A. Introduction

1. As in other regions, the overall goal of the population and housing censuses for various members of the United Nations Economic and Social Commission for Western Asia (ESCWA) is to provide reliable, high-quality and timely information on the population size as well as on basic selected topics for establishing a comprehensive profile of the country. The process is planned to ensure the confidentiality of individual records and cost-effectiveness within the context of established national priorities. For several countries, the process is expanded to include censuses for establishments and dwellings and in a few cases to provide information on agriculture or to be linked to an agriculture census.

2. The history of population censuses differs significantly among various countries of the region, from being a continuous process for over a century (Egypt carried out 12 censuses between 1882 and 1996) to being completely disregarded and compensated for by large-scale sample surveys (Lebanon), or being just recently initiated (Oman, 1993 and Occupied Palestinian Territory, 1997).

3. The scope and content of such population and housing censuses is to be given due consideration since they are the core component of the national statistical system. Complementarities and interlinkages with other components, such as administrative records and regular or ad hoc sample surveys, allow for securing basic information for planning and decision-making, not only at the national level but also for subnational geographic areas and groups.

4. The aim of this paper is to review briefly the progress in undertaking population and housing censuses in the ESCWA region, according to the United Nations principles and recommendations for the 2000 census round, and to elaborate the current position vis-à-vis the key issues that will be covered in the Symposium.

B. History of census taking in the region

5. Fielding of population and housing censuses became universal for almost all countries of the ESCWA region in the 1980s round. Whereas Oman undertook its first census in 1993, other countries fielded censuses between 1985 and 1994. Only Lebanon and Occupied Palestinian Territory did not participate because of their special conditions. For the following round (1995-2004), seven countries have already carried out a population and housing census (Bahrain, Egypt, Iraq, Kuwait, Occupied Palestinian Territory, Qatar and the United Arab Emirates) and another five countries are scheduling their censuses during the years 2003-2004 (Jordan, Oman, Saudi Arabia, Syrian Arab Republic and Yemen).

6. Since the early 1990s several recommendations emphasized the importance of fielding a comprehensive population and housing census for Lebanon to reach a consensus regarding the size of the population as well as to establish a factual demographic and socio-economic profile for the country. However, because of internal and regional factors there is no indication that this can materialize in the near future (Table 1).

7. National statistics organizations are mandated to carry out the census in collaboration and full consultation with other stakeholders. The only exception is Lebanon, where such responsibility is assigned to the Ministry of Interior. The Lebanon Central Administration of Statistics is, however, concerned with the other component of the national statistical system, which is currently being rehabilitated after a long period of political instability.

8. With these varying levels of experience in carrying out censuses, there is no documented evidence about the progress in achieving the goals of complete coverage and capturing high-quality information. Comprehensive evaluation plans are not given due attention within the framework of the census programme. Accordingly, the absence of basic measurement for the level of coverage and data quality would not allow users to assess the success of the census operations.

9. A few countries (Egypt, Occupied Palestinian Territory and Yemen) have carried out a post-enumeration survey (PES) to evaluate coverage and content errors. The utilization of their results for adjusting census findings raised several points about the approaches to be adopted in this case. The first considered that these results are mainly for evaluation purposes and future projections while the census data are to be kept unchanged (Egypt, 1976). On the other hand, some argue that these measurements should be used to modify the census findings accordingly. The most recent PES was carried out in conjunction with the 1997 Palestinian Census, and its results indicated the high level of coverage (only about 2 per cent omissions). PES results were taken into consideration in reporting the final results of the census (population size).

10. With the exception of Bahrain, all countries of the ESCWA region opted for a de facto census, with a number of exceptions that differed from one country to another based on its conditions. Generally, this was adopted for practical reasons and to avoid the possibility of double counting. However, a major problem in that respect is the estimation of native population abroad, which is usually carried out through special sources. Various arguments were stated in support of such a directive for a clear understanding of family composition and an accurate estimate of flow indicators.

C. Content of the population and housing censuses

11. The main topics included in the list recommended for the 2000 round of censuses were covered by almost all ESCWA members. However, in most cases income was not included in the census questionnaire. Partial coverage was noticeable for two topics, namely, international migration and disability. Table 2 shows that the first was not covered by Egypt, Iraq, Kuwait and the United Arab Emirates, while disability was not covered in the censuses of Iraq, Kuwait, Saudi Arabia and the Emirates.

