The Microscope

Transparency and Casualty figures: Those who seek transparency must also be transparent

Harshula <>

Last updated: August 16, 2019

Table of Contents

1 Introduction

Sri Lankans in general are very skeptical about politics and politicians. This skepticism now appears to extend to the United Nations (UN) and international non-governmental organisations (INGOs), for a number of reasons. One of those reasons is the lack of transparency.
The casualty figures for the last five months of the separatist war in Sri Lanka are mired in controversy. The UN did not initially attempt to publish their methodology.

2 Summary

3 United Nations

John Holmes, Under-Secretary-General for Humanitarian Affairs and Emergency Relief Coordinator of the United Nations, in a CNN interview on December 14, 2009, states:
AMANPOUR: Let me ask Sir John Holmes. There are, obviously, huge allegations about the – the force that the government employed to crush the Tamil Tigers. What are you saying? The ambassador is denying that there were 7,000 or more deaths.
HOLMES: Well, we always said that we didn’t know exactly how many people have been killed. Seven thousand was an internal estimate. The number may be lower; it may be higher. What is clear is the civilian casualties were unacceptably high. [2]  [2]
A number of months earlier, an anonymous UN source claimed that 20,000 civilians died:
Confidential United Nations documents acquired by The Times record nearly 7,000 civilian deaths in the no-fire zone up to the end of April. UN sources said that the toll then surged, with an average of 1,000 civilians killed each day until May 19, the day after Velupillai Prabhakaran, the leader of the Tamil Tigers, was killed. That figure concurs with the estimate made to The Times by Father Amalraj, a Roman Catholic priest who fled the no-fire zone on May 16 and is now interned with 200,000 other survivors in Manik Farm refugee camp. It would take the final toll above 20,000. “Higher,” a UN source told The Times. “Keep going.” [3]  [3]
The figure and the methodology reported by The Times (London) was criticised by others within the UN according to an article in The Guardian (London):
Citing a UN source, the Times reported that 20,000 people were killed in the final throes of the civil war, mostly by government shelling. However, one UN official described the method used to calculate the toll – which appeared to involve multiplying the figure recorded by doctors by five – as dangerous extrapolation. The Sri Lankan government has rejected the Times report.
Officially, the UN spokesman in Colombo, Gordon Weiss, would say only that "we have always said many thousands of people died during the conflict". But privately, UN staff admitted they were puzzled by the methodology used to achieve the new death toll.
"Someone has made an imaginative leap and that is at odds with what we have been saying before," one official said. "It is a very dangerous thing to do to start making extrapolations." [4]  [4]
It was also reported in Le Monde that the 20,000 figure was an “unaudited” “extrapolation”:
But Vijay Nambiar has been informed that the final toll "may exceed the 20 000 dead". This figure, unaudited, is an extrapolation of the limited information available. [5]  [5]
To add to the confusion, the former UN spokesman in Sri Lanka, Gordan Weiss claimed in an interview with SBS:
GORDON WEISS: I believe that between 10,000 and 40,000 is a reasonable estimate. I think most likely it’s somewhere between 30,000 to 40,000. [6]  [6]
The controversial circumstances in which Weiss’ employment with the UN was terminated politicises the issue further.
The United Nations has estimated between 80,000 and 100,000 deaths during the entire duration of the separatist war. This has been widely reported by international media organisations. e.g.
“The UN’s humanitarian co-ordination office says 80,000 to 100,000 people were killed in the war between the government and Tamil Tiger rebels.” [7]  [7]
“According to the UN, between 80,000-100,000 people may have been killed in the country’s civil war, including at least 7,000 ethnic Tamil civilians killed in the last five months of the conflict.” [8]  [8]
On 12th of March 2007, the United Nations News Centre published:
“The UN estimates that some 70,000 people have been killed and 465,000 displaced by Sri Lanka’s more than two decade-long conflict, including nearly 205,000 uprooted since fighting escalated in April 2006 despite a ceasefire signed in 2002.” [9]  [9]
Therefore, the United Nations has estimated that between 2007 and May 2009, when the war ended, that an additional 10,000 (80,000 - 70,000) up to 30,000 (100,000 - 70,000) have died.

