How many troops does Sadr have, exactly?

by Daniel on August 10, 2004

Another entry in my occasional role as co-ordinator of the Campaign For Real Body Counts; grateful for any comments that might help me make sense of these numbers:

As recently as the April uprising, the Sadrite Al-Mahdi militia was estimated by Iraqi experts to be between 3,000 and 10,000 strong, with the Pentagon suggesting that the hard core of fighters could be as small as 1,000.

In the May offensive against Sadr in Najaf and Karbala, it was once more credibly estimated that 1,500 of the Al-Mahdi Army were killed (note that this reference suggests that, as of the beginning of May, only 1,000-2,000 of the militia were located in or around the city of Najaf).

In the more recent episodes of fighting, official sources have told us that the Najaf branch of Sadr’s forces have taken a further 300 casualties, and lost a further 1200 men captured or surrendered.

So to recap … a force which was meant to have only 1,000 serious fighters, has had 1,800 of them killed and continues to fight on. Sadr had about 2,000 fighters in Najaf, has lost 3,000 of them and continues to fight[1]. Something doesn’t add up (or to put it another way, nothing does add up). Either:

  • original estimates substantially underestimated the size of Sadr’s forces, or
  • we have substantially overestimated the amount of Sadr’s force which has been neutralised, or
  • Sadr has managed to recruit very substantial amounts of force indeed over the last three months.

To be honest, each of these three possibilities looks as bad to me as each of the others. Someone wake me up when it finally becomes acceptable to make comparisons to Vietnam.

[1] Even allowing for the likelihood that the Najaf militia would have been reinforced after May from Sadrite forces elsewhere in the country, I still can’t get this one to pass the laugh test. I’d also note that the 1,500 figure refers to Sadrite casualties in the whole of Iraq and probably shouldn’t be conflated with the Najaf figure of 300, but the qualitative conclusion is unlikely to be affected.



John Quiggin 08.10.04 at 12:39 am

It’s always a sign of impending disaster when the two sides in a conflict stop exaggerating their own casualties (caused by the unprovoked aggression of the other side) and start exaggerating the casualties they’ve inflicted (showing that victory is at hand).


John Quiggin 08.10.04 at 12:51 am

OTOH, if the US has overestimated the number “neutralised” that’s a good thing, since each such overestimation corresponds to one less dead body.


thehim 08.10.04 at 12:52 am

I’d venture to say all three of those things are likely true, except that we’ve likely overestimated the percentage of al-Sadr’s forces that have been neutralized. There’s no other way to describe the wide gaps in what we’re told and what the reality is pointing to.


minnesotaj 08.10.04 at 2:31 am

Of course, I’m reminded of the old Russell-Fr. Copleston debate on Pascal’s Wager: As Russell showed, there’re always more than two sides to every dichotomy! Could be that Sadr’s forces ARE small(ish) and that the US is accurate in the number of dead bodies it has policed up in after-action street/hospital sweeps: the extra numbers are civilians killed by indiscriminate fire. Just a thought on a very likely answer to this puzzle.


dsquared 08.10.04 at 2:34 am

The approach to casualty measurement that minnesotaj suggests was known in the Vietnam War as the principle of “If It’s Dead And It’s Gook, It’s VC” and was known by serving troops to be about as bad news as John suggests at the time.


Tom Slee 08.10.04 at 3:10 am

I have no idea what the number is, but the arithmetic of guerilla casualties is certainly at fault. That is, if there are 3,000 guerillas, and 1,000 are killed by an invasion force, that does not leave 2,000 guerillas. The mileage may vary from place to place, but the it may, in fact, leave more than 3,000.


