Eskinder Debebe

Seeking Accountability in Sri Lanka

In November 2011, Sri Lanka’s Defence Secretary Gotabhaya Rajapaksa said that discovering how many civilians “died or went missing during the last few months of the conflict” would be “the first step towards reconciliation.” The government’s Enumeration of Vital Events (EVE) attempted to answer that question by collecting information about people who have died, disappeared and emigrated from Sri Lanka since 1982. The survey was overseen by the Ministry of Defence but was conducted by the Department of Census and Statistics (DCS).

More recently, Sri Lanka’s Permanent Representative to the United Nations (UN), Ambassador Ravinatha Aryasinha, said the following at the UN’s Human Rights Council (HRC): “Addressing concerns relating to accountability, the only authoritative and credible source of information relating to the demographics and those killed and untraceable in the Northern Province during the period 2005 – 2009, the Enumeration of Vital Events 2011 in the Northern Province (EVE 2011) by the Department of Census and Statistics (DCS) conducted during the months of June and July 2011, has estimated that the total number dead from 1 January to 31 May 2009 is 8998, including deaths caused due to old age/sickness, natural deaths, deaths due to accidents/homicides/suicides and other causes.”

Ambassador Aryasinha’s recent remarks, the present administration’s hostile stance towards accountability or a proper recounting of the war’s final phases and the total collapse of the rule of law have compelled TSA to revisit Sri Lanka’s Enumeration of Vital Events.

The Methodology

In 2011 the Department of Census and Statistics (DCS) held population enumeration activities specifically for the Enumeration of Vital Events (EVE) in the country’s Northern Province (Jaffna, Kilinochchi, Mannar, Mullaitivu and Vavuniya). A form designed by the DCS was used for the survey. The form included the general census information, like the names and information (including passport numbers) of people living in their house permanently or temporarily. There are sections for family members who live in other districts or for those who have emigrated. Under that section, the form asks for the personal information of the people who have left, where they have gone (domestically and/or internationally), and why they went.

There is a section, special to the Northern Province, about deaths and disappearances. Respondents are only supposed to discuss their family members, and not domestic aids/laborers or boarders (tenants) who may have died or disappeared while living with them.

The form asks for the following information for the deceased/ disappeared: ID card number; Relationship to the Respondent; Sex; Whether they died or disappeared; Age when they died or disappeared; Date that they died or disappeared; Whether a death certificate was obtained for the person. For deaths, there is a section for the cause of death. The options are natural causes, the 2004 tsunami, other natural disasters, an accident, or other. There is no option for death due to war. For disappearances, there is a field to state where the person disappeared.

The Department of Education (DoE) was willing to provide teachers with leave to attend one-day training and also to collect data. However, the DoE was only willing to give teachers short leave or duty leave to conduct the survey. This suggests that the lack of time given to teachers made it difficult for all of them to conduct the survey completely. According to the official report, “field data collection required for the project covering the entire Northern Province was planned and executed within the period of two weeks from 3rd July to 17th July 2011.” However, the survey was extended an additional ten days and ended on July 27, 2011. The survey was extended because some teachers were unable to finish their work during the allotted time period. For its own survey, why could the DCS not get its numbers right?

Additionally, it appears that not all teachers were compensated for their work – in spite of promises that had been made by DS officials working in the five districts. Teachers were promised that they would be paid 3,000 Sri Lankan rupees for completing the first hundred forms. Teachers who completed more than one hundred forms were paid extra. While it remains unclear how many teachers have been paid, some definitely have.

Those who have been paid were asked to come to the closest DCS department – where they were paid in cash. Teachers were paid at varies times, ranging from September to December 2011. This occurred in all five districts in the Northern Province. Anecdotally, one teacher from the Vanni (who completed more than one hundred forms) was paid 5,215 Sri Lankan rupees. Teachers were paid thirty rupees for the first hundred forms they completed. For any additional forms completed, they were paid thirty-five rupees per form.

Data Collection

Government schoolteachers collected the data by going door-to-door and using database forms. As noted, prior to collecting data, schoolteachers participated in a training program. In addition, on June 17, 2011, a one-day “Master Training” seminar was held at the Jaffna District Secretariat for Tamil staff of the Department of Census and Statistics (DCS) based in Jaffna, Mullaitivu, Kilinochchi, Mannar, and Vavuniya. Since Sri Lanka had not conducted an island-wide census since 1982, this was a capacity-building exercise for DCS staff. Army officials visited the DCS in Jaffna during that one-day training. Staff from the DCS were brought to work on the project from other parts of the country (outside the Northern Province) on June 22 and 23, 2011.

Significantly, most teachers had no prior experience with data collection or survey work. Once teachers collected the data, those database forms were collected by officials working in the DCS’s Colombo office. In order to collect the data and bring it back to Colombo, the DCS sent five separate teams of Colombo office staff to the Northern Province’s five districts. The size of these teams varied according to the number of DS divisions located within each of the five districts, with one additional person being designated as the “team leader” for each group. For example, the team that traveled to Kilinochchi (a district with 4 DS divisions) was comprised of five people working in the DCS’s Colombo office. The remaining teams are numbered as follows: Mullaitivu (6 people); Vavuniya (5 people); Mannar (5 people); Jaffna (15 people).

