The Whys and Wherefores of Caste Census

This copy provides a thorough analysis of the debate over caste enumeration in India's Census. It examines historical origins, recent developments, socio-economic implications, and arguments for and against caste enumeration.
The Whys and Wherefores of Caste Census

In 2010, the Parliament passed a unanimous resolution (with both Congress and BJP on board) calling for caste to be enumerated as part of the 2011 Census. As per the 1931 Census when caste was last enumerated, there were 4147 castes in India apart from the Depressed Classes/Untouchables (as they were called then). The Anthropological Survey of India has identified 6325 castes. Unfortunately, the UPA2 government made a hash of the Socio-Economic Caste Census (2011) which threw up a ludicrous figure of 46 lakh castes and the results were never released.

In 2018, the Government of India (GOI) promised that the next Census would collect the population data of the Other Backward Classes (OBCs). However, in 2021, GOI did a volte face and announced that it would not enumerate caste as part of the next Census. GOI reiterated this stand before the Supreme Court in a case filed by the Maharashtra government in 2021. At the same time, the BJP supported the Bihar Caste Survey of 2023 while the results of the Karnataka Caste Survey of 2015 conducted by the Congress government are yet to be released. 

Such flip-flops by leading political parties show that the enumeration of caste in the Census is a highly contentious issue. Powerful forces are at work trying to make it happen and to sabotage it. There is as much vote-bank politics in not counting caste as in counting caste. 

The Whys and Wherefores of Caste Census
In Conversation with Devdutt Pattanaik: Exploring Mythology, Religion, and Culture | Interview

To count or not to count:

The British conducted the first Census between 1865 and 1872, at different times in different parts of the country. The first simultaneous Census across the country was done in 1881. Thereafter, the British conducted the Census every 10 years except in 1941 (due to the Second World War). The Censuses conducted between 1881 and 1931 enumerated caste.

In the first Census conducted by independent India in 1951, GOI ordered that caste should not be enumerated. However, an exception was made in respect of the 1234 castes in the SC category and the 698 tribes in the ST category. These have been enumerated in every Census from 1951 onwards. 

In his essay “Thoughts on Linguistic States” (1955), Dr. B. R. Ambedkar wrote as follows about the folly of not enumerating caste in the 1951 Census:

“I cannot illustrate these points by reference to facts and figures. The Census which is the only source of information on these points fails to help me. The last Census omits altogether the caste tables which had been the feature of the Indian Census since its birth. The Home Minister of the Government of India who is responsible for this omission was of the opinion that if a word does not exist in a dictionary, it can be proved that the fact for which the word stands does not exist. One can only pity the petty intelligence of the author.” 

Without detailed caste-wise Census data, we cannot look at India’s socioeconomic reality as it is, and plan for inclusive development. We cannot monitor whether India’s development has been equitable across castes and communities and take remedial measures. In the absence of Census data, we are forced to rely on the occasional caste-based surveys conducted by government and non-government agencies. It is not a pretty picture. For a quick overview of India’s highly inequitable development across castes, please see the 8 charts shown in the BOX. 

Non enumeration of caste in the Census, even if it was well-intentioned to begin with, has massively shortchanged the various marginalised castes and subaltern groups which constitute the overwhelming majority of India’s population. And, it has helped the elites, namely, the upper castes and dominant OBCs, to corner a disproportionate share of the nation’s assets, incomes, sources of credit, and positions of power and influence under the cloak of caste-anonymity. 

Let us now consider some of the common arguments advanced against and for the enumeration of caste in the Census.

The Whys and Wherefores of Caste Census
Oppenheimer Biographer Kai Bird Discusses the Film, AI, and Politics | Interview

Arguments AGAINST enumerating caste in the Census:

Argument 1: It would accentuate caste consciousness, lead to caste conflicts and fuel demands for increased reservations. 

A 2017 study by MIT scholars Evans Lieberman and Prerna Singh, evaluating over 1,300 Census questionnaires in 156 countries over two centuries, found that enumerating racial, ethnic and religious identities does increase the possibility of competition and conflict between social groups. 

This is not unexpected. But this cannot be a ground for not conducting the decennial Census or not enumerating essential social categories in the Census. If the deprived sections of the society demand their legitimate share of development, is it a bad thing? The bogey of ‘caste conflicts’ has been exaggerated by vested interests. It may be noted that the enumeration of the 1234 castes in the SC category and the 698 tribes in the ST category in the Censuses since 1951 has not led to any conflicts among these castes or tribes.