Donald Johnson 08.10.04 at 3:13 am

There’s also the possibility that 300 bodies were counted, but not all were militia.


minnesotaj 08.10.04 at 4:00 am


Yes, of course, the accurate count is probably something like:

(Dead Guerillas) + (Dead Civilians) = (Official US Body Count)


((Prev. Guerilla Count) – (Dead Guerillas)) + ((Perceived Justice of Attack on Guerillas) * (Enraged Iraqis/(Despair * Hope)) = (New Guerilla Count)

Which is why, in absence of anything like a sense of justice among ordinary Iraqis, our continued aggressive attacks on Sadr’s men are probably counterproductive (unless, and what I don’t know: his death would bring end of insurrection and hope for ordinary Iraqis of some chance or order/peace in their country).


dsquared 08.10.04 at 4:12 am

FWIW, I think that this is probably a terminating process; Sadr’s initial resources may have been larger than expected but were finite, and I don’t doubt that the US Army’s ability to destroy his troops outstrips his ability to recruit new ones in the near term. He’s recently been quoted on the newswires saying that he will “fight to the last drop of blood”, which when said by a military commander is usually a sign that things are going pretty badly. But we ought to be very worried by “c)” above; the fact that, if I am correct, Sadr managed to largely rebuild his militia in three months suggests that somebody else could quite easily take his place. Hopefully not al-Quaeda.


Rajeev Advani 08.10.04 at 6:20 am

Body count comparisons to Vietnam are out of the question, for obvious reasons. So I’m guessing, Daniel, that you’re itching to start making comparisons of a more political dimension? How does Sadr compare to Ho Chi Minh, one might ask?

Uncle Ho was committed to democracy and was supported by the vast majority of the Vietnamese population. The US had ignored the Geneva Accords and forstalled elections, while Ho Chi Minh pushed and pushed to get the United States to acquiesce to national unification followed by a national vote. To my knowledge, Muqtada al Sadr has done quite the opposite, rejecting all of Allawi’s overtures to enter the political process (please correct me here if I’m wrong). What comparison, I ask, do you want to make between Iraq and Vietnam? Thanks, and I commend your efforts toward acquiring accurate body counts.


Motoko Kusanagi 08.10.04 at 7:13 am

Riverbend, October 2003:

“He is frightening and I don’t think his influence should be underestimated. He easily has over a million followers (some say it’s up to 4 million) and they practically revere him.”

Juan Cole, May, 2004:

“One speaker claimed that Muqtada has only 10,000 men. In fact that is the size of his formal militia. Muqtada’s movement is like the layers of an onion. You have 10,000 militiamen. But then you have tens of thousands of cadres able to mobilize neighborhoods. Then you have hundreds of thousands of Sadrists, followers of Muqtada and other heirs of Muhammad Sadiq al-Sadr. Then you have maybe 5 million Shiite theocrats who sympathize with Muqtada’s goals and rhetoric, about a third of the Shiite community. The Sadrists will now try to shift everything so that the 5 million become followers, the hundreds of thousands become cadres, and the tens of thousands become militiamen.”


Motoko Kusanagi 08.10.04 at 7:36 am

One more. Raed Jarrar, April 2004:

“AsSadr is NOT reflecting a minority of Iraqis, this is a stupid big lie. Whether we liked him or not, he is the political and religious leader for MILLIONS of Iraqis in the southern region… There are 15 million Iraqis living in the south, and another 5 million in Baghdad, I can say that 5 to 7 millions of them can be considered as AsSadr followers… AsSadr is THE GOVERNMENT in most of the cities of the south: Amara, Kut, Nasryya and Diwanyya and Simawa partially, and Najaf partially…”

If this is true it doesn’t seem impossible that he can rebuild his militia quickly after heavy losses. It’s hard to imagine, but perhaps we should reckon with the possibility that US intelligence on the strength of Sadr and his militia was somewhat colored by wishful thinking.


VJ 08.10.04 at 10:51 am

Give Motoko Kusanagi a cigar! This was exactly the experience of Vietnam. In the end we were fighting the birth rate of the North. Even if only 1-2% of the population are actively resisting & fighting, this is a very large number indeed, and can indefinitely sustain a guerrilla war for decades perhaps. We are chest deep in the big muddy and don’t know it yet.


dsquared 08.10.04 at 12:17 pm

Rajeev, you wrote:

Uncle Ho was committed to democracy and was supported by the vast majority of the Vietnamese population.

This is simply not true. Ho was committed to Stalinism, not democracy. He did not support elections unless he thought he could win them and regularly used violence against people who thought they were his nationalist allies. He also did not have the support of even a plurality of the South Vietnamese population at any point during the war. The USA was entirely right to regard him as anti-democratic, Communist and totalitarian, entirely correct when they said that they were fighting against someone who would be very bad for the average Vietnamese, but nevertheless entirely wrong to pursue the Vietnam War.