4 International Crisis Group

Then there is the International Crisis Group’s estimate which claims that the death toll is “certainly high enough to triple the UN’s internal figure of 7,000”. Though they state it is based on “credible evidence” that there were 330,000 internally displaced persons (IDPs), no sources are provided:
“While determining a conflict death toll is always a task to be approached with caution, there are multiple reasons to believe that an estimate of tens of thousands of civilians killed is reasonable.
First, a comparison of the number of civilians known to be in the Vanni during the fighting with the number known to have reached government internment camps by the end of the fighting provides a rough estimate of those killed or missing.
Crisis Group has credible evidence that there were 330,000 displaced civilians in the second NFZ and adjacent areas as of mid- to late-February. At that same time, according to the government, there were between 33,000 and 38,000 internally displaced persons (IDPs) from the Vanni already in camps in government-held areas. That brings the total of known civilians to about 365,000.
By the end of the fighting, the highest number of survivors registered in government-run camps was 290,000 as of 26 May 2009. It is not clear to what extent this number includes the more than 10,000 who had surrendered or were detained on suspicion of involvement with the LTTE. There may also have been several thousand people who fled by boat to India or otherwise escaped the war zone but avoided being detained in government camps.
Accepting these uncertainties, there is a plausible case that as many as 75,000 persons remain unaccounted for. Even if the figure of 330,000 is reduced by as many as 30,000, or some adjustment to the difference between 330,000 and 290,000 is made to account for civilians who may have been killed lawfully because they were directly participating in hostilities at the time they were targeted – a number Crisis Group believes is very low – or to account for some number that may have avoided the government camps, it is still difficult to arrive at a figure for the killed or missing that is lower than 30,000.
Moreover, Crisis Group has evidence from various individuals who were in the NFZs until the very end of the fighting to suggest that the scale of death was much higher than reported at the time, and certainly high enough to triple the UN’s internal figure of 7,000. Crisis Group also believes that all but a small portion of these deaths were due to government fire. Eventually, an independent and impartial survey of those still living in the Northern Province will be needed to establish this part of Sri Lanka’s history.” [10]  [10]

