Search Climate Insights Directory

The authors argue that traditional country-level conflict analysis fails to account for spatial and temporal variations, as conflicts often occur in peripheral regions and are triggered by sudden events. By combining a global logit regression model with a subnational risk index—incorporating population, socioeconomic status, ethno-political exclusion, distance from the capital, and conflict history—the study maps latent conflict risk across twelve Asian countries. The results demonstrate significant internal variation, highlighting high-risk areas in India, Nepal, and the Philippines, while suggesting that irregular regime changes and previous conflict history are powerful predictors of future violence.