Correlates of Armed Violence Within the Nation-State: Predicting Future Conflict with Varying Data

• Activity Number: 630
• Type: Contributed
• Date/Time: Thursday, August 5, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Government Statistics
• Author(s): Michael Kisielewski*+
• Companies: StatAid
• Keywords: predictive modeling ; armed conflict ; human rights

Abstract

Predicting armed conflict in a country or territory is challenging. Research on the application of correlates of armed conflict to such events remains in its infancy, and uniformly updated or real-time data for analysis are lacking. Based on research conducted for the Small Arms Survey, this paper discusses statistical models for predicting armed conflict (e.g., civil war, armed, gang-related, or domestic violence) that impede the a country's development. Particular emphasis is placed on harmonizing datasets that contain valuable predictors for armed conflict, but which vary over how recent they are. Emphasis is given to adjusting for missing or incomplete data. The author refers to predictive models used in the human rights and international development fields to illustrate the power of such models in studying the well-being of populations experiencing or threatened by armed conflict.