The Case for Caste Census in India
Prasenjit Bose, Samiran Sengupta and Soumyadeep Biswas
After the publication of a caste survey in Bihar, which found that 63% of Bihar’s 13 crore population belong to castes listed under the Extremely Backward Classes (EBC) and Other Backward Classes (OBC) categories, Prime Minister Narendra Modi criticised and rejected the demand for a nationwide caste census made by Congress and several other Opposition parties.
What do socio-economic data indicate?
The intersection between class and caste-based deprivation in Indian society is evident in a gamut of socio-economic statistics. The average monthly per capita consumption expenditures (MPCE) of Scheduled Tribes (ST), Scheduled Castes (SC) and OBC households in rural areas were, respectively 65%, 73% and 84% of the MPCE of the ‘Others’, i.e. the general category, as per the National Sample Survey (NSS), 2011-12. In urban areas the average MPCE of ST, SC and OBC households were 68%, 63% and 70% of the general category in 2011-12.
The persistence of inequality across caste categories in India can also be seen in the multidimensional poverty estimates based on the National Family Health Survey (NFHS-4/2015-16). Estimates by scholars from the Oxford Poverty and Human Development Initiative (OPHI) show that while almost 28% of all Indians were multidimensionally poor in 2015-16, the poverty headcount ratio among STs was 50.6%, SCs 33.3%, OBCs 27.2% and Others 15.6% [Alkire, Sabina and Oldiges, Christian and Kanagaratnam, Usha, Examining multidimensional poverty reduction in India 2005/6–2015/ 16: Insights and oversights of the headcount ratio (February 19, 2021). World Development, Vol. 142, 2021]. As per this estimate, while STs, SCs and OBCs taken together comprised around 77.6% of the Indian population in 2005-06, this share increased to almost 84% of the country’s poor in 2015-16.
The multidimensional poverty headcount ratio among Muslims at 31% was also higher than that of Hindus (28%), Christians (16%) and other religious communities (15.7%). The Sachar Committee report (2006) had estimated that 31% of Muslims were ‘Below Poverty Line’ on the basis of the NSS 61st round, 2004-05, while the poverty headcount ratio among SCs & STs together was 35%, Hindu OBCs 21% and other Hindus (general category) 8.7%. Methodologies of poverty estimation have changed significantly in the past two decades, suggesting a significant decline in the poverty level.
However, the disproportionate concentration of poverty among the STs, SCs, OBCs and Muslims in India have remained stable over time. This clearly indicate that discrimination and exclusion based on caste via-a-vis STs, SCs and OBCs as well as religion, particularly with regard to Muslims, have a causal relationship with poverty and deprivation.
What about education and employment?
The pattern of caste-based deprivation can be clearly seen in official data on education and employment indicators. The general category has a much higher proportion of literates, secondary and high school pass outs, graduates and post-graduates than OBCs, SCs and STs.
As per NSS 75th round (2017-18), while only 3% of STs, 4% of SCs and 6% of OBCs are graduates, the proportion of graduates in the general category is over 12%. The proportion of post-graduates within the general category is over 3%, among OBCs around 1% and among SCs and STs, below 1%.
In terms of employment status (PLFS 2021-22), over 30% of the workforce in the general category had a regular job, while the proportion of regular or salaried workers among OBCs and SCs was around 20% and among STs just over 12%. In contrast, almost 29% of STs, 38% of SCs and 20% of OBCs were casual labourers, against only 11.2% of the general category. This indicates that the informal sector is largely populated with STs, SCs and OBCs while the general category has a disproportionately large share of formal employment.
The Central government itself is one of the largest employers in the formal sector, employing over 18.78 lakh persons as on January 2021, as per Union government data, 52.7% of whom belonged to the general category. The proportion of employees in the general category is much greater at over 64% in the most qualified and highest paid cohort, i.e. Group A employees. This indicates that the persistence of caste-inequality in educational outcomes is reproducing a similar pattern of caste-inequality in skilled, formal employment, even three decades after the official implementation of the Mandal Commission recommendations. The demand for a nationwide caste census has gathered momentum in this backdrop.
When did caste surveys begin?
Caste wise enumeration of the population was introduced under the British colonial administration in 1881 and continued till the 1931 census. Independent India’s governments abandoned full caste enumeration on the apprehension that it would strengthen caste divisions and perpetuate the caste system.