This is a terribly prevalent tendency on the left; the rewriting of history to turn the Vietnamese Communist party into some version of Tolkien’s elves. They weren’t; they were a communist party like any other, and equally disastrous in power. The real lesson of the Vietnam War is that even good, well-justfied humanitarian interventions often go very badly wrong.

(I apologise in advance for the tone of this post, Rajeev; I suspect that you have copped some flak mean for other people with whom I am having this argument)


Kevin Donoghue 08.10.04 at 1:53 pm


Since you like to annoy the war party with references to Vietnam, may I remind you of the sad story of Sam Adams, the CIA analyst who discovered that Viet Cong numbers had a tendency to rise faster than US forces could destroy them? He thought the top brass would be interested in his research, but then as now, the true function of intelligence was to come up with the figures the Pentagon wanted.

As far as honest appraisals go, AFAICT the best source of information on Iraqi Shiites is Juan Cole.

BTW, your reading of Uncle Ho may be inaccurate (which is not to say you are being unfair). It seems to me he was a nationalist first and a communist second. Also, references to “a plurality of the South Vietnamese population” don’t really carry much weight; South Vietnam was essentially an American creation.


Anna in Cairo 08.10.04 at 1:55 pm

I would just like to add to the very astute comments above that mention Sadr’s populist appeal that Iraq has a male population that has virtually 100% served in the military and knows how to use weapons. Also it was a police state meaning that a lot of former ARmy and former cops would have kept their weaponry. Meaning that the male population of Iraq already has military experience and a lot of them probably have the guns to go with it. So the numbers of militias could continue to increase and probably the heavy handed tactics the US and Allawi are using are only going to make the problem worse.


a 08.10.04 at 1:56 pm

You mean Uncle Ho was actually Uncle Joe?


Anna in Cairo 08.10.04 at 1:58 pm

Also although this was really before my time I have read some quotations of Ho Chi Minh that seem to suggest that he was indeed a nationalist first and a communist second and that he admired the founding fathers and the declaration of independence and stuff like that. I think I read about this in the Blum book about the CIA, all of them were fact checkable quotations taken from news sources from that time period.


dsquared 08.10.04 at 3:10 pm

Ho did admire the founding fathers of the USA, and indeed, the Vietnamese Declaration of Independence borrows from the American one. But he didn’t really see communism and nationalism as separate ideologies; he was radicalised during his period under the name Nguyen Who Hates The French, and he spent a lot of time in China before the Vietnamese revolution, which is quite hard to square with a nationalist view rather than an ideological one.


praktike 08.10.04 at 3:17 pm

At the end of the day, though, these guys pretty much suck at fighting. They’re getting killed by the bucketload. Escalate rapidly and prove that joining these guys is not a smart career move, and you’ll go a long way toward solving the problem.


nick 08.10.04 at 3:48 pm

Escalate rapidly and prove that joining these guys is not a smart career move, and you’ll go a long way toward solving the problem.

Ah, yes. The ‘exterminate all the brutes’ approach to grand military strategy, which has proved so successful in times gone by. Believe it or not, I don’t think that many Iraqis regard armed resistance as a ‘career’, no matter how facetious your metaphor. And that’s the point. (Nor are those American soldiers who signed up as a ‘career move’ likely to be treating their current service in Iraq as ‘a career’.)

Let’s just say that the ‘set of people likely to be shooting a gun on behalf of al-Sadr’ is equal to the intersection of the sets ‘young’, ‘poor’, ‘male’ and ‘Shi’a’. And let’s just say that the egregious underestimates of Sadr’s resource pool were designed to keep a rather dumb segment of the pro-war American populace happily making crosses through silhoutted stick figures with turbans, okay? And let’s move on from there…


Jim Henley 08.10.04 at 3:54 pm

I admire Daniel’s tenacity here. Based on my reading into the conflicting reports on last winter’s attacks in Samarra, I think it’s safe to reduce all US casualty estimates that don’t involve actually lining up corpses and tagging toes by a factor of five. This is probably the divisor to apply to the highest-end civilian casualty estimates too.

