UNIVERSITY TEACHERS FOR HUMAN
Special Report No: 34
Date of release:
Let Them Speak
The Population Game: Disappeared on Paper and Killed with Cannon
For a governments claim to have accomplished the hostage rescue with zero civilian casualties to carry conviction, it should have had some idea in advance of how many hostages it had to rescue and where they were. Minimally it had to ensure that they did not starve. It was basic intelligence, and indeed administrative work, to determine how many there were, and where. After all it is the government of these people, with administrators in the area whom it regularly met formally and informally and who would have told them where matters stood even when local records were flawed. Tamil administrators were worthy of at least that little respect. By rejecting their word the Government was deliberately or through incompetence preparing to act blindly.
The Governments cavalier attitude to the lives
of the trapped people is revealed by its unacceptably low figures for the displaced
population, and is further illustrated by its claim on 17th May before
the final free-for-all that 50 000 civilians had come out of the NFZ and all the civilians
had been rescued. After the final bash it announced on
18th May: Despite the speculations
of a bloodbath and a humanitarian catastrophe at the
final military push Sri Lankan soldiers were able rescue about 70,000 people
within the last 72 hours without causing any harm to the innocent
(defence.lk). In fact 29 000 civilians were transported from the battlefield
to Chettikulam Zone 4 from 18th May and 1400 injured
civilians to Padaviya Hospital. Civilians were coming
out of the war zone until at least
The vexed question of how many were killed in the Vanni by the Government and the LTTE is closely tied to how many there were originally. After all, we do know that from the time the welfare camps were set up in 2008, the number that had been detained there after the close of the war was 290 000. This was practically all who had survived the war. In 2008, the task of aid agencies was to ensure that the people were fed. Thus the practical way out for them was to err on the higher side and in late 2008 estimated the population at 430 000. Many of us groped in the dark and judged from past experiences or past biases. The truth was unknowable within a wide margin.
When enumerating the displaced on 4th November, the Government Agents of Mullaitivu and Kilinochchi gave the total as 348 103. The same month the UN High Commissioner for Refugees estimated there are 230,000 displaced in the Vanni. The Consortium of Humanitarian Agencies currently estimated the number to be 300 000, which HRW pointed out the UN also at times quoted inconsistently.
Curiously, the Government which was the employer of the GAs of Kilinochchi and Mullaitivu, was not sitting down with them to arrive at a working consensus, which was its duty to the people. Instead, it quoted an unrealistic figure of 100 000 IDPs and stuck to it. The war in the Vanni reached its more destructive phase for the civilians in January 2009. By early March about 37 000 civilians had escaped to the government-controlled area. The Government lowered the number of IDPs remaining in the Vanni to 70 000. It stuck to this figure until the next exodus of 103 000 civilians during 20th to 22nd April, 33 000 more than the Governments total.
Then by its curious arithmetic, the Government insisted that
only 15 000 to 20 000 people were left in the NFZ (e.g. defence.lk,
There was no consistency or any genuine information behind the Governments figures. These were just pulled out of the hat, and on the basis of these food and medicines to the IDPs were curtailed to ridiculously low quantities, irregularly delivered, causing starvation and extreme hardship. It did not treat the Tamils as citizens of this country. In a more sinister vein, by deliberately understating the IDP numbers the Government was preparing for the eventuality where it could dismiss any later suggestion of high civilian casualties by pointing out that, according to its statistics, the dead persons never existed. It was as though they had been disappeared on paper in preparation for their extinction by cannon fire.
The Government had in a way turned the discourse to suit its aims, where the international agencies, if they wanted to avoid confrontation that would be fatal to their operations, were best advised to moderate their figures. In January 2009, the UN spoke of 230 000 IDPs on the move and the ICRC of 250 000. When the final NFZ was created in February, the Government stuck to 70 000 IDPs in the NFZ. The UN said in a briefing to the diplomatic community in Colombo on 9th March 2009 that on the basis of satellite imagery there were at least 100 000 to 150 000 in the NFZ. The UN also pointed out that the food being supplied was less than a quarter of the monthly 3000 tons needed to feed 200 000 people (i.e. barely adequate for 70 000). The International Crisis Group (ICG) in a statement in early March quoted the ICRC as saying that there were 150 000 people in the NFZ and got a bloody nose from the Governments Peace Secretariat.
