The first Agricultural Census (AC) undertaken in Bangladesh was carried out in 1960. The second AC was conducted in 1977, followed by the 1983/1984 AC, the 1996 AC and the 2008 AC, to which the metadata and data presented here refer
The practice of conducting a census of agriculture at an interval of five or ten years has now become an established national and international routine activity. The first world census of agriculture was promoted by the International Institute of Agriculture (IIA). The FAO of UN, successor of IIA, has continued espousing implementation of the census programme in various member countries. Bangladesh carried out agriculture censuses in 1960, 1977, 1983-84, 1996, and 2008. The censuses were designed to provide a broad range of statistics on both structure and flow aspects of agriculture according to defined government national policy guidelines and also FAO advocated standard formats warranting international comparability.
The broad objective of the census was to determine structure and operational characteristics of
agriculture holdings managed by dwelling households. The specific objectives were to determine:
I. number of agriculture holdings, area of holdings, average size of holding,
tenure, equipment, and farming system
II. number and distribution of agriculture households
III. agriculture employment
IV. number of agriculture wage labour by gender
V. irrigated area under crops
VI. stock of livestock and poultry
VII. agriculture loan
VIII. agriculture implements
In addition, the specific objective was also to use all census bench-mark data as basis for improving current crop and livestock statistics as well as for formulating and implementing various policies and programmes of agriculture development.
Kind of Data
Census/enumeration data [cen]
Unit of Analysis
The scope od the study includes:
- land ownership and land tenure
- land unit and sub-division of land
- land utilization
- crop acreage and production
- livestock and poultry
- employment in agriculture
- agricultural population
- agricultural power and machinery
- irrigation and drainage
- fertilizer and soil dressing
- wood and fishery products
- agricultural credit
- agriculture and sericulture
- fruits and vegetable products
Agriculture & Rural Development
Land (policy, resource management)
Food (production, crisis)
Forests & Forestry
Access to Finance
The statistical unit was the agricultural holding, defined as a production unit under single management (both technical and economic) and engaged in agriculture activities. The AC 2008 covered only the agricultural holdings in the household sector. All holdings were classified according to size as:
· small - holdings with a total land area between 0.05 and 2.49 acres (1 acre = 0.4047 ha);
· medium - holdings operating a total area between 2.5 and 7.49 acres; and
· large - holdings operating more than 7.5 acres of land.
Producers and sponsors
Bangladesh Bureau of Statistics (BBS)
Ministry of Agriculture
Ministry of Planning
Department of Fisheries
Directorate of Forest
Food and Agriculture Organization
The census was carried out in accordance with policy recommendations of NSC and broad guidelines of FAO to maintain international comparability. The full-count of all households both in rural and urban areas was adopted for the census. For the census taking, the entire country was divided into 1,978 zones. For each zone, a zonal officer was appointed from among the manpower of BBS. On average, a zone consisted of about 78 EAs. The zonal officers carried out various functions before doing the actual data collection work. These were:
i. forming and demarcating zones and enumeration areas (EAs)
ii. preparation of EA sketch maps
iii. selecting and making enumerator and supervisor lists
iv. selecting training centers for imparting trainings to enumerators and supervisors
v. updating lists of restricted areas, special areas, and VIPs
vi. contacting local administrative authorities
vii. making arrangement for safe keeping of census materials
viii. preparing lists of bank branches for disbursing honoraria
ix. forming census committees at union, upazila/thana, and district/zila levels.
The census was conducted in two phases:
· Phase I: complete enumeration of all households, both in rural and urban areas.
· Phase II: sample enumeration with a long questionnaire was adopted to obtain detailed data on agriculture.
For Phase II, 10 percent of the 155 578 EAs were selected; subsequently, 10 percent of the households involved in agricultural activities were drawn in each selected Enumeration Area, or EA (a total of about 155 000 households).
Dates of Data Collection
Data Collection Mode
Data were collected by means of two questionnaires:
(i) a short questionnaire in the first phase (full count)
(ii) a long questionnaire in the second phase (sample-based)
The AC 2008 questionnaires covered 13 items of the 16 core items4 recommended for the WCA 2010 round, namely;
0001 Identification and location of agricultural holding
0002+ Legal status of agricultural holder
0003 Sex of agricultural holder
0004 Age of agricultural holder
0005 Household size
0006 Main purpose of production of the holding
0007 Area of holding according to land use types
0008 Total area of holding
0009 Land tenure types on the holding
0010 Presence of irrigation on the holding
0011 Types of temporary crops on the holding
0012 Types of permanent crops on the holding and whether in compact plantation
0013 Number of animals on the holding for each livestock type
0014 Presence of aquaculture on the holding
0015+ Presence of forest and other wooded land on the holding
0016 Other economic production activities of the holding's enterprise
The following items were not covered:
(i) "Legal status of agricultural holder"
(ii) "Age of agricultural holder"
(iii) "Main purpose of production of the holding
(a) Data Processing
The data processing of the census was originally conceived to be done with help of ICR (intelligent character reader) machine use. The aim was to cut-down substantially the usually long data- entry time of the
census. But some practical limitations led the steering committee to recommend abandoning plan of ICR data entry for the census. Therefore, the manual keying of data into micro computers was adapted. Here, the following rigid measures were taken up to expedite error-free manual data entry for the census:
i. establishment of local network system
ii. ensuring correctness of data right at the stage of data entry process; this included checks on completeness, possible errors, and consistency
iii. incentives for data-entry operators and supervisors on the basis of volume of work done
iv. constant monitoring of progress.
(b) Manual Editing
As soon as the data collection phase was completed, the documents were stored systematically in order of the computer geo-code list to ensure their easy retrieval as and when needed. Systematic storage of
documents was a very important step that ensured smooth processing. An elaborate list of editing checks was developed and the editors were thoroughly trained. For every 10 editors, one supervisor was engaged to ensure accuracy of editing. Also, for strict control of data-edit one supervising officer had overseen work of every two supervisors. The data entry program was developed in such a way as to reject any record with wrong or inconsistent data. In order to speed-up data entry, it was necessary that the manual edit was comprehensive and correct. However, data entry operators and their supervisors were also trained in manual edit in order to enable them to carry out necessary corrections when needed. Apart from accuracy of data sets, manual edit ensured completeness of enumeration-books and their geo-code identifications.
(c) Data Entry and Computer Edit
A special data entry program was developed in FoxPro with exhaustive checks of every kind including duplication of records. These checks, although slowed down the speed of data entry operation, ensured
accuracy and consistency. As a results, no further edit or imputation was necessary before embarking upon tabulation process. At the end, products of operators were merged together to make a district file. At this stage, it was ensured that the relevant number of records of all EAs were in the file and that no EA was missing. Completeness, accuracy, and consistency were the main elements of computer edit.
Post Enumeration Check (PEC):
Upon completion of field work of full count census, a post enumeration check (PEC) was done in order to assess the census data quality. The PEC findings are available in a report published separately.
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses