Linear programming methods, indicators of nutritional adequacy from the Indian Council of Medical Research and household expenditure survey data from the National Sample Survey Organization were used to construct poverty lines for India. Poverty ratios were calculated for 1993–1994 and 1999–2000 on the basis of nutritional adequacy poverty lines and compared to official estimates of poverty based on energy requirements.

Nutritional adequacy poverty lines are higher than official poverty lines, particularly in rural areas. The application of nutritional adequacy poverty lines points to greater rural–urban poverty differences than in official estimates. Declines in rural poverty during the 1990s were also slower under the nutritional adequacy definition, especially in south India. There is a greater degree of rural–urban and regional bias in nutritional adequacy poverty reduction than suggested by official data.

Inter-state variations in changes in nutritional poverty and official poverty in the 1990s are largely explained by differences in assumptions on overall price movements. However, relative price movements in food items also played a role, particularly the slow increase in prices of cereals and edible oils in comparison to the prices of pulses, and in some southern states, compared to milk and vegetable prices as well.