Wednesday, July 18, 2012

Critique of “Testing The Surge,” Article Misses Major Factors That Reduced Violence In Iraq

The summer 2012 edition of International Security had an article entitled “Testing the Surge, Why Did Violence Decline in Iraq in 2007?” by Stephen Biddle, Jeffrey Friedman, and Jacob Shapiro. The article asked which events caused the end of Iraq’s civil war. It critiqued the ideas that it was the cleansing of Baghdad, the Anbar Awakening or the Surge alone. Instead, it argued that it was a combination of the Surge troops and new counterinsurgency tactics along with the Anbar Awakening and the Sons of Iraq program. The piece had several problems. First, the authors misconstrued the nature of the fighting in Baghdad as an unrelenting battle for territory when it was more about local groups trying to impose their will on each other. Second, it claimed that security did not improve until mid-2007. That ignored the fact that while attacks increased when the Surge started Iraqi deaths had already peaked in December 2006, and declined after that showing that there was another dynamic going on besides just the troop increase. Finally, it failed to consider the impact of Sunni militants feeling that the Shiite militias had beaten them as a turning point in the war. Overall, the main point of “Testing the Surge” has been made before, and there were simply too many holes in the argument for it to stand up.

The first thing that “Testing the Surge” did was attempted to disprove the counter theses starting with the ethnic cleansing argument. As the civil war took off, Baghdad became the center of the fighting in Iraq, as it was the seat of government, and had the largest population of the country’s 18 provinces. In 2003, the capital itself was a largely mixed city of Sunnis and Shiites with some Christian areas as well. That co-mingling of communities led to insecurity as each group came to feel threatened by the other. Sunni and Shiite militants began fighting for control, which evolved into pushing their rivals out of neighborhoods. The article believed that the ethnic cleansing argument holds that as Baghdad became more homogenous the warring parties felt more security, and violence then went down. The authors felt that was false for two reasons. First, they wrote that for the cleansing theory to be true Baghdad had to be the main area of fighting to explain the decrease in violence. Second, attaks in other areas therefore had to be secondary. Biddle, Friedman, and Shapiro went through Significant Activities (SIGACTs) collected by the U.S.-led Coalition from February 2004 to December 2008. From 2005-2006 most of those occurred in Anbar province, meaning that the fight for Baghdad was not the center of the conflict. Within the capital, most SIGACTs occurred in Sunni areas, not mixed ones, and incidents went down in the latter before it did in the former. Also, violence did not decline in neighborhoods when they became more homogenous. This debunking had several problems. First, the article claimed that Anbar was the main battleground in Iraq from 2005-2006, and no sectarian cleansing happened there, because it was a Sunni province. That’s not true. There were several hundred Shiite and Kurdish families in the governorate that were forced out, and Anbar also received several hundred Sunni families that were pushed out of Baghdad, so it was not untouched by the civil war. In October 2008 for instance, the government was trying to close down a refugee camp in Najaf that housed 230 Shiite families, mostly from Anbar. There were also Shiite families from Anbar that fled to Sadr City, and of 8,623 displaced families interviewed by the International Organization for Migration in Anbar in October, 76.9% were from Baghdad. (1) Finally, there were Kurds who resided in northern Anbar that were forced out as well. (2) Second, in dismissing the sectarian cleansing argument, Biddle et al. seemed to mischaracterize the nature and intent of the warring parties. For one, the article said that there were more security incidents in Baghdad’s Sunni areas than the mixed ones. How does that disprove the thesis? One of the first neighborhoods to see mass forced removals was the Sunni majority neighborhood of Hurriya in northwestern Baghdad, which was conquered by 2006. It is only because the article seemed to think that mixed areas would be the central battlefront that saying Sunni areas had more fighting could be an issue. Biddle has also been arguing for years that the sectarian conflict in the capital was for complete control. (3) When one area was cleansed, the fighting simply moved to the next, because the goal was to completely force out the other from the entire city. He’s written that since there were still Sunni areas of Baghdad left such as Adhamiya on the northeastern side of the Tigris River, yet violence fell in 2007 that the sectarian cleansing thesis must be false. This misses the