12. On average, the census questionnaire included around 50 questions, out of which 64 per cent aimed at collecting data on individual characteristics while the balance was distributed between housing (14 per cent) and households (22 per cent). The smallest number of questions was in Saudi Arabia (35), and the largest was in the Occupied Palestinian Territory (59). Generally, however, the size of census questionnaire fluctuated between 47 and 55 questions (Table 3).

13. National priorities and the suitability of the topics are considered the main factors determining the selection of topics to be included in the census. However, this was clearly linked with the need to ensure dissemination of findings within a timely, reliable and cost-effective framework. Moreover, regional and international comparability as well as the adoption of international standards were taken into consideration in the formulation of topics for the census instruments.

14. The usefulness of census data was also enhanced in some countries through the fielding of large-scale sample surveys to further investigate specific issues. This should allow better responses and higher-quality data from a smaller number of interviewers as well as reduce the overall cost. Two countries of the ESCWA region adopted such an approach in the last round of censuses, namely, Jordan and the Syrian Arab Republic. In both cases, the survey was carried out in conjunction with the census framework, although the two activities were independent; around 67,000 households were sampled in Jordan, compared with only around 20,000 in the Syrian Arab Republic. Their results are linked to the census and their findings were disseminated simultaneously. The sample design was usually based on the updated frame obtained from the accompanying census and it provided additional detailed information on the demographic, social and economic characteristics of the Jordanian population.

15. The validity of such an approach stems from its ability to provide timely, accurate and cost-effective data that can also be linked to the main findings of the census. At the same time it eases the pressure from users to expand the census questionnaire, which should be kept at the optimum size.

16. This approach is also expanded to other topical surveys that are carried out within the context of the national statistical system. A clear example is the set of questions on economic characteristics that are collected through the census and complemented by detailed information obtained from the labour force surveys. Table 4 outlines samples of both standard and additional questions included in the population and housing censuses during the 1990 and 2000 census rounds. For intercensal or labour force surveys, supplementary questions are added to collect detailed and explanatory information pertaining to the status of employment. Examples of such questions are:

· Status of employment

· Type of work

· Reasons for not seeking work

· Reasons for absenteeism

· Wages/salaries

· Actual hours of work

D. Integrated national statistical systems

17. Successful population and housing censuses are considered a country’s most comprehensive, accurate and reliable source of information. They provide benchmark data that can also be integrated with other components of the national statistical system (administrative registers and sample surveys, for example) to provide a full assessment of the demographic and socio-economic characteristics of the population as well as information on the living conditions and their impact on the environment. Censuses also provide the basic data for the measurement of various indicators required for planning and decision-making at the national, regional and local areas as well as for different subgroups.

18. The adoption of such an integrated framework was built in within the design of the population and housing censuses for three ESCWA members (Bahrain, Kuwait and Occupied Palestinian Territory), although with different approaches. In the cases of Kuwait and Occupied Palestinian Territory, the identification number was among the data collected through the census questionnaire to allow linkages with the population registers at a later stage. Moreover, in the case of Bahrain, selected factual information, extracted from the registers, was printed on the census questionnaire for validation by the interviewer.

19. The analysis of the pilot census of Bahrain showed some interesting inconsistencies between the information obtained from both data sources (registers and census) that required clear guidelines for their handling. These should be linked to the special characteristics of each source and provide clear instructions for the imputation and cleaning of census data. Inconsistencies in the reported census data can be partly attributed to the different methodologies adopted in collecting the data for each source:

· Information on the registers is based on individual supporting documents whereas respondents to the questionnaire provide information for all members of the household. The accuracy of such reported data is dependent on the respondent’s level of knowledge and his/her relationship to various members;

· Registers document the information according to the official rules and regulations adopted in the country, while census respondents are reporting the actual status (e.g., business licenses); and

· Delay in updating individual information on the registers, especially for secondary-level data, might lead to such discrepancies.

20. Reaching an acceptable level of consistency between the two sources, although required, should be based on the basic principles guiding the process of reporting and registering the data. To this end, the adopted guidelines included:

· As much as possible, respondents should be the head of the household or a clearly knowledgeable person;

· Further validation of the inconsistent information should be carried out by the interviewer and different responses are to be documented on the census form;

· Inconsistent records on the registers are flagged for further validation actions. Moreover, a special field was added to the register to include the different answer reported in the census. However, the validation of such inconsistencies will be a time-consuming and costly operation.

21. In both Bahrain and Kuwait, the need for undertaking the census was questionable in view of the comprehensive civil registration system. The question was also linked to the complexity of the census process and its cost, while “e-government” applications are being planned and implemented in these countries in the forthcoming years. Within that context, there is an indication that the 2001 census of Bahrain would be the last one. Current policy anticipates that the census will be replaced by a continuous e-government registration system.

