Caste of Votes: Need for a Caste Census?

[Kartik Shrivastava writes about the government’s reluctance to conduct a caste census and argues that there is a pressing need to conduct the same. Increasing Diversity by Increasing Access to Legal Education (IDIA), Hyderabad Chapter had partnered with Law and Other Things to organize an Essay writing competition. We are publishing the winning entry to the competition selected by IDIA.]

On 23rd September this year, a writ petition was filed by the Maharashtra government in the Supreme Court which argued for the collection of data about the “Backward class of citizens” in the upcoming 2021 census. This writ petition is a part of the demands which have, like those before previous censuses, arose for including caste in the nationwide census. This demand has been mostly about enumeration and classification of the Other Backward Castes. The government in the affidavit filed in response to this writ petition has denied conducting a caste census.

Caste, as a form of identity, pervades many spheres of the life of an Indian citizen. The political leaders across ideologies have recognised this truth. Despite this, we haven’t had a caste census since 1931. What are the reasons behind this? Would conducting a caste census create more divisions? Or would it allow the state to reduce existing divisions? This piece concerns itself with these questions. The answers to these questions answer the ultimate query of do we really need a caste census right now.

Why not a Caste Census? Arguments against Enumeration of Caste in the Census

The idea of having a caste census has not been welcomed by everyone. The present government has denied the demands of a caste census for majorly administrative reasons. It has argued that enumeration of caste in the census would be administratively cumbersome and unfeasible. The government cites the 2011 social economic caste census (“SECC”) as one of the reasons for the denial. The 2011 SECC was conducted when the demands from political parties in the alliance and the opposition became irresistible. The results of the SECC revealed the existence of more than 46 lakh castes in India and 99% of the castes had less than 100 people belonging to them. This made the data unviable for policy matters. The government also argued that the questionnaire for the census is made sometime before the actual conduction of the census and the 2021 questionnaire has been prepared and reviewed. Thus, the column of caste cannot be included now.

Identity politics is a major concern against the caste census. With the enumeration of caste, it is argued that the political parties at power will try to appease the groups which are favourable vote banks. The caste census can ‘re-mandalise’ Indian politics. Counting people on the basis of caste may further divide the regional political parties according to caste-based interests, entailing unstable governments, like in the 1990s. In EV Chinniah, the Supreme Court ruled down the sub-classification of Scheduled Castes (SCs) and Scheduled Tribes (STs) by stating that the governments might do the same for political reasons.

Some scholars argue that policy should not be driven by caste identification, as they argue that it is a colonial construct. According to this argument, c     ensus and caste, to the extent that it is recognised by the state, have origins in colonial India. The colonial officials saw the different Jatis as sub-parts of the four varna system. Thus, classifying people in different homogenised but arbitrary groups, which came to be known as caste. The Jatis unlike the varnas are not strictly based on occupation but can be linguistic, regional, religious as well as gendered. By conducting a caste census, it is claimed that the government would be perpetuating these ‘arbitrary’ divisions, thus, further distorting the conception of the Indian citizen.

Essentiality Of A Caste Census

The above mentioned arguments paint a picture of the Indian state as one which abhors the idea of classifying people on the basis of caste. This isn’t true; the census enumerates people belonging to the SC and ST community. The central government and state governments also provide reservations to the Backward castes other than the SCs and STs. Thus, there is some merit in recognising caste officially.

The caste census, in the age of IT and AI, is far more administratively feasible than it was beforehand. The concern for the existence of a large number of castes having similar spellings can be solved by software that can club similar sounding names as one caste, which can be later verified by experts. The questionnaire being fixed should not be a concern as the caste census does not need to be started along with the decennial census but can be done separately like the 2011 SECC. Rather it is possible that the caste data can pose a threat to the Hindu majoritarian view of the BJP making it deny its conduction. Nevertheless, the task will be daunting, which can only be justified by firm socio-legal imperatives.

The caste census would provide statistical justification for caste-based reservations. The caste data which the governments use is from the 1931 census. After independence and after the implementation of Mandal it is known that certain marginalised castes from the SC/ST and OBC communities have been denied the benefits. Fresh and accurate data is needed for the reformulation of policies according to present needs. In M Nagraj and Jarnail Singh, the Supreme Court has ruled that the government should collect quantifiable data before providing reservations for any community. Identity politics has been going on for years and, certainly, caste census would not make a big difference.

According to data collated by the World Inequality Database, the upper castes, on average, earn 47% more than the national average. While, the ST, SC, and OBCs earn 34%, 21%, and 8% less than the national average. Even if the caste was a colonial construct, it has contributed to vast socio-economic inequalities. The government has the duty to reduce these inequalities according to Article 38(2) of the Constitution. Framing policies to pursue these aims require the government to have accurate data, which will be obtained from the caste census. The claim of caste categories not being homogenous is true. The report by Justice G Rohini Commission has revealed, inter alia, that 25% of the reserved jobs are taken by just 10 dominant castes, while 2400 backward castes are left with 25% of them. In Davinder Singh, the Supreme Court has recognised that certain backward castes among the SCs need extra protection. The caste census will help in gauging the differential among the castes and enumerating the people belonging to the extremely backward ones.


A caste census may not be a policy imperative for a state striving for a casteless society. However, this sense of a ‘casteless’ society doesn’t eliminate the inequities that arose because of caste. To remove those inequities, the government needs to take active steps, whose direction can be correct only after using accurate data.

To conclude, an extract from an article by Pratap Bhanu Mehta, where he criticises the caste census, is illuminating:

 First, a caste census condemns us to the tyranny of compulsory identities. The premise of   enumeration is that we can never escape caste. Our identities are not something we can choose; they are given as non-negotiable facts which we can never escape. The state has legitimised the principle that we will always be our caste. In the name of breaking open prisons, they imprison us even more.

The metaphoric prisons which Mr. Mehta talks about are not bolstered but merely located by the caste census, which the lost liberators need to do, before anything else, to break these prisons.

Kartik Shrivastava is a student pursuing BA LLB (Hons.) from NALSAR University of Law, He is the author of the winning entry to IDIA Essay Competition. 

Join the discussion

This site uses Akismet to reduce spam. Learn how your comment data is processed.