The document summarizes Nepal's approach to linking its population and housing censuses with its agricultural censuses. It describes how Nepal has conducted agricultural censuses every 10 years since 1961/62. For its 2011/12 census, Nepal used a two-stage sampling design, with wards and enumeration areas as primary sampling units and agricultural holdings as secondary units. It derived its sampling frame from the 2011 population census and collected data on land area and livestock ownership. The census covered all of Nepal and collected data on crops, livestock, and community characteristics through paper questionnaires.

1. Roundtable on
World Programme for the Census of Agriculture 2020
Bangkok, Thailand
19-23 September 2016
Ambika Bashyal
Director
Central Bureau of Statistics
ambikabashyal@hotmail.com
Linking Population and Housing Censuses with
Agricultural Censuses: the Case of Nepal
Technical Session 3
1

2. Outline
2
•Introduction
•Methodological Approach Used in Census of
Agriculture
•Linkage between Population and Agricultural
Census
•Sample design
•Sampling frames
•Lessons learnt and Major challenges
•Way forward

3. Census of Agriculture
3
Agricultural census is the main source of agricultural statistics regarding
the structure of agriculture in a country. The CBS is responsible for
carrying out agricultural censuses in Nepal.
•Nepal has been conducting decennial agricultural censuses since
1961/62.
•The six agricultural censuses conducted in the country so far are:
•first (1961/62),
•second (1971/72),
•third (1981/82),
•fourth (1991/92),
•fifth (2001/02) and
•sixth (2011/12).
•Each of these censuses was undertaken in line with the recommendations
and framework of the WCA 2010 prepared and promoted by the FAO.
•All of these censuses were conducted on a sampling basis.

4. Legal basis
4
•The legal provision of the census and survey is one of the important
aspects for empowering data collection operation in Nepal.
•The Statistics Act 2015 is very explicit in the provision of confidentiality
to encourage participation in surveys by individuals, firms and others.
•In accordance with the Statistics Act 2015, the Government may also by
‘notified order’ (order published in the Gazette) issue directions for the
collection of statistics of any matter.

5. Census organization
5
•The agriculture census project was implemented by the Crops and
Livestock Census and Survey Section of the Economic Statistics Division
of the CBS .
•It was the responsibility of the Director General of the CBS to implement
the agricultural census.
•Under the direct control of the DG, project coordinator and census
officers, the enumerators carried out the field enumeration with the
supervision and direction of the field supervisor.
•A hierarchical an ad-hoc organization structure was made to successfully
implement the census CBS .

6. Organization Chart
6
SteeringCommittee
CentralBureauofStatistics TechnicalCommittee
AgriculturalCensus2011-12Project
AdministrationSection FinancialAdministrationSection DistrictAgriculturalCensusOffice
OooooOfficeOffice

7. 7
To control the enumeration activities in the district, the following provision was made:
Director General
Project Coordinator
District Agricultural Census Coordination Committee
Coordination Committee
District Agricultural Census Officer
Supervisor
Enumerator
at the central
level
at the district
level

8. International collaboration/assistance
8
•The technical assistances received by the Government from agencies like
FAO and ADB has enhanced the capacity of the CBS to undertake
agricultural census and surveys.
•Technical/financial support from FAO up to 1991/92 census. In 2001/02,
technical support for data processing was by ADB.
•In 2011/12 census, there was a technical assistance for sample design
review by FAO.
•The CBS has been regularly supplying the agricultural statistics to FAO as
required.
•Assistances will be required for the paradigm shift such as
• migration from PAPI to CAPI,
• use of area frame or combination of both area and list frame instead of list
frame alone of the agricultural holdings,
• policy analysis of the census and survey data, etc.

9. Linkage between Population and Agricultural Census
9
•The sampling frame for NSCA 2011/12 was derived from the population
census 2011 with complete enumeration for the items:
• Area of the holdings and
• Number of livestock and poultry by type.
• The information collected though this listing form was used to identify the
number of agricultural holdings in each ward.
• Population Census 2011 project had prepared EA maps for 58 municipalities
and for village development committees (VDCs) of 12 district headquarters.
Further, EA (block) maps were prepared for a few large VDCs of the country.
Those EA maps were used in the fieldwork of the 2011/12 census of
agriculture.
• Furniture, laptop computers, computer server networks, and other logistics
used in the population census were used in the Agricultural Census 2011/12
data processing which reduced the cost, time and effort of the latter.

