Population and Housing Census 1996

In general, the mode of data collection was face-to-face, but in some cases it was done by mail.

The proposed Superintendent Areas were drawn up and sent to respective Superintendents along with information on the number of Supervisors and Enumerators to be hired. This was followed up by visits from a senior official of the census office, who during organised meetings in the various divisions, further briefed the appointed Superintendents of the whole census operation. During these meetings the Superintendents were urged to begin identifying suitable Supervisors and Enumerators in their areas. The importance of hiring suitable 'locals' was emphasised in view of the importance of local area knowledge in the enumeration exercise, and this was generally followed in the rural areas.

A Superintendents Briefing Conference was held in Suva which basically was their training for the census. Superintendents Areas and expected expenses were finalised and Superintendents appointment letters were given during the Superintendents Conference. Following the Superintendents Briefing Conference, Superintendents were instructed to finalise recruitment of supervisors and enumerators as well as training arrangements. Divisional Liaison Officers were mobilised to direct and assist in overseeing this exercise. This was followed by the enumerators training which was conducted by their respective supervisors under the direction of Superintendents.

A total of 38 Superintendents, 295 Supervisors and 1679 enumerators were deployed for the main census enumeration. The enumeration period had to be extended by a week and several enumerators and supervisors were retained for the mop-up exercise to ensure maximum possible coverage of households in the country. This was necessary since the enumeration week coincided with the second school term holidays, a period when Fiji's population is generally mobile. Major religious and sporting activities and the biggest festival in Fiji are held during this period, and a lot of people flock to Suva for these functions. The importance of having school teachers as enumerators and supervisors was again an overriding factor in having the Census enumeration during this period. It was evident that the timing of the enumeration needs to be seriously reconsidered during the next census, however it is important that school teachers continue to form the majority of the enumerators and supervisors. They are a specially trained lot of individuals, therefore more receptive to census instructions. They also have a respectable standing in society and this is invaluable, particularly in rural areas.

Census Circulars providing instructions for any special arrangements necessary were issued to Census Superintendents and relevant people such as; Officials of Government Institutions, Ship Owners/Masters, School Principals/Head teachers and Owners/Managers of Hotels and Other Lodging houses. The circulars provided detailed information covering things such as; guidelines on the; appointment of Superintendents, Supervisors and Enumerators; the industrial guidelines governing their employment; allowances payable and remuneration; training arrangements and schedules; publicity and enumeration procedures.

CENSUS MAPPING

Preparations began in January 1994 with the appointment of field teams to revise the existing Enumeration Area [EA] boundaries or census collection units. The EA boundaries to be revised were basically similar to those used in 1976 with minor revisions in the urban areas, done in 1986. Some EAs had become too large and needed to be subdivided considering the underlying criteria of enabling one enumerator to canvass the area in one week [which is the normal census enumeration period]. Also some natural or cultural features which the EA boundaries followed had disappeared, hence boundaries needed to be redrawn to enable identification on the ground. Identification on the ground was an important criterion, however, where changes were needed minimal alteration was brought about so that meaningful comparison could be possible.

The field-workers appointed were mostly those involved in the three inter-censal household surveys carried out by the Bureau. Their familiarity with EA boundaries was an advantage and this assisted greatly in the whole revision exercise. Two field-workers and a driver were based in the Western Division, similarly in the Northern Division and four were based in the Central Division.

The new computer based mapping technology available at the Fiji Land Information System (FLIS) was utilised as it provided the Fiji Islands Bureau of Statistics (FIBOS) with exciting new potential outputs. However, it was in its development stage with only the cadastral data base captured in the system. This was quite useful for producing urban EA maps. FIBOS required topographical maps in most rural and some peri-urban areas, therefore the 1990 1: 50,000 topographical series maps were utilised as a base where available. Where these were not available, like most of Vanua Levu, the 1950 1: 50,000 series was used. Features such as roads that had come up since the maps were drawn, had to be included manually. A total of 1346 Enumeration Area maps were prepared. In urban EAs, computerised maps produced by FLIS were mostly used.