Whether we like it or not, caste has been the indubitable reality of Indian social life for the past 3000 years and continues to be so even today. The use of caste surnames and the display of caste marks on the body reinforce caste identities on a daily basis. A 2016 study by NCAER showed that only 5% of Hindu marriages are inter-caste. Residential segregation by caste persists in most villages, and also in many towns. Choices of candidates for elections and ministers for Cabinets are dictated largely by caste considerations. Not a day passes without news of caste atrocities somewhere or other in the country. If we are serious about abolishing caste, then these are the issues the government and the civil society should address. 

It should be noted that India enumerates religion, language, and region in the Census which are as divisive as caste, if not more. Casteism is not going to wither away merely by not counting caste in the Census any more than communalism, linguistic fanaticism and regionalism will disappear by not enumerating religion, language and region in the Census. 

Regarding “demands for increased reservations”, the real problem is with governments’ propensity to implement reservations for political considerations without quantifiable empirical data. The availability of caste-wise Census data will actually help curb such arbitrary decision making. We will be able to objectively and transparently debate and dispose of the claims of the Marathas, Patidars, Jats, Vanniars or any other caste for reservations/increased reservations. We will also be able to identify precisely the dominant OBCs, SCs, and STs which are hogging the lion’s share of their quotas and recalibrate reservations suitably. 

Argument 2: Enumerating caste in the Census is an administrative nightmare.

No Hindu has any existential doubt as to what his caste is and to which category his caste belongs. When GOI has been able to smoothly enumerate the 1234 castes in the SC category and the 698 tribes in the ST category in all the Censuses since 1951, it is difficult to understand why the enumeration of the 4000-odd other castes most of which are specific to particular States should pose an intractable problem.

The reasons for the UPA2 government’s botched up Socio-Economic Caste Census (SECC) of 2011 were:

  • It was not conducted under the Census Act, 1948. Disassociating the Census Commissioner made the exercise casual and perfunctory. 

  • It was done through the Union Ministries of Rural Development and Urban Development which did not have prior experience of conducting sociological/anthropological surveys.

  • GOI should have enlisted sociological/anthropological experts to draw up a draft list of castes (together with alternative caste names and the names of sub-castes and larger caste groups) specific to each State, published the draft list on the Web inviting suggestions and comments from the public before finalising it, and given only that list to the enumerators. 

  • The questionnaire was very poorly designed and asked open-ended questions about caste. The enumerators couldn’t distinguish between genuine castes, alternative caste names, larger caste groups, sub-castes, surnames, clan names, gotras, etc. 

So, the SECC 2011 threw up an absurd figure of 46 lakh castes. 

In contrast, the Bihar government did a neat job of its recent Caste Survey by giving enumerators just the list of 214 caste names specific to Bihar, with the 215th category labelled “Other Castes”. If the British could do it and Bihar could do it, why can’t GOI do it?

The Whys and Wherefores of Caste Census
Democracy, EVMs, and the Role of Governors: Interview with Former ECI S Y Quraishi

Arguments FOR enumerating caste in the Census:

Argument 1: You cannot remedy caste discrimination without taking caste into account.

It is counter-intuitive and yet true. You cannot solve the problems of special groups that have been historically discriminated against without taking group identity into account and gathering data group-wise. The special groups may be caste groups, racial groups, ethnic groups, women, transgenders, the disabled, orphans, nomads, or whatever. 

It is important to note that the demand for enumeration always comes from the victims of discrimination, not the perpetrators. For example, Germany does not enumerate people by race in its Census. This has actually worked to the disadvantage of the Black peoples who have started a private, country-wide, online survey called Afrozensus. The results of Afrozensus show that anti-Black racism is widespread in Germany and entrenched in institutions. 

The 2021 report of the Office of the High Commissioner for Human Rights (OHCHR) of the United Nations called for express recognition of race, ethnic origin, sex and other factors in national Censuses and the making of the data public to increase understanding of discrimination and to monitor the effectiveness of policy measures to counter it.

Argument 2: Enumerating SCs, STs and OBCs is a legal and administrative imperative. 

The Constitution of India provides for reservation for SCs and STs in the Lok Sabha (Article 330), State Legislative Assemblies (Article 332), Panchayats [Article 243D(1)], and Municipalities [Article 243T(1)]. After the 73rd and 74th Amendments (1993), the Constitution also provides for reservation for OBCs in Panchayats [Article 243D(6)] and Municipalities [Article 243T(6)]. For reservations to be made in electoral constituencies, caste-wise, area-wise Census data of SCs, STs and OBCs are essential.