10. Linkage between Population and Agricultural Census (cont…)
10
10) Does your household operate any land for agriculture purpose?
Yes/ No
If yes, the total area of the land in
11) Bigha 12) Katha 13) Dhur
14) Ropani 15) Ana 16) Paisa
17) Does your household raise livestock/poultry for agriculture purpose?
Yes/ No
If yes, the total number of livestock/poultry for agriculture purpose
Livestock
18) Cow/Ox 19) Buffalo 20) Yak/ Nak/ Chauri 21) Horse/Ass
22) Sheep 23) Goat 24) Pig/pork 25) Other
Poultry
26) Chicken 27) Duck 28) Other
(Serial numbers of questions have been preserved same as in the questionnaire)
Questions asked in population Census 2011

11. 11
Statistical unit• The statistical unit used in the census was the agricultural holding.
• An agricultural holding was defined as an economic unit of agricultural
production
under a single management comprising of all livestock and poultry kept, and
all land
used wholly or partly for agricultural production purposes.
• In the case of Nepal, an agricultural holding is generally the same as an
agricultural
household.
• Any one of the following conditions was essential for the identification of
an agricultural holding:
 Having area under crops greater than or equal to a quarter of a Ropani
(or
four Anaa) in the Hills or Mountains (0.01272 hectares), or greater than
or equal to eight Dhur (0.01355 hectares) in the Tarai; or
 Keeping one or more heads of cattle or buffaloes; or
 Keeping five or more heads of sheep or goats; or

12. Methodological approach
12
• The sampling frame for NSCA 2011/12 was derived from the population
census 2011 with complete enumeration for the items:
—Area of the holdings and
—Number of livestock and poultry by type.
• For agricultural census, all the core items recommended by WCA 2010 and
major and relevant supplementary items were collected in a sample basis by
using single questionnaire in a single visit.

13. 13
Sample design
• A two stage stratified sampling design was used in the census
• Each of the 75 districts was considered as a stratum
• Rural areas - Wards as Primary sampling Units (PSUs)
• Urban areas – Enumeration Areas (EAs) as PSUs
• Agricultural holdings as Secondary Sampling Unites (SSUs)
An EA was defined as either a ward or part of a ward or a combination of
wards, the lowest administrative division within VDC/Municipality

14. 14
Selection of PSUs
• Selection of PSUs at the first stage was done using probability
proportional to size (PPS) systematic sampling with number of
agricultural holdings in PSUs as a measure of size based on the 2011
Population Census.

15. 15
Allocation of PSUs
•The total number of PSUs to be selected in the NSCA 2011/12 was set at 5,200.
• The domains of study (districts) varied widely with respect to agricultural
importance.
• It was considered better to have larger samples in the more important districts.
• The measure of importance of each district was based on the total area under
nine major crops in the district obtained from MoAD. Past three-year averages
of these crop-areas were used for the purpose.
• The predetermined number of sample PSUs were allocated to 75 strata
(districts) by compromise power allocation method.
• Samples were allocated to different strata in proportion to xλ, where x is a size
measure and the parameter λ = 0.4. (The measure of size for importance of the
district being the three year's average area under nine major crops.)
• Minimum number of EAs was taken as 50 from each district (livestock was a
prime consideration)

16. 16
Allocation of PSUs Across Districts
Allocation of PSUs, NSCA 2011/12, Nepal
Number of PSUs
Total number of
Districts PSUs
Less than 60 PSUs 27 1,415
60 – 70 PSUs 16 1,046
70 – 80 PSUs 10 740
80 and more PSUs 22 1,999
Total 75 5,200

17. 17
Selection of SSUs
• Agricultural holdings were considered as “second stage units” (SSUs).
• The target number of holdings to be selected from each sampled EA was set
at 25.
•The number was adjusted to ensure the self-weighting nature of the sampling
design.
(based on the ratio of number of holdings at the time of listing and 2011
Population Census)
• The actual number of sampled households varied between 20 and 30.
• The selection was based on equal probability linear systematic sampling
method. In each of the sampled PSUs, a list of holdings was prepared.

18. 18
Selection of SSUs (Cont…..)
• Before selection, the agricultural households within the sampled EA
were classified into following four strata to ensure an implicit
stratification:
operating less than 1 bigha/10 ropani,
operating 1 bigha/10 ropani or more but less than 3 bigha/20
ropani,
operating 3 bigha/20 ropani or more,
not operating land but keeping livestock and poultry.

19. Geographical Distribution of SSUs
19

20. 20
Sampling frame
The major sources of information used for the allocation and selection
processes were :
• The list of village development committees (VDCs), municipalities and
wards published by the Ministry of Local Development,
• The total area under nine major crops (paddy, maize, wheat, millet, barley,
potato, oilseeds, sugarcane and vegetables) published annually by the
Ministry of Agricultural Development,
• Number and area of holdings in each ward obtained from the
“House/Household Listing Form” of the 2011 National Population and
Housing Census, and
• Listings of agricultural holdings in the selected PSUs (the listing was
carried out in the field prior to the census enumeration of the NSCA 2011/12
agricultural census itself).