fact that the Mahdi Army, which was the main militia fighting at this time was not an organized force, but rather a collection of local militias operating with only minimal control by Moqtada al-Sadr. The article itself even mentions the factionalism that emerged within Sadr’s followers. There was no plan therefore to completely conquer the city since each local unit was largely acting upon its own initiative. One could have been completely satisfied with taking a few neighboring areas rather than continue fighting until it controlled every single part of Baghdad. Finally, the article failed to contemplate the possibility that the civil war was not about controlling territory, but perhaps an attempt to destroy the Iraqi state and dominating rival groups. Al Qaeda in Iraq wanted to start a sectarian conflict not only because they believed Shiites were apostates, but starting a civil conflict could bring down the entire country, and thus defeat the Americans’ plans. After the 2006 bombing of the shrine in Samarra, Salahaddin, the Shiite militias felt unrestrained in their desire to strike back. Taking territory was a way to punish and impose their will on their enemy, and didn’t appear to be an end in itself as Biddle characterized it. By 2007 it was apparent that the Shiites were winning with their superior numbers, implicit support from the government, and the dramatic demographic changes in the capital. That led the Sunnis to begin to give up, which was shown by the fact that the majority of the insurgency was willing to switch sides and work with the Americans. The article even mentioned this fact, and Biddle acknowledged it as well in earlier pieces. Altogether that would point to the sectarian cleansing thesis as being far more important than Biddle and company gave it.
In 2003 Baghdad was mostly made up of mixed Sunni and Shiite neighborhoods (yellow areas) (Columbia Univ.)
By 2007 the capital had turned into a majority Shiite city (green areas) because of the sectarian civil war (Columbia Univ.)
Next, “Testing the Surge” dealt with the Anbar Awakening and Sons of Iraq. It starts by claiming that the reason why Sunnis were willing to work with the Americans was because of money. The U.S. had been offering money to Sunni groups for years, and had not been able to make any substantial headway, so this was an odd thing to say. This ignored plenty of other literature on the topic that offered a variety of other reasons such as being fed up with the extremist policies of Al Qaeda in Iraq, and realizing that they were losing to the Shiites. The article then went through several examples of tribes in Anbar trying to rise up against the insurgency, but failing to prove that the Awakening alone did not turn the tide in the fighting. Those included the Albu Nimr Tribe in early 2004, the Albu Mahal in the spring of 2005, American Special Forces attempting to organize tribes into the Desert Protectors in the fall of 2005, and the Anbar People’s Council formed by the Albu Fahd and former members of the 1920 Revolution Brigades insurgent group in late 2005. In all those cases, the attempts to switch sides failed, because the Americans were unwilling or incapable at the time to provide any kind of support or protection. This was a good counter-argument, and one of the strongest parts of “Testing the Surge.” It showed that some Iraqis were becoming tired of the Islamists, and willing to fight them, but were simply outnumbered and out gunned at the time to be successful. The Awakening that would be formed in 2006 therefore would have likely gone down the same path if the United States had not changed course, and decide to back it.
An April 2009 map showing the number of Awakening and Sons of Iraq fighters. "Testing The Surge" argues that the Awakening and the spread of the Sons of Iraq program would not have been possible without the Surge (Special Inspector General for Iraq Reconstruction)
The last thesis that the authors tried to deal with was whether the Surge alone reduced violence in Iraq. That argument believes that the 30,000 additional troops that entered Iraq beginning in February 2007, the offensives they were able to launch, the use of new counterinsurgency tactics that emphasized protecting the population, and the leadership of General David Petraeus were able to turn the tide. Here, the article used a statistical analysis to show a correlation between the formation of the Sons of Iraq (SOI) and a general decline in violence to prove that the Surge alone would not have been enough. It found that in 24 of 38 areas where an SOI unit was put together security incidents declined. Several others have argued that the extra troops alone were not enough. The only problem with this section was that it missed the fact that a population centric strategy using tactics like security stations in Baghdad actually started at the end of 2006 with General William Casey before the Surge started.

