SECC: More than a database5 min read . Updated: 08 Jul 2015, 01:58 AM IST
Effective use of SECC to identify beneficiaries is not only crucial for effective delivery of subsidies but also for the quality and effectiveness of the data itself
Last week, preliminary tabulations of the Socio Economic Caste Census (SECC) were released by the government. This nationwide census of all households in the country was conducted in 2011 and 2012 but largely ignored by the media even though the data as well as details have been available online for the past three years. The National Democratic Alliance (NDA) must be congratulated for not only taking ownership of the SECC but also publicizing it. Unlike the population census data, which too covered similar questions but remains confidential at the individual/household level, the SECC relies on transparency—based as it is on a public audit of the data in the gram sabha.
Unfortunately, most of the media attention has focused on indicators used in the survey which serve as a useful measure of the extent of deprivation in rural areas. A preliminary analysis reveals a grim picture of rural areas with three in four rural households earning less than ₹ 5,000 per month and almost 90% of households having incomes of less than ₹ 10,000 per month. Overlooked by the media, these numbers are very close to the estimates of poor and vulnerable derived from other estimates based on the consumption surveys of the National Sample Survey Office (NSSO). ₹ 5,000 per month per household with an average household size of five would also mean an income of ₹ 33 per person per day in the rural areas.
These estimates of vulnerability in rural areas at 2011-12 prices would be closer to the findings of the Arjun Sengupta committee, which identified 77% of the population as vulnerable. While there was a hue and cry about the low poverty line of ₹ 26 per person per day in rural areas as implicit in the Tendulkar committee, these numbers have not evoked the same reaction even though they are grimmer than what the official poverty estimates would suggest. But then, reading the SECC as an estimate of poverty would neither be statistically appropriate, nor do justice to the reason for which SECC was designed. It was designed as an informational database to identify households for various government benefits such as the public distribution system, social pensions and so on.
But unlike the previous experiments (the worst being the Below Poverty Line Census, 2002, which was designed, conducted and implemented during the previous NDA regime), this census of households is not meant to estimate poverty. The SECC itself does not identify any household as poor, excluded or included for any benefit from the government. It is only meant to provide information on multiple indicators, chosen carefully after a pilot study of more than 250 villages spread across the country. The job of deciding the final criteria as well as actual identification has been left to various line ministries and the states, wherever applicable.
However, based on initial analysis of the data as well as random checks, it is heartening to know that it is of good quality. The estimates of various indicators should not surprise anyone since the data itself is based on questions which are readily available in the population census house listing exercise, except for income and land. It roughly compares with the population census estimates of housing, household size, literacy, disability and other such characteristics reported in the Census by the Registrar General of India (RGI). Part of the reason for the good quality data is that the data collection was linked to the caste census of the RGI. With RGI as an integral partner, not only was the entire process done with public resources and reliable government machinery, it also meant that very few households were excluded. More importantly, the data was made available publicly and villagers were free to correct any discrepancies. Finally, the village council had to sign off on the use of the data.
It is here that the claims of the NDA are not only hypocritical but also show a lack of commitment to using the data for the purpose it was designed. Not only has the process of the finalization of data been delayed inordinately, with only 299 of the 630 districts having completed the process, it also leaves out a large majority of poorer states and districts. The total coverage in the rural area of the final list is then less than 40%—three years after the survey was concluded. For urban areas it is worse, especially since the questionnaire was designed without any rigorous pilot. The quality of data for most big cities is suspect.
The claim by the government regarding the conduct of a caste census after seven decades also sounds hollow as it has not disclosed any timeline for the release of the caste data. Though the caste census was a promise made to Parliament by the United Progressive Alliance, it was expected that the NDA would kick off the process of analysing the data. Given the complexity of the caste data, it is unlikely to be available even for internal use in the next five to 10 years.
While the Union government has shown no urgency in using the data, states too are not utilizing the data for which it was made available. Very few states use the SECC for identification of beneficiaries for the National Food Security Act (NFSA) even though it is two years since the law was enacted. The NDA is guilty of deferring the date of implementation of the NFSA; officially, its deadline expired in July 2014. Significantly, some states where the final lists have not been readied are part of the NDA. This raises serious questions on the sincerity of the government in using the SECC data for the purpose it was designed for.
The effective use of the SECC for identification of beneficiaries is not only crucial for effective delivery of subsidies but also for the quality and effectiveness of the data itself. The same holds true of the criteria for identification of beneficiaries, which was indicative in the initial stages but was subsequently examined by various committees, including by a panel chaired by Abhijit Sen. Neither has the ministry not made the report public nor has it initiated a wider consultation with the state governments—the biggest stakeholders in the process.
Disclaimer: The author was the convenor of the core group set by the ministry of rural development which designed and analysed the pilot survey as well as the final SECC questionnaire.
Himanshu is an associate professor at Jawaharlal Nehru University and visiting fellow at Centre de Sciences Humaines, New Delhi.