5 University Teachers for Human Rights (Jaffna)

Lastly, the University Teachers for Human Rights (Jaffna) (UTHR(J)) reports the following:
3.5 A Note on Casualty figures
Almost any discussion of figures has been fraught with so much polemics that the truth lay buried. In the first place the Government failed to allow in international observers and severely underestimated population figures to justify harsh military action. The Government claimed the population in the No-Fire-Zone created on 12th February 2009 was 70 000. Most others agreed this was an underestimate, but in their own estimates tended to err on the side of caution. When the International Crisis Group in early March quoted the ICRC figure of 150 000, Prof. Rajiva Wijesinghe of the Government Peace Secretariat launched an acerbic attack on the ICG. When the Army breached the NFZ on 19th April resulting in about 100 000 civilians escaping from the LTTE, the Foreign Minister said that only 15 to 20 thousand people remained in the truncated NFZ. When this too was breached about 80 000 survivors came out. Against the 70 000 originally stated by the Government there must have been nearly 200 000.
Challenged by government obfuscation, anyone interested in what was going on had to take casualty figures given by the LTTE seriously as a first approximation knowing that its intention was to play up the charge of genocide. Apart from this there was a real problem with gauging civilian casualties because the LTTE figures or figures it influenced appeared to conflate civilians, conscripts and combatants. Following the end of the war, the Government claimed that nearly 23 000 “terrorists” were killed. Taking this figure, along with how the LTTE was making up its fighters, it would not be wide off the mark to say that more than 13 000 of this number were made of recent conscripts who did not want to fight and family men forced to dig bunkers. This would account for a huge chunk of the total dead. (This is why we said that a responsible government should have relied primarily on political measures that have not even been tried.)
Thus although figures attributed to the UN have been considered authentic, there are many imponderables, such as who gave the figures from the ground and whether they adequately distinguished between civilians and (enforced) combatants. From the third week of January 2009 until the end of March civilian casualties were relatively high because there was constant fighting as the Army advanced and the LTTE fired from among civilians.
The ICG citing unofficial figures from the UN gave the number of civilian dead from late January to the first week of March as 2300. One did not find it easy to reconcile this with other indicators. During the same period the ICRC transferred 1500 surgical cases from the war zone to Trincomalee Hospital. This would suggest naively about 800 civilian dead during this period. For another indicator, TamilNet reported on 11th April that hospital records showed that 599 civilians died over nearly the past month (from 13th March) and went on to comment that the total may exceed 1500 taking into account deaths not reported to the Hospital. That gives a wide margin. This is why it is viciously stupid for the Government to impugn and detain the doctors who were giving fairly objective reports with conservative casualty figures.
The Timesonline [10] report of 29th May citing UN sources, which gave the total killed for the year as 20 000, with 13 000 of the dead in May, said “the numbers killed by shelling doubled from March to April, with an average of 129 every day.” In contrast our Bulletin No.47 published on 17th April stated, “Persons in regular touch with those who have escaped confirm that an average of 15 to 20 people die each day; either killed by shells or shot by the LTTE attempting to drive fear into would-be escapees.” That huge discrepancy cannot be bridged even if one adds combatant deaths. Also averages are misleading. April was a relatively quiet month until the Army entered the NFZ on 19th April. TamilNet said its correspondent personally witnessed 300 bodies while fleeing the area and an email from one of the doctors said that 700 injured had been taken to the Mullivaykkal temporary hospital. TamilNet reported again on the 29th April and 2nd May that 300 were feared killed in the first instance and 64 in the second.
To get some idea of the casualties, after 19th April, the ICRC removed nearly 1000 patients and carers in two trips until 7th May and TamilNet reported that a further 500 injured were waiting for transportation. The next two major incidents involving civilian casualties from army action were on 9th and 15th May. On the 9th TamilNet estimated 2000 dead with 257 dead bodies and 814 injured brought to hospital. Dr. Shanmugarajah told AP that the dead and dying brought to hospital accounted for 430 deaths but the total was likely closer to 1000. He called it also the bloodiest day he had witnessed. On the 15th, a volunteer doctor reported seeing 800 bodies in the area where he moved. We quoted the TamilNet only as a useful indicator, although figures give by the doctors were, when available, significantly lower.
We may say that there was a fairly regular pattern until the end of March while the Army advanced and the LTTE and the people were pushed east. There was in the first 18 days of April a lower level of deaths due to army shelling, while the LTTE shooting at escapees continued. From 19th April there were as indicated above days on which deaths were high, 300 or more, notably on 9thand 10th May when it touched nearly 1000. On the 14th many escapees died due to LTTE shelling. On the 15th there was heavy fighting until about mid-day, resulting in heavy casualties among combatants, as well as civilians caught up in it. Allowing for some exaggeration by TamilNet, the civilian dead was probably below 1000. As discussed earlier, the heaviest civilian casualties were during the last battle from the 17th evening to the 18th morning. Wile many civilian casualties resulted from LTTE shelling, the Army too by its harsh approach on this occasion caused many deaths totalling 1000 to 4000.
In conclusion, the civilian casualties after 19th April need careful research. Speaking of averages is meaningless and of a hidden massacre is misleading. Such an expression immediately puts the government forces in mind. We know that on the May 14th and May 17th night, the LTTE was to a large extent responsible for civilian deaths. When an organisation fell apart in that manner, one section joined the civilians and escaped, another section feeling abandoned by the leadership and facing certain death from a Government that did not want to accept surrender, lost its balance and was angry with the others seeking to protect their life. We also pointed out that in giving casualty figures, the distinction between civilians, conscripts and cadres has not been clearly made.
We do not wish to underplay numbers, but focusing on numbers may result in unnecessary polemics to the cost of the deeper issues. If we avoided a worse catastrophe, it was due to unplanned factors. For the Government to launch the operation in this way was wrong and utterly irresponsible. No one could think well of a government that bombs and shells its own civilians. That point need not be made by quoting high figures of casualties. The only accurate means of finding out casualties is to count and alternatively to have a clear idea of what was happening on the ground. In their absence, technology and statistical formulae may turn out to be very misleading. Another important indicator is that the people who escaped during the last week of the conflict blame largely the LTTE, towards which their anger is directed.
What gives ground for satisfaction is that international pressure had a benign effect in protecting civilians. In the context of present war which took a heavy toll on the lives of soldiers, these ordinary men have shown remarkable restraint towards civilians when they come to contact with them. The civilians are uniformly scathing about the LTTE, and frequently found the Army helpful and considerate. So far we know of no major discrepancy in the civilian population before the fall of Killinochchi and after the war’s end. The population before was estimated at 300 to 350 thousand. Presently nearly 300 000 are accounted for in IDP camps. It would be some time before all are registered and detailed checks could be made. There is no evidence of genocide. It is hard to identify any other Army that would have endured the provocations of the LTTE, which was angling for genocide, and caused proportionately little harm. [11]  [11]

