Relative lethality

Survival odds for civilians and occupiers in Afghanistan and Iraq

by Marc W. Herold
Departments of Economics and Women's Studies
Whittemore School of Business & Economics
University of New Hampshire


Before dying in combat (or from an 'accident'1), a U.S. occupation soldier in Afghanistan will have participated in the killing of 16-19 Afghan civilians. Although data for Iraq is not strictly comparable (insofar as reported civilian deaths include those caused by more than just U.S. military action and the conflict in Iraq has increasingly taken on the character of a civil war), the ratio of total civilian deaths to that of U.S. military deaths is 18.5-20.5. Another way of putting this is that civilians bear a rising and overwhelming burden of modern war (so-called precision munitions notwithstanding).2 Gino Strada, war surgeon and founder of Emergency Italia, with extensive first-hand experience in many modern war theaters, has argued that over 80% of casualties in modern wars are civilians.

The following Table 1 contrasts U.S. military deaths and estimated civilian deaths resulting from the Afghan and Iraq invasions through December 2006. The data on Afghan civilian casualties killed by U.S. or NATO actions only is derived from three data bases constructed by the author and available on the Internet in disaggregated format.3 The data for total Iraqi civilian deaths comes from the website of Iraq Body Count. The data on U.S. and NATO military deaths is derived from the Iraq Coalition Casualty Count website.4 The important message here is more the order of magnitudes, less the specific numbers. In all three wars, the ratio is relatively similar.

Table 1. Civilian and Military Deaths in Afghanistan, Iraq and Lebanon

War theater Military in-theatre deaths Civilian deaths Ration of Civ/Mil
Afghanistan U.S: 296 + 5 = 301* 4,851-5,684 16.1 18.9
Iraq 2003-6 U.S: 3,000 55,44460,992** 18.5 20.3
Gulf War 1991 U.S: 148 (combat deaths) 2,000 3,000 13.5 20.2
Lebanon (2006) Israeli military killed:119 Leb. civ: 1,200 10.1

* The Associated Press reported (December 30, 2006) that at least 296 U.S. soldiers had died in the Afghan theater (Afghanistan, Pakistan, Uzbekistan) - 192 due to hostile action - to which needed to be added one military civilian death and four CIA officer deaths. ** Preliminary totals for 2006 provided to the author privately by Iraq Body Count.

Source: data on the Gulf War comes from John G. Heidenrich, "The Gulf War: How Many Iraqis Died?" Foreign Policy No. 90 (Spring 1993): 108-125 and "Comments" in Foreign Policy No. 91 (Summer 1993): 182-92.

The case of Iraq, however, as Table 2 shows, is increasingly one of internal civil war. The high ratio of civilians killed relative to U.S. military deaths in 2003 (the year of the U.S. invasion) is followed during 2004-5, by smaller figures. As civil war erupted in Iraq during 2006, the ratio almost doubles.

Table 2. Iraq: U.S Military and Iraqi Civilian Deaths

Years Iraqi civilian deaths U.S. military deaths Ratio of civ/U.S.
2003 10,007 11,932 486 20.6 24.6
2004 9,664 10,494 848 11.4 12.4
2005 12,774 14,019 846 15.1 16.6
2003-5 32,445 - 36,445 2,180 14.9 - 16.8
2006 22,999 24,547 820 28.0 29.9
2003-6 55,444 60,992 3,000 18.5 20.3

The data dramatically reveals that in modern American wars of the twenty-first century civilians perish by a multiple of at least 18 compared to the aggressor nation's military personnel (data on Taliban or Iraqi resistance deaths is unavailable). This figure (of 18) represents a gross underestimate for a variety of reasons: it includes only impact deaths and omits injured civilians who later die; it does not include large numbers of civilian deaths which simply have gone unreported; it omits deaths caused by exposure to depleted uranium, etc.5