DCS teams in Mullaitivu, Kilinochchi, Mannar, and Vavuniya all stayed in army camps during that time. While those four teams had originally planned to spend three weeks in those districts, DCS staff spent an entire month at those army camps. More specifically, in Vavuniya, the DCS staff who came from Colombo stayed at the Kudiyiruppu Army Camp while the survey was conducted. In Jaffna, DCS staff who came from Colombo stayed at a lodge; this was coordinated by the Ministry of Defence. DCS staff stayed in Jaffna for two weeks. To be clear, all five Colombo-based DCS teams were in the Northern Province during the entire time that the survey was conducted for their respective districts.

Evaluating, collating, analyzing and finalizing the data were done exclusively by people working in the DCS’s Colombo office. No district level government officials were involved in this process. Importantly, no real information about this survey was kept in any of the DCS’s district offices in the Northern Province. The only information remaining in those district offices was a note about the number of forms that were used in each district to conduct the survey.


There are numerous reasons to be concerned about this survey. The timing of the survey – and the military’s heavy involvement – suggests that this was (and remains) an obvious effort to deflect international pressure at the HRC and other international fora. When and why did the DCS begin to work so closely with the Defence Ministry? The survey may also have been an attempt to discredit the UN’s Panel of Experts report. Indeed, the government furtively released the EVE in February 2012 but did not hesitate to mention it on the opening day of the HRC’s 19th session.

Usually, a data collector in the field will write a summary of the information that they collected at the end of the survey period. But, this wasn’t done for the EVE – thereby creating an opportunity for the government’s Colombo representatives to manipulate data as they see fit.

The conspicuous omission of war as an option for the cause of death is perhaps another sign that the government is merely using the EVE as a tool to underreport the number of deaths that took place during the war. For the survey form which was written in Tamil, in the “Reason for Death” Category (Column 11, Section 4) the direct translation cites “instantaneous accident” as an option. However, in the English version of the survey, that same part is categorized as “Accident/Homicide/Suicide.”

This awkwardly worded portion of the survey form in Tamil is confusing and unclear. Evidently, the GoSL has major problems when it comes to translating official documents into Tamil. The ambiguous phrasing provides ample opportunity for respondents to omit deaths which occurred as a result of the war.

Furthermore, the form itself invites manipulation. In the sections related to people residing in the home and people who are living in other districts or abroad, the respondent can fill in information for their extended family – no matter how far away they are living. However, for deaths and disappearances, the form says that the respondent can only report on deaths/disappearances from their primary family. This invariably leads to the underreporting of deaths – compared to people who have left the region and are remaining there.

Doing a survey in this manner would unvaryingly lead survey information being both inaccurate and massively underreported. Was a thirty-year war not deserving of its own category? The people who are in the districts now are not necessarily the same people who have been there since 1982. Whole families were killed, displaced to other districts in Sri Lanka, or emigrated from the country entirely. When it comes to the death toll, none of these families will be counted in the survey. This inevitably results in thousands of deaths and disappearances not being reported.

It is also highly unlikely that all respondents gave completely honest answers to the data collectors. In the North, there is a justifiable fear of the military and, by extension, the government. Many people are afraid to admit that a family member died as a member of the LTTE. Some might have lied about the circumstances of the death(s), or even that family members died at all. Moreover, many community members in the Northern Province might prefer to say that their loved ones have disappeared, instead of acknowledging that they have passed away. Many women do not want to accept the fact that their husbands are gone forever if they are not completely sure. Some are fearful that, by making a death official, they are killing their husbands.

Community members are especially afraid to share information about family members that have emigrated. They are afraid that the government will assume that the emigrating family member is LTTE and that they may be also. They believe that if they admit that a family member left they will be targets of harassment, extortion, and/or investigation by the military. These fears are well-founded as the heavy monitoring of ex-LTTE cadres continues in the country’s North and East. The fact that the government collected the passport numbers of people in the survey points to an effort to identify those who would potentially travel abroad and prevent them from doing so. It may also be an attempt to identify possible LTTE supporters.


The processes surrounding the Sri Lankan government’s EVE are deeply flawed. The EVE’s survey and subsequent statistical analyses lack rigor and are shrouded in opacity. The EVE has resulted in the production of highly questionable information. More than four years after the conclusion of the war, the present administration maintains that far fewer deaths (less than 8,000) occurred during war’s final phases than most people think. As TSA noted more than a year ago, the government’s dubious claim lacks merit and should not be taken seriously.

This article has been adapted from a previous work which TSA circulated anonymously in February 2012, “Behind the Numbers: a Closer Look at Sri Lanka’s Enumeration of Vital Events – 2011 – Northern Province.”