Insurgent/militia control in Iraq

**mweaver** · 12-12-2008

My name is Michael Weaver, and I am political science undergraduate the University of Chicago. I am currently working on a project in which I hope to investigate whether theories about the use violence in non-ethnic civil wars also hold in purportedly "ethnic" civil wars. In particular, I hope to evaluate these hypotheses by looking at Iraq. One of the theories I would like to test is Stathis Kalyvas's argument about how insurgent and incumbent control determine the use of selective or indiscriminate violence. I plan to test this hypothesis by examining variation in violence across both neighborhoods in Baghdad and across cities, towns, and villages throughout Iraq between 2005 and 2007. Unfortunately, I am unaware of any systematic collection of data on the degree to which militia or insurgent groups exercised control over territory (either at the neighborhood or town/village level) in Iraq.

Does anyone know if such systematic data exists, and if so, who I should contact to either obtain it, or attempt to obtain it (seeing as it is highly probable such information would not be publicly available)?

Assuming either the non-existence or unavailability of such data on control, I plan to collect data myself. I am currently using media reports from both major newspapers and Iraqslogger.com to come up with a rough coding scheme for control over neighborhoods in Baghdad. Are there any other resources that anyone would recommend for information on the degree to which militias/insurgent groups exercised control over particular neighborhoods in Baghdad or towns/villages/regions across Iraq?

(Note: I do have intermediate Arabic reading skills and potential access to someone familiar with Iraqi dialect)

On a related note: I am also interested in obtaining access to disaggregated figures on both violent attacks and civilian deaths in Iraq from the SIGACTS database. I initially contacted the MNF-Iraq press desk seeking information about what was available and what was not. After 2 months of back and forth I was told that they did not have the disaggregated data I was looking for. If anyone has suggestions about who I should contact about the possibility of obtaining this data, I would greatly appreciate any leads, particularly if I could be directed to people in the military involved in the data collection operation or in data analysis.

**Shek** · 12-12-2008

Originally Posted by mweaver

My name is Michael Weaver, and I am political science undergraduate the University of Chicago. I am currently working on a project in which I hope to investigate whether theories about the use violence in non-ethnic civil wars also hold in purportedly "ethnic" civil wars. In particular, I hope to evaluate these hypotheses by looking at Iraq. One of the theories I would like to test is Stathis Kalyvas's argument about how insurgent and incumbent control determine the use of selective or indiscriminate violence. I plan to test this hypothesis by examining variation in violence across both neighborhoods in Baghdad and across cities, towns, and villages throughout Iraq between 2005 and 2007. Unfortunately, I am unaware of any systematic collection of data on the degree to which militia or insurgent groups exercised control over territory (either at the neighborhood or town/village level) in Iraq.

Does anyone know if such systematic data exists, and if so, who I should contact to either obtain it, or attempt to obtain it (seeing as it is highly probable such information would not be publicly available)?

Assuming either the non-existence or unavailability of such data on control, I plan to collect data myself. I am currently using media reports from both major newspapers and Iraqslogger.com to come up with a rough coding scheme for control over neighborhoods in Baghdad. Are there any other resources that anyone would recommend for information on the degree to which militias/insurgent groups exercised control over particular neighborhoods in Baghdad or towns/villages/regions across Iraq?

(Note: I do have intermediate Arabic reading skills and potential access to someone familiar with Iraqi dialect)

On a related note: I am also interested in obtaining access to disaggregated figures on both violent attacks and civilian deaths in Iraq from the SIGACTS database. I initially contacted the MNF-Iraq press desk seeking information about what was available and what was not. After 2 months of back and forth I was told that they did not have the disaggregated data I was looking for. If anyone has suggestions about who I should contact about the possibility of obtaining this data, I would greatly appreciate any leads, particularly if I could be directed to people in the military involved in the data collection operation or in data analysis.

Mike,

You're not going to be able to get the disaggregated SIGACTS database, as that will be classified. I had a cadet last year who was looking to do a senior thesis using the same data your looking for and it's the same story. The best you're going to be able to do is code the info yourself, trying to piece together the info from open source reports and recognizing the large measurement error that will exist (and therefore the limitations of the inferences that you could draw).

I've got a colleague who's co-authored a paper that looks at price dispersion and violence levels across the governates, but I don't think she had data down to neighborhood level in Baghdad. I'm not sure whether she can release the data since she got the price data through the firm that tracks CPI in Iraq, but if you're interested in it, PM or email me and I can ask her (and do so quick so I'll have a chance to face-to-face with her prior to the holiday break).

**Cavguy** · 12-12-2008

I'll second Shek that these studies will be difficult for awhile. I get pinged about once a week for that kind of data for someone doing a master's/doctoral thesis. Every researcher wants into the CIDNE databases where all this is stored - but the event details remain classified because often they detail our methods, tactics, vulnerabilities, and how certain casualties happened - and these vulnerabilities are still real threats.

So all you'll get from CENTCOM is aggregate info. Frustrating for the researcher if they want to publicize the results (I will say a number of people are studying the results, but not for public release), but necessary for the safety of our soldiers in ongoing ops.

**MikeF** · 12-12-2008

I'd recommend starting with Kimberly Kagan's Iraq Reports I-X. The first one is linked here http://www.weeklystandard.com/Conten...3/343fthms.asp. Not only are her reports fairly thorough, but they are fairly accurate as well (IMO).

Mark Kukis from Time Magazine wrote some very accurate articles from 2005-2007 ranging from Basra to Fallujah to Diyala.

If you want to read of an account of Al Qaeda using Mao's techniques and tactics to occupy and establish the Islamic State of Iraq, you can read my article The Break Point published for SWJ. It parallels Bob Andrews The Village War: Vietnamese Communist Revolutionary Activities in Dinh Tuong Province, 1960-1964.

Regardless of any data collected, I'd urge you to be highly critical of the results. I have yet to see any game theory model or systems data accurately reflect the actions on the ground. In our desire to measure activities to define metrics of success, we often fail to be critical of the source and motives.

With that said, these are good start points for you. Good luck with your research.

v/r

Mike

**mweaver** · 12-12-2008

Thanks for your resource recommendations. I've seen both Kimberly Kagan's work (and the rest of the resources on her website) and a few articles by Mark Kukis, but I hadn't thought of using them for this part of my project.

I also read your article, and I found it very helpful. I'd be interested, if possible, in asking you a few questions via email about sectarian violence in Zaganiyah. The rich, detailed account that seems possible in this case provides a useful qualitative way in which to evaluate (as you mentioned in your post, it is important to ensure that data and theory reflect what is actually happening) the more generalized statistical models I intend to do.