Shifting to examining the relative lethality for U.S. occupation soldiers in Afghanistan versus Iraq, my analysis will reveal that the numbers are generally comparable. Contrary to popular perception, Iraq is not a much more deadly place for U.S. soldiers than Afghanistan. In the initial phases of the Afghan war, U.S. casualties were minimal as the U.S. fought using a new type of military operation that relied upon highly mobile Special Forces, airpower, and purchased Afghan allies.6

An historical note is worthy of mention: since 1865 wars have killed fewer soldiers as a percentage of the deployed combat force than was the case in previous wars.7 The main reason for this trend is the increasing dispersion of forces across the battlefield in the face of more lethal weaponry. For example, for the World War I the ratio was 12.0 soldiers killed per 1,000 per annum and for World War II it was 9.0.8 Table 3 below shows that for Iraq it is about 5.7 currently. Also, fewer injured soldiers are dying. Whereas in the Vietnam War (1961-73) and the Persian Gulf War (1990-91), the percentage of U.S. soldiers who died from wounds was 24%, in the current Afghan/Iraq conflicts it is 10%. A study in the New England Journal of Medicine noted that

Though firepower has increased, lethality has decreased. In World War II, 30 percent of the Americans injured in combat died. In Vietnam, the proportion dropped to 24 percent. In the war in Iraq and Afghanistan, about 10 percent of those injured have died. At least as many U.S. soldiers have been injured in combat in this war as in the Revolutionary War, the War of 1812, or the first five years of the Vietnam conflict, from 1961 through 1965. This can no longer be described as a small or contained conflict. But a far larger proportion of soldiers are surviving their injuries.9

On the other hand, though lethality for soldiers has decreased over the twentieth century, no other century has matched it in terms of numbers of people killed, all, between 167 million and 188 million people died because of organized violence in the twentieth century..."10

The following Table 3 compares the lethality for U.S. occupation troops of Iraq and Afghanistan for the years 2005-6. The lethality ratio is defined as soldiers killed per 1,000 in-theater troops. The figures for the in-theater total levels of U.S. troops represent estimated annual averages. The data on U.S. military deaths come from the Iraq Coalition Casualty website.

Table 3. Lethality Ratios for U.S. Soldiers in Afghanistan and Iraq

2005 In theater troops In theater killed Ratio of killed/1,0000 in-theater troops
Afghanistan 18,000 99 5.50
Iraq 140,000 846 6.04
Afghanistan 22,000** 98 4.45
Iraq 142,000** 820 5.77

*In June 2006, U.S. troop strength in Afghanistan was estimated at 23,000 before the handover to NATO.

**In December 2005, U.S. troop strength in Iraq was 160,000, in June 2006 it was 127,000, in November 2006, 152,000 (source).

A number of interesting observations emerge from Table 3. First, the level of apparent lethality for U.S. troops is higher in Iraq (about 25 percent higher in 2006). Secondly, in both war theaters the level of lethality has been falling for U.S. occupation forces.

As regards Afghanistan, the primary reason for this apparent decline is that the United States has successfully "convinced" NATO member countries (especially Canada and Britain) to increasingly bear the brunt of the combat in southern Afghanistan, experiencing far greater lethality ratios. NATO's occupation force rose from 11,000 in 2005 to 18,500 in 2006.11 NATO casualties rose from 31 to 93 between 2005 and 2006. When one adds NATO to U.S. casualties for both years, assuming that U.S. forces would have continued to do the fighting in southern Afghanistan, the lethality ratio (defined as killed in-theater/1000 troop level in-theater) would rise to, respectively to 7.11 (2005) and 8.7 (2006) compared to respectively 6.0 and 5.7 in Iraq. In other words, lethality in the Afghan theater rose significantly during 2006 compared to 2005, and surpasses that of Iraq.12 The figures for Iraq, on the other hand, reveal an approximately constant number of non-U.S. military deaths during the same period (51 to 49). The lethality ratio for British troops in Afghanistan during 2006 was 6.3 9.8 with Britain's Afghan adventure increasingly resembling a quagmire.13

