Commentary on the Census on human and property damages due to conflict 2013 by CPA
Executive Summary
The results of the ‘Census on human and property damages due to conflict - 2013’ are due to be published in March 20141, in time for the 25th session of the UN Human Rights Council in Geneva from 3 - 28 March 2014. However, despite the Director General of the Department of Census and Statistics claiming that this census will “help arrive at an exact decision on the persons missing and died during the period of conflict in the country”, it is evident from the methodology juxtaposed against the fear and intimidation people feel in the Northern Province alone, that the ‘Census on human and property damages due to conflict - 2013’ will not reveal the true extent of death, disappearance and damage caused by the war.
Highlights from this report
•Standard census methods were utilised for this census which had a different objective to that of a traditional census. By using standard census methodology the census department would have failed to meet the primary objective of this census - which was to ascertain the exact total number of people who died and went missing, as well as the extent of damage to property due to conflict. In order to avoid double counting, a household could only provide information of their nuclear family and of relatives or other persons who permanently reside in that household and do not have a permanent household elsewhere. This meant that one could not provide information about another family member who had a permanent residence elsewhere, which is important in a standard census. However, in the ‘Census on human and property damages due to conflict - 2013’ what happens to the information about entire nuclear families that died or went missing in the war? If they lived in a separate permanent residence, any relatives who are still alive and able to provide their details would not be able to do so.
For example, if members of household X included father and mother, and household Y included their married daughter, daughter’s husband and their 2 children, this is counted as 2 family units living in 2 permanent residences, even if the 2 houses were next to each other. If all members of household Y died during the war, then there is no one at present to answer on their behalf. Their family who belong to household X cannot answer in their own census schedule either as household Y was a separate permanent residence. Therefore, the 4 people who died in household Y would not be counted in any census schedule and would not be represented in the total death
!toll as per this census.
Attempting to capture information such as this poses several challenges and there is never a complete guarantee that every single affected persons’ information is recorded. It is vital to get as close as possible to an accurate figure without compromising methodology and quality, and every effort should be made to do so in any meaningful census. The ‘Census on human and property damages due to conflict - 2013’ clearly did not.
•However, it is important to note that even if the methodology was more successful in including those affected by the war, the intimidation and fear still faced by people in the North and East, the militarisation in the North and the harassment and violence faced by families of ex-combatants over the years would have still prevented respondents from providing complete and accurate ! information about family members.
While some households chose to not even mention the deaths and disappearances of certain family members for fear of reprisals from military or intelligence officials and left them out of the census altogether, other households held back on information such as on who was responsible for the deaths/disappearances. Even those who knew the answers to these questions for certain, selected ‘don’t know’ or ‘other’ as they were not sure about who had access to this information or who would see the completed forms and did want any information that might cause them problems in the future to be mentioned in the census schedules. Grama Niladharis who were the data collectors for this census also expressed concern about being privy to sensitive information that many of them were not aware of prior to the census - some even stated that they went to the extent of selecting ‘don’t know’ or ‘other’ answer categories when the answer given by the household was different, in order to safeguard themselves as well as the families.
•The bias evident in certain sections in the L2 census schedule was cause for criticism by people as well as Grama Niladharis. In the sections that asked about death, disappearance, injury or disability of family members, one question that is asked is about who is responsible for it. Answer option number three is ‘groups suspected to be security forces’ and is the only option that can be selected even if the household is completely certain that the death or injury was caused by the military. Some individuals even know specifically that it was the Army, Navy or Air Force who was ! responsible - however there is no provision to capture this information.
When asked if specific answers like ‘Army’, ‘Navy’, ‘Air Force’ etc would have made a difference in people reporting the truth - the Grama Niladharis said that there still would have been households reluctant to answer the questions directly due to fear but that the inclusion of these answer categories would have been a sign of impartiality of the Government, which would have given this exercise more credibility in peoples eyes.
