As the Bihar assembly elections draw near, Bihar Chief Minister Nitish Kumar and Rashtriya Janata Dal leader Lalu Prasad have called on the central government to release the caste data from the Socio Economic and Caste Census immediately. Nitish Kumar raised the demand in a recent meeting of the Niti Aayog, and last week, Lalu Prasad observed a fast and called for a one-day bandh in Patna to reiterate the demand. Both leaders contend that the data, when released, will show that the Other Backward Classes make up higher numbers than the reservation policy accounts for.

The Socio Economic and Caste Census, initiated in 2011, aimed to survey rural and urban populations on their socio-economic characteristics such as ownership of land, houses, vehicles, farm equipment, as well as their caste.

Attempts to map the social and economic indicators of India’s population are not uncommon. The central government carries out a population census every 10 years. Besides this, it conducted censuses to count people living below the poverty line in 1992, 1997 and 2002 with the aim of identifying households for various social schemes. The Socio Economic and Caste Census, or SECC, was the fourth such exercise, but with a more comprehensive methodology of enumerating the households than used so far.

More contentious, however, is the effort to map the caste composition of the population. So far, the reservation policy for various caste groups has been based on data from the 1931 caste census. After that, the government stayed away from enumeration of caste groups. However, regional parties such as the Janata Dal, the Samajwadi Party and the Rashtriya Janata Dal, which rose to power through the struggle for benefits for OBC communities, fought hard for caste to be included in the SECC.

Four years after the SECC started, the Centre released provisional data on socio-economic indicators on July 3 for only rural population, with little word on the data on urban population and on caste enumeration. Amid criticism, it announced an expert group under Niti Aayog vice chairperson Arvind Panagariya, with members from the Social Justice Ministry and Ministry for Tribal Welfare. This group will analyse and consolidate over 46 lakh caste and sub-caste names the caste data has thrown up before this information can be released, the government has said.

The provisional rural data shows Scheduled Castes at 18.46%, Scheduled Tribes at 10.97%, Others at 68.52%, and 2.04% (or 36.57 lakh) as “No Caste & Tribe” households. Political parties such as the Rashtriya Janata Dal and Janata Dal (United) in Bihar say this “Others” figure may reveal that OBCs have a higher proportion than was identified in the Mandal Commission report in 1979, which said that OBCs made up 54% of the population.

While opposition parties have accused the government of holding back caste data for political reasons, there could possibly be administrative reasons too. Letters exchanged by the Office of Registrar General and Census Commissioner under the Home Ministry and the Ministry of Rural Development last year – while the caste enumeration was on – point to wide discrepancies in Scheduled Caste, Scheduled Tribes population in the census data.

Discrepancies in SECC

In a letter on April 15, 2014, the joint director in the Office of the Registrar General and Census Commissioner wrote to the Economic Advisor supervising SECC in the Ministry of Rural Development on the problems in caste enumeration. In a large percentage of cases, the letter states, families provided incomplete information. Instead of providing caste or tribe names, households preferred to provide “General, Others, OBC, Muslim” as their caste name.

The Office of Registrar General and the Ministry of Rural Development had to organise a revised survey of 1.02 crore households as the SECC figures for Scheduled Caste, Scheduled Tribe and even the total population did not match the numbers recorded in the Population Census of 2011. As against the population census, the figures recorded by the SECC for the Scheduled Caste population was 2.96% lower, for the Scheduled Tribe population was 4.08% lower, and the total Indian population was 3.34% lower.

The Office of Registrar General stated that the differences in total numbers were a result of a number of factors:

* The Population Census is conducted under the Census Act 1948, which makes it mandatory for citizens to provide information. In the case of SECC, information disclosure is voluntary.

* The military and the para-military were kept out of the SECC.

* As the SECC had gone on for over three years, it may have missed the seasonal migration as those who had migrated out of a state had been struck off but the in-migrants from other states had not been accounted for.

* The population of rural homeless had dropped by half in SECC. The Population Census recorded 11.65 lakh rural houseless people, while in SECC their numbers were only 6.1 lakh. “In Census, the enumeration of homeless population was conducted on the entire night of February 28-March 1, but no such night enumeration has been reported by state governments for SECC,” stated the Office of Registrar General.

After a revised survey in a number of enumeration blocks, discrepancies in the Scheduled Caste, Scheduled Tribe still persisted.

In a second letter to the Ministry of Rural Development on May 21, 2014, the Office of Registrar General stated that while the Population Census had recorded the Scheduled Caste population as 15.38 crore, in SECC this was higher at 15.88 crore. The population of Scheduled Tribe recorded was lower. The Population Census had recorded the Scheduled Tribes population as 9.38 crore, but in SECC this was only 9.27 crore.

A senior official who supervised the SECC said the differences in the two databases may still remain. “The Population Census has categories for SC, ST, Others, while in SECC we had provided SC, ST, Others, and even No Caste,” said a senior official of the Ministry of Rural Development. “Field surveyors said in several places, people said they did not know their caste or did not want to specify it.”

In the provisional rural data released last month, the proportion of no caste and no tribe categories is as high as 35.2% in Assam and 30.6% in Goa, and varies from 8% in Jammu and Kashmir to 3.6% in Gujarat, 2.1% in Uttarakhand and more than 1% in Karnataka and Jharkhand.

To what extent is this a sign that at least a section of Indians are leaving behind their caste categories can only be known when government releases the full data on caste enumeration.