Cordesman: Iraqi civil war

All the dirt thrown in the air over the Lancet report on Iraqi civilian casualties is really pointless.  Whether the actual number is 200,000 or 600,000 does not really affect the broader point that Iraq has become a deeply insecure place where a lot of civilians live in fear.  The violence has morphed from an anti-occupation insurgency into a combined insurgency and ethnic civil war.  In a report released yesterday titled "Is there a civil war in Iraq?", CSIS analyst Anthony Cordesman gives a good sense of the scope of this violence and its political meaning.

Cordesman reproduced several sobering graphics from the August 2006 Department of Defense briefing of Congress - not from nutty lefties or partisan MSM filters or whatever is the insult du jour.   First, "sectarian incidents and violence, May 2005-July 2006":

Cordesman clarifies that this "is based on a massive undercount of  actual violence, since it does not [include] many, if not most, low level incidents where the source of the attack cannot be confirmed, and makes no effort to estimate  'softer' forms of ethnic and sectarian violence like intimidation and non-violent ethnic cleansing.  While not meaningless, this count is so narrowly defined as to grossly understate the level of civil conflict in Iraq."   In addition to the dramatically increasing death toll, Cordesman points to what political scientists quaintly call "ethnic unmixing":  internal displacement as individuals and families flee to "safer" (i.e. ethnically more homogenous) neighborhoods.   Cordesman cites a June 2006 UN estimate of 150,000 Iraqis who had fled their homes due to direct or indirect threats to them or their families. 

Cordesman also reproduces another DOD chart, this one based on series of State Department polls asking Iraqis whether they are concerned that a civil war might break out in Iraq.

The chart shows that from the pre-sovereignty period (April-June 2004) to the establishment of the current Iraqi government (May-August 2006), the percentage of Iraqis concerned about civil war breaking out has increased from just over 40% to just under 80%.   (The surveys didn't ask whether a civil war was already occuring.) 

None of this should come as a surprise to anyone paying attention to Iraq, or who has just seen the endless parade of headlines of sectarian bloodshed.   But it is worth reiterating lest it get lost in pointless partisan squabbling over the Lancet study.  "This is a civil war," concludes Cordesman.    That - not whether the US will spark a civil war by withdrawing -  should be the starting point for political debate.

UPDATE:  Cordesman messed up his figure 1, transposing a different chart about attacks.  Here (thanks to WH) is the correct figure of polling data on expectations of civil war:

Comments

[Craig]

I don't think it's a civil war, though. I think it's ethnic cleansing. Or, at least, Sectarian cleansing. The "insurgency" was by defintion a civil war, from the moment Iraq regained sovereignty and the occupation ended. But as you point out, the insurgency period is over.

[Andy Vance]

I agree that it's pointless to squabble over point estimates. But Cordesman's unqualified statement that "the results of the Lancet study--present serious credibility problems" is deeply unhelpful.

The study is what it is. It uses accepted methodology and has a wide error band because of limitations imposed by the security situation. That is not a "credibility problem," and it's wrong and irresponsible for Cordesman to cast such aspersions.

[wh]

The “Measuring Stability and Security in Iraq” reports Cordesman refers to are worth reading...one of the things congress has done to promote a bit of accountability is to require the reports and make them public.

http://www.defenselink.mil/home/features/Iraq_Reports/Index.html

[Andy Vance]

I failed to mention the other gem Cordesman gave the AP: "This is not analysis, this is politics."

[Rodger]

Coincidentally, I was looking at that same Cordesman graphic earlier today. I wanted to reproduce it for a talk I was giving, but ultimately didn't.

In any case, in the text he notes the 10 to 12 fold increase in sectarian violence this year. And that's DoD data.

Fareed Zakaria, btw, is saying that the violence in Iraq is both "civil war" and "ethnic cleansing." The policy has simply failed.

[WH]

is that really the right figure 1? it's labeled with the number of attacks not the percentage of Iraqis concerned about civil war breaking out

[aardvark]

WH- I think it's a mislabled figure, the more I look at it the stranger it is. I think it's actually transposed from Cordesman's Figure 2, which is labeled attacks. In the text, he does talk about the State Dept survey data that I mentioned, which shows a big increase in expectations of civil war as I said in the text. But that figure looks screwy.