Iraq: the Human Cost


Violence in the Iraq War

by John Tirman, Executive Director, MIT Center for International Studies

Random killings, human bombs, dozens of violent groups, and a deepening sense of insecurity gnaw Iraq. The evidence of pervasive and persistent mayhem is everywhere, from the formal statistics of mortality to broader estimates of numerical outcomes, from the reports of families evacuating dangerous neighborhoods by the tens of thousands to the paralysis of governance. The deadly violence is omnipresent, but without a visible front or an apparent strategy---and for those reasons, among others, it is poorly understood.

It is for this reason that the mortality study conducted by Burnham et al was commissioned by the MIT Center for International Studies. Understanding the scale, the sources of violence, the demographical profiles of the victims, and the geographic dispersion of killing---all recorded in the household survey of the Iraq mortality study---provides an indispensable tool in coming to terms with the violence in Iraq.

Explanations for and Effects of the Violence

For all the attention given to the insurgent attacks and suicide killers, scholars and policy professionals---including those in the U.S. Government---have not grasped the roots of violence or its consequences. Violence has shaped not only the sense of dread and despair that now beleaguer the country, but its politics, social organization, economy, and culture. Its wider implications, which include the stirring of global jihad, regional instability, and Shia versus Sunni bloodshed, are disturbing in every conceivable sense.

The recent spate of books1 and ongoing news coverage only partially address these matters of violence. These accounts are necessarily incomplete, given how dangerous it became to report from Iraq soon after Saddam Hussein was deposed. The violence in effect has become a silencer.

Academic studies, including interviews with jailed insurgents, suggest that violent actors are creating an enormous, self-generating feedback loop in which violence begets violence. This cycle, significantly initiated by the aggressive rules of engagement of the U.S. military and its counter-insurgency strategy, better explains the durability and decentralized nature of the Sunni Arab insurgency than any other cause. The violence from Shia militia came later in the war, and is likely a reaction to coalition's deadly force, Sunni insurgent actions, and the provocations of non-Iraqi jihadists.

That the violence is horrific and worsening is broadly acknowledged, and it is increasingly evident that mortality is very high. (The effect of the U.S. troop surge in 2007 is still being assessed; violence has decreased in Baghdad partly because it has been ethnically cleansed---the "peace of the grave.") The death rate is important because it is a pivotal explanation for the insurgency. There are a few commonly cited reasons for the Sunni Arab insurgency. In the succinct words of Barry Posen, Professor of Political Science at MIT, there are "four sources of energy":

the anger of a group that once stood at the pinnacle of power and privilege in Iraq and now has been thrust to the bottom; tight kinship ties among this group . . . that draw new family members into the insurgency as others are killed, captured, or roughed up; nationalism, both Iraqi and pan-Arab; and the recent flowering of religious fundamentalism in the Islamic world.2

The reason most frequently cited to explain Sunni Arab violence is the loss of the status they enjoyed under Saddam Hussein, as Posen notes, but this seems an unlikely engine to generate such widespread disruption for more than five years. Sheer resistance to occupation (regardless of prior status) is also at work. Anti-occupation nationalism is evident in nearly all opinion surveys over the last three years in Iraq, with majorities nationwide stating their opposition to U.S. troops despite growing concern about insecurity.3

This stout and sizable opposition to the United States' policies and presence in Iraq can be explained, potentially, by seeing the large numbers of casualties allegedly due to coalition military actions as a source of perceived insecurity. Another motivation is watching the humiliation and shame of one's compatriots, tribesmen, and family, especially vexing at the hands of a non-Muslim occupier, and intensified by the loss of life. These emotions stir moral outrage, as psychologists studying this have found, which can be expressed in violence. Virtually all the accounts of the American invasion and occupation describe harsh practices that have stirred a ferocious response.

One of the striking characteristics of the insurgency is how durable and decentralized it is. The research on this (with captured fighters, for example, or failed suicide bombers) indicates that the major motivation is "defense," i.e., they believe they are defending their family, village, or tribe from assaults. The fighters are volunteers, new to insurgency or jihad. Their religious convictions or knowledge are thin, leading researchers to believe that insurgents are volunteering more for secular reasons---nationalism, defense, moral outrage---than as part of a holy war.4 Yet mosques play a role. Sunni imams of the local mosques are not connected to each other as are the Shia clerics, who follow Ayatollah Sistani's instructions; the Sunni sermons grow from local circumstances, and a process of bidding up the resistance is at work as a result.

As a result, the kinship/tribal networks so integral to the ratcheting up of violence in reaction to counter-insurgency are morally validated by mosque leaders; both, importantly, are largely autonomous and decentralized. Research also shows high local esteem for fighters. This indicates how difficult it will be, and has been, to reverse the tide of violence. The "self interest" of the Sunni Arabs to give up arms and participate politically thus has had little traction. The "Awakening councils" of the last few months are purchased, and may be motivated not only by money but by a desire to regain control of local areas from foreign jihadis and to create a new bulwark, with the U.S., against Shia militia. With whom these now well-armed Sunni fighters will cast their lot in future years is anyone's guess.

What the Mortality Study Contributes

While no one is insisting that the large numbers of fatalities reported in the Burnham et al study are indisputable---their mid-range estimate as of July 2006 was 655,000 dead due to the war, 600,000 by violence---the study provides an order of magnitude estimate, the only scientific calculation of the war dead at the time (see analyses of other surveys published since, on this site), and invaluable demographic data. All surveys, including one released in January 2008 by the Iraq Ministry of Health and showing 400,000 excess deaths, are largely congruent in their main conclusions, with their differences reflecting the difficulty of gathering data during wartime.