During the 1980's the annual lethality rate for Soviet occupation forces in Afghanistan was 12.5.14 A statistician of Britain's Royal Statistical Society, Professor Sheila Bird, has calculated that since May 2006 (through August) for every 1,000 NATO soldiers in Afghanistan (out of a total NATO commitment of 18,500 troops), 12.9 were dying on an annualized basis, giving a lethality ratio of 12.9, that is, higher than the Soviet one in the 1980s.15 While the lethality ratios are comparable, the levels of in-theater troops differ markedly: the Soviets had 120,000 occupation soldiers in Afghanistan compared to NATO's 18,500 at the end of 2006. The overall annual lethality rate during 2006 is 5.0 for NATO forces, but this is drawn down because most NATO forces are not in the turbulent south and east. This compares to a NATO lethality ratio of 2.8 for 2005.16 Bird found that attacks by insurgent forces have raised the fatality rate (ISAF) to an average of five a week - more than twice the death rate coalition forces sustained during the battle for control of Iraq in 2003.17

Anecdotal evidence from U.S. and British soldiers supports the finding that Afghanistan is "tougher" than Iraq.18 In Afghanistan, the United States has been very successful at shifting the burden of military casualties upon its NATO allies, precisely at a time when the lethality of combat there has been rising substantially and surpasses that in Iraq. The lethality ratio for NATO occupation forces in 2006 was over three times that for U.S. troops. The French seem to have understood the dynamic and announced that by the end of the year 2006, they will have withdrawn their ~200 Special Forces combat troops from southern and eastern Afghanistan (where they had been deployed since July 2003).

Table 4. Lethality Ratios in Afghanistan, 2006

United States 4.45
Britain 6.3 - 9.8
Canada 14.4
NATO 5.0 - 12.9
Soviets (1980s) 12.5

The level of lethality for U.S. occupation forces is actually much higher than the figures above suggest. The reason is because of the "long tail phenomenon" as applied to the U.S. military (the number of support personnel (the tail) required to support combat troops (the tooth)). The tail has also been greatly lengthened as the U.S. military has contracted out to private military contractors (e.g., Halliburton-KBR, DynCorp, Triple Canopy, Blackwater, Executive Outcomes, etc.) for support services (with all kinds of problems regarding lack of oversight, corruption, over-billing, etc.).19 Such outsourcing has been driven less by cost considerations than by a desire to reduce military casualties which are politically costly in the United States. In 1998, the U.S. military in its TTA-2003 analysis estimated it required a ratio of 2.5 for support personnel to combat personnel, that is, over 70 percent of the in-theater troops are involved in support rather than in combat.20 Others cite an even higher number of 10:1 for the support to combat ratio.21

If one reduces the in-theater troop universe by 70 percent, the lethality ratios rise enormously. A problem by so-doing is that this neglects that many U.S. military deaths are of what would technically be considered support troops (e.g., convoy protection, other transportation activities, etc.), all the more so as the domestic resistance has chosen to primarily strike softer targets (e.g., support troops, non-governmental organizations, etc.).22 The true lethality ratio is also significantly raised by the long-term effects upon U.S. military forces caused by weaponry employing depleted uranium.23

The death profile (measured in terms of monthly U.S. deaths) of the Iraq war far surpasses that of the Vietnam War at a similar time point in each war's political lifetime, as demonstrated in Table 5 below.24

Table 5. U.S. Deaths in Vietnam and Iraq

But this comparison while visually dramatic is fundamentally flawed as a measure of lethality because one needs to compare soldier deaths to the in-theater number of troops. When one does that for Vietnam, one discovers very high lethality ratios. In 1966, monthly reported U.S. deaths were 500 soldiers out of a total in-theater force of 300,000, giving a lethality ratio of 20; in 1967, 1,000 soldiers died per month when the in-theater force was 400,000; and in 1968-69, some 1,500 troops died monthly when the in-theater force was 500,000 (giving a lethality ratio of 36).25 The average annual lethality ratio in the three-year Korean War (1950-53) was a very high 53.26 Oblivious of the need to make a relative comparison and typical of the claptrap bandied about in conservative circles today was the view expressed in 2005 by Victor Davis Hanson, senior fellow at the Hoover Institution

