India - Employment and Unemployment Survey: NSS 43rd Round : July 1987 - June 1988, Fourth Quinquennial Survey

It may be mentioned here that in order to net more households of the upper income bracket in the Sample , significant changes have been made in the sample design in this round (compares to the design of the 38th round).

SAMPLE DESIGN AND SAMPLE SIZE
The survey had a two-stage stratified design. The first stage units (f.s.u.'s) are villages in the rural sector and urban blocks in the urban sector. The second stage units are households in both the sectors.
Sampling frame for f.s.u.'s : The lists of 1981 census villages constituted the sampling frame for rural sector in most districts. But the 1981 census frame could not be used for a few districts because, either the 1981 census was not held there or the list of 1981 census villages could not be obtained or the lists obtained from the census authorities were found to be grossly incomplete. In such cases 1971 census frame were used. In the urban sector , the Urban Frame Survey (U.F.S.) blocks constituted the sampling frame.
STRATIFICATION : States were first divided into agro-economic regions which are groups of contiguous districts , similar with respect to population density and crop pattern. In Gujarat, however , some districts have been split for the purpose of region formation In consideration of the location of dry areas and the distribution of the tribal population in the state. The composition of the regions is given in the Appendix.
RURAL SECTOR: In the rural sector, within each region, each district with 1981Census rural population less 1.8 million formed a single stratum. Districts with larger population were divided into two or more strata, depending on population, by grouping contiguous tehsils similar, as for as possible, in respect of rural population Density and crop pattern. (In Gujarat, however , in the case of districts extending over more than one region, even if the rural population was less than 1.8 million, the portion of a district falling in each region constituted a separate stratum. Further ,in Assam the old "basic strata" formed on the basis of 1971 census rural population exactly in the above manner, but with cut-off population as 1.5 million have been retained as the strata for rural sampling.)
URBAN SECTOR : In the urban sector , strata were formed , again within NSS region , on the basis of the population size class of towns . Each city with population 10 lakhs or more is self-representative , as in the earlier rounds . For the purpose of stratification, in towns with '81 census population 4 lakhs or more , the blocks have been divided into two categories , viz . : One consisting of blocks in areas inhabited by the relatively affluent section of the population and the other consisting of the remaining blocks. The strata within each region were constituted as follows :