However, the caste system has persisted and flourished in independent India — even without the caste census — along with its discriminatory and exclusionary consequences, as revealed by official surveys and statistics. The report of the Backward Classes Commission headed by B.P. Mandal submitted to the President of India on December 31, 1980, had the following dictum inscribed at the outset: “There is equality only among equals. To equate unequals is to perpetuate inequality.”
The Mandal Commission report had extrapolated from the caste/community-wise population figures of the 1931 census and estimated the combined population of Hindu and non-Hindu OBCs to be 52% of India’s population. The commission had also arrived upon a crucial conclusion, that “caste is also a class of citizens and if the caste as a whole is socially and educationally backward, reservation can be made in favour of such a caste on the ground that it is a socially and educationally backward class of citizens within the meaning of Article 15(4).”
Article 15(4) of the Constitution enables the State to make special provisions for the advancement of “socially and educationally backward classes of citizens.” While the population share of OBCs was estimated at 52%, only 27% reservation for OBCs in all government services, technical and professional institutions was recommended by the Mandal commission, in order to keep the overall reservation for SCs, STs and OBCs together below the 50% ceiling set by the Supreme Court.
What did Indra Sawhney judgment ensure?
The V.P. Singh government implemented 27% OBC reservation for public services in 1990 and following legal challenges, the Supreme Court upheld it in the Indra Sawhney & others versus Union of India judgment in 1992 with the significant observation: “Identification of a group or collectivity by any criteria other than caste, such as, occupation cum social cum educational cum economic criteria ending in caste may not be invalid.” This cleared the way for OBC reservation in public employment and educational institutions in India.
The demand to resuscitate full caste enumeration in the national census has arisen as a corollary of these historical developments. Numerous government policies, including reservations and other benefit schemes are currently implemented on the basis of the claimed population share of caste groups, even though the actual population size of caste groups in the OBC and general category remain unknown.
Why is there opposition to a caste census?
Opposition to a nationwide caste census has been aired from some quarters on the ground that the revelation of the exact population share of OBCs greater than or equal to 52%, as estimated by the Mandal commission, would trigger demands for enhancing the 27% reservation quota for OBCs. Such demands, however, were already set into motion following the enactment of the Constitution (103rd Amendment) Act, 2019, which provided 10% reservation to “economically weaker sections” (EWS) within the general category — excluding SC/ST/OBCs — in admission to public and private educational institutions as well as civil posts and services. The law breached the 50% ceiling on reservation and faced legal challenges, but was ultimately upheld by a majority judgment of the Supreme Court in November 2022. The Chief Justice of India also dismissed a review petition against the judgment in May 2023.
With the Supreme Court itself validating reservation quotas beyond the 50% threshold, the demand for expanding OBC reservation beyond the 27% Mandal commission threshold has naturally arisen, because the 27% figure was derived as a residual from the 50% threshold after adjusting the quota for SCs and STs. It never had any basis in population estimates.
What should be the way forward?
Given the wide divergence of the OBC population thrown up by various official sample surveys ranging from 41% to 46%, between 2015-16 to 2021-22, only a full caste enumeration can help in ascertaining an accurate number. More importantly, ascertaining the numbers and proportions of individual castes within the OBC category have assumed much significance in the light of concerns regarding reservations and other benefits for OBCs getting concentrated among certain dominant caste groups. The Bihar caste survey for instance, shows that 112 small caste groups combined together in the EBC category, accounted for 36% of the State’s population, which was the largest share among social categories.
Justice Rohini Commission, which was constituted in 2017 by the Modi government to examine the sub-categorisation of OBCs in the Central list, submitted its report in August 2023. A nationwide socio-economic caste census is necessary to evolve scientific criteria for such sub-categorisation. This would also be necessary for all States, which have their own State-level OBC lists, given the wide variety in caste compositio
(Prasenjit Bose is an economist and activist. Samiran Sengupta and Soumyadeep Biswas are data analysts at CPERD Pvt Ltd, Kolkata. Courtesy: The Hindu.)
Caste Census: Why the Liberal Intelligentsia’s Opposition Is Misplaced
Sharik Laliwala
The Bihar government’s release of its caste survey has brought to the fore a fundamental fault line in India’s political representation: the 85-15 divide between India’s marginalised and upper castes.
