A Burundi army soldier raises his gun as he joins demonstrators celebrating what they perceive to be an attempted military coup d’etat, in the capital Bujumbura, Burundi Wednesday, May 13, 2015. (AP Photo/Berthier Mugiraneza)

Recent political events in Burundi are a legacy of a conflict that lasted for almost a decade. It is well known that civil wars affect countries’ political stability. In a recent study, we found that civil war experience can dramatically affect post-conflict democratic politics at the local level. More specifically, we studied new data on armed groups’ location and violent episodes about the civil war that occurred in Italy during the last phase of World War II to learn how local mobilization and violence during the war impacted later elections.

The Italian civil war between the partisan bands and the Nazi-Fascist forces was a major conflict, though somehow ‘lost’ in the general history of World War II. It took place between September 1943 and May 1945, caused 117,000 battle deaths and 10,000 victims of one-sided violence. The Italian case allows us to study democratic elections both before and after the conflict.

Civil wars can lead to political change in conflict-ridden democratizing states in three ways: changing the rules of the game, i.e., affecting institutions; affecting the political preferences of individuals; and influencing the organization and the strategies of the parties that come out of the war. The first sets the limits and the stakes of the nonviolent political competition, while the latter two concern the ways politics actually works on the ground.

The first explanation focuses on changes in people’s behavior, hence a change in the “demand side” of the electoral process. In fact, civil conflicts can influence local politics on the ground affecting people’s political preferences through violence. In the aftermath, when people can freely express their opinions, individuals who suffered from the violence of a warring party usually support the opposite side.

The second explanation focuses on changes in new parties’ skills and behavior as they transform once the conflicts ends. If they accept the rules of the game, the armed actors transform into political parties and continue their political action using non-violent strategies. Thanks to the skills and relationships acquired during the civil war, the parties can actually shape local post-conflict politics.

In fact, these political parties that have emerged victorious from a civil war can count on organized members from the areas where they actively fought. Civil wars often show very different features in a short-range distance; armed groups have more or less control from one area to the other. The predominant group in each location has a clear interest in strengthening the political boundaries of that location, once it transforms into a political party.

The parties emerging from previous armed groups have both a keen interest and the opportunity to shape the local political market in their favor. Former combatants make up the backbone of the party’s local branch and are perfect for mobilizing voters’ support. Thanks to the experience acquired during the war, these “political entrepreneurs” know the geographic, social and economic context perfectly; they know the population and how to contact the voters; they are trained to work together and can count on their reputation and local loyalties. In short, political entrepreneurs deriving from victorious armed groups enjoy a strategic advantage in post-conflict elections because of what happened during the conflict.

In our study we show that former armed groups  — transformed after the conflict into political parties – influence post-conflict elections even more than the memory of past violence.

Since in Italy most of the partisan bands during the civil war were organized by the Communist Party, after the war, left-wing parties enjoyed a remarkable election advantage in the areas where the partisans were active. Matching locations with similar socioeconomic and political features, we found a 12 percent mean difference in left-wing parties’ votes between locations where partisans had mobilized and where they had not. That’s a significant difference in a proportional electoral system.

Moreover, areas that had an average left-wing vote in 1921 will show a 44 percent vote share for the left-wing parties in 1946 if they had partisans, whereas locations with average left-wing voting in 1921 but without a partisan presence will show only 36 percent of votes for the left-wing bloc (See figure). Quite interestingly, there is a stunning difference among areas that were more conservative before the civil war.  Previously conservative areas that did not host partisans gave left-wing parties 18 percent of the vote in 1946, while previously conservative areas that did host the resistance gave left-wing parties on average 37 percent of the vote. That’s a significant 19 percent difference.


Previous research found that experiences during conflict influence citizens’ degree of political participation. Our research reveals that the conflict can also strengthen new parties’ political capacity and organization. Contemporary conflicts may often be fought along ethnic or religious lines, but they are likely to have similar local dynamics that can influence post-conflict electoral politics and, ultimately, peace and stability.

Stefano Costalli is Isaac Newton Fellow and member of the Michael Nicholson Centre for Conflict and Cooperation, Department of Government, University of Essex.

Andrea Ruggeri is Associate Professor of International Relations, Department of Politics and International Relations, University of Oxford and Research Fellow of the Michael Nicholson Centre for Conflict and Cooperation, Department of Government, University of Essex.