Census data should also be the basis for implementing, monitoring and recalibrating reservations for SCs, STs and OBCs in admissions to educational institutions (Article 15) and public employment (Article 16). Caste-wise Census data is necessary to correct what statisticians call Type 1 error (wrongful inclusions of undeserving castes) and Type 2 error (wrongful exclusions of deserving castes), and also to ensure that a few dominant castes do not corner all or most of the benefits. 

As per Article 340 of the Constitution, the National Commission of Backward Classes is mandated to investigate the conditions of OBCs and make recommendations as to the steps to be taken to improve their conditions. There is also a Backward Classes Commission in each State. Evidently, these Commissions cannot perform their duties satisfactorily without proper caste-wise demographic data of OBCs. 

The reason advanced by GOI for enumerating only SCs and STs from the 1951 Census onwards was that these categories enjoyed reservations in electoral constituencies, admissions to educational institutions and public employment. This was shaky logic because OBCs also enjoyed reservations in admissions to educational institutions and public employment in many parts of the country long before Independence, and do so throughout India now. Post-1993, OBCs are entitled to reservations in electoral constituencies in Panchayats and Municipalities. In the Indra Sawhney case (1992), also known as the Mandal judgment, the Supreme Court ruled that that the OBC list (which was based on the 1931 Census) should be revised periodically. So, GOI should have enumerated OBCs in the 2001 Census. But it did not.

Whenever States like Uttar Pradesh, Madhya Pradesh, Gujarat, Maharashtra, Karnataka, Odisha and Jharkhand tried to implement reservations for OBCs in elections to Local Bodies, the High Courts and the Supreme Court stayed the same on the ground that there was no caste-wise data of OBCs. On the one hand, GOI doesn’t include caste as a parameter in its decennial Census. On the other hand, the Courts demand caste-wise data of OBCs for upholding reservations in elections to Local Bodies.

Hence, the enumeration by caste of persons coming under the OBC category is a legal and administrative imperative. It is not a “policy matter in which Courts should not interfere” as claimed by GOI. 

However, the 10% reservation for the economically weaker sections (EWS) among the upper castes was implemented and upheld by the Supreme Court in the complete absence of any supporting empirical data. The Census should enumerate persons coming under the EWS category also. 

While Census is a Union subject, the Collection of Statistics Act, 2008 empowers States and even Local Bodies to gather the necessary statistics. So, individual States can always do Caste Surveys like Karnataka (2015) and Bihar (2023) did. But Census data carry more authority and are less contested. GOI’s reluctance to enumerate caste as part of the Census is legally indefensible and administratively unwise. 

Concluding Remarks:

According to the eminent sociologist Satish Deshpande, “As a modern republic, India felt duty-bound to ‘abolish’ caste, and this led the State to pursue the conflicting policies of social justice and caste-blindness.” It turned out to be a case of falling between two stools. The policy of social justice cannot be pursued effectively without caste-wise data. The policy of caste-blindness is pointless when various rulings of the Supreme Court have held caste to be a ‘relevant criterion’, ‘sole criterion’ or ‘dominant criterion’ for defining a socially and educationally backward class. The decision not to enumerate caste in the Census was a corollary of the policy of caste-blindness. In retrospect, this has proved to be one of independent India’s biggest mistakes. It has perpetuated and exacerbated caste inequalities while shielding the Indian elites’ privileged position from the glare of public scrutiny. It’s time to correct this mistake so as to facilitate a more just allocation of collective resources. 

Author : K. Ashok Vardhan Shetty

K. Ashok Vardhan Shetty, a former Vice-Chancellor of the Indian Maritime University in Chennai, previously served as a member of the Indian Administrative Service (IAS), Tamil Nadu Cadre, since 1983. Throughout his tenure, he held various pivotal roles including Registrar at the University of Madras, Director of Collegiate Education, District Collector of Viluppuram, and Managing Director of the Tamil Nadu State Marketing Corporation (TASMAC), among others. Renowned for his adeptness in successful project implementation, he received commendations from the Government of Tamil Nadu on multiple occasions. Shetty is also an accomplished author, having contributed numerous articles to esteemed English and Tamil publications on subjects ranging from public administration and management to E-Government, popular science, and popular mathematics.


No stories found.
Vikatan English