...the nearly 2,000 U.S. combat fatalities in Afghanistan and Iraq, while tragic, are a fraction of the 292,000 American battle deaths in World War II -- about 0.6 percent, in fact.27

Conclusion: "Those who cannot learn from history are doomed to repeat it" - George Santayana (1905)

The twenty-first century wars of the United States kill large numbers of innocent civilians relative to combatants (conforming with the trend begun in the twentieth century28) and exhibit high lethality ratios for U.S. troops, particularly in Afghanistan. Factoring in medical, doctrinal, and technological improvements,

...some supporters may continue to refer to [Iraq] casualties as "light," noting that typically tens of thousands of Americans must die in war before domestic support crumbles. Both miss the point. The casualty statistics make clear that our nation is involved in a war whose intensity on the ground matches that of previous American wars...Today's grunts are patrolling a battlefield every bit as deadly as the crucible their fathers faced in Southeast Asia.29

The combination of elevated civilian casualties (Table 1) and the reality of asymmetric war have strengthened the resolve and effectiveness of the resistance. The outcome is predictable: the United States losing more wars but with increasing lethality ratios, adding another chapter in history's lesson as told by the prominent American historian Gabriel Kolko,

(The Pentagon) is superb at spending money but its way of warfare is now in a profound and perhaps terminal crisis. It has lost all its wars against persistent guerillas armed with cheap, light weapons that decentralize and hide.30

Sophisticated, expensive arms are no match for AK-47 wielding, highly mobile rebels who need little training and know the local terrain better, very much the case in rural Afghanistan and urban Iraq. Some call this the new reality of small conflicts. As Larry Kahaner noted, such a sentiment was expressed by Maj. Gen. William Livsey, the commandant of Fort Benning, Ga., in the early 1980s, when the military was first integrating computer chips into smart weapons. "Despite all the sophisticated weapons we or the Soviets come up with,'' he warned, "you still have to get that one lone infantryman, with his rifle, off his piece of land. It's the damn hardest thing in the world to do.''31

The first U.S. occupation soldier to die by hostile fire in Afghanistan on January 4, 2002 - Sgt. 1st Class Nathan Ross Chapman, 1st Special Forces Group (Airborne) - was killed by a teenager shooting an AK-47. Almost exactly five years later, the 301st U.S. occupation soldier to die in the Afghan theater, Staff Sgt. Joseph E. Phaneuf II, was killed in his armored Humvee on December 15, 2006 by an improvised explosive device, another low-tech weapon. On December 27, 2006, an Afghan child was killed by a NATO occupation soldier who fired upon a civilian vehicle near the Pol-e-Tarnak Bridge outside Kandahar city, becoming the approximately 5000th Afghan civilian death caused by U.S/NATO military action (Table 1) since the initial U.S. bombing began on October 7, 2001.

-- 30 --


1. A difficulty exists here insofar as the U.S. military tries to minimize combat deaths and classifies some combat deaths as accidents. Some have questioned the accuracy of officially reported U.S. military deaths, as in Brian Harring, "US Military Report: The High Death Rates Exposed," (updated June 16, 2005), and in "Still Lying about Real Iraq Deaths," (June 14, 2006).

