A. Strategies for involvement of stakeholders in census activities

1. In Bangladesh, the planning commission, policy makers, researchers, international bodies and the Bangladesh Bureau of Statistics are the potential stakeholders of census results. Seminars, technical back-stopping from technical committees and application of new technologies were simultaneously arranged for the Bangladesh population and housing census of 2001.

Potential issues of interests are :

· Two seminars and expert consultations were made to review the results of the last census and to make recommendations for the 2001 census;

· The time frame of 3:2 years (three years for preparation and two years for processing and dissemination) was adopted for this census;

· Aerial photographs at the scale 1:25,000 were taken to prepare digital enumeration area (EA) maps for data collection and Geographic Information System (GIS) for storage, spatial distribution and dissemination;

· Technical committees were formed and were consulted in each stage of preparation;

· Five Optical Mark Readers (OMRs) and four Optical Character Recognition (OCR) machines were procured, and questionnaires were designed and printed locally so that they would be readable by both machines;

· Three national censuses—population, agriculture and economic—are being conducted to develop a benchmark database for all the sectors of the economy at the beginning of the new millennium;

· Census committees were formed at the local level to ensure the participation of the local government in the census;

· Enumerators and supervisors were selected from the local educated unemployed youths;

· An extensive publicity campaign was done for public motivation;

· Six pre-tests were done to limit the questions in the short questionnaire and accommodate them in the long questionnaire;

· Vital issues like housing structure, slums and disability status have been included in the census; and

· A unique code for each individual has been inserted.

B. Strategies for choosing among data-collection methods as sources of demographic and social statistics

2. There are three potential data sources in Bangladesh:

1. Census. The census provides benchmark data for the terminal year of each decade;

2. Sample vital registration system. This system provides data on population composition and changes on an annual basis;

3. Sample surveys. These surveys provide ad hoc data on all other socio-economic characteristics needed for national income accounting, poverty monitoring and so forth.

3. The following practices are followed in Bangladesh:

· For the census, the time frame set is three years for preparation and two years for processing and dissemination of results;

· For sample surveys, the time frame set is one year for preparation and six to nine months for processing and dissemination of results;

· Principles of official statistics have been adopted;

· A pilot has been done for a general data dissemination system under the conceptual framework of the International Monetary Fund (IMF); and

· A relational database and home page technology are being used to ensure timeliness of data and accessibility to users.

4. With the advances in technology, it is time to consider an alternative approach to census taking. One alternative that has been suggested is doing a listing of households with just a headcount (three questions only) of the population on a 100-per-cent basis and conducting detailed sample surveys for sociodemographic characteristics.

C. Adapting new technologies to census operations

5. It is difficult to anticipate all the administrative and methodological problems in the application of new technology in the conduct of the census.

6. Aerial photography was undertaken with the technical assistance of the United Nations Development Programme (UNDP) by an Australian company to prepare digital EA maps for the census. But because of repeated failure to produce perfect photographs by the vendors and faulty administrative decisions, it was not possible to derive the benefits of this technology.

7. Both the OMR and OCR technologies were adopted to improve the quality of information and control the timeliness in the dissemination of results. Because of insincere dealing by the vendors, the performance was disturbed but subsequently resulted in success.

8. The use of mobile telephones in the census operation was a success. Development of a home page and producing CDs were also successes.

9. Outsourcing of the mapping operation and the publicity campaign are possible, but outsourcing processing and dissemination should be avoided.

D. Maintaining census-related activities during the intercensal years

10. Follow-up surveys, census analysis, preparation of monographs, database development and archival arrangements are the effective practices in Bangladesh. All the technical staff members of the Bangladesh Bureau of Statistics are mobilized during the mapping and census operations. After the census other officials (except those working on data processing and report writing) are engaged in other censuses and surveys. This is very effective in Bangladesh.

E. Identifying and resolving problems of census mapping

11. New aerial photographs were taken to prepare a numerical mosaic and EA maps for urban areas through outsourcing. This effort did not work well because of repeated failure in the preparation of good photographs and faulty administrative decisions in the preparation of EA maps. The census results were affected because the traditional maps were inaccurate.

F. Post-enumeration surveys: are they worth it or not?

12. What is more acceptable, census or post-enumeration survey (PES) results? In the sample areas, PES results are definitely better than the census for two reasons:

· The PES is managed with more care and uses more experienced enumerators; and

· The PES provides for review of unmatched cases.

13. Of course, there are problems of independence with a PES, but if handled properly it provides the census planners with an estimate of coverage and content errors.

* This document was reproduced without formal editing.

** Bangladesh Bureau of Statistics, Bangladesh. The views expressed in the paper are those of the author and do not imply the expression of any opinion on the part of the United Nations Secretariat.