agcensus.htm

FAO STATISTICAL DEVELOPMENT SERIES
No 6
CONDUCTING AGRICULTURAL CENSUSES AND SURVEYS
FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS
ROME, 1995

INTRODUCTION

This publication is intended for national statisticians who are responsible for conducting an agricultural census or survey. Within the census organization there will be many specialists involved in particular aspects of the census. For these specialist, this publication will also provide an insight into all the main aspects of census taking and enable them to better understand their own role within the census organization. It may also be of interest to the government officials involved in deciding if and when to organize an agricultural census, particularly for those who have limited experience of such a major statistical exercise.

Special endeavours have been made to make the document 'user-friendly' by:

  • adding a summary at the beginning of each chapter to help the reader select those subjects of specific interest;
  • adding specific references at the end of most chapters, providing further information on specialized fields, and;
  • the insertion of a glossary of the main terms used in the document.

In different countries the agricultural census may be organized in various ways depending upon the resources available, the importance of agriculture and the tradition in census organization. An agricultural census should be part of an integrated system of agricultural statistics with the objective of providing primary data on the structure of the agricultural sector, such as size of holdings, land use, land tenure, etc., which do not change quickly over time. Detailed data on agricultural production and inputs are part of the system of agricultural statistics, called current statistics, and are collected through specialized agricultural surveys and other sources.

It is recognised that many countries do not have sufficient resources for conducting a series of surveys and collect the most important data through a single survey called an agricultural census. It is also important to note that, by 1960, collecting census data on a sample basis had become an accepted practice and made possible the organization and conduct of censuses in countries lacking the resources required for a complete enumeration. Therefore, there is now no clear distinction between agricultural censuses conducted on a complete enumeration basis and those conducted on a sample basis. Because of this, the title "Conducting Agricultural Censuses and Surveys" was given to this publication.

The Programme for the World Census of Agriculture 2000 is the eighth publication for promoting a global approach to agricultural census taking. The first action towards a World Census of Agriculture was initiated by the International Institute of Agriculture (IIA) in Rome in 1924, which made an effort to persuade member countries to provide data based on a uniform plan prepared by the Institute. The main objective was to obtain internationally-comparable data for 1930 through the enumeration of crop and livestock production in one operation.

Since then, every ten years there has been international action to promote World Censuses of Agriculture, first by IIA (for the 1930 and 1940 censuses) and then by FAO. Concepts, definitions and methodology of census taking evolved and became more uniform with the scope increasingly focused on the structural aspects of the agricultural sector. FAO Statistics Division had a major role in promoting the use of sampling in agricultural statistics in general and in the agricultural census in particular. This has been done throughmethodological publications, technical assistance and training activities. The agricultural holding became a standard unit of enumeration. The emphasis in international action shifted from the objective to collect internationally comparable data to assisting countries to meet more closely their national objectives for the census.

Once the decision to conduct an agricultural census is taken, many different activities have to be planned. These activities are described in the present document in eighteen different chapters, which are presented in chronological order, to the extent possible.

One of the first steps to be undertaken is to ensure that the legal basis for the census organization exists. This is the subject of Chapter 1 "Census legislation". Related activities may be very different from country to country, ranging from simply ensuring that satisfactory legislation exists, as in some countries even the obligation to organize a census is governed by law, to the situation where the preparation of complete legislation is necessary in order to designate the responsible office for census taking, usually the Central Statistical Office or the Ministry of Agriculture, and to ensure the required budget.

The next step is the establishment of the census organization. This includes establishment of an inter-ministerial steering committee, described in Chapter 2 "Census committee", deciding on a detailed plan of action, described in Chapter 3 "Work plan, budget and expenditure control" and planning the staff recruitment, in Chapter 4 "Census staff".

It is essential to start early preparations of the cartographic material and the census frame, which are required for the organization of the field enumeration as described in Chapter 5 "Cartographic preparations" and Chapter 6 "Preparation of frame". Closely related to these issues is the question whether the census could be organized on a sample basis, instead of, or in addition (for a part of the data) to, a complete enumeration. This is discussed in Chapter 7 "Use of sampling techniques".

Basic census documents are the questionnaire, which specifies the exact data to be collected, and the tabulation plan, which must be in line with the questionnaire and prepared simultaneously with it. These two subjects are covered in detail in Chapter 8 "Census questionnaire" and Chapter 9 "Tabulation plan". These two documents are crucial for the whole census organization: the questionnaire for the organization of field work and the tabulation plan for defining data processing activities and for planning a dissemination programme.

Final preparations for the conduct of the census are described in Chapters 10 to 13. An effort should be made to encourage holders to cooperate in providing data, this is done through a publicity campaign which is described in Chapter 10 "Census publicity". One of the major census activities is training the enumerators and supervisors, which is normally done just before the census. For this purpose, detailed manuals are required which will also be used as reference material during the enumeration. The relevant information on these topics is provided in Chapter 11 "Instruction manuals", and Chapter 12 "Training programme". It should be kept in mind that the training of technical staff, such as training of trainers, data processing and sampling specialists, etc., has to be organized very early, so that these staff are available to work on the census preparations. All preparatory work described in Chapters 5 to 10, as well as preparations for data processing, will have to be tested extensively in the field prior to embarking on the enumeration. Methods of testing are described in Chapter 13 "Pre-test surveys and pilot censuses".

The most important census activity is the field enumeration. This involves many topics including decisions on methods of enumeration and organization of the supervision. These topics are covered in Chapter 14 "Census enumeration" and Chapter 15"Organization of field work". It is important to have an independent evaluation of the data collected, before they are made available to users. Methods of data evaluation are described in Chapter 16 "Quality checks and post-enumeration surveys".

The activities after the enumeration include processing the data collected, which is described in Chapter 17 "Data processing". It should be kept in mind that although the actual processing is undertaken after the enumeration, the technical part, i.e. preparation and testing of the computer programs and procedures, must be completed before data start arriving from the field. The presentation to the public of the final product of the census operation, i.e., the census results, is described in Chapter 18 "Dissemination programme".

The following sketch (see next page) gives an overview of an agricultural census, involving first the holders for the collection of primary information, which is basically aimed at providing statistical data required for development planning and for constructing frames for other surveys, and also increasingly for individual uses, in private monitoring of agriculture for example. The organization of an agricultural census involves many different resources: personnel of different specializations, equipment for transport and data processing, and various expendable materials, as well as long preparations including cartographic preparations, development of manuals, pilot censuses, etc. Such an operation cannot be done quickly and may take from three to six years from the initial planning to delivery of the final results, depending upon the size of the country, existing resources and experience.

This publication is complementary to other FAO publications which are listed below, and which include reports and examples of census methodology as applied in different countries, proposed census items and their definitions, and the role of the agricultural census in the national system of food and agricultural statistics.

Suggested reading

  • FAO (1977). Report on the 1970 World Census of Agriculture, FAO Statistics series no. 10.
  • FAO (1986). Food and agricultural statistics in the context of a national information system.
  • FAO (1992). 1980 World Census of Agriculture: Methodological review.
  • FAO (1995). Programme for the World Census of Agriculture 2000

An Agricultural Census is a large-scale, periodic, statisticaloperation for the collection of quantitative information on the structure of Agriculture.

The word "census" implies a complete enumeration of all agricultural holdings. However, by extension, it can be conducted by a sample enumeration, provided the sample is large enough to generate sub-national data.