The Dynamic Conflict Risk Model (DCRM) assesses the risk of internal conflict at the subnational level in the next six months.
Using up to 30 predictor variables, the model estimates the likelihood and potential severity of internal conflict across four categories.
The DCRM covers 140 countries across Africa, Asia, Eastern Europe, Latin America, and the Caribbean and is updated on a monthly basis.
To access the latest DCRM results, click here.
For guidance on exploring and navigating these visualisations, please visit the Help & Learn section of the Science4Peace Portal.
The DCRM provides an overall risk score for internal conflict (Any conflict), which includes three sub-categories:
- State-based conflict (SBC): Armed conflict involving at least one state government,
focused on government power or territorial disputes, resulting in a minimum of 25 battle-related deaths within a calendar year.
- Non-state conflict (NSC): Armed conflict between groups unaffiliated with a state government.
This includes clashes between rebel groups, militias, political party supporters, and ethnic, religious, or social groups,
also resulting in at least 25 battle-related deaths annually.
- One-sided violence (OSV): The direct and deliberate killing of civilians by either a state government or a
non-state armed group, leading to at least 25 deaths in a calendar year.
These conflict definitions are aligned with the Uppsala Conflict Data Program (UCDP).
The DCRM’s primary output is a combined conflict risk score ranging from 0 to 10.
This score reflects both the likelihood and severity of conflict in a given area over the next six months and is updated monthly.
The DCRM uses separate models for each conflict type and world region, incorporating region- and outcome-specific input variables and parameters.
Unlike the GCRI, which covers conflicts with at least 25 battle-related deaths annually, the DCRM risk
scores also cover low-intensity conflicts with fewer fatalities.
The DCRM relies on two types of input variables to forecast conflict:
- Event data: This includes past conflict events, protests, and related incidents within each administrative unit and its neighbouring areas.
- Structural variables: These encompass local demographic and geographical data, along with political and economic variables at the national level.
The DCRM is developed by the Joint Research Centre in collaboration with the Conflict Prevention and Mediation Support Division within the
EU External Action Service (EEAS),with funding from the Service for Foreign Policy Instruments (FPI).
The DCRM is continually evolving, with ongoing efforts to enhance modelling techniques,
revise input variable selection as new data sources become available, and improve the explainability of model outputs.