Thus far, this book has examined the relationship between relative power and identity cleavages in terms of alliance formation and group fractionalization in the Afghan Jihad and the intra-mujahedin war. Actors have included alliances, warring groups, and their subgroups. The book's decision on the relevant level of analysis was made based on an argument about minimum effective actors – an argument that a unit of analysis aggregated at the levels of warring groups and their subgroups would adequately capture alliance shifts and group fractionalization throughout the conflict's trajectory. That approach is fully consistent with the general historiography on the Afghan Jihad and intra-mujahedin war, which has been almost exclusively presented on the level of mujahedin parties and their elites.

The intent of this chapter is to see whether the anticipated behavior, as reflected in the theory, also holds for a lower level of analysis than the subgroup. As referenced in the first chapter of this book, there is a strong trend in the civil war literature of focusing either at the macro (state or rebel group) or micro (individual or village) level with little attempt to link the two. Having examined the macro-level of warring groups and then the lower level of subgroups, the goal of this chapter is to move another level lower and to explore whether local level behavior indeed matches the macro-level predictions of the theory. To the extent that the theoretical predictions hold at multiple levels of analysis, we have greater confidence in the theory's explanatory power and specified mechanisms. In the Afghan context, the relevant lower level is arguably that of the wartime commander.