•Cause of death/ injury/ disability - Interviews revealed that the fact that ‘air strikes’ was not one of the answer categories for this section had been problematic and Grama Niladharis had filled either ‘(4) Other bomb attack’ or ‘(8) Other’. The fact that air strikes had not been included as an answer category made people question as to whether the Government was trying to cover up events as air strikes had caused many deaths and injuries during the war in the North. A good example is the air strikes on 9th July 1995 by the Sri Lanka Air Force where 65 people were killed and 150 wounded while seeking refuge at the Church of St Peter and St Paul in Navaly, Jaffna3. This is just one example of many air strikes that took place and an answer category that should have been ! included if the census schedule went into details such as ‘set fire to’ or ‘acid attack’.
•The lack of public awareness about the census and it’s objective at the time it was conducted is cause for concern. While people in war affected areas are used to information being collected from them regularly from multiple State actors, including military, this census captured a lot of detailed information which caused a lot of distress to respondents who had to revisit and recall traumatic memories. Furthermore, it was reported by some Grama Niladharis that the easiest way to get information from households had been to inform them that there would some form of relief or compensation made by the Government using the information provided by them - raising the hopes of many. If people had been more aware about the census and why it was being conducted, if there had been more information and assurances available about how the information would be used and who had access to it - respondents may have been less fearful or suspicious and more forthcoming.
•Another cause for concern is in the data collection itself. The data was collected in the North and East by Grama Niladharis (with no reports thus far of any military involvement in data collection) who got just a one day training and had no prior experience of carrying out such extensive surveys of this nature. Even though the detailed handbooks are extremely useful and the methodology and each item in the census schedules are explained in detail, the time period allocated for data
3!ICRC Resource Centre, ‘Sri Lanka: displaced civilians killed in air strike’, 11 July 1995 collection did not leave much time for Grama Niladharis to consult the handbook when in doubt. The data collection period was 20 days and a Grama Niladhari in the North and East had on average 500 - 600 affected households to gather information from. The time spent at each household being a minimum of 30 minutes due to the complexity of information gathered and recalled, there was not a lot of time for them to familiarise themselves with every aspect of the handbooks. Given that this was not a standard census and is one that relies heavily on collective and subjective memory (a lot of which is traumatic for most of the respondents), a more rigorous training process over a longer period would have been ideal, in addition to the data collection
period being at least a month.
In a paper titled ‘Conducting Censuses under Challenging Situations, Crisis and post Conflict’4 that reviewed the different challenges facing census operations in situations of post crisis and post conflict, drawing from the experience of operations supported by the United Nations Population Fund (UNFPA) in different countries, there are certain measures that had been taken by countries such as Sudan and Kenya that Sri Lanka could have also adopted or revised to suit this census. Referring to types of adaptions taken in census taking exercises, it highlights the need to “build into census strategy, management and planning, some mechanisms that enhance credibility and secure acceptance of census results in a social and political environment where controversy can seriously damage the legitimacy, acceptance and utilisation of the statistical data”. These mechanisms include vigorous publicity and communication campaigns during the months preceding the census, leaders from all political parties seen publicly supporting the census and advocating for full and accurate information to be provided and engaging national as well as international census monitors, to provide independent, unbiased assessments on the reliability of the census process and its ability to produce accurate results.
The numbers that will emerge from this census is crucial to the Government to counter the 40,000 death toll quoted in the report commissioned by the UN Secretary General Ban Ki-Moon5 in 2011, and even the number of disappearances. P.B. Abeykoon, Secretary, Ministry of Public Administration which supervised the census, at press conference in November 2013 stated “A lot of people have come out with various accusations with their own figures. We will come out with the real facts,”6. In the lead up the release of the preliminary data, State media quoted a Census Department official who claimed that the data for the number of disappearances used by the TNA and NGOs has been exaggerated and now that the Government has the correct figures, it will be able to prevent any adverse effect on the country7. This will probably be the strategy adopted by the Government and State media once the census results are released as the data will prove to be very useful to counter against accusations of war crimes, civilian deaths caused by the Government forces in the last stages of the war, military involvement in disappearances.
See Full report here