21. 21
Sampling frame of SSUs
• Frame for the selection of holdings in each PSU was prepared during the
listing phase of the agricultural census.
• In each of the sampled PSUs, a list of holdings was prepared.
• Holdings within each sample PSU were grouped into four strata (on the
basis of operated land area or number of livestock kept) to ensure an implicit
stratification.
• Selection of SSUs was done in the field using equal probability linear
systematic sampling.
• In each PSU, the required number of agricultural households was selected
systematically with a given random start. Finally, the list of holdings to be
enumerated was prepared.

22. Statistical and geographical coverage
22
•The agriculture census 2011/12 covered whole Nepal, including urban
areas.
•Only household sector was included; agricultural activities undertaken
by government organization, businesses, etc. were excluded.
•It is also suggested that large private and institutional farms may be
included in the Agricultural Census.
•Such large enterprises being only few could be surveyed separately and
incorporated in the census results.
•Geographical scope of reporting is National (1), Ecological belt (3),
Development Region (5) and District (75) level.

23. Statistical and geographical coverage cont…
23
Units Number
Ecological Belt 3
Development Regions 5
District 75
Agriculture Holding 5,385,271
Total No. of VDC/Mun 3,960
Selected VDC/Mun 3,440
No. of wards 36,022
No. of PSUs 36,257
Sample PSUs 5,200
Sample SSUs 152,137

24. Number and type of questionnaires used
24
•The guidelines of the FAO were accommodated while designing the
questionnaire of the agricultural census of 2011/12.
•Most of the questions were carried over from the agricultural census of
2001/02. This facilitated comparison, uniformity and standardization.
•There were three types of household questionnaires used in the census. In
addition to this, a community questionnaire was administrated to collect
community (PSU level) characteristics.
Schedule 1A (List of Selected Agricultural Holdings)
Schedule 2 (Questionnaire for Individual Agricultural Holdings)
Schedule 1 (Agricultural Holdings Listing Form for the selected PSUs)

25. Enumeration period
25
•The census enumeration work was conducted in two phases.
•In the first phase during January to March 2012, enumeration
was carried out in 59 districts in the Tarai and Hills districts.
•The enumeration for the remaining 16 Mountains districts was
conducted in April to June 2012 when the climate was more
favorable.

26. Reference period and date
26
•The data was collected in respect to the twelve-month reference
period of January to December 2011.
•However; land area and livestock were collected in respect of
the day of enumeration.
•House/Household listing was done before the 2011population
census enumeration (15-31 May)

27. Enumeration method
27
•The enumeration method used in agriculture census 2011/12
was Paper and Pencil Interview (PAPI).
•The enumerators interviewed the respondents face-to-face and
data is collected by the enumerators using paper
questionnaires.

28. Census innovations
28
•The census included the production of the major temporary
and permanent crops.
•In addition to this, a community questionnaire was
administrated to collect community (PSU level) characteristics.
•The census was carried out by the Government budget fully.
•Data processing was done in-house by giving an incentive to
data entry operators on key-stroke basis.

29. Census budget
29
•The cost of the 2011/12 National Sample Census of
Agricultural was about US$ 1.43 million funded by the
Government.
•It excludes the salary of regular staff.
•The sample design review was done with the technical support
of FAO.

30. Census Reports/Micro-data
30
Census Reports
The main census results are being published by the CBS in a series of 85
reports and each of the reports contain the same set of 32 tables.
• one national level report;
• reports for each of the five development regions;
• reports for each of the three ecological belts;
• reports for each of the 75 districts;
• and a report presenting a summary of district level data.
Availability of census micro-data
In addition to the census reports, for the first time census micro-data are
also have been made available for sale. The pricing is as follows:
• Nepalese individuals: NRs. 3,000;
• GoN, and NGOs: NRs 5,000;
• Other users: US$ 300.

31. Relevant links to census instruments
31
•The documents and reports related to agricultural census of Nepal can be
accessed through the CBS website.
•The links are as follows;
http://cbs.gov.np/sectoral_statistics/agriculture
http://cbs.gov.np/sectoral_statistics/agriculture/district_summary
http://cbs.gov.np/image/data/Agriculture/National%20Level/NATIONAL%20REPORT%202011
_12.pdf
http://cbs.gov.np/image/data/Agriculture/National%20Sample%20Census%20of%20Agricultur
e%202011-12/National%20Sample%20Census%20of%20Agriculture%20Nepal%202011-
12.pdf
http://cbs.gov.np/image/data/Agriculture/Metadata%20Agriculture/Metadata_Agriculture.pdf
http://cbs.gov.np/sectoral_statistics/agriculture/agriculture_monograph_preface_and_contents

32. THANK YOU
32