Caste, Religion and Ethnicity in Indian Politics
Modernisation theory predicted that caste and religion would fade as politics matured. The opposite happened, and the reason is that democracy gave numbers value, which made ascrip…
Modernisation theory predicted that caste and religion would fade as politics matured. The opposite happened, and the reason is that democracy gave numbers value, which made ascriptive identity the cheapest available currency of mobilisation.
This is chapter 30 of the PSIR Optional Notes, from the part on Indian Government and Politics in the Paper I syllabus. The complete book is a free download.
UPSC syllabus
Caste, Religion and Ethnicity in Indian Politics.
In one page
- Rajni Kothari‘s formulation is the standard starting point: it is not politics that gets caste-ridden, it is caste that gets politicised. Democracy secularises caste even as caste organises democracy.
- Sanskritisation (Srinivas) is mobility by emulating the ritual practices of a higher caste; secularisation of caste is its conversion into a political interest group; both operate simultaneously.
- Three phases of caste in politics: upper-caste dominance to 1967; the rise of the middle peasant castes from 1967; and the Mandal moment from 1990 that made OBC mobilisation national.
- Mandal (report 1980, implemented 1990) recommended 27 per cent reservation for OBCs in central services, upheld with the creamy-layer exclusion and a 50 per cent ceiling in Indra Sawhney (1992).
- The caste census question is live: the last full caste enumeration was 1931; SECC 2011 produced unusable caste data; Bihar’s 2023 survey and the 2025 decision to include caste in the next census have reopened it.
- Communalism is the politics of treating religious community as the primary political unit with irreconcilable interests, and it is a modern phenomenon, not an ancient one.
- Ethnicity in India operates most powerfully in the North-East and in tribal regions, where the axis is indigeneity, land and autonomy rather than caste or religion.
- The pattern common to all three: constitutional recognition of group identity for protective purposes coexists with a prohibition on appealing to it electorally, which is a tension the system has never resolved.
Caste in politics
The theoretical frame
Rajni Kothari‘s Caste in Indian Politics (1970) rejected both the modernisation view that caste would disappear and the alarm that its persistence made democracy impossible. His argument: in a society organised by caste, politics must work through it, and in doing so it changes caste itself. Caste associations become interest groups, ritual hierarchy loses its salience relative to numerical strength, and the caste that mobilises must ally with others, which forces bargaining rather than dominance.
M.N. Srinivas supplies the two concepts that describe the mechanics. Dominant caste is a caste with numerical strength, economic power in land, and a not-too-low ritual position, which enables it to control a locality regardless of its varna rank. Sanskritisation is upward mobility through emulation of the practices of a higher caste, which is mobility within the system; Westernisation and education offer mobility outside it.
Ambedkar‘s position, from Chapter 17, is the necessary corrective: analysing caste as a resource for mobilisation risks treating as a political asset what is in origin a system of graded degradation.
Three phases
- To 1967: dominance and the vote bank. Congress worked through locally dominant castes, who delivered blocs of votes in return for patronage, the arrangement Srinivas named the vote bank. Upper castes held a share of legislative seats and offices far above their population.
- 1967–89: the middle peasant castes. The Green Revolution created a prosperous cultivating class among Jats, Yadavs, Kurmis, Patidars, Marathas, Vokkaligas, Lingayats and Reddys. Lohia’s preferential-opportunity programme, from Chapter 18, gave it an ideology; the 1967 elections and the Charan Singh phenomenon gave it a vehicle. The Kalelkar Commission (1953) had already reported on backward classes and been shelved.
- 1990 onward: the Mandal moment. Implementation of the Mandal recommendations nationalised OBC politics, produced the Samajwadi Party and Rashtriya Janata Dal as OBC-anchored formations, and prompted a counter-mobilisation on religious lines that Chapter 33 covers. The Bahujan Samaj Party, founded in 1984 out of Kanshi Ram’s organising, gave Dalit politics an independent electoral vehicle for the first time since Ambedkar.
The caste census
The 2025 question asked whether a caste census would address people’s aspirations, so the facts must be exact. The last comprehensive caste enumeration was the census of 1931; caste data other than for Scheduled Castes and Scheduled Tribes has not been collected in a census since. The Socio-Economic and Caste Census 2011 collected caste data but it was never released in usable form, having produced an unmanageable number of returned caste names. Bihar conducted a caste survey in 2023, and the Union government announced in 2025 that caste would be enumerated in the forthcoming census.