There had in fact been well above 250 000 in the NFZ, taking into account the 37 000 in IDP camps in early March and 290 000 on 25th May. How much more, is a question we need to answer as part of determining how many died in a straightforward manner? It is interesting that INGOs who were planning for a Vanni population of 430 000 in late 2008, also talked about 300 000 and 250 000, finally came down to 150 000 in the NFZ and 37 000 in welfare centres. We were ourselves guilty of understatements and mix ups, the sum of which allowed the Government to manipulate the discourse.
Sadly, not many of us took seriously the figures given by the abused and abased GAs and AGAs, who were trying to do an honest job within their constraints and now dare not speak. We were given a jerk by a senior community leader who had been in the NFZ and told us of a conversation he had in early March 2009 with a high ranking government administrator. The latter told him that from the checks he had made through the village headmen (GSs) of the Vanni who were nearly all there, there were 330 000 civilians in the NFZ. The community leader asked him, With or without ghosts? The administrator assured him, Without ghosts. Ghosts refer to padded numbers on official lists enabling the siphoning off of part of the supply to feed the LTTE. The community leader told us, I cannot prove it, but I believe the administrators figure was correct.
The community leader gave us another jerk by assuring us that a minimum of 6400 civilians were killed up to the end of March. (The UN briefing referred to above suggests 4800.) What we gathered is that we must be prepared for much higher casualty figures than are commonly talked about. The UN briefing gives us a hint.
We must return to the question we began with. The surest way to find out how many died, a question the Government does not want to investigate, is to compare the number of those who arrived in IDP camps with an authoritative estimate of the original Vanni population. The best figure we have for the latter is about 365 000 given by government administrators in the Vanni to persons they trusted. These figures have by far shown themselves credible in comparison with the official Colombo figures.
Given the huge vacuum in information, we need to proceed carefully in giving any estimate of the dead. The two sets of reliable figures we have are, one, those collected by the local office of the UN OCHA for the dead and injured in January, February and early March. This is an important indicator but is subject to interpretation before we could draw any conclusions. The second important set of figures are the casualties shipped by the ICRC, obtainable from the ICRC web site and the Indian High Commission web site (http://www.hcicolombo.org/events_Detail.cfm?id=66 ). Local information from the ground suggests that two persons died for every casualty shipped by the ICRC.
Another important figure is the 330 000 given for the number of people on the run by a Senior Government Official (SGO). The gap between this figure and those eventually in IDP camps is more than 40 000. As to whether these give an indication of the total dead requires careful consideration. Persons interested in such matters who were in Manik farm tell us that a number got away by paying money or through sneaking away without leaving a record with the Government.
The Government originally maintained from September 2008 that there were only 100 000 civilians in the war zone and by March reduced it to 70 000. Food and medicine could be brought only with government approval. It was clear that the numbers were grossly understated and the UN and INGOs were helpless. They asked the SGO how many there really were so that they could make a case. It was sometime in February that the SGO began working through his subordinates to compile a census of those in the war zone.
Persons who moved closely with SGO told us that he did a thorough job and have no doubt about the integrity of his figures. But they said that the figures need to be qualified by two factors. One was that even as the figures were being compiled, people were dying and escaping. The other is that the figures included LTTE cadres and conscripts who were on family lists.
The second and important matter is that the Government was totally denying civilian casualties and there were no expatriates in the war zone. Mr. G, a political science graduate of the University of Jaffna, was the head of the OCHA office in the war zone, which was originally close to PTK Hospital and was later shifted close to Putumattalan Hospital. Those who moved with him and spent time with him in bunkers attest to his absolute impartiality and the professional integrity of casualties he compiled.
Persons who moved with Mr. G told us that he went beyond the call of duty, actively to maintain contact with all the makeshift hospitals and collect casualty figures on a regular basis. These figures were the basis for the briefing given to diplomats in Colombo on 9th March 2009.
OCHA gave figures of the dead as nearly 3000 from January 20th to March 8th. But the figures bear closer examination.
