December 17, 2010

Afghanistan Trip Report, Part IV: Graphing the Conflict

Over the past several days, I have shared both several observations from my most recent trip to Afghanistan as well as a few things I think that policy makers in Washington can do to help the war effort. Today, I am going to return to a few themes I have dealt with in the media and on the blog but in a different format. A picture is worth a thousand words, and maybe these graphic representations, taken from my field notebook, can help explain how I view the conflict.

Most of these graphs were drawn in the course of a conversation I had with Col. Joe Felter, who is one of my heroes. Joe led a platoon from 3rd Rangers in Panama before becomming a Special Forces officer in 1st Group and earning his doctorate in political science from Stanford. (Here (.pdf) is an example of Joe being really smart.)

(Fig. 1) I drew this graph, which will be familiar to many, to illustrate a dynamic that frequently takes place in international interventions. The x axis is time, while the "$" on the y axis represents the financial committment of the international community, and the "C" represents the capacity of the host nation to effectively absorb and administer international aid. As we progress along the x axis, funding drops off while capacity increases, leaving a shortfall toward the tail end of the intervention. The shaded part, meanwhile, demonstrates where corruption, waste and fraud is likely to take place. In Afghanistan, we are in the shaded section of the graph at the moment, which is one of the reasons I wish we could decrease international aid and bank some of it for the future.

(Fig. 2) This represents the "normal" theory for violence in an insurgency. The x axis maps control, and the y axis represents levels of violence. When an area is under the control of either the government or the insurgency, there is very little violence. It's when an area is contested that we have violence. This is one of the reasons why violence can be a poor metric for success or failure. Kandahar, Helmand and Kunar Provinces account for 65% of the violence in Afghanistan, but the former two provinces have arguably been in the process of falling toward government control. Ghazni Province, by contrast, is not that violent these days, relatively speaking, but I think that is in part due to the fact that it is, based on reports, in large part under Taliban control. (Mea culpa, I have admittedly used the lack of violence in northern Afghanistan as part of an argument that things might not be as bad there as reports suggest, but of course there are other things, such as a large Tajik population, that serve to limit Taliban gains. And yes, I do know the Taliban has enlisted support from some Tajiks, but this is limited as far as I can tell. My thoughts on northern Afghanistan are probably a whole 'nother blog post.)

(Fig. 3) This is my big worry for the "normal" theory as it applies to Afghanistan. What if, even after getting an area under the control of the government, we never really pacify it because the behavior of the government itself is in part driving the conflict? Insurgent sanctuaries in Pakistan might also distort the normal theory by thwarting security gains. (I should also say something about the fact that the normal theory assumes the conflict is binary and that the insurgency and government are both unitary actors, which is not the case in Afghanistan.)

(Fig. 4) Ignore this. Joe and I were merely discussing some stuff in Iraq that has only minimal relevance to Afghanistan. Oh, and those big blue blocks are where I censored some of my notes, which includes names and phone numbers and such.

Afghanistan Notebook