It’s actually a lot harder to kill someone than we tend to think, something I wrote about at length early in my blogging career. And military forces tend to quit the field rather than continue to take KIAs and even wounds. The sort of potted military theory one gets from Gene Wolfe books suggest that a force loses cohesion under fire at a casualty rate of 10-33%. Sadr’s militia being a militia, I expect that to happen a lot closer to the low end than the high end. So take one fifth of the US estimates of Sadrist casualties at the end of any victorious engagement, and multiply that by 10 to get a ROUGH estimate of the force the Sadrists were fielding.


Dept. of Made-Up Statistical Thresholds 08.10.04 at 5:13 pm

The policy of the U.S. Government that if twice the total number of enemy is killed, the remaining number of enemy should drop by one half.

As only 150% of the enemy was killed, it is not unexpected that the number of enemy should increase.

BTW, the increase in fighting demonstrates that they are desperate and we are winning.



dsquared 08.10.04 at 5:19 pm

Praktike: blah blah Vietnam blah blah didn’t work blah … tell me when this gets boring.

Jim: I also tried to look up that number, but was also unable to find any source for it other than rulebooks for playing games with lead figures. I’ve got a couple of Liddell Hart books downstairs but I can’t find anything useful in them.

All readers: It really is worth making the effort to look into the archives of Jim’s blog for last winter for his posts on the Samarra casualty count; I read them at the time and remember them as being quite excellent.


Jane Galt 08.10.04 at 5:22 pm

Kevin, South Vietnam as a separate region, culturally and administratively, has a much longer history than a “unified” Vietnam. Though the Vietnamese began expanding south along the Mekong delta in the 12th century, they didn’t fully assimilate the territory of the south until the mid-18th, and by then Vietnam had already been split into two administrative districts, under two separate ruling families, for roughly 200 years–the South was basically only truly a part of a unified Vietnam from 1802, when Nguyen Anh consolidated the two regions, until 1858, when the French invaded to revenge the slaying of a bunch of French missionaries. Not that that has anything much to do with the topic at hand, but while I’m for anything that has to do with getting rid of the French, who were really appalling colonial masters even by the standards of western imperialism, Vietnamese nationalism of the sort that demands a unified north and south has somewhat shaky roots. Or at least that’s my understanding from my rather shallow dive into Vietnamese history.


Dan Hardie 08.10.04 at 5:43 pm

A general point is that soldiers generally tend to overestimate enemy deaths- not just when, as John Q suggests, the war is going badly for them. (Just to be clear: the war is, I think, going very badly for the Coalition right now.)Any history of the Battle of Britain has the RAF pilots overestimating enemy ‘kills’ by 300% or more: a more accurate figure was privately arrived at by RAF intelligence who collected shot-down planes on the ground, but the higher estimates were published to aid morale. On the ground, I can think of lots of examples. Martin Middlebrook, in ‘The battle for the Malvinas’, recounts the Royal Marine patrol on (I think) Mt Harriet, who bumped an Arg unit and shot them up before getting the hell out: they reported at least 18 enemy deaths, but Middlebrook dug out the Argentine papers and witnesses and found it was three. His whole book is replete with such examples of British over-reporting: and remember, the men reporting the inflated body counts were not cowboys but some of the best light infantry in the world.(They also had little motive to exaggerate: there was no ‘body count’ policy as per Vietnam.) It’s not hard to see why squaddies exaggerate: a lot of terrified, hyped-up men all firing at individual targets, which may hit the ground because they’re dead or because they are taking cover- who knows how many get killed.

The only thing one can trust, I think, when compiling casualty figures, is an actual count of dead bodies on the ground. And this isn’t happening in Iraq: Iraqi civilians and militiamen alike have both customary and prudential reasons for burying their dead asap without notifying the local Coalition forces.

Btw, Praktike’s point about killing lots of militiamen being good is bunk: all accounts of Iraq agree that the blood feud is a live-and-kicking custom for many locals, so if you kill a large number of militiamen without then paying blood money to his family and tribe, you get all his kinsmen taking revenge potshots at you.