2. A well-established point wonderfully documented historically in Walter C. Clemens Jr. and David J. Singer, "The Human Cost of War," Scientific American 282, 6 (June 2000): 56-7 and also in William Eckhardt, "Civilian Deaths in Wartime," Bulletin of Peace Proposals (Great Britain) 20, 1 (1989): 89-98, which analyzes the number and causes of civilian deaths in wartime during the period 1700-1987. I have documented the rising trend in civilian casualties from Vietnam to the Iraq in "Urban Dimensions of the Punishment of Afghanistan by US Bombs," in Stephen Graham (ed), Cities, War, and Terrorism. Towards an Urban Geopolitics (Oxford: Blackwell Publishing, 2004), Table 17.2 on p. 316. During the Soviet war in Afghanistan, an estimated 15-30,000 Soviet troops were killed, but the number of Afghans who perished has been estimated at 1.3 million out of a population of 17 million (Lawrence G. Kelley, "Afghanistan Revisited," Parameters (U.S. Army War College Quarterly) (Spring 2000): 132-41). As of the Vietnam War, body counts have become extremely politicized, for an excellent analysis see Margot Norris, "Military Censorship and the Body Count in the Persian Gulf War," Cultural Critique No. 19 (Autumn 1991): 223-245. Today, the civilian death count in Iraq is hotly debated between Iraq Body Count and the Lancet. A thorough critique of the Lancet study and methodology has been made in Hamit Dardagan, John Sloboda and Josh Dougherty, "Reality Checks: Some Responses to the Latest Lancet Estimates" (London; Iraq Body Count Press Release 16, October 2006).

3. At my website.

4. Located here and here. The website is maintained by Michael White (details in "U.S. 'Joe Blow' Keeps Track of Iraq War Dead," Reuters (December 27, 2006).

5. A more severe problem in Afghanistan than in Iraq insofar as much of the U.S/NATO bombing and ground assaults take place in isolated regions, and in Afghanistan, the aggressor nations have a much more perfected system of news management in place.

6. Analyzed in Richard B. Andres, "Winning with Allies: The Strategic Value of the Afghan Model," International Security 30, 3 (Winter 2005/6): 124-160

7. From Richard A. Gabriel and Karen S. Metz, "Chapter 6. Lethality and Casualties," in A Short History of War. The Evolution of Warfare and Weapons (Carlisle Barracks, PA.: Professional Readings in Military Strategy, Strategic Studies Institute, U.S. Army War College, June 30, 1992).

8. Ibid. By way of comparison, in the battle on July 1, 1863 at Herbst's Woods during the Gettysburg campaign, a combined lethality ratio of 77 percent was recorded for the combined Union and Confederate forces (see Steven J. Eden, "The War's Hottest Half Hour," Civil War Times 42, 3 (2003): 56-65).

9. Atul Gawande, "Casualties of War Military Care for the Wounded from Iraq and Afghanistan," The New England Journal of Medicine 351, 24 (December 9, 2004): 2471-2475.

10. From Niall Ferguson, "The Next War of the World," Foreign Affairs 85, 5 (September/October 2006)

11. Carlotta Gall, "As NATO Forces Ease Role of G.I's in Afghanistan, the Taliban Steps up Attacks," New York Times (December 11, 2005). In mid-2006, foreign nations were contributing 3,100 troops to "Operation Enduring Freedom" and another 12,000 troops to the NATO force, for a total of 15,100.

12. A point I had made some three years ago in "The Taliban's Second Coming," (February 24, 2004). See also "Afghanistan Said to be More Dangerous for US Troops than Iraq in 2005," Associated Press (February 25, 2005) at Cooperative Research website.

13. In-theatre British troops during 2006 averaged 4,000 mostly in southern Afghanistan. During 2006, 25 British occupation forces were killed, to which should be added the 14 airmen who died when a British spy plane crashed. British casualties' data is from here. During 2006, some 2,500 Canadian troops were stationed in Afghanistan. During the year, 36 Canadian soldiers were killed in hostile action, giving a lethality ratio of 14.4, that is, over three times that for U.S. soldiers. An outstanding analysis of Britain's evolving role in Afghanistan may be read at Bahlol Lohdi, "Blair's Folly. Britain's Afghan Quagmire," (December 28, 2006).

