The publication, in February 2024, of a household consumption survey by the National Sample Survey (NSS) Office, after a gap of over a decade, has made it possible to estimate the poverty rate in India. One such estimate, released by the World Bank in April 2025, has received the most attention. It points to the poverty rate being very low by now. To quote the World Bank, "Over the past decade, India has significantly reduced poverty. Extreme poverty (living on less than $2.15 per day) fell from 16.2 percent in 2011-12 to 2.3 percent in 2022-23...." ('Poverty and Equity Brief: INDIA', 2025). If this is indeed an accurate description, it would be a source of satisfaction, for it suggests that extreme poverty has virtually disappeared from the country.

The 'thali meal' as a consumption metric

The conventional approach to poverty measurement, pioneered by the Government of India over half a century ago, entails first determining the income that would enable food intake of a specified calorific value, and then classifying those with less as poor. This is a physiological approach, and has some merit. But there could be other approaches, based on the consumption of goods, for instance. One such approach would recognise that humans are likely to approach food from an angle wider than just its calorie content, taking into account the energy it provides, which calories measure, nourishment, and the satisfaction it gives. We believe the thali meal reflects this thinking, making it a natural choice to measure food consumption in real terms.

As a combination of carbohydrates, protein and vitamins, the thali is a balanced and self-contained unit of food consumption in south Asia, even if the nomenclature may vary across it. With this in mind, we have estimated the number of thalis the monthly expenditure reported in the Household Consumption Expenditure Survey of 2024 would translate into. The rating agency, Crisil, has estimated the cost of a home-cooked thali, comprising rice, dal, vegetables, roti, curd and salad, as ₹30. Adopting this price, we found that in 2023-24, up to 50% of the rural population and up to 20% of the urban population could not have afforded two thalis per day at the food expenditure recorded. If two thalis a day is taken as the minimum acceptable standard of food consumption, our estimates point to much greater food deprivation in India than conveyed by the poverty figures from the World Bank.

A crucial reason why our findings differ is that we do not assume that all of a household's income is available for spending on food. A household needs to spend on rent, transportation, telephony, health maintenance and education to just remain in the workforce. Now, expenditure on food ends up as the residual. Therefore, we have based our estimates on the actual expenditure on food.

It tends to be assumed that the Public Distribution System (PDS) deals effectively with food deprivation. To assess this, we have computed the value of food consumption including the imputed value of supplies received via the PDS -- both purchased and free of cost. With the value of consumption thus adjusted, the proportion of the population that cannot afford two thalis declines to 40% in rural areas and 10% in urban areas. Notably, food deprivation remains very high in the former even with subsidised food.

The role of PDS

To understand how the PDS can be leveraged to alleviate food deprivation, we estimated the subsidy per person across expenditure classes. What is interesting is that both PDS purchases and free food availed of by sections of the population that can afford more than two thalis a day are high. For instance, in rural India, the subsidy received by an individual in the 90%-95% fractile is 88% of the subsidy received by an individual in 0%-5% fractile, even as the first has a consumption expenditure more than three times greater, and going by our own thali index does not require further support. By contrast, in urban India, the subsidy regime is strongly progressive. But here too about 80% receive subsidised sales from the PDS and also free food, even when they too can afford more than two thalis per day.

Based on the data on food deprivation and the structuring of the food subsidy regime we make a proposal on how policy should evolve. This data point to what needs to be done and how it can be achieved. First, there is scope for restructuring the food subsidy, raising it at the lower end of the distribution and eliminating it altogether at the upper. However, we learn from the most recent consumption survey that there is a constraint to be faced: cereals consumption is almost identical for individuals in the 0%-5% fractile and the 95%-100% fractile.

This suggests that the desired level of consumption of cereals, both rice and wheat, has been reached, for the richest can afford to purchase all the cereals that they desire. While it points to the success of the PDS, in that it has equalised the consumption of a staple food, at the same time, it points to the limits to using the PDS in its present form to end food deprivation. Not only has cereals consumption very likely reached its desired level across the population but also it constitutes only 10% of the average household's expenditure.

It is unreasonable, from considerations of both logistics and expense, to expect that a government can distribute the entire food basket to any section. There is a middle path though, and that is to expand the distribution of pulses through the PDS. In a further comparison of the consumption patterns at the two ends of the distribution, we find that unlike in the case of the cereals, the per capita consumption of pulses in the 0%-5% fractile is exactly half of that in the 95%-100% fractile.

Pulses consumption

The PDS can be leveraged to equalise the primary food consumption across the population. The expansion of the PDS to ensure desirable levels of consumption of pulses -- for many Indians, the only source of protein and a very costly food item -- is both desirable and feasible. The financial aspect can be addressed by restructuring the PDS. The per capita consumption of rice and wheat in the 0%-5% fractile implies that the PDS entitlement of rice and wheat is well above what is necessary for a significant number.

Expanding the subsidy regime to supply cereal to 80 crore people, as done by the central government in January 2024 and the particularly large entitlement of rice given to those below the poverty line in some States, do not reflect need. Also, they come at a cost to the economy, given the alternative uses of public funds. Trimming the current entitlement of cereals to levels indicated as needed by the recent consumption survey at the lower end of the distribution and eliminating it altogether at the upper end would also require lower stocking requirements for the Food Corporation of India, with substantial gains.

We have proposed an expansion of the PDS in the direction of the food most needed by the least well-off, namely pulses, accompanied by the elimination of subsidies for those whose food consumption exceeds a reasonable norm, such as two thalis a day. Right now, the PDS is both unwieldy and ineffective, as it spreads resources thin. Our proposal will render it compact, enabling the equalisation of primary food consumption in India by raising that of the poorest household to the highest level observed in the economy, a globally significant outcome.

Pulapre Balakrishnan is Honorary Visiting Professor, Centre for Development Studies, Thiruvananthapuram. Aman Raj is an independent economist based in Patna, Bihar

Equalising primary food consumption in India