In the most acute periods of violence (2005-07), the growing numbers of deaths---which all observers agree occur broadly throughout Iraq---were found in the survey to be caused mainly by gunshot. This matches the widespread observation of assassinations, single shots to the head either in groups or individually. It also explains the high numbers, as Professor Juan Cole, a leading expert on Iraq, explained in House testimony. (Hypothetically, for example, if just three people were killed daily in the major urban centers of Iraq, of which there are more than 80, something like 300,000 deaths would accrue over the 40 months of the study.) The study showed that the deaths were mainly among young men, which also tracks with anecdotal evidence. More than two-thirds of fatalities take place outside Baghdad, which is indirectly confirmed by figures from Multinational Force reports.

So the violence is widely dispersed, mainly involving young men, and principally caused by gunshots. The respondents to the survey---some 1,850 individuals---also attribute the deaths in such a way that, over time, we see sectarian violence at work and escalating. However, a large proportion of deaths are attributed to coalition forces, between 14 and 21 percent over time directly attributed, a belief that is reflected, as explained in footnote 3, in other surveys of Iraqis. In the mortality survey, it is not clear how many deaths can fairly be attributed to coalition forces, since respondents' knowledge of the source of death is far less reliable than the fact of death. But the responses suggest between 100,000 and 200,000.

These numbers provide significant evidence backing the explanation of a cycle of violence. The U.S. forces, perhaps understaffed and taken aback by the early resistance to the invasion and occupation, applied deadly force broadly in Iraq, especially Sunni Arab regions, which actually spurred a larger insurgency. Deaths in the dense kinship networks activated other male members of the families to join or form violent groups. The dispersed and headless nature of the Sunni Arab insurgency would lead to precisely the kinds of killing we have seen: assassinations of "traitors" or "collaborators," revenge killings of various kinds, hit-and-run attacks on U.S. forces and the Iraqi security forces, and the like. This has escalated, as the survey clearly demonstrates, as sectarian violence has grown. The killing has become an increasingly Shia-Sunni fight, or between factions within sects, in addition to anti-occupation and counter-insurgency violence.

Only a fraction of killing occurs in Baghdad, yet the capital is the nearly exclusive venue for the international news media, which sees the violence there as predominating. (One difficulty of understanding violence in Iraq stems from the news media's tendency to use "passive surveillance" in counting the dead---i.e., using only Baghdad morgue numbers, mainly---which is a result of being so concentrated in Baghdad alone; this is true of U.N. and other statistics as well.) The Baghdad-centric focus has several distorting effects: first, the news media is essentially blocked from reporting violence from around the rest of the country. Second, it makes car bombs and other spectacular acts of violence the center of attention. Third, it creates information dependencies on others who are Baghdad bound. And fourth, it lends credence to the idea that the violence is now between Sunni groups and Shia militia alone.

By seeing the violence in Iraq as significantly a reaction to occupation and fitful, uneven, or even corrupt attempts at state building, we would come to different conclusions about the problem solving, including diplomacy, needed to stabilize Iraq. We would gain many insights about the nature of insurgency, and the ways organized political violence proliferates and sustains itself. And, perhaps most important, we would realize the human costs of the war in Iraq.


1 These include, most prominently, Michael R. Gordon and Bernard E. Trainor, Cobra II: The Inside Story of the Invasion and Occupation of Iraq (Pantheon, 2006), George Packer, The Assassins' Gate: America in Iraq (Farrar Strauss Giroux, 2005), Thomas Ricks, Fiasco: The American Military Adventure in Iraq (Penguin Press, 2006), among several others.

2 Barry R. Posen, "Exit Strategy: How to disengage from Iraq in 18 months," Boston Review (January-February 2006), on line at

3 Three different polls with sizable samples indicate strong reactions to the U.S. presence. In a State Department poll, publicized in September 2006, three-quarters of the respondents said "they would feel safer if U.S. and other foreign forces left Iraq, with 65 percent . . . favoring an immediate pullout." In a similar poll, conducted by a survey unit at the University of Maryland, also in September 2006, "78 percent of Iraqis believe the U.S. military presence causes more conflict than it prevents." Amit R. Paley, "Most Iraqis Favor Immediate U.S. Pullout, Polls show," Washington Post, Sept. 27, 2006, p. 22; and David Alexander, "Most Iraqis favor withdrawal of U.S. forces soon: poll," Reuters, Sept. 27, 2006, available on-line.

In a 2007 poll conducted for several news organizations, including ABC News and BBC, which sampled some 2,200 people, "More than eight in 10 Shiites (as well as 97 percent of Sunni Arabs) oppose the presence of U.S. and other forces in their country. (Kurds, again, differ powerfully; 75 percent support the U.S. presence.) More than seven in 10 Shiitesand nearly all Sunni Arabsthink the presence of U.S. forces in Iraq is making security worse." As with the other polls, the violence was mainly attributed to U.S. forces. "Ebbing Hope in a Landscape of Loss Marks a National Survey of Iraq," March 19, 2007, available on-line:

A large-sample British survey (>5000 respondents) released in March 2007 did not ask about blame, but did gauge mortality in astounding numbers: 26% reported that someone in their family had been murdered during the war; an additional 12% reported knowing a friend who had been murdered. This Opinion Research Business poll can be found at:

4 See the workshop report on Transnational Violence convened by the MIT Center for International Studies at

5 As Gordon and Trainor report, "Frustrated with the insurgent attacks and unprepared to deal with the complexities in Iraq, there was 'a default to meet violence with violence on the part of U.S. forces,' [Brig. Gen. John] Kelly observed, which led to civilian casualties and hardened the attitudes of many Iraqis against the Americans." (310)

6 John Tirman, "Diplomacy and the Iraq War," Strategic Insights, Volume VI, Issue 2 (March 2007);