6 Darusman Report

The Darusman report states:
“E. The number of civilian deaths
132. There is no authoritative figure for civilian deaths or injuries in the Vanni in the final phases of the war. Several factors make it very difficult to calculate a reliable casualty figure: (a) the number of persons in the conflict area remains uncertain, although it was likely to have been as many as 330,000; (b) the lack of an accurate count of the number of persons who emerged from the Vanni, due to the lack of transparency in the screening process; (c) lack of certainty on the numbers of LTTE combatants, complicated further by the increase in forced recruitment in the final phase; and (d) the fact that many civilians were buried where they fell, without their deaths being registered, in some cases, unobserved.
133. Some have developed estimates based on the statistics of the injured and dead collected by the doctors, which were collated by the hospitals and the District Disaster Management Unit. One estimate is that there were approximately 40,000 surgical procedures and 5,000 amputations performed during the final phase. Depending on the ratio of injuries to deaths, estimated at various times to be 1:2 or 1:3, this could point to a much higher casualty figure. Others have put the estimate at 75,000, a figure obtained by subtracting the number of people who emerged from the conflict zone (approximately 290,000) from the estimate of the number thought to have been in the conflict zone (approximately 330,000 in the NFZ from January, plus approximately the 35,000, who emerged from the LTTE-held areas before that time).
134. The United Nations Country Team is one source of information; in a document that was never released publicly, it estimated a total figure of 7,721 killed and 18,479 injured from August 2008 up to 13 May 2009, after which it became too difficult to count. In early February 2009, the United Nations started a process of compiling casualty figures, although efforts were hindered by lack of access. An internal “Crisis Operation Group” was formed to collect reliable information regarding civilian casualties and other humanitarian concerns. In order to calculate a total casualty figure, the Group took figures from RDHS as the baseline, using reports from national staff of the United Nations and NGOs, inside the Vanni, the ICRC, religious authorities and other sources to cross-check and verify the baseline. The methodology was quite conservative: if an incident could not be verified by three sources or could have been double-counted, it was dismissed. Figures emanating from sources that could be perceived as biased, such as Tamil Net, were dismissed, as were Government sources outside the Vanni.
135. The number calculated by the United Nations Country Team provides a starting point, but is likely to be too low, for several reasons. First, it only accounts for the casualties that were actually observed by the networks of observers who were operational in LTTE-controlled areas. Many casualties may not have been observed at all. Second, after the United Nations stopped counting on 13 May, the number of civilian casualties likely grew rapidly. Due to the intensity of the shelling, many civilians were left where they died and were never registered, brought to a hospital or even buried. This means that, in reality, the total number could easily be several times that of the United Nations figure.
136. It is worth noting that the United Nations raised casualty figures in private entreaties with the Government, but never publicized its specific estimates. Government officials strongly refuted the figures provided by the United Nations, stating that the numbers were fabricated and that this was not the business of the United Nations. Publicly the United Nations referred to the “heavy toll” of the fighting on civilians, or that the casualty figures were “unacceptably high”, but that the actual figures were not verifiable. The decision not to provide specific figures made the issue of civilian casualties less newsworthy. However, this position was maintained by senior United Nations officials until 13 March 2009, when the High Commissioner for Human Rights publicly stated that 2,800 civilians may have been killed and more than 7,000 injured since 20 January, many of them inside the NFZs. Pressure from the Government of Sri Lanka and fears of losing access may have resulted in a general under-reporting of violations by United Nations agencies. Some have criticized the failure of the United Nations to present figures publicly as events were unfolding, citing it as excessively cautious in comparison with other conflict situations.