While the headcounts of Scheduled Castes and Tribes are available in the decadal population census, the caste survey is notable because, for the first time in Independent India, it has supplied concrete numbers for the strength of the Other Backward Classes: about 63% of Bihar’s population.
Opposition parties have rallied around the demands of a nationwide caste census for long, particularly after Congress leader Rahul Gandhi’s “jitni aabadi, utna haq” pronouncement in April this year, a slogan that borrows from the progressive caste movement’s leaders and translates representation equalling population.
A nationwide caste census is essential to enable proportional representation and equitable claims over resources. The upper-caste dominance is an open secret of Indian society: they own 40% of the country’s wealth, occupy nearly 90% of media leadership positions and almost all top university faculty positions and continue to disproportionately occupy seats of power, especially in the Hindi heartland. In fact, the resurgence of Hindu nationalism in the last decade has been characterised as a “revolt of the upper castes” against democratising socio-political movements.
This time, too, the backlash to the caste census demand comes from some intellectuals. For example, Pratap Bhanu Mehta, public commentator and former vice-chancellor of Ashoka University, called the caste census demand “snake oil in the name of social justice” that mirrors the logic of Hindu majoritarianism. In 2006, when the Congress-led Union government extended the Other Backwards Classes quota in Central educational institutes, Mehta termed it a move to “inject an insidious poison” that defeated India’s interests.
In protest, Mehta, along with sociologist Andre Beteille, had resigned from the National Knowledge Commission. Others like Atul Kohli, a political scientist at Princeton University, have attributed the accentuated caste competition in Uttar Pradesh’s politics to worsening “neo-patrimonial” governance tendencies, whereby elected representatives lack a public purpose and exploit state resources for private gains.
Liberal intelligentsia, lower-caste political movements
A broad look at the liberal intelligentsia’s abhorrence of lower-caste political movements is revealing. Mehta’s critique, for instance, is focused on the elusive concept of state capacity: for him, excessive attention on caste takes away focus from inclusive governance and institutional structures that can empower marginalised groups. This perspective presents a far too simplistic, institutionalist vision of governance where social relations are divorced from state capacity. Definitions of “state capacity” can vary but the term generally refers to the ability of the state to implement policies and achieve intended outcomes.
Social anthropologist Jeffrey Witsoe, in his study of lower-caste politics in Bihar, Democracy Against Development, provides an elaborate account of how state-directed development in postcolonial India enabled dominant-caste hegemony in politics and society, despite the abolition of the zamindari system. Leaders such as Ram Manohar Lohia and later Karpoori Thakur capitalised on the vacuum of political representation for Other Backward Classes and built a bottom-up democratic movement that directly confronted the state-directed idea of development.
By the time Lalu Prasad Yadav of Rashtriya Janata Dal became Bihar’s chief minister in 1990, the bureaucracy remained in the hands of upper castes, while an agrarian conflict prompted the landless to take up arms against the state and the landlords formed private militias to perpetrate horrendous caste violence.
How is state capacity to be improved in this precarious situation? Certainly not by strengthening state institutes helmed by upper castes. As Witsoe shows, Yadav’s party, not without its flaws, either put in place sympathetic officers in some institutes or fully dismantled systems where the hold of dominant castes could not be done away with.
For proponents of liberal democracy, deinstitutionalisation is anti-development and promotes lawlessness, but from the place of an astute politician, it created immense short-run gains while assuring the dignity of left-out groups. Indeed, development was not even the agenda of backward-caste politicians: “Vikaas nahin, samman chahiye” (we want dignity, not development) was Yadav’s war cry in the early 1990s.
The liberal democratic vision of state capacity reflects a deeper misunderstanding of the politician-bureaucrat relationship vis-à-vis caste. Recent work by political studies scholar Poulomi Chakrabarti shows something more nuanced about Indian politics.
Even as the Mandal moment of Indian politics, a “silent revolution” (a la political scientist Christophe Jaffrelot), did not lead to better social spending, in places where both politicians and bureaucrats came from marginalised caste backgrounds, the bureaucracy was far more likely to spend for social welfare.
In fact, Chakrabarti’s empirical work highlights a key problem that elected representatives from non-dominant backgrounds face: bureaucrats from dominant caste backgrounds would not obey orders from Yadav in his initial years as the state’s chief minister. De-institutionalisation was a low-hanging fruit to attack the entrenched hold of dominant castes over the much-loved “system” and trump anti-incumbency for over a decade.