D-squared: for estimates of things like combat effectiveness, unit cohesion etc a good place to start is S.L.A. Marshall’s studies of WW2 and Korea. These are out of date. Military institutions- in Britain, for example, the Royal College of Defence Studies- produce updated versions, but so far as I know the UK documents are classified. Possibly current Pentagon estimates are out there. I would say that applying ‘unit cohesion’ to the Sadrist insurgency, as Jim Henley does, is a bit of a category error. As Juan Cole pointed out- and as, before him, the Brigadier running British forces in Basra noted- if the Iraqi Shi’ites kick off an Iranian style ‘crowd revolution’, then it’s game over. ‘Unit cohesion’ and other narrowly military measures are less important than measures of broad political support and alienation, in this war. Kill enough Shi’ite civilians and we will have all the survivors at our throats. As Clausewitz might have told us, no possibility of divorcing the military from the political. On those grounds, I am very much a pessimist.


Rajeev Advani 08.10.04 at 6:16 pm


What I wrote was based off the conclusions of William Duiker’s biography, Ho Chi Minh – A Life. What many tend to do is mix up Ho’s views with those of his underlings: Truong Chinh, Vo Nguyen Giap, Le Duan, and so on. When it comes to Ho himself, he was a nationalist first. I won’t deny that he was a fervent Communist, but he was by no means totalitarian. Some objections to characterizing him as Stalinist:

1) When the North Vietnamese attempted a land reform program in the 60s, Ho Chi Minh repeatedly urged caution — he had learned from the mistakes of Stalin and Mao.

2) When the Vietnamese people were suffering from famine, instead of callously disregarding the problem like Stalin and Mao, Ho Chi Minh fasted himself 1 day out of 10.

3) Unlike Stalin, who consolidated all power, Ho devolved most of his power to his underlings in the Party, and only kept an interest in foreign affairs.

4) While Ho was alive the Party never engaged in any Stalin-esque totalitarian purges, nor were any Vietnamese Gulags constructed, nor

4) In Ho’s later years, when he actually had power, he called for freedom first, everything else second.

5) In his will, Ho called on Le Duan to extend a one year moratorium on all taxes to peasants, knowing that they had suffered the brunt of the pain from reunification and land reform. Not exactly Stalinist.

As William Duiker summarizes, Ho was not Stalinist, but one part Lenin, one part Gandhi, with a dash of Confucism. Unfortunately today many people conflate what happened in Vietnam with him directly, when many of the North Vietnamese atrocities were either spillover effects from an uncontrolled revolution or the results of rabidly Communist local policies put in effect by lessers in the party.

Regarding the point about democracy: I recall reading somewhere (but I’ll have to verify this) that 90% of the population supported him for President, and that’s including the South. What I do know for sure is that the South Vietnamese revolution was not started by Ho Chi Minh — he simply guided what was an upstart, national revolution against the Diem (and the successor) regime.

Regarding his time in France and China. While in France under the name Nguyen Ai Quoc (Nguyen the Patriot!) he was only attracted to Communism becuase of its obsession with anti-imperialism (and admitted this much in later interviews). So that’s easy to square away with nationalism.

Regarding his time in China: he spent most of that time arguing against the hard-liners in the party like Ha Huy Tap, who called for a tight and vicious Communist revolution. Ho Chi Minh believed — and this is anti-Stalin — in a united front revolution that would bring the Communists together with lesser bourgeois elements. Granted of course, that this was to be followed, eventually, by a Communist revolution — but nevertheless it shows Ho putting nationalism first.

(It’s ok about the tone, blogging has thickened my skin… and I expect somebody to bash me for painting too rosy a picture of Ho Chi Minh here)


dsquared 08.10.04 at 6:30 pm

Rajeev: fair enough. I have to confess that I gave up half way through the Duiker book because of the failure of my third attempt at a coloured highlighting system to keep track of all the Nugyens. Though perhaps worth emphasising that the Americans were for the most part fighting the (second) Vietnam War against Ho’s heirs rather than he himself.


Rajeev Advani 08.10.04 at 6:47 pm


I agree with your emphasis, and would second it because by the mid 1960’s Ho’s heirs had already marginalized him (they thought the old man was losing it). In that case I suppose it becomes possible to make a comparison between Sadr and Le Duan, but I don’t know enough about either.


