DCJ Dataset

The During Conflict Justice (DCJ) dataset was collected with support from the National Science Foundation (SES# 1227985) and the United States Institute of Peace.

An introduction to the data is forthcoming: Cyanne E. Loyle & Helga Malmin Binningsbø. “Justice during armed conflict: A new dataset on government and rebel strategies” Journal of Conflict Resolution. (Forthcoming).


Justice and rule-of-law measures have often been touted as a solution to prevent the resumption of violence following conflict; yet, little work has been undertaken to understand the prevalence of these practices while conflict is ongoing. The DCJ dataset provides an evidentiary basis to determine the conditions under which implementation is more likely, to facilitate the development of these conditions in countries that may benefit from justice measures and to prevent the development of circumstances detracting from these goals. This data greatly improves our understanding of the presence of such processes and expands the current state of research on both the implementation of during-conflict justice and the dynamics of conflict more broadly.


The During Conflict Justice (DCJ) dataset includes 2205 justice processes implemented during 204 internal armed conflicts in 108 different countries between 1946 and 2011. The DCJ dataset includes information on six forms of addressing wrongdoing: trials, truth commissions, reparations, amnesties, purges, and exiles. Each DCJ process is a separate observation in the dataset. A given country-year can have multiple observations as more than one DCJ processes can be implemented in that context. 76% of the countries that experienced conflict used at least one form of DCJ.

A trial is defined as the formal examination of alleged wrongdoing through judicial proceedings within a (quasi-) legal structure. Trials can include domestic prosecutions, military courts (by either the government or rebel group), ad hoc tribunals, or international prosecutions.

Truth commissions are defined as temporary investigative bodies that focus on a pattern of abuse over a particular period of time. This includes formal truth commissions as well as commissions of inquiry.

Reparations are compensation given to an individual or group who was harmed in some way during the conflict. Reparations are broadly defined in the DCJ dataset, including monetary reparations, property transfer, skills training, education and employment assistance, and/or compensation to war-affected communities as a whole.

Amnesty is defined as a promise (or in some cases formal legislation) on the part of a party to the conflict to not prosecute or punish (alleged) violators of human rights or domestic laws now or in the future.

A purge is defined as the act of removing politicians, members of the armed forces or judiciary, or other members of society for their (alleged) collaboration with or participation in a conflict and limiting their influence accordingly.

Exile or expulsion is defined as a period of forced absence from the country of interest.

Below we present some snapshots of the data.

Figure 1 shows the total number of each DCJ type included in the dataset separated between the parties to the armed conflict that initiated the DCJ process (the “sender”). Note- "Other" category includes international organizations (such as the UN), NGOs, and countries not party to the conflict.

Regime type

Figure 2 shows the relationship between regime type and those governments and challengers which choose to pursue DCJ versus those which abstain.

Conflict intensity

Figure 3 shows the relationship between DCJ and the overall intensity of the conflict. DCJ usage for both governments and rebels is more likely in medium-intensity conflicts of 1,000 to 4,999 battle deaths. While the pattern is starker for governments, DCJ is less likely to be used at both extremes, e.g., low-intensity and high-intensity conflict years.

Balance of power

Figure 4 shows the relationship between the strength of the rebels (relative to the government) and DCJ processes.