The last section of the article went over the authors’ idea that it was a mix of the Surge and the Anbar Awakening and Sons of Iraq program that reduced violence in Iraq. They argue that without the Surge, the Awakening would not have spread across the country with the SOI. That deprived the insurgents of numbers as the majority of militants switched sides by the end of 2007, provided intelligence on their activities, and changed the security environment. With the reduced Sunni threat, the Shiites were more willing to end the fighting, and allowed the Americans to change their focus to the militias. The growing criminality of many of the militiamen, the power struggle between the Sadrists and the Supreme Council, the breaking apart of the Mahdi Army into smaller factions also meant that they were losing support. The new Surge tactics helped reduced violence as well. The authors write that 30,000 extra troops on top of 155,000 Coalition forces and 323,000 Iraqi ones would not have been enough to change the environment in Iraq alone. They believed that violence would have gone down, but then erupted again when the Surge was over in 2008 if it hadn’t encouraged the creation of the Sons of Iraq. They point to mid-2007 as the turning point in the conflict when they claimed violence started going down. There is no denying that the Surge and turning of the insurgency were important events in the Iraq War. For most of 2003-2006, the U.S. was more focused upon withdrawing from Iraq rather than winning. It seemingly had no answers when the conflict went from an insurgency to a civil war. The new counterinsurgency tactics and extra troops were definitely needed. More importantly the realization that many Sunnis were not willing to keep up the fight anymore might have been more significant, because it allowed the U.S. to isolate the militants, and gain support from the population. That didn’t mean there weren’t problems with this part of the article. The main one was that it missed an important element of the war. Attacks increased throughout 2007 as U.S. troops launched more offensives and spread out amongst the community. In the first quarter of 2007, there were an average of 1,815.3 attacks per month according to the United Nations. That went up to 2,525.6 in the second quarter as the Surge troops arrived. They then went down to 2.254.2 in the 3rd quarter, and 2,595.3 in the fourth. By 2008, the average number of attacks dipped to below 2,000 with 1,937.3 in the 1st quarter, 1,788.3 in the second, 1,494.0 in the third, and 1,003.0 in the fourth. What the article missed was that while security incidents increased in 2007, Iraqi deaths actually declined. The number of Iraqi casualties increased through 2006 as the civil war grew in intensity, but then peaked at the end of the year. Iraq Body Count had an average of 1,666.0 in the first quarter of 2006, going to 2,134.3 in the second, 2,836.3 in the third, before reaching its highest point ever at 2,921.6 in the last. The number of Iraqis killed then steadily went down throughout 2007 with an average of 2,719.0 in the first quarter, 2,479.6 in the second, 2,098.3 in the third, and 1,082.3 in the fourth. This goes a long way to undermine a main point of the article that the Surge turned things around in the middle of 2007. This was especially strange since the article did use Iraq Body Count as a source, but it seemed to focus upon the attack statistics the organization collected rather than the number of Iraqi deaths. If the number of Iraqis killed was actually consistently going down throughout 2007, even before the Surge really got under way, it would point to something else such as the Shiites winning the civil war, and the Sunnis beginning to give up as perhaps a leading causal factor for changing the nature of the conflict.

“Testing the Surge” presented an argument that has been made before that an increase in troops and new tactics alone were not enough to change the war in Iraq. A mix of the Surge along with Sunni willingness to switch sides and help the Americans seemed like a sound thesis. The real problem with the article was its dismissal of the sectarian cleansing of Baghdad. Since Iraqi deaths were going down even before the Surge started meant that something else was playing a leading element in the conflict. Control of the capital started in earnest in 2005, and took off in 2006 after the bombing in Samarra. Shiites went on the offensives afterward, and had mostly won by the middle of 2007. That was likely a leading cause of Sunnis willing to work with the Americans, because they could see that they were not going to beat the Shiite militias and switching sides might have been their only way to escape annihilation. That was another element that the article missed. Perhaps the authors’ emphasis upon American actions and sources, while largely ignoring the Iraqi perspective could explain this oversight. In the end, Biddle, Friedman, and Shapiro did not add that much to the Surge debate, and by dismissing the sectarian cleansing argument hurt their cause more than helped it.


1. International Organization for Migration, “IOM Emergency Needs Assessments,” 10/1/08

2. International Organization for Migration, “Anbar, Baghdad & Diyala, Governorate Profiles,” December 2008

3. Biddle, Stephen, “Iraq after the Surge,” Committee on Armed Services Oversight and Investigations Subcommittee, United States House of Representatives, 1/23/08


Biddle, Stephen, “Iraq after the Surge,” Committee on Armed Services Oversight and Investigations Subcommittee, United States House of Representatives, 1/23/08

Biddle, Stephen, Friedman, Jeffrey, and Shapiro, Jacob, “Testing the Surge, Why Did Violence Decline in Iraq in 2007?” International Security, Summer 2012

Biddle, Stephen McMahon, Robert, “Stephen Biddle: Teleconference with CFR,” Council on Foreign Relations, 11/20/07

Fadel, Leila, “Displaced Iraqis, now told to go home, fear for their lives,” McClatchy Newspapers, 10/5/08

Gwertzman, Bernard, “Biddle: Security, Political Improvements Seen in Iraq in Recent Months,” Council on Foreign Relations, 11/15/07

Inter-Agency Information and Analysis Unit, “Attacks and Casualties in Iraq”

International Organization for Migration, “Anbar, Baghdad & Diyala, Governorate Profiles,” December 2008
- “IOM Emergency Needs Assessments,” 10/1/08

