Oslo Summer School in Comparative Social Science Studies 2008

The Quantitative Study of Armed Conflict

Lecturers: Professor Scott Gates, Centre for the Study of Civil Wars, PRIO, and;
Senior Researcher Håvard Strand, Centre for the Study of Civil Wars, PRIO, Norway

Main disciplines: Political Science, Economics

Dates: 28 July - 1 August 2008
Course Credits: 10 pts (ECTS)
Limitation: 20 participants

The quantitative study of armed conflict has developed enormously over the last decade, in very many ways. Improvements in measurements and estimation techniques along with a heated theoretical debate over the core causes of conflicts have led us to a situation where a once unified field has become extremely diversified.

The aim of this class is to introduce students to the many different faces of quantitative conflict research. We will show how they differ, why the differences are important and how we still might reach valid conclusions. Two lectures will be held in computer labs, with a hands-on approach to issues of data, data structures, estimation techniques, and interpretation of results.

The requirements for the paper will be to produce an empirical test of a relevant research question. This can either be as a critical replication study or an independent paper.

Highly recommended readings:


Lecture 1: The Quantitative Study of Civil War – an overview.
We will introduce the lecture series and discuss efforts to quantify civil war. We will discuss four different quantifiable aspects of conflict: Onset, Incidence, Duration, and Severity in relation to both time and space.


Lecture 2: Onset of civil war I (Baseline Probabilities of Conflict Onset).
These quantitative studies examine the age-old questions regarding the causes of war. The first lecture will discuss the state of the literature, focusing on the factors that facilitate conflict.



Lecture 3: Onset of civil war II (Shocks and Triggers).
Continuation of the previous topic. Having discussed the more static causes of conflict, we move on to discuss if there exists classes of triggering events.



Lecture 4: Duration of civil war.
A striking feature of civil wars and armed conflicts is that some lasts for few hours while others lasts for decades. Why are some conflicts more durable than others?



Lecture 5: Post-Conflict instability.
Once conflicts have been terminated, the focus must shift to understand how peace can be stabilized. Many of the factors that we have discussed as contributing to the risk of conflict in the first place are reinforced by the presence of a conflict, which leads to the notion of a conflict trap. The nature of the conflict settlement and the presence of third-party security forces are factors that can explain how countries can escape this trap.



Lecture 6: Severity of civil war.
The temporal duration or spatial distribution of a conflict does not necessarily inform us as to how deadly these conflicts are.



Lecture 7: Conflict on both sides of the equation?
In several settings, the presence of armed conflict will be seen as both a cause and as a consequence, which raises the question of simultaneity. In other circumstances, the onset of conflict is a requirement for the observation of the properties of conflict, which raises the issue of selection mechanisms. We will review some examples of these problems and discuss the implications of both the problems and the remedies.



Lecture 8: Geography and conflict.
Just because part of a country is in conflict it does not mean that the whole country is. Like politics, all conflict is local.



Lecture 9: Lab: The importance of data structures.
Comparative research, which civil war studies are a part of, is based on the basic principle of comparing something with something else. The practice of this principle is more difficult than the amount of focus received would indicate. We will discuss various data structures and how these differ in the way they allow us to create a control group, and we will discuss the consequence of these differences.



Lecture 10: Lab: Estimation and Interpretation; Discussion of paper ideas.
A large selection of possible statistical methods is available for conflict research, which on paper seems to fit well with the problems we encounter in this field of research. Yet, many fancy methods should be used with utmost care. A sledgehammer will open a peanut, but it might not be the best tool at hand. We will review some of these, and illustrate the dangers of violating core assumptions.




The Lecturers
Scott Gates is Research Professor and Director of the Center for the Study of Civil War, PRIO and Professor of Political Science, Norwegian University of Science and Technology (NTNU). Scott Gates received his MA and PhD degrees in Political Science from University of Michigan and an MS degree in Applied Economics from the University of Minnesota. From 1989-2003, Gates was at the Department of Political Science at Michigan State University. Gates’ current research interests include: civil war, organization theory, international relations theory, democratization, and governance.

Håvard Strand is PhD in Political Science from the University of Oslo and Senior Researcher at the International Peace Research Institute, Oslo (PRIO). His articles have appeared in American Journal of Political Science, International Studies Perspectives, Journal of Conflict Resolution, and Journal of Peace Research. His main research interests are the study of democracy, civil war, and the relationship between elections and civil violence.


Main page - Current Oslo Summer School Program