General Population Census of Cambodia 1998

The Demographic Survey of 1996 marked the beginning of the preparation for the census. One of the most important pre-census activities was to develop maps at various administrative levels. For the purpose of enumeration, each enumerator was to be given an enumeration area (EA) with definite boundaries. An EA, with an average size of about 100 households, might be a village or part of a village (in the case of large villages). Starting from December 1996 the work of preparing village sketch maps and delineating the Enumeration Areas was completed in about a years time.

De Facto enumeration
The census was conducted on a de facto basis. In other words, all persons staying in Cambodia on the Census Night were enumerated in the place they were found present. These persons included foreigners but excluded foreign diplomatic corps and the like and their families. Officials of Cambodia's missions and their families living in other countries were enumerated by mailing questionnaires to them with the help of the Ministry of
Foreign Affairs.

Canvasser Method
The method of enumeration adopted was the canvasser method. An officially appointed enumerator visited every household in the allotted EA and collected for every person the information required. This information was recorded on the census questionnaire.

Geographical Frame
The Geographical frame for the 1998 census followed the defined structure of province, district, commune and village. There are 24 provinces in the country which include the municipality of Phnom Penh and the "Krong" or "Towns" of Preah Sihanouk, Kaeb and Pailin. These provinces are divided into 183 districts which are subdivided into 1,609 communes containing 13,406 villages. Census enumeration covered the entire inhabited
geographical areas with the exception of a few areas which were inaccessible during the census due to military operations. These areas were: (i) Whole districts of Anlong Veaeng in Otdar Mean Chey province, Samlot in Bat Dambang province and Veal Veaeng in
Pousat province (ii) Ou Bei Choan village of Ou Chrov district in Banteay Mean Chey province. The population in these excluded areas is estimated to be about 45,000. The refugee population who immigrated to Thailand at the time of the census were not included in the census as it was conducted on a de facto basis.

Census Decree
The 1998 Census was conducted under the authority of the Royal Decree (Kret) No. JS/RKT/02-96/08 dated February 29, 1996 on the organization of the General Population Census of Cambodia. According to this Decree, the Ministry of Planning is responsible for the General Population Census with the National Institute of Statistics (NIS) as the implementing agency. The Decree inter alia ensures confidentiality of personal information collected in the census.

Organization Structure
The National Committee for the General Population Census of Cambodia constituted by Sub-decree No. 12 dated January 31, 1997, under the chairmanship of HE Deputy Prime Minister and Co-Minister, Ministry of Interior, had the responsibility for instructing, deciding and endorsing all general items of work pertaining to the census. The Census Technical Committee with HE Minister of Planning as chairman dealt with technical issues relating to the census. In the conduct of the census, the population census office of the National Institute of Statistics (NIS) under the Director of Census was the central unit that served as the monitoring, communication and action centre for the census. It had seven divisions each in charge of a specific activity, like administration, planning and monitoring etc. The Province Director of Planning acted as the Province Census Officer (PCO) for his / her province and was responsible for the operations in the province. A middle level officer of the NIS was posted at each provincial headquarters as Regional Officer (RO). Trained junior officers of NIS were posted in provinces as Assistant Regional Officers at the rate of one for three or four districts. The Regional and Assistant Regional Officers acted as technical advisers to the province census officers. A Provincial Census Committee, with the Governor of the Province as chairperson, was established in each province to assist in implementing the census successfully.

Recruitment and Training
About 25,000 enumerators and 8,350 supervisors were appointed to do the enumeration by visiting each and every household. They were drawn mostly from teachers. Training on the census procedure and filling-in of questionnaires and other relevant census forms (with practical exercise) was imparted at three levels. Training materials were prepared carefully in advance. These materials included instruction manuals to enumerators/supervisors, training guides for trainers, flip charts and training forms. At the first level, training was given to about 70 trainers' trainers, who were the core census staff and the staff of the Ministry of Planning and NIS earmarked for the census work. They were trained at NIS for three to four weeks.Training of enumerators and supervisors was the third level training held for a week at training centers situated in district headquarters / selected communes. In all, there were 505 training centers in which 961 training sessions were held.

Data Processing
The main post-census activity of the Data Processing Division was the manual coding and editing of census documents, data entry from the questionnaires, computer editing and tabulation and the generation of various dissemination products.

Field Supervision
The census was closely supervised by census supervisors and commune, district and province officials. National and International staff as well as the Minister of Planning and UNFPA Representative visited several areas to oversee the census operations. A team consisting of three international observers who inspected the census operations in different parts of the country, concluded that the census coverage and quality was generally good.

Post Enumeration Survey
The census was closely followed by a Post Enumeration Survey (PES) conducted in sample areas to evaluate the census. This was necessary as in a large-scale exercise like the census, it is inevitable that some inaccuracies arise from deficiencies and errors in coverage and response. A separate report on the findings of the survey will be published. The most important finding of the PES was that the net under-enumeration in the census was 1.78 percent. This is considered to be good.