Civil War and Armed Conflict

Human Rights and Democracy

Civil War Mediation (CWM) dataset

Karl DeRouen, Jacob Bercovitch and Paulina Pospieszna

In collaboration with Prof. Karl DeRouen and Prof. Jacob Bercovitch, Dr. Pospieszna has been involved in the creation of Civil War Mediation (CWM) dataset, which provides detailed information about mediation efforts between 1946 and 2004 for each civil war episode identified by the Uppsala Armed Conflict Termination data. The CWM dataset is the first to focus solely on civil war mediation in contrast to other datasets on international crises and provides information about mediation for smaller civil wars with a fatality threshold at least 25. Future plans include the extension of the dataset beyond 2004.

Mediation and Civil Wars Involving Terrorism

Karl DeRouen and Paulina Pospieszna

While mediation in internal conflicts has been studied extensively, few scholars have examined mediation as a possible factor reducing the use of terror in civil war, thus leaving us with underdeveloped answers to important research questions about the impact of mediation on conflict intensity. In collaboration with Prof. Karl DeRouen at the University of Alabama, Dr. Pospieszna aims to fill this gap by investigating the link between mediation and terrorism during the internal conflict. Mediation might decrease the rate of terrorism if it successfully offers alternatives to waging war. However, mediation also can lead to more terror if the process is exploited by the type of peace spoilers. On the other hand, terror affects mediators’ decision to intervene. Mediators may be more likely to interfere in violent wars in order to protect civilians or to prevent the spread of terrorism, but there also is a chance that mediators are discouraged by the severity of the conflict. Some preliminary results show that the terror attracts mediation which in turn increase the use of terror by rebels in civil wars.

Power Sharing in Civil Conflicts

Paulina Pospieszna and Gerald Schneider

In collaboration with Prof. Gerald Schneider at the University of Konstanz, Dr. Pospieszna is involved in two kinds of project exploring power sharing arrangements and mediation in civil conflicts. The first project focuses on power sharing provisions in mediated peace agreements concluded between warring parties, the second one, however, on power sharing institutions that were established in post-conflict countries. Within the first project, Pospieszna and Schneider aim to understand conditions under which some mediated peace agreements lead to a durable peace (long-term success of mediation) and other not, and they take a closer look at the provisions included in agreements. Placing the analysis in the broader framework, preliminary results show that political and territorial provisions for power-sharing in mediated peace agreements translate into the success of mediation in the long term. Within the second project, Pospieszna and Schneider investigate different factors affecting the adoption of power sharing institutions by post-conflict societies. For example, the study explores whether constitutional choices are a consequence of the conflict outcome and mediation efforts. Especially whether the adoption of power sharing institutions such as proportional representation becomes more likely when a war ends in a stalemate, ceasefire or with mediated peace agreements. Paper entitled “Power Sharing Provisions and Long-Term Success of Mediation in Internal Conflicts” was presented at APSA 2011. A second paper “The Illusion of ‘Peace through Power Sharing’: Constitutional Choice in the Shadow of Civil War” was presented at ISA 2012.

What You See Is What You Get: Description and Inference in Micro-Level Conflict Research

Working Paper by Anita Gohdes and Jule Krüger

In 2008, Kalyvas prominently debated the ‘promises and pitfalls of an emerging research program’, the micro-dynamics of civil war. Seven years down the line, we attempt to take stock of what we have learnt about micro-level determinants of violence through disaggregated conflict analysis. We review the existing sub-national studies, comparing commonalities and contradictions of past empirical research findings. As we uncover an obvious empirical puzzle regarding the significance of micro-level determinants of violence, we suggest that these disagreements are heavily contingent on the data-generating process of the data sources used. To illustrate our claim, we specifically look at the dynamics of three recent conflicts – Sierra Leone, Kosovo and Syria. In a first step, we present descriptive evidence that reveals how multiple data sources disagree over magnitudes, and temporal and spatial trends of observed violence that occurred during these armed conflicts. Building on the current literature, we then test simple empirical models that seek to explain these dynamics in Sierra Leone, Kosovo and Syria. We run these models with multiple lists on casualties that were collected by reliable institutions in all conflicts. We find that the directional and substantial effects of our explanatory variables are contingent on the data source used. Our results demonstrate that challenges of ecological fallacy and selection bias deserve increased attention in conflict research.