137. In the limited surveys that have been carried out in the aftermath of the conflict, the percentage of people reporting dead relatives is high. A number of credible sources have estimated that there could have been as many as 40,000 civilian deaths. Two years after the end of the war, there is still no reliable figure for civilian deaths, but multiple sources of information indicate that a range of up to 40,000 civilian deaths cannot be ruled out at this stage. Only a proper investigation can lead to the identification of all of the victims and to the formulation of an accurate figure for the total number of civilian deaths.
78 On 15 February 2010, the United Nations Country Team in Sri Lanka released a statement “for the record”, “UN Statement on former Spokesman views”, www/, stating: “The UN repeatedly and publicly said that there were unacceptably high civilian casualties from the fighting in the last months of the war, as a result of the LTTE forcibly preventing people leaving and the Government’s use of heavy weapons in areas close to thousands of civilians. While we maintained internal estimates of casualties, circumstances did not permit us to independently verify them on the ground, and therefore we do not have verifiable figures of how many casualties there were.”
79 The Government responded that it was “very disappointed and dismayed at the unprofessional nature of the press release” and that it “categorically” rejected the allegations which were “unsubstantiated, unverified and vague” and reflected LTTE propaganda. Sri Lanka Government website,, “Archives”, “Government rejects OHCHR statement that supports LTTE propaganda”, 15 March 2009. The United Nations Country Team spokesperson in a public statement on or after 20 April 2009, referred to a “bloodbath” but this was similarly disputed by the Government.
80 After the war the Government expelled the spokesperson for UNICEF who had been vocal about violations against children.
81 The section below on credible allegations relating to events outside the conflict zone and in its aftermath will be dealt with thematically rather than chronologically.
82 Throughout the final phase from January until May 2009, IDPs fled the area, although until 20 April the numbers were still relatively low (at around 50,000).”  [12]  [12]
The report concludes:
“VII. Conclusions
1. Nature and scope of the allegations
421. Both parties to the armed conflict in Sri Lanka conducted military operations with flagrant disregard for the protection, rights, welfare and lives of civilians and failed to respect the norms of international law. There is a reasonable basis to believe that large-scale violations of international humanitarian and human rights law were committed by both sides. As a direct consequence, up to tens of thousands of Sri Lankan civilians were killed and hundreds of thousands suffered immensely, including through the loss of loved ones, serious injuries, displacement and loss of homes and livelihoods. In the aftermath of the armed conflict, many were forced to endure further hardships and humiliation. ” [13]  [13]
The report clearly states that the United Nations had an internal estimate of 7,721 civilian deaths during the period of August 2008 to 13 May 2009. It then explains why this should be considered a lower-bound. The period from January until May 2009 is described as “final phase” of the war.
The report states “40,000 surgical procedures and 5,000 amputations” were made, which indicates 45,000 injured. A death to injured ratio of “1:2 or 1:3” leads to a possible 22,500 to 15,000 deaths. The report causes confusion by referring to a “ratio of injuries to deaths” when the correct term is ’ratio of deaths to injuries’.
Interestingly, the Advisory Panel also refers to:
“Others have put the estimate at 75,000, a figure obtained by subtracting the number of people who emerged from the conflict zone (approximately 290,000) from the estimate of the number thought to have been in the conflict zone (approximately 330,000 in the NFZ from January, plus approximately the 35,000, who emerged from the LTTE-held areas before that time).”
Presumably, the Advisory Panel did not find the 75,000 figure credible since it draws the upper-bound at 40,000. The 75,000 assumes that the LTTE did not suffer any deaths, that the LTTE did not forcibly conscript children and/or civilians, that the LTTE did not inflate civilian numbers and that all displaced civilians were registered after the war.
The section concludes that “as many as 40,000” and “a range of up to 40,000 civilian deaths cannot be ruled out at this stage” because “multiple sources of information indicate”. The report makes no attempt to calculate a death toll, instead it simply refers to estimates made by the United Nations in 2009 and unnamed sources.
The Advisory Panel refers to a possible civilian death toll range from approx. 7,721 to 40,000 in the final 5 months of the war.