No. of civilians injured in January 2208 Killed in January 420 Dead /Injured = 1/5
No. of civilians injured in Feb. 4756 Killed in Feb. 1914 Dead /Injured = 1/2.5
No. of civilians injured in Mar. 1072 Killed in Mar. 536 Dead /Injured = 1/2
The above is reflected in the way the figures were collected. During January and February the people were very much on the move and many of the dead would not have been reported to the hospital and hence the low dead/injured ratio. The figure is higher for February and stabilises at 1/2 when conditions are more stable. Thus the figures collected for the dead are unreliable and it is from the injured that we must try to get some idea of the number of dead. We must also make corrections for the injured who die for the lack of adequate medical care and the major casualties who did not receive transport to a proper facility in time (there was no ICRC transport from 27th January to 10th February). Making these corrections, the total dead for January and February comes to over an estimated 6000. 
The dead for March estimated from ICRC shipping data comes to more than 2000. This corresponds closely with the OCHA figure of an average of 67 persons being killed daily during early March.
This makes the dead from January to March, excluding those killed by the LTTE, more than an estimated 8000. We compare this with other figures given to us.
The senior community leader quoted above gave the dead for this period as a minimum of 6400
A TRO official told a friend confidentially that based on dead bodies they buried (60 to 80 a day) they have a record of just about 4000 dead for the same period. The source also told us that TRO did not reconcile this with other sources.
At our request through a mutual friend, a lady doctor in the LTTEs Medical Corps, who had a lot of field experience during this period gave her estimate of the total dead as 35 000 to 37 000. Her breakdown was January to March 10 000, April 10 000 and May 15 000.
We find her estimate for April to be high with regard to estimates from casualties shipped by the ICRC together with perhaps another 1500 killed in the April 20th rescue operation, but her other estimates are quite plausible, especially the one for May.
9th to 20th May is the most difficult period. The one day about which we have some information is the shelling of 9th to 10th May. Dr. Shanmugararah estimated the number of dead at 1000. The fact that about 430 injured died in hospital suggests that the dead could have been as high as 1500. The Defence Ministrys web site defence.lk told us that the remaining area to be captured on the 17th May evening was ¼ sq.km. This means 35 000 people (as could also be checked from (http://en.wikipedia.org/wiki/Sri_Lankan_IDP_numbers,_May_2009 ) were in that small area receiving the full force of the Armys cannon. If one makes a naļve comparison with late January where about 300 000 people in a safe zone of 35 sq. km were being pummelled by the Armys cannon, the casualties during the latter stages in May would have been several thousands. Persons on the ground estimated over 4000 dead on 17th/18th May, in agreement with a naļve calculation.
Also the number killed by the LTTE would also run into several thousands. LTTE cadres and conscripts killed based on a working estimate of 50 a day from those who had LTTE contacts, works out to nearly 7000 from January to May 2009.
We are so far not able to obtain random samples from most communities who were displaced and even from the same area, there are wide variations. Subregional variations arise from the fact that they stuck together when displaced and often made their break for army custody together. Those from Mannar District generally made their exit during January and February. Unlike those in Killinochchi and Mullaitivu districts, they are long experienced in displacement and are acclaimed by those who know them to be better survivors. The heaviest casualties were among those from Mullaitivu and Killinochchi Districts.
Metaphorically, the talk among IDPs is that those who got out in January and February are the Grade Eights in terms of experience, those who got out from Puthumattalan, the O. Levels, those from Valaignarmadam the A. Levels and those from Mullivaykkal the university graduates.
Another group that exited early in January and February comprises those from areas like Chundikulam that are a continuation of Vadamaratchi East. This group along with those from Mannar have been the first to be resettled as deaths among them were significantly less, making it easier for the Government to put up with casualties. The Government would be much more reluctant to allow free access to those from Killinochchi and Mullaitivu.
Resettled in Jaffna
A social activist having contact with those resettled in Jaffna confirmed that most of them exited from LTTE control in January and February. He suggests that there are at least 10 dead among 20 families. Among the 10 dead, he places 7 as due to army firing and the remaining 3 as killed by the LTTE when trying to escape.
Given there were 60 000 families in the Vanni, the sample suggests that the total dead are numbered in tens of thousands.
Resettled in Mannar District
The samples below show significant differences in the same district.
Of the 60 families settled in Seevivinayagar Kulam, Adampan, there have been 22 deaths from August 2008 to April 2009. Of the dead, three were LTTE conscripts. The rest were killed by shelling mostly in 2009. Two families suffered most or several members killed (see footnotes below).
44 families resettled in Manayankulam report 8 civilians killed during the war along with 15 conscripts from the same village.
The 144 families resettled in the Kovilkulam and Athimoddai area report 20 dead in the war and in addition, five conscripts killed.
