Counting carrots and sticks : a data analysis of insurgent activity in Afghanistan from 2008-2014

Abstract

After nearly two decades of conducting history’s most expensive counterinsurgency campaign, the United States left Afghanistan in the wake of the government’s collapse and the fall of Kabul. This project seeks to understand the underlying reasons behind the failure to stabilize and rebuild Afghanistan, by identifying what factors affect levels of insurgent activity. To that end I apply a multi-level mixed effects model for the testing of counterinsurgency hypotheses, with a specific look into the US engagement in Afghanistan from 2008 to 2014. The analysis involves security, economy, and governance variables, and over 400,000 individual incidents of insurgent activity. I further test my results by conducting a qualitative comparison of a select province pair. My findings demonstrate that structural factors impact insurgent activity in the medium to long term, while strategic and tactical adaptation have a greater effect on insurgent activity in the short term. I also propose a new theoretical model on the structures and elements of insurgencies. It describes how such elements are fundamentally hierarchical within a porous structure subject to feedback loops. The model highlights the importance of infrastructure for an insurgency. It illustrates the primacy of security as the foundation upon which infrastructure, economy and governance factors must be built. It also stresses that the goal for insurgents remains the control or denial of popular support at the uppermost, governance level. Finally, my model suggests a counterinsurgency approach based on this structure of insurgency.