferences & behavior. Researchers performed lab games to estimate altruism & attitudes towards risk & patience. The authors find people directly affected by the civil war were more altruistic; riskier; & more interested in an immediate payoff.

Link to Full Study

Category: Conflict and Violence

Tags: conflict, risk, altruism, Burundi, violence

Date of Publication: Wednesday, November 18, 2015

EGAP Researcher: Maarten Voors

Other Authors: Eleonora E. M. Nillesen, Philip Verwimp, Erwin H. Bulte, Robert Lensink, Daan P. Van Soest

Geographical Region: Africa

Research Question:

How does exposure to violence affect preferences about altruism, risk and patience?

Preparer: Seth Ariel Green



Burundi is a poor, sub-Saharan African country with a population of about 10 million people. Over 90% of Burundians are engaged in agriculture, mostly at the subsistence level. From 1993 to 2005, the country experienced a devastating civil war between the country’s two main ethnic groups, Hutus and Tutsis. Violence began when a democratically elected Hutu president, Melchior Ndadaye, was assassinated by Tutsi extremists in 1993, and ended with the 2005 election of Pierre Nkurunziza, who is still president as of 2015. The BBC estimates that 300,000 people were killed during the war and many more were internally displaced. Peace remains fragile, as the recent election violence demonstrates.

Research Design:

The researchers study 300 respondents in 35 communities that experienced a wide range of violent events. To identify past violence, the authors measure how many deaths communities experienced during the war, and also interview individual households about their experiences during war.  To measure attitudes, the authors play a series of games in which people are faced with choices about sharing money with co-villagers; whether to accept a fixed amount of money or participate in a simple gamble; and whether to accept some money in the immediate future or more a higher amount later. 

To establish that violence caused these changes in attitudes, the authors show that there is no underlying variable systematically related to both exposure to violence and attitudes. If no such factor is present, then exposure to violence will, on average, be the only major difference between exposed and non-exposed communities, and thus can be linked conclusively to changed behavior.

To demonstrate this, the authors provide both qualitative and quantitative evidence. They first provide analysis from NGOs arguing that violence was random, and not determined by particular community characteristics. They then further bolster their claim by conducting statistical analyses of relationships between community and individual characteristics and violence. The authors find that while most factors show only weak relationships, distance from the capital, Bujumbura, and altitude, are both negatively related to exposure to violence. They also find that that households headed by males during the civil war were disproportionately less likely to be present during the study. When the authors conduct additional analyses to correct for these three factors, the results are substantively the same. 


Outcome Effect of Exposure to Violence
Altruism Strongly more altruistic
Attitudes towards risk Strongly more risk-acceptant
Patience Weakly less patient


The first game gave participants a choice between two options of how to distribute a sum of money between them and another, randomly assigned and anonymous member of their community. The authors find that people who were exposed to violence were more altruistic, meaning that on average, they chose to give their partners more money than did those not exposed.

The second game asked people to choose between a sure gain of a certain amount of money or a risky bet where they could gain either more money or nothing. Participants chose across 3 different bets, where the guaranteed amount of money was either less than, equal to, or more than the expected value of the bet. The authors find that those exposed to violence had a higher tolerance for risk, meaning more likely to take the bet rather than the sure gain.

The third game asked people to choose between a payoff in the immediate future and a variety of augmented payoffs 15 days after the experiment. The authors find that those who lived in communities exposed to violence were more impatient, meaning more likely to choose the immediate payoff. This set of results was less conclusive than the other two.

In addition, the authors also find that people exposed to violence were more likely to be involved in community organizations and to farm risky cash crops, bolstering the first two results.


Policy Implications:

The main finding of this paper is that exposure to violence is perhaps not as damaging to people's social capital as was once assumed. The results regarding altruism are particularly encouraging. However, the research also suggests that people who are exposed to violence weigh present interests, relative to the future, higher than do people who were not exposed to violence. This might confound development projects that seek to encourage long-term investment, or might make those which require people to make investments less likely to succeed.