The difference is explained by Manayankulam being a poorer area and Kovilkulam one of relatively rich farmers, who managed to get many of their children away and protect them. We also note that basing conclusions on resettled families is misleading. Many families who suffered heavily may never come back at all.  Others have come back considerably depleted and fragile. 
We have at present little to go on, but there are also other indications of a high level of distress. A resident of a block in Manik farm having 214 families told us that there are 56 widows. Finally we go back to the SGOs figure of 330 000, which may represent something higher as people were moving out during the compilation. On 9th February 2009 there were already nearly 15 000 in IDP camps, which rose to 36 000 by 25th February. The highest number recorded in IDP camps is 290 000 on 25th May. This leaves a minimum of 40 000 to be reconciled as dead or missing.
Anything that deserves to be called a government should leave no stone unturned to measure the scale of a distress affecting its own people in its various aspects; such as the numbers dead, disabled, widowed, psychiatrically affected and children impaired or handicapped in a variety of ways. This is key to rehabilitation. This so-called government has been sitting on the problem, jailing the IDPs, tying the hands and sealing the lips of government officers who served in the Vanni, and keeping out foreigners and INGOs purely to suppress.
Given direction and allowed a free hand, experienced government administrators who served in the Vanni have the capacity to account for the population and make records of what happened to every individual family, with the help of local universities and NGOs. And these people can be rehabilitated only under robust political autonomy and not by political zamindars from Colombo or foreign agencies who would inevitably become patronising when operating in a political vacuum.
Apart from the problem of quantifying the suffering, there are also humanitarian issues of an unprecedented nature. Under previous governments, even when the defence establishment was ruthless with the civilians in war, other ministries dealing with food, rehabilitation and medicine were often understanding. There was a tacit acknowledgment that they were dealing with people of their country. Even leaders like J.R. Jayewardene undersood the political cost of starving war affected people.
The UN briefing on 9th March said, We need to send to the NFZ at least 3,000 MT of food per month for a caseload of 200,000 people. The UN had been delivering more than this by road until mid-January. From mid-February to mid May, the ICRC was allowed to deliver (ICRC statement of 12th May) just 2350 tonnes of food instead of the nearly 10 000 tonnes required for an average population of 200 000 over the final three months. The justification for this was apparently the Governments fictitious population figures. Only 50 tonnes of food reached the NFZ in May because the shelling did not permit the ICRC to make more deliveries or pick up more casualties. The civilians themselves had confirmed that they were starving during the latter days. Even the quantities of medicines appear to have been determined by the Governments position of zero civilian casualties and a grossly underestimated population.
There were supposedly independent ministers in charge of these subjects who should have done their homework. They must answer. There must also be a transparent and credible inquiry into how many there were in the Vanni and a breakdown of what happened to them. Unless this step is taken and greater openness is forthcoming, rehabilitation would remain a fraud, however much money foreign governments pump into to it.
Content / Part I/ Part II/ Part III/ Part IV/ Part VI/ Part VII/ Part VIII
 Estimates from the ground suggest that about 25% of the injured (particularly
with head or stomach injuries) died in the hospitals and surroundings. These
sources also said that about 40% of the casualties are normally shipped by
the ICRC. 5/8 (62.5%) of casualties shipped are major casualties and
 This is a family that largely died out. Muthusamy from Seevivinayagar Kulam, Adampan was killed by an elephant some time ago. His wife Mallikadevi was in Puthumattalan with their four children. During a shell attack in April 2006, Mallikadevi perished along with three children. The one child who survived is with the grandmother.
 Munian Sivalingam and his wife Chandra of Seevivinayagar Kulam in the Mannar District had six children. Chandra and three children were killed in a shell attack in Puthumattalan in April 2009. Three children survived and are with the father.
Nadarasa (57) and Paratharani
(51) is a couple from Vamadevapuram, Adamban. Their children were Revathy
(30), Kanesamoorthy (28), Priya
(26), Brintha (24), Premila
(20) and Vanimalar (19). Vanimalar
was conscripted by the LTTE in October 2008 and was killed 4 months later.
Paratharanis mother Pattiyamma
went missing during the displacement. Revathys
husband was shot dead by the LTTE in Mullivaykkal
Home | History
| Briefings | Statements
| Bulletins | Reports
| Special Reports | Publications
Copyright © UTHR 2001