India - Household Consumer Expenditure, NSS 51th Round, Schedule 1, July 1994 - June 1995

As usual, a stratified two-stage sampling design has been adopted. The first-stage units are the villages (Panchayat wards in case of Kerala) in the rural areas and urban blocks in the urban areas. The second-stage units are the enterprises for the survey of unorganised manufacture and households for the survey on household consumer expenditure.

Sample Size (FSU's) : The total sample size (i.e. total number of villages and blocks to be surveyed) at all-India level was fixed as 14,072 for the central sample and 16,080 for the state sample.

Stratification : Each district generally forms a separate stratum. However, for Gujarat where NSS regions cut across some district boundaries, parts (viz. group of taluks) of each such district belonging to different NSS-regions from separate strata. If any district (or part thereof lying in an NSS region in case of Gujarat) has a small number of manufacturing enterprises, it is clubbed with the neighbouring district(s) within the same NSS region to form a stratum to ensure minimum allocation of 8 villages at the stratum level as far as possible.

Sampling Frame : Latest available lists of Urban Frame Survey (UFS) blocks have been used as the sampling frame. However, for a few newly declared towns of 1991 census for which UFS lists of blocks are not received, lists of 1991 census EBs are used as the sampling frame.
Two different weights have been provided in each file in the data set. Details are as follows:-

1. Weight for each sub sample is stored in the variable name : Wgt_SubSample
2. Combined subsample weight is stored in the variable name : Wgt_Combined