Multipurpose Production Survey, Urban 1990

The complexity in the sampling design of the MPS (Urban) was further simplified by the formation of smaller enumeration blocks from the large municipality wards. A single stage stratified sampling was adopted by maintaining the sample fraction of 1/10 of urban enumeration blocks for all strata. Sub-division of wards was done by distributing the number of dwellings in each ward into blocks consisting of 150-200 dwellings.

For an appropriate area sampling it is necessary that enumeration area be more or less equal in size in population and the characteristics to be investigated be homogeneous to the extent possible. However, existing size of municipality wards are not fit to be considered as an enumeration unit due to various reasons. First of all, high variation in the size of population among the wards is noteworthy. Secondly, most of the urban wards, belonging to big towns like Kathmandu, Biratnagar, Birgunj, etc. are too large and are not manageable even for field operation. So it was decided to form the sub-wards with 15-200 dwellings through an intensive field work in order to prepare a sampling frame for the MPS (Urban). Despite several constraints of resources and lack of experience in such activity, altogether 1500 sub wards (to be considered as an enumeration block) from the entire 33 towns were formed and subsequently maps for these blocks were prepared.

Different level of urbanization of towns was another aspect to be considered for sampling design. In Nepal, some of the towns seem nothing more than an administrative center or major district settlement and are either partly urbanized or yet to be urbanized. The others possess urban characteristics to a large extent but still include some rural type of settlements. Only Kathmandu and Lalitpur can be considered as urbanized municipalities. Hence, the towns were divided into three groups for stratification purposes.

1. Urban - Kathmandu, Lalitpur (2)
2. Semi Urban - (towns with more than 40,000 population according to 1981 census) i.e. Biratnagar, Birgunj, Bhaktapur, Dharan, Pokhara and Mahendranagar (4)
3. Partly urban - Ilam, Bhadrapur, Damak, Dhankuta, Inarwa, Rajbiraj, Malangwa, Kalaiya, Lahan, Janakpur, Jaleswore, Hetauda, Bharatpur, Banepa, Dhulikel, Bidur, Tansen, Bhairawa, Butwal, Taulihawa, Trivhuvan Nagar, Nepalgunj, Birendra Nagar, Dhangadi, Dipayal (25)

After formation of sub-wards and subsequent mapping operation it became possible to avoid complicated sampling procedure. A single stage sampling of enumeration areas was adopted for all three strata. The details of the sampling scheme are found in the Report.

Selection of sub-wards was made according to the method for linear systematic sampling where the towns were arranged in the order of economically active population.

The stratification adopted here has been vindicated by results of the survey also. Percentage of households engaged in these four sectors of economic activities is higher (26.3%) in the towns under the category "Urban" and are lower in other categories with 21.9% in the "Semi-urban" and 20.7% in the "Partly urban" as shown by the survey results.