PILOT CENSUS

A pilot census was conducted for all intents and purposes as a trial run of the whole census enumeration with findings to be used in formulating strategies for the main census enumeration. This, it should be noted, was a scaled down version of what is normally a complete pilot run which would include questionnaire design and finalisation, system design and programming, field enumeration arrangement and operation, tabulation of results and the evaluation of the whole exercise. Due to time constraints analysis of the pilot census was not possible and it is recommended that it should be done in the next census. Given the benefit of hindsight there would have been valuable lessons learnt if the pilot exercise was completed. The Pilot Census questionnaire together with the Instructions Manual was ready and pre-tested within the Census Office in Mid February 1996 with interviews conducted in English, Fijian and Hindi. This was to ensure a good flow as well as check for ambiguities. Slight modifications were brought about prior to the Users Conference which was held on the 29th of March, 1996

Publicity for the Pilot Census was through newspapers and the radio with special mention of the areas where the pilot census was to be carried out. This was effective as feedback from the enumerators indicated that all households enumerated were aware of an enumerators pending visit.

Training was conducted by Census Office personnel at various District Administration Offices with the enumerators who were identified by relevant District Officers (DO's). These DOs by then had been informed of their upcoming role as Census Superintendents for the main census enumeration.

The pilot census did expose a few weaknesses which was taken care of prior to the main census enumeration. It also provided a valuable lesson for Census team members of the Bureau who had no previous experience in census taking. It was evident that training and supervision were critical if good data was to be collected. Changes were brought about to the census questionnaire and this was useful because possible complications and confusion were averted.

Our inability to process data collected during the Pilot Census was a major failure and was due to data processing programmes not being ready. This definitely is a major lesson for the next census. The results, however would have been of high quality given the level of supervision and support provided to field workers during the pilot census. It simply was not possible to provide similar levels of support during the main census operation. Enumerators for the pilot census were also instructed to note down any difficulties as well as any unusual incidents encountered during the enumeration. This was to ensure that contingency plans were put in place for the main census enumeration to safeguard against such difficulties and incidents.

The EAs selected for coverage in the pilot census were as varied as possible i.e. high class in urban areas, squatters, HA, Fijian Village and rural EAs.

TRAINING

For the Population Census 1996, training procedures were kept simple and clear and easy to understand written instructions were provided to all trainers. The training was based on the questionnaire and the instructions to enumerators manual.

The main thrust of the training was towards ensuring that enumerators were fully aware of the importance of the census exercise as well as having a clear understanding of the Instructions Manuals contents and its link to the questionnaire. Mock Interviews was a major component of the training. Wherever possible training was centralised in Suva.

PUBLICITY

A formidable publicity programme was part of the census awareness programme. The main thrust of the publicity programme was towards creating an awareness amongst the population of the importance of the Census so that public co-operation could be forthcoming during the census enumeration. Related to this was the highlighting of the sort of information to be collected during the census and their use. Publicity leaflets containing the census questionnaire and other relevant information were distributed through school children and electricity bills. Posters were also distributed and these were placed at suitable places.

Towards the census night, publicity intensified with regular radio messages and daily newspaper advertisements in Fijian, Hindi and English. For the first time TV was also used and overall the campaign was very effective.

CENSUS ENUMERATION

Following the Superintendents Conference work intensified towards preparatory measures in the districts i.e. appointment of Supervisors and enumerator and their training arrangements. Divisional liaison officers were appointed and they were given the specific instruction of following closely all preparatory measures in the superintendent areas of their allocated divisions. Two liaison officers were assigned for the Central Division two for the Eastern Division, two for the Western Division and one for the Northern Division. The number of liaison officers appointed to each division reflected their population size and geography.

Census Superintendents were instructed to recruit as many car owning Supervisors as possible. Their cars were used during the census and a commuted allowance to cover fuel costs was provided.

In the final few days before the Census night, EAs were allocated to Enumerators by their respective Supervisors and their boundaries were shown to them on the ground. Boundary identification on the ground is always a problem because it is not possible for Census Office staff to identify all EA boundaries for enumerators. So in a lot of cases much would depend on how the Superintendents or Supervisors interpret the EA boundary maps and descriptions provided.

Forms for Hotels, Institutions and Ships were delivered before the Census night in order that particulars of those staying in such places on the reference night are filled by appropriate people. Arrangements were made to have their forms collected or delivered to the Census Office immediately following the Census night.

Identification Cards were provided to each Supervisor and Enumerator and they were instructed to carry it with them always while involved in the enumeration.