On April 30, the BJP-led central government reversed the decades-long practice of excluding caste enumeration for the significantly-delayed, upcoming census. The last time caste data was recorded in the census was 1931. In 2011, the UPA government agreed to record caste data, but not as part of the main decennial exercise conducted by the Registrar General and Census Commissioner of India under the home ministry; instead, it was undertaken as a separate Socio-economic and Caste Census supervised by other departments. However, the caste data collected in 2011 was never released.
In May 2010, RSS sarkaryawah Suresh Bhaiyyaji Joshi took a position against the caste census in accordance with the principle of privileging samrasta (social harmony) over samta (social equality). However, by September 2024, the RSS had endorsed the caste census, with its leader Sunil Ambekar cautiously adding the caveat that the data should be used only for developmental purposes, and not for politics. In July 2021, the BJP-led central government took a stance against caste census in parliament, only to reverse it recently.
The rationale for this change of stance is twofold. Firstly, at the high-stakes upcoming assembly elections in Bihar, there is significant traction for a caste census due to sustained popular campaigns led by organisations like Janhit Abhiyan since 2009. Secondly, there are pressures on the BJP to blunt the explicit social justice turn of the Congress, particularly Rahul Gandhi, who has been campaigning for a nationwide caste census for the last couple of years.
The 2024 general elections vindicated the Congress’s changed stance on caste, as it increased its seats to 99 from the 52 it held in 2019, thereby also ensuring that the BJP missed the majority mark on its own. Of the four states that have conducted caste surveys so far—Karnataka (2015), Bihar (2023), Telangana (2024), and Andhra Pradesh (2024)—the first three were Congress-led or -supported governments.
However, the conversation on caste census must be rescued from the pragmatic electoral ethno-clientelism and fortunes of political parties. It must be anchored in a firmer normative ground of crafting a decent nation through substantive democracy. In other words, the deepening of the varied interpenetrative logics of justice—redistribution, recognition, representation, and reparation. A caste order of birth-based hierarchies absolutely militates against the idea of democracy premised on the equal moral worth of all human beings and perpetual contestation and contraction of unfreedoms and inequalities in all domains of social life.
Babasaheb Ambedkar viewed castes as anti-national “because they bring about separation in social life” and “generate jealousy and antipathy between caste and caste”. Ram Manohar Lohia argued that caste led to the “shrinkage and contraction of opportunity and ability” for most marginalised castes, leading to historical slavery and backwardness. Since caste, as the storehouse of social power for the privileged elite, worked against the efforts of crafting a meaningful nation and democracy, it needed to be named, frontally engaged with, and annihilated. Ambedkar lambasted the caste-blindness of the home minister in 1951, and by extension, the entire governing class for dropping caste category from the census as “if a word does not exist in a dictionary, it can be proved that the fact for which the word stands does not exist”.
The census must be visualised as a cartographic exercise that dynamically mirrors India’s immense cultural diversity and socio-economic inequality, including caste. Three important points need to be made here. Firstly, as legal scholar Upendra Baxi reminds us, there is a need to move from the imagination of caste as an ‘identity’, with a fixed essence, to the dynamic process of ‘identification’, conceived as a flow.
Caste is intimately and dynamically connected with privilege and access to material resources. A properly conducted caste census will yield disaggregated jati-level data, which will be instrumental in adjudicating the varied demands for categorical revisions and sub-classifications within the statutory SC, ST, OBC and EWS categories, periodically reviewing the inclusions/exclusions of communities, expansion of quotas, the delimitation of electoral constituencies, and overall crafting of evidence-based redistributive policies. At this point, the political classes are weaponising and profiteering from the socio-cultural cleavages through rumour mongering in the absence of credible social-scientific data about caste.
Secondly, religion has been broadly used as a proxy by the numerically challenged pan-religion dominant castes to secure their interests in a democratic game that privileges numbers. The caste census will reveal the jati-based segmentation and power differentials within all religions, thereby scrutinising the fiction of religious monoliths so instrumental in the competitive staging of the majoritarian-minoritarian conflict.
For instance, the Bihar caste survey results revealed data for more than 35 Muslim castes and computed the Muslim population share as 17.7 percent, with Ashrafs at 4.8 percent and Pasmandas 12.9 percent. The Telangana survey estimated data for around 60 Muslim castes and determined the Muslim population share to be 12.56 percent (general Muslims at 2.48 percent and backward Muslims at 10.08 percent). The officially dumped Karnataka survey, too, revealed 99 castes within Muslims, with about 17 enlisted in BC Category 1, 19 in Category 2A, and all the others, privileged as well as peripheral castes clubbed unjustifiably in an exclusive Muslim quota in Category 2B.
Historically, caste—a secular category that organised pan-religion symbolic, erotic, and material life—was religionised and transformed into a Hindu phenomenon through the orientalist-colonial knowledge-administrative regime. This orthodox but dominant view renders caste in putatively egalitarian non-Hindu communities, particularly Muslims and Christians, as illegitimate despite the strong anthropological evidence of caste-based discrimination leading to adverse policy consequences.
Finally, the pan-religion caste elite is understandably extremely anxious about getting its privilege publicly scrutinised. Even during the run-up to the 1931 census, both the Hindu Mahasabha and the Muslim League had vigorously campaigned and asked their community members to mention only Hindu and Muslim in the caste column. The citizenry must be extremely vigilant during the 2027 census process to ensure it remains as free from political interference as possible.
Khalid Anis Ansari
Associate Professor of Sociology, Azim Premji University, Bengaluru
(Views are personal)