7 United States of America - Assistant Secretary, Bureau of South and Central Asian Affairs

“ASSISTANT SECRETARY BLAKE: The LTTE, the so-called Tamil Tigers, have been on our terrorist list since 1997. One of the most brutal, lethal terrorist organizations in the world.
As the Sri Lankan army was pushing north into the Tamil areas, the predominantly Tamil areas that were controlled by the LTTE for more than two decades, they displaced, the Sri Lankan army displaced a large number of Tamil civilians and they all began to move northwards. The LTTE systematically refused international efforts to allow those internally displaced persons to move south. To move away from the conflict areas where they could have been given food and shelter and so forth. So they systematically basically refused all efforts and in fact violated international law by not allowing freedom of movement to those civilians. So had the LTTE actually allowed people to move south, none of this would have happened in the first place, so it’s important to make that point. I think that often gets lost in the debate on this.
Secondly, the LTTE often deliberately put its heavy artillery in the midst of civilian encampments, precisely to draw fire so that people would get killed in the hopes that there would then be international outrage and there would be essentially demands on the Sri Lankan government to stop the fighting and [agree to] some sort of negotiated settlement.
The Sri Lankans, not without reason, argued that the LTTE was really never interested in peace and that they had always used ceasefires as a way to regroup and rearm themselves, so they essentially refused any efforts to resume the peace process.
So we faced this very very difficult situation. On the one hand we wanted to see the defeat of a terrible terrorist organization that had been responsible for hundreds if not thousands of civilian casualties. On the other hand we wanted to ensure that there were not going to be civilian casualties as a result of this. I have to say, both sides were guilty of massive human rights violations that caused the deaths of many many civilians. I think, just to say what I said earlier, which is for this country now to recover from this experience I think there needs to be a reconciliation process, there needs to be new elections that are held in the north so that a new indigenous leadership can emerge, and I think there also needs to be some sort of accountability mechanism so that the Sri Lankan nation can put this episode behind them and that they can be confident that those who were responsible for the deaths that took place will be held accountable.”  [14]  [14]

8 Doctors

It was reported in The Guardian (London) that the UN relied on Tamil Government doctors working within the Liberation Tigers of Tamil Eelam (LTTE) controlled areas for accurate casualty figures:
Gordon Weiss, a UN spokesman, said reliable reports from inside the war zone had dried up after the "courageous" doctors who had been working out of the last makeshift hospital at Mullaivaikal East primary school were forced to abandon the building in the face of heavy fighting on Friday. [15]  [15]
Later it was reported by the BBC that the doctors claimed the casualty figures they reported were the result of pressure by the LTTE, however, this claim was made whilst they were in Government custody:
"Every day the LTTE (Tamil Tigers) people came to the hospital, they gave the list," he said. "This amount got injured, this amount dead, this area shells fell. We had to tell that list. Read it out. The list was wrong, exaggerated number."
Dr V Shanmugarajah said that on one day, some 60 people were killed but they were instructed by the rebels to say 1,000 were dead. [16]  [16]
It is completely plausible that the doctors were threatened by the LTTE to mislead the UN, but it is also plausible that they were under duress by the Government to recant their previous statements. Or the truth could be a combination of both.

9 Conclusion

Until the UN can provide some clarity about the methodology and sources used to arrive at the estimated death toll, one should not be surprised if Sri Lankans question the credibility of the UN.

10 Version History