Table (1.2) : Composition of urban strata
--------------------------------------------------------------------------------------------------------
Stratum population class of town
number
-----------------------------------------------------------------------------------------------------
(1) (2)
-----------------------------------------------------------------------------------------------------
1 all towns with population less than 50,000
2 -do- 50,000 - 199,999
3 -do- 200,000 - 399,999
4 -do- 400,000 - 999,999 ( affluent area)
5 (other area)
6 a single city with population 1 million and above (affluent area)
7 " (other area)
8 another city with population 1 million and above
9 " (other area)
-------------------------------------------------------------------------------------------------------------
Note : There is no region with more than one city with population 1 million and above. The stratum number have been retained as above even if in some regions some of the strata are empty.
Allocation for first stage units : The total all-India sample size was allocated to the states /U.T.'s proportionate to the strength of central field staff. This was allocated to the rural and urban sectors considering the relative size of the rural and urban population. Now the rural samples were allocated to the rural strata in proportion to rural population. The urban samples were allocated to the urban strata in proportion to urban population with double weight age given to those strata of towns with population 4 lakhs or more which lie in area inhabited by the relatively affluent section. All allocations have been adjusted such that the sample size for stratum was at least a multiple of 4 (preferably multiple of 8) and the total sample size of a region is a multiple of 8 for the rural and urban sectors separately.
Selection of f.s.u.'s : The sample villages have been selected circular systematically with probability proportional to population in the form of two independent interpenetrating sub-samples (IPNS) . The sample blocks have been selected circular systematically with equal probability , also in the form of two IPNS' s.
As regards the rural areas of Arunachal Pradesh, the procedure of 'cluster sampling' was:- The field staff will be supplied with a list of the nucleus villages of each cluster and they selected the remaining villages of the cluster according to the procedure described in Section Two. The nucleus villages were selected circular systematically with equal probability, in the form of two IPNS 's.
Hamlet-group and sub-blocks : Large villages and blocks were sub- divided into a suitable number of hamlet-groups and sub-blocks respectively having equal population convent and one them was selected at random for surveys.
Hamlet-group and sub-blocks : Large villages and blocks were sub- divided into a suitable number of hamlet-groups and sub-blocks respectively having equal population convent and one them was selected at random for surveys.
Selection of households : rural : In order to have adequate number of sample households from the affluent section of the society, some new procedures were introduced for selection of sample households, both in the rural and urban sectors. In the rural sector , while listing households, the investigator identified the households in village/ selected hamlet- group which may be considered to be relatively more affluent than the rest. This was done largely on the basis of his own judgment but while exercising his judgment considered factors generally associated with rich people in the localitysuch as : living in large pucca house in well-maintained state, ownership/possession of cultivated/irrigated land in excess of certain norms. ( e.g.20 acres of cultivated land or 10 acres of irrigated land), ownership of motor vehicles and costly consumer durables like T.V. , VCR, VCP AND refrigerator, ownership of large business establishment , etc. Now these "rich" households will form sub-stratum 1. (If the total number of households listed is 80 or more , 10 relatively most affluent households will form sub-stratum 1. If it is below 80, 8 such households will form sub-stratum 1. The remaining households will 'constitute sub-stratum 2. At the time of listing, information relating to each household' s major sources of income will be collected, on the basis of which its means of livelihood will be identified as one of the following : "self-employed in non-agriculture " "rural labour" and "others" (see section Two for definition of these terms) . Also the area of land possessed as on date of survey will be ascertained from all households while listing. Now the households of sub-stratum 2 will be arranged in the order : (1)self-employed in non-agriculture, (2) rural labour, other households, with land possessed (acres) : (3) less than 1.00 (4) 1.00-2.49,(5)2.50-4.99, (6) 5.00-9.99 and (7) 10.00 or more. 2 sample households will be selected from sub-stratum 1 circular systematically with a random start. In the case of sub-stratum 2,8 households will be selected from the arranged frame, also circular systematically with random start.
Selection of households : Urban: At the time of listing, information on the major sources of income will be ascertained from each household, based on which its mean of livelihood class will be identified as one of the following : (a) self-employed , (b) regular wage/ salaried employee , (c) casual labour and (d) others. Further , the average monthly consumer expenditure and the household size will also be ascertained from each household from which its (average monthly ) per capita expenditure (pce) will be obtained . Now , in all towns except those with population 10 lakhs or more all households with pce Rs. 800/-or more will from sub-stratum 1 ( the affluent or rich section) and the rest , sub-stratum 2. In towns with population 10 lakhs or more , households with pce Rs,1000/-or more will constitute sub-stratum 1 and the others, sub-stratum 2 Households of sub-stratum to m.1. classes (a) and (b) above will be assigned to one of the three pce classes : (1) less than 'A' , (2) 'A' to 'B' : A and B have been determined at states / U.T. level such that the above three pce classes will roughly include , respectively , the lowest 30 %, the middle 60% and the top 10% of the population for state/U.T. The values of A and B are given state wise in Section 2 ). The households of sub-stratum 2 will be arranged by means of livelihood x pce class thus : a-1, a-2, a-3, b-3,b-2, b-1, c, d where a b,c and d stand for the m.1. classes and 1,2 and 3 for the pce classes as given above . Now, in the affluent strata of towns with population 4 lakhs or more, households will be selected from sub-stratum 1 and 6 from sub-stratum 2 . In the remaining strata , 2 household will be selected from sub-stratum 1 and 8 from sub-stratum 2 as in the rural sector . The required number of sample households will be selected separately from each sub-stratum circular systematically with a random start .
Sample size ( central sample ) : The all India sample in respect of the central sample consisted of 8518 villages and 4648 blocks


Three different weightsare provided in each record of filess in the data set. Deatils are as follows:-

1. Weight at stratum level is stored in Variable name : Wgt1_Statm.
2. Weight for each sub-round (Sub-round wise weight) is stored in Variable name : Wgt2_SS
3. Weight for each subsample (Sub-sample wise weight) is stored in Variable name : Wgt3_SR.
4. Weight for all Subrounds pooled and combined subsample weight is stored in Variable name : Wgt4_Pooled