The only possible counterfactual to de-institutionalisation is the following: Dalits and Other Backward Classes would have continued to be humiliated, thrown out of government offices, not allowed to sit at village chowks or walk on the main streets and so on. Perhaps, that is a counterfactual that must be avoided.
Later, the politics of dignity in the 1990s and early 2000s fed into a “revolution of rising expectations” among the marginalised groups that clamored for economic gains. In turn, voters rewarded Janata Dal (United) leader Nitish Kumar’s susashan, or good governance, discourse in the mid-2000s that tapped into material gains in a rapidly growing, liberalised Indian economy.
To many, Yadav, along with other lower-caste politicians from northern India, embodies the chaos of democracy and all that is worrying with the country’s corrupt politicians. There is no denying that criminality flourished and bureaucracy deprecated under him in Bihar as he continued to govern the state from prison and his proxies in the late 1990s. But to suggest that state capacity can be developed by breaking free from the social realities of caste requires a “willing suspension of disbelief,” to use a term by British poet Samuel Taylor Coleridge.
In the political science conception, sociologist Max Weber says that the state has an absolute monopoly over violent ends. In that sense, perhaps Yadav’s Bihar, despite a strong backlash from upper-caste militias in the form of caste violence, was successful in a key area: a full stop on Hindu-Muslim violence, a task that very few chief ministers have achieved.
Why counting castes matters
Enumerating group membership, albeit fundamental to any census, inevitably reinforces some identities. Nicholas Dirks’s Castes of Mind argues that British colonial practices of census enumeration and other ethnographic studies made caste, as it is known today, a more salient category of identification and contestation. This feature is not particular to caste, though: census exercises lead to stronger group identities across the world.
A 2017 study by scholars Evans Lieberman and Prerna Singh evaluating over 1,300 census questionnaires in 156 countries over two centuries posits that enumerating ethnic identities increases the possibility of more relevant groups emerging and the subsequent ethnic competition.
Given this possibility, the state’s moral prerogative about what to enumerate and ignore comes into focus. It leads to an important ethical question: why must the Indian state, other political actors, and its concerned citizens not have information about something as fundamental as caste and jati, just as it collects data about region, language, religion, and gender? More specifically, what is in the nature of caste that makes hiding it more acceptable than other categories?
To reiterate Mehta’s critique, the new social justice agenda of “caste majoritarianism” feeds into, instead of fighting, Hindu majoritarianism – hence, let caste and jati not be counted. But Mehta, first, mistakenly lumps together the Brahminical impulse to crush India’s Muslims by relying on a mythical glorious past free of Muslims with a qualitatively different movement that seeks justice and equity for marginalised castes.
Second, by presenting caste-based political mobilisation against the idea of welfare-based programmes of various state governments like Delhi, Rajasthan and Odisha, Mehta presents a false dichotomy implying that both ideas cannot be reconciled – that states, given their limited resources, can either advocate caste-based mobilisation or roll out welfare programmes.
Bihar under Yadav was one of the first states to introduce menstrual leave for women employees in the early-1990s, and Raghuvansh Prasad Singh, a trusted aide of Yadav, was one of the key architects of the rural employment guarantee programme, the world’s largest of its kind.
Finally, Mehta’s critique wastes the intellectual energy to counter Hindu supremacy in the present political moment. Instead of preventing a debate about caste diversity in Indian society, a more productive engagement will be to ask: how should the religious right be prevented from using the caste census for majoritarian aims? Compulsive contrarianism, which blinds intellectuals to social realities they do not come from, helps no one.
Indeed, caste remains the weak point of India’s intellectual elite. Ironically, some of India’s intellectuals are liberal on individual rights, but conservative on questions of group rights. Their preference for caste-blind state institutionalism unwittingly borrows from conservative intellectual traditions: scholars like Samuel Huntington and Francis Fukuyama have emphasised the need to develop governance structures before mass participation in politics to avoid state failures.
In other words, improvements in state capacity must precede democratisation. However, the fact that many of India’s state institutes could not withstand caste-based democratisation highlights a notable feature of upper caste dominance: exceptions aside, upper caste leaders either reinforced their privileges in state institutes or failed to create capable institutes.
(Sharik Laliwala is a PhD scholar of political science at the University of California, Berkeley. His research interests include identity politics and political economy in India. Courtesy: Scroll.in.)
