Early Warning?

A Review of Conflict Prediction Models and Systems

In attempting to redress this weakness in international interventions, practitioners and scholars have developed a variety of early warning models. These monitor longer-term, society-wide, structural variables to help forecast the complex dynamics that can result in conflict or instability, with the goal of providing policymakers with more time to take preventive actions before conflicts erupt.

National governments, multilateral organizations, universities, NGOs, think tanks, and private enterprises have created several dozen models designed to predict conflict and instability throughout the world. With a diverse set of methodologies, ranging from qualitative studies to complex regression and systems dynamics analyses, most identify factors, causes, or variables that are linked to the occurrence of conflict and instability.

For this report, the PCR Project at CSIS surveyed 30 publicly-available models and strategies that share a common objective—predicting conflict and instability. All are general in scope. The survey includes a cross-section of models from different types of sources—although not classified military or intelligence models—and attempts to digest the information into an accessible form for a broader audience.

Download Database of Early Warning Indicators (Excel file)

Download Large-scale Matrix of Indicators



Rick Barton, Karin von Hippel, Sabina Sequeira, and Mark Irvine