22. Meanwhile, for other countries it is important to coordinate the main concepts and definitions that are adopted by the census and other various components of the national statistical system. Moreover, it is essential to specify the rules that ensure cost-effectiveness, especially in view of the current economic situation in most developing countries and the size of the budget for the census. Funding the census process is mainly from public sources, with possible financial and technical support from various donor agencies or bilateral country agreements. The scarcity of such public funds should lead to the adoption of new strategies for private financing and cost recovery. Statistical data and indicators should be viewed as valued goods that are to be considered as an investment.

E. Future directions

23. The contribution of basic census findings to the overall planning process, the development of sound policies and programmes and the effective management of national affairs are becoming clearly noticeable and the results will be reflected in the welfare of the country. To this end, expanding dissemination and full utilization of census findings is given due attention and, accordingly, national priorities are important in determining the content of the census. Involving the different stakeholders in such a process, through various consultative mechanisms, is essential to establish ownership and to ensure wider responsiveness to the stakeholders’ priorities.

24. A set of census committees at various levels is to be formulated to allow stakeholders to express their views at various stages of the census process. Assessment of data needs according to national priorities is, however, linked to factual discussion about the cost and the suitability of the census for providing such information.

25. The development of a master statistical plan including all components of the national statistical system would clearly identify the role of various components and reduce the demands on the census. Reaching an optimum questionnaire size for the data to be collected in such a process would ensure the dissemination of timely, accurate and cost-effective data.

26. In addition, plans for the census operations should consider various options that can be adopted to ensure the efficiency of various activities including mapping, data collection, dissemination and utilization of census findings.

27. The utilization of the Geographic Information System (GIS) for assisting in the planning, control of census operations, tabulation, presentation, analysis and dissemination of census results is an important step in that direction. This was clearly demonstrated in the experience of Qatar in undertaking its last census in 1997. Similarly, the Global Positioning System (GPS) was used for the Kuwaiti census (1995) and is planned to be used for the Syrian census (2004). Besides long-term cost-effectiveness, the full utilization of such techniques would enhance the quality of the census and the timely dissemination of results. Moreover, it provides decision makers with an accurate profile of the country, various administrative units and localities, for formulating customized policies and programmes.

28. The type of instruments to be used for capturing the information is another important point to be discussed. Up until now, all countries of the region adopted regular paper questionnaires in fielding the census, and this procedure might also be adopted for planned censuses, especially since the experiments of both Egypt and Kuwait in using optical mark readers (OMR) were not successful, although they proved to be faster in data processing. The main reasons for abandoning OMR usage were:

· The difficulty of field and office editing of the questionnaires and the need to ensure the accuracy as well as the reliability of the collected data;

· High probability of being unable to read all sections of the questionnaire because of the sensitivity of the special papers; and

· The need for costly specific paper and the sensitivity of the paper to environmental conditions.

29. Other alternatives, however, are to be considered for the planned censuses for both Oman (2003) and Saudi Arabia (2004). The high cost of using electronic media for collecting and processing the data in the field can be reduced if coordination among countries of the region is established and even if they are to be used for other data-collection activities.

30. Dissemination and utilization of census findings is still lagging behind for several reasons. Among these the key elements are confidentiality, delay in data processing and lack of resources for printing. New technological advances have demonstrated efficiency in solving such constraints. As an example, the Palestinian census was disseminated through the traditional paper approach as well as using the Internet. It is also notable that the findings of Lebanon’s large-scale sample survey (68,000 households), were disseminated on a CD-ROM.

31. New trends to ensure transparency and integration into the global economy are giving momentum to the timely dissemination of data. A clear example is the recent decree by the Council of Ministers in Saudi Arabia to allow the release of statistical data, and accordingly the last census (1992) is to be put on the Internet.

F. Conclusions

32. With the exception of Lebanon, all countries of the ESCWA region have already fielded or are planning to field a census in the 2000 round. This should be taken within the context of the national statistical system to ensure full integration among the various components of the system, and this would allow a clear identification of roles, especially since the census is still considered the core element of the system.

33. Dissemination and utilization of census findings is to be expanded to enhance planning and decision-making at various levels. Various technological approaches are to be used in that direction to increase accessibility and secure transparency.

34. Regional coordination is to be strengthened to enhance national capacity and to encourage the adoption of measures to produce timely, accurate and cost-effectiveness data. This would also allow the exchange of experiences between countries and the exploration of new techniques in the planning, collection of data, analysis, dissemination and utilization of findings.