Francisco 08.10.04 at 7:43 pm

Paraguay lost 95% of its male population on the triple alliance war (
I think it’s safe to say there’s not an upper number beyond which a country stops fighting, valid for all cases.


John Quiggin 08.10.04 at 9:47 pm

Dan, I didn’t mean to say that overestimation implies that war is going badly.

What I meant is that, in the leadup to a full-scale war, both sides tend to exaggerate the casualties by enemy action and minimise or deny any harm they have done themselves. The idea is that “we are the defenders of peace, they are wanton aggressors”.

It’s a bad sign when this pattern is replaced by exaggeration of casualties inflicted precisely because this implies that all-out war has begun.


Jim Henley 08.10.04 at 10:49 pm

I think it’s safe to say there’s not an upper number beyond which a country stops fighting, valid for all cases.

Francisco: The Paraguayan example doesn’t seem comparable. The scale is wrong – a whole war when we’re talking about individual battles – and the deaths are, as even the brief Wikipedia account notes, compounded by disease, malnutrition and no doubt not a few massacres. Under discussion is the tendency of forces in being to retreat from a tactical situation, when they can retreat at all. That’s a whole ‘nother thing.


Dan Hardie 08.11.04 at 9:16 am

John- fair enough. I feel, and no doubt you do too, that the Americans needed to start thinking, from 1st May 2003 if not before, that success would be measured by things like the number of hours of electricity available to major urban centres, and the number of Iraqis prepared to offer intelligence on jihadists- not on any body count. As you say, banging on about the latter is a very bad sign, for all sorts of reasons.


Dan Hardie 08.11.04 at 9:31 am

‘Under discussion is the tendency of forces in being to retreat from a tactical situation, when they can retreat at all. ‘

Jim- I agree with the point you’re making to Francisco re the War of the Triple Alliance. But more generally I think you’re badly wrong: Sadr’s militia is not a ‘force in being’ in the sense in which, say, the 82d Airborne is a ‘force in being’. Sadr’s forces are those proportion of the (already heavily armed, already firearms-trained) Iraqi population prepared to back him against US and/or other Iraqi forces. Kill x percent of a battalion of US soldiers and they will lose unit cohesion- with x being lower if there is no system of battle casualty replacements, low unit morale, poor leadership, etc, and higher if those conditions are reversed.

But the battles in Najaf and elsewhere are a whole other ballgame: there has to be a risk that if one kills x percent of Sadr’s gunmen one degrades the ‘unit cohesion’ of his existing militia but incites currently neutral or anti-Sadr Iraqis to join in the struggle against the Americans.

‘Unit cohesion’ for conventional forces is one thing: can a formally structured, lengthily-trained body of men continue to carry out demanding orders for an extended period of time. Sadr’s forces don’t have to have ‘unit cohesion’ in that sense: their goals are a mix those of a street gang, a protest movement and a political party.


tom beta 2 08.11.04 at 2:34 pm

In his post, Daniel notes:
…the Sadrite Al-Mahdi militia was estimated by Iraqi experts to be between 3,000 and 10,000 strong, with the Pentagon suggesting that the hard core of fighters could be as small as 1000.
(emphasis mine)

He then concludes:
So to recap … a force which was meant to have only 1,000 serious fighters

Note this was a minimum estimate from the Pentagon.

has had 1,800 of them killed and continues to fight on.

Nobody claimed to have killed 1,800 of Sadr’s core fighters. 1,800 out of a force estimated to have up to 10,000 fighters isn’t that much of a stretch.

Sadr had about 2,000 fighters in Najaf, has lost 3,000

Whoever heard of groups of guerrillas based in different parts of a country gathering for an attack? Novel idea, that.

Something doesn’t add up

You’re right about that.


gijoe 08.12.04 at 11:35 pm

stand with your troops you freedom-freeloader


George 08.13.04 at 9:02 pm

Late to the party here, but the answer is likely to be either (a) Sadr is successful at recruiting lots of young, inexperienced (and easily killed) fighters, or (b) initial estimates were low.

Comments on this entry are closed.