We discuss characteristics of the spatial distribution of poverty and calorie and protein deficiency in India. Two units of analysis are considered - states and NSS-defined agro-climatic zones. The data used are the NSS Expenditure Surveys of the 43rd, 50th and 55th rounds. Results on stochastic dominance as per these criteria are also reported.
An important caveat regarding the results on calorie undernutrition is that we do not report a unique value for it. The reason for this is that there is imperfect correlation between work categories and calorie requirements, i.e., there is no straightforward way in which many categories of work in the rural sector can be regarded as requiring \"sedentary\", \"moderate\" or \"heavy\" work. This problem becomes most acute in the data for the 55th round (1999-2000) where employment data is sparse and a considerable mismatch for the consumption expenditure data. In view of this we follow the procedure of computing three indices for calorie undernutrition assuming, seriatim¸ that everyone in the rural sector does sedentary, moderate and heavy work. Hence our computations should be viewed as providing bounds (with the \"heavy\" norm providing the upper bound and the \"sedentary\" norm providing a lower bound with the \"moderate\" norm providing an intermediate figure). In this part of the report we provide, in the main, results using the \"moderate\" norm. Some results using the other two norms have been provided in earlier papers.
The analysis in this paper is done at two levels - states and NSS agro-climatic zones, labelled as \"NSS regions\". We have used the following 75 regions in this analysis. We note below the state, the name of the region, the NSS code and a simplified code that we use for enumeration.
ASARC Working Paper 2004-2, 76 pp.