Iraq Body Count

Ollivant, Douglas, “Countering the New Orthodoxy,” New America Foundation, June 2011
- "Reassessing The U.S. Surge, And Recognizing Iraqi Agency In Ending The 2005-2008 Civil War In Iraq, An Interview With Douglas Ollivant," Musings On Iraq, 2/27/12

PBS Frontline, “Gen. George Casey,” End Game, 6/19/07
- “Interview Col. Kalev Sepp (Ret.),” End Game, 6/19/07
- “Interview Col. William Hix,” End Game, 6/19/07
- “Interview Frederick Kagan,” End Game, 6/19/07
- “Interview Gen. Jack Keane (Ret.),” End Game, 6/19/07
- “Interview Lt. Col. Andrew Krepinevich (Ret.),” End Game, 6/19/07
- “Interview Michael Gordon,” End Game, 6/19/07
- “Interview Philip Zelikow,” End Game, 6/19/07
- “Interview Thomas Ricks,” End Game, 6/19/07
- “Timeline Struggling to Find a Strategy for Success,” End Game, 6/19/07

al-Shara, Basim, “Anbar’s Displaced Stay Away,” Institute for War & Peace Reporting, 10/8/08

Thiel, Joshua, “The Statistical Irrelevance of American SIGACT Data: Iraq Surge Analysis Reveals Reality,” Small Wars Journal, 4/12/11


DJ Elliott said...

Every report I read miss the point and objectives of the surge because they leave out the Iraqi component.

We sent 5 Brigades, at the same the IA formed 20 new Brigades that those 5 US Brigades partnered with to train them.

When the US surge brigades left - they left 20 additional trained IA Bdes that haven't gone away.

When the surge was announced, the IA had 10 Divisions, 36 Brigades, 112 Battalions. Today the IA has 14 Divisions, 56 Brigades, 182 Battalions - almost all the extra was formed and trained by the US surge brigades.

There were corresponding increases in the Iraqi MoI forces as well.

The real purpose of the surge was to expand the permanent forces in Iraq and it was a success. It had its problems - at least 5 Bdes of the IA had to be scrapped and rebuilt during that period, the FP [NP] had to be completely retrained, etc.

This is what all the critics ignore...

Joel Wing said...


The original "Testing The Surge" article mentioned that there was a large increase in Iraqi forces by the time the Surge was over. I'd have to check with the piece, but I believe they said there were around 30,000 additional Iraqi forces by the end.

Joel Wing said...


As for your main point about the purpose of the Surge, I would say that it had many and those changed as well over time. The Surge did not start for example with the idea of starting the Sons of Iraq program to turn the insurgents, but it became one. Creating more Iraqi forces was another goal, but I personally don't think it was the main one.

DJ Elliott said...

Actually, Maliki announced those increases as the Iraqi part of the surge - it was a primary from 2 months before the first US surge Bde arrived...

Were there additional taskers that popped up - yes.

3,000 per Bde, plus DivTroops. The IA increased over 70,000 in that period. They were graduating a new IA Bde from USF at Besmaya at the peak.

Also, Malaki authorized the 5 northern divs to recruit and man up to 125% as part of this. Since then, the excess has been being converted into Artillery and Support Bns - IA combat bns peaked in Oct 2008 at 212.

DJ Elliott said...

Joel, it was announced as part of it 2 months before the first US surge BCT arrived - by Malaki...

The numbers went from 125,000 in the IA to just under 200,000 during the surge. And since the IA Brigades each had a US Battalion partnered with it we know the numbers are accurate for that time.

As to the awakening, look on the FP page of my OOB. Under Emergency Police. Except for the Region I [KRG], those elements and 2 of the FP Divs did not exist prior to the surge - they did by the end of it. The entire Anbar PSF Division is personnel from the awakening. Not to mention that the entire FP existing was purged and retrained at the same time then partnered with a US Bn per each FP Bde.

They have a large GIGO factor in that article - primarily due to the US centric focus...

We went in with enough forces to provide a partnered Bn with each IA/FP/DBE Bde and changed the NTC training/embed practices just to optimize ISF training.

That is all on the record - not opinion.

Anonymous said...

You should write this up and send it to International Security. There are some great points here that need to be published, and IS publishes short(er) response pieces. Just a thought.

Joel Wing said...

Anon thanks for the suggestion, I'll do that.

Steve Donnelly, AICP said...


Yours and Elliot's points: There were more things going on that have yet to be considered.

Like chaos theory. It all looks random unless you know what all the drivers and interactions are.

Anonymous said...

Joel, have you ever been to Baghdad?

Joel Wing said...

No, never been to Baghdad or Iraq. Had a chance a couple years ago for a job, but couldn't take the offer.