Tables

Table 1:

Dates of Population and Housing Censuses

in ESCWA Member States

				2000 round (1995-2004)
Country	1970 round (1965-1974)	1980 round (1975-1984)	1990 round (1985-1994)	Completed	Planned	2010 round (2005-2014)
Bahrain	1971	1981	1991	2001		-
Egypt	1966	1976	1986	1996		2006
Iraq		1977	1987	1997		-
Jordan		1979	1994		2004	-
Kuwait	1970	1975,1980	1985	1995		2005
Lebanon	1970			1996,1997	(…)
Oman			1993**		2003
Occupied Palestinian Territory				1997**		-
Qatar	1970		1986	1997		-
Saudi Arabia	1974	…	1992		2004	-
Syrian Arab Republic	1970	1981	1994		2004	-
United Arab Emirates		1975,1980	1985	1995		2005
Yemen			1986,1988,1994		2004	-

* Large-scale sample survey.

** First national census.

- No information about the date for fielding census.

(…) No plans up till now for fielding a census.

Table 2:

Some Characteristics of Population and Housing Censuses

Carried out in ESCWA Member States

Country	Date of census		Type of census	Geographical & internal migration	Household & family	Demographic & social	Topics addressed
	Month	Year					Fertility & mortality	Educational	Economic	International migration	Disability
Bahrain		2001	De jure	X	X	X	X	X	X	X	X
Egypt		1996	De facto	X	X	X	X	X	X	-	X
Iraq		1997	De facto	X	X	X	X	X	X	-	-
Jordan	12	1994	De facto	X	X	X	X	X	X	X	X
Kuwait		1995	De facto	X	X	X	X	X	X	-	-
Lebanon*
Oman	11	1993	De facto	X	X	X	X	X	X	X	X
Occupied Palestinian Territory	12	1997	De facto	X	X	X	X	X	X	X	X
Qatar	3	1997	De facto	X	X	X	X	X	X	X	X
Saudi Arabia		1992	De facto	X	X	X	X	X	X	X	-
Syrian Arab Republic	9	1994	De facto	X	X	X	X	X	X	X	X
United Arab Emirates	12	1995	De facto	X	X	X	X	X	X	-	-
Yemen	12	1994	De facto	X	X	X	X	X	X	X	x

* No census carried out since 1932.

Table 3:

Number of Questions Related to Housing, Household and Individual Persons

in the Population and Housing Census of some ESCWA Countries

Country	Housing ⁽¹⁾	Household ⁽²⁾	Persons	Total
Bahrain	9	11	32	52
Jordan	9	15	29	55
Oman	4	15	36	55
Occupied Palestinian Territory	7	13	39	59
Qatar	9	7	30	55
Saudi Arabia	2	10	23	35
Syrian Arab Republic	9	11	31	51
Yemen	9	9	29	47
AVERAGE	7	11	32	50

(1) A housing unit is a building or part of a building for the residence of one or more households, regardless of its occupancy status at the time of the census. A housing unit has an independent entrance leading directly (or through a road, passage or stairs) to a public road without passing through other housing units.

(2) Household refers to one person or a group of persons living in a separate housing unit or part of it.

Table 4:

Samples of questions on economic characteristics included in population

and housing censuses of ESCWA countries (1990 – 2000 rounds)

Standard questions:

For persons 10, 12 or 15 years and above (reference period = one week):

- Relation to labour force:	- Employed - Unemployed previously employed - Unemployed not previously employed	- student - housewife - retired - income recipient - unable to work
- Main occupation (current)	As is (in details)
- Type of economic activity	As is or using ISIC
- Employment status	- employer - employee - own account	- unpaid work - others
- Employment sector	- public - private (national, foreign)	- others

Additional questions:

- Reasons for being unemployed (Syrian Arab Republic):	- Salary not appropriate - Work not appropriate - End of project	- Desire to start own work - Others
- Duration of unemployment (Bahrain and Jordan)	As is (in months)
- Duration of work in the main current occupation (Yemen)	In years
- Secondary occupation (Yemen)	As is (in details)
- Place of work (Oman, Palestine and Syrian Arab Republic)	- At home - In the same governorate - In other governorate	- Others
- Hours of work for wage employees & persons with unpaid work (Syrian Arab Republic)	As is (in hours)
- Training in the domain of main occupation (Syrian Arab Republic)	- Yes - No

* This document was reproduced without formal editing.

** Chief, Statistics Division, Economic and Social Commission for Western Asia (ESCWA). 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