14. I calculate that during the ten-year war, about 1,500 soldiers died each year. The size of the in-theater Soviet force was 118,000 (1985) 120,000 (1987), giving a lethality ratio of 12.5.

15. She found that five NATO soldiers were being killed each week, see "Afghanistan Deadlier for Coalition Troops than Iraq: Study," CBC News (September 7, 2006).

16. In August 2005, ISAF troops in Afghanistan numbered 11,000. During that year, 31 NATO occupation soldiers were killed.

17. About 15,000 Soviet soldiers died in Afghanistan while 900,000 served there (see Matthew Fisher, "Afghan War Still Haunts Russian Vets," (December 27, 2006)).

18. As in Declan Walsh, "'I Knew Afghanistan would be Tough, but I didn't think it would be this Tough'," The Guardian (December 3, 2006), and in Catherine Philp, "They Expected an Easy Ride, Then the Enemy Struck Back," The Times (July 30, 2005).

19. An analysis of military combat outsourcing is provided in David Shearer, "Outsourcing War," Foreign Policy No. 112 (Autumn 1998): 68-81. The general topic of private military contracting is analyzed in Joshua Kurlantzick, "Outsourcing the Dirty Work. The military and its reliance upon hired guns," American Prospect 14, 5 (May 1, 2003), and in P.W. Singer, Corporate Warriors: the Rise of the Privatized Military Industry (Ithaca: Cornell University Press, 2003), 330 pp. On DynCorp, see Tod Robberson, "DynCorp has Big Role, Little Oversight in War Efforts," Dallas Morning News (December 24, 2006).

20. Data from the Congressional Budget Office, "Chapter Two. The Army's Force Requirements for Various Missions," in Structuring the Active and Reserve Army in the Twenty-First Century (Washington D.C.: Congressional Budget Office, December 1997).

21. For example in John Robb, "Global Guerrillas. Long-Tail Counter-Insurgency," Global (September 9, 2005).

22. See for example for the Afghan case, David C. Isby, "Soft Targets in post-election Afghanistan," Jamestown Foundation Terrorism Monitor 2, 23 (December 2, 2004).

23. See Chalmers Johnson, "The Real Casualty Rate for America's Iraq Wars," (May 2, 2003).

24. The source for this comparison is "U.S. Deaths in Iraq vs. Vietnam: The Handoff," (November 5, 2006).

25. From Jim Lindgreen, "1300 US Deaths Would be a Fairly Bad Month in Vietnam," The Volokh Conspiracy website (December 26, 2004).

26. Derived from "Statistical Data on Strength and Casualties for Korean War and Vietnam" (Historical Manuscripts Collection (HMC), Office of the Chief of Military History, November 17, 1965). The average annual in-theater U.S troop strength in Korea was 228,883 (December 1950 July 1953), whereas the average annual in-theater military deaths were 12,172 for that time period (or a monthly average of over 1,000).

27. Victor Davis Hanson, "Today's Politicos Invent the Past. When references to history are totally wrong," Tribune Media Services (August 1, 2005).

28. For a more long-run perspective, see Eric Hobsbawm, "Barbarism: A User's Guide," New Left Review 206 (July-August 1994), John V. Denson, "A Century of War," (December 5, 2005), and also Gwynne Dyer, War. The Lethal Connection (New York: Carroll & Graf Publishers, revised edition 2005).

29. Details in Phillip Carter and Owen West, "Iraq 2004 Looks Like Vietnam 1966. Adjusting Body Counts for Medical and Military Changes," (December 27, 2004).

30. Gabriel Kolko, "Rumsfeld and the American Way of War," (December 26, 2006). See also Jim Lobe, "A Bad Year for Empire," (December 23, 2006).

31. In "Weapon Changed Pattern of Modern War," (December 29, 2006).


Professor Marc Herold's Afghan Canon