It has flaws, but start using it to restrict subsidies
The Socio Economic Caste Census (SECC) may well have serious flaws, as FE columnist Surjit Bhalla has argued on Wednesday. According to Bhalla, the SECC data which tends to over-estimate income in comparison with the NSS still ends up showing 56% of Indians as landless as compared to 26% as per the NSS data; similarly, while the SECC show that 51% of rural Indians are landless labourers, the comparable figure is 35% in the NSS data. Obviously, this huge difference throws up important questions of methodology, and the government needs to decide which method is going to be used—there is little point using both sets of data simultaneously if they throw up contradictory results. Since the SECC is a census, it should logically be more reliable than a sample, but a final choice has to be based on what the analysis throws up. Also, in case the SECC data is to be used—appropriately corrected for the kinds of flaws that Bhalla alleges—a method has to be found to ensure the information is updated regularly. There are also issues in interpretation that need to be looked at carefully when it comes to caste. A casual look at SECC data suggests that, for instance, poverty is concentrated among SC/ST which have roughly 85% of households where the chief earner is earning under R5,000 per month as compared to 70% for non-SC/ST households. Yet, there are 88.8 million non-SC/ST households in this category versus 44.6 million for SC/ST, suggesting poverty is not determined by caste alone.
What is more important than the quality of the data, however, is that this is the first time there is a commonly agreed list of poor across the country. Given that, it is important to use this list. After all, if, based on the SECC data, there are 7.1 crore households that come under 14 exclusion criterion like having a 2/4-wheeler, why should they be getting subsidies? All told, if, according to the SECC, roughly a third of rural Indians are poor, why is the government planning to spend Rs 1.25 lakh crore in order to give huge subsidies on rice and wheat to 67% of the population? (The subsidy works out to roughly 88% of the market price for wheat and rice.) Similarly, what is being done to ensure the Rs 22,000 crore of LPG subsidy and Rs 8,000 crore of kerosene subsidy reaches only the poor? It is true that seeding the SECC data with Aadhaar numbers will take anywhere between 6 months and a year, but all efforts have to be made to do this. The Supreme Court verdict on whether Aadhaar can be made mandatory for availing government subsidies is obviously critical, but the government would do well to find ways to incentivise states to cooperate with it. After all, based on the SECC proportion of poor people, India should be spending just Rs 42,000 crore on giving subsidised rice and wheat to the poor—the National Food Security Act envisages 5 kg of either per month per person—instead of the Rs 125,000 crore budgeted for this year. If a fourth of the savings are shared with the states, that’s a sizeable incentive to come on board and help quickly seed the SECC data with Aadhaar numbers.