The case for. Reservation policy currently rests on 1931 proportions and on estimates; Indra Sawhney itself contemplated periodic revision of backward-class lists, which requires data. Sub-categorisation of OBCs, examined by the Rohini Commission from 2017, is impossible without knowing which groups within the category actually benefit. And the constitutional requirement of quantifiable data for reservation in promotions, from Nagaraj (2006), makes enumeration a legal necessity in some contexts.
The case against. Enumeration may harden caste identity by giving it official recognition and a statistical existence; it will generate immediate demands to raise quotas beyond the 50 per cent ceiling; caste names are not uniform across regions, so the exercise is technically difficult; and it risks converting a policy of redress for the worst-off into a competition among groups for proportional shares.
The measured answer. Data is necessary but not sufficient. Enumeration will identify who is where; whether that addresses aspirations depends on whether it is used to target the genuinely excluded, through sub-categorisation and creamy-layer refinement, or simply to reallocate quotas among already-mobilised groups.
Religion in politics
Communalism
Communalism in the Indian usage means the belief that people of one religion share secular, political and economic interests as a community, that those interests differ from the interests of other communities, and, in its extreme form, that the interests are irreconcilably hostile. Bipan Chandra‘s three-stage formulation of this is the standard scheme, and it is worth reproducing because it distinguishes the mild from the murderous.
Its origins are modern. The instruments were the colonial census and the classification of the population by religion, separate electorates from 1909, communal representation in services and legislatures, and the competition for a small pool of educated employment. The two-nation argument, whose antecedents Chapter 16 traces to Sir Syed’s later politics, culminated in the Lahore Resolution of 1940 and partition.
Post-independence, communal politics has operated through the mobilisation of majority and minority identity alike: the Ram Janmabhoomi movement from the 1980s, the demolition of the Babri Masjid in December 1992, and the riots of 1969, 1984, 1992–93, 2002 and 2020 are the reference points, as are the personal-law controversies from Shah Bano (1985) to Shayara Bano (2017).
Secularism in practice
The constitutional design is set out in Chapter 23. What belongs here is the political practice. Representation of the People Act 1951, Section 123(3), makes an appeal on the ground of religion, race, caste, community or language a corrupt practice, and Abhiram Singh (2017) read this to prohibit appeals to the religion of the candidate or the voter. In practice the provision is nearly unenforceable, since appeals are made obliquely and election petitions take years.
The competing critiques of Indian secular practice, from the right, from the left and from Ambedkar’s position, are given in Chapter 23 and should be cross-referenced rather than repeated.
Ethnicity
Ethnicity is the third axis and the least well covered in most answers. Its politics turn on indigeneity, land and cultural survival rather than on ritual rank or on faith.
- The North-East. Autonomy demands under the Sixth Schedule, statehood movements producing Nagaland (1963), Meghalaya (1972), Mizoram and Arunachal Pradesh (1987), and insurgencies negotiated into accords, the Mizo Accord of 1986 being the most successful and the Naga framework agreement of 2015 still unconcluded.
- Insider–outsider politics. The Assam movement of 1979–85 over illegal migration, the Assam Accord of 1985, and the subsequent NRC exercise; similar dynamics in Tripura, where demographic change made the indigenous population a minority.
- Adivasi politics in central India. Land alienation, forest rights and displacement by mining and dams; the Jharkhand movement culminating in statehood in 2000; the Forest Rights Act 2006; and the Fifth Schedule and PESA framework from Chapter 26.
- Sub-regional identity elsewhere, such as Bodoland, Gorkhaland and Vidarbha, where ethnicity, language and development grievance combine.
Debate: Is the politicisation of caste good or bad for Indian democracy?
Good. It has made democracy substantive. Caste mobilisation brought into legislatures and offices groups that upper-caste-dominated parties never selected; it converted a hierarchy of ritual status into a competition of numbers, in which the numerous have an advantage they lacked; and it produced the redistributive politics of reservation, welfare and land. Kothari’s insight holds: politics has secularised caste more than caste has communalised politics. Bad. It has made ascriptive identity the stable currency of politics, crowding out programmatic competition on health, education and employment. It rewards group leaders rather than group members, so the benefits concentrate in dominant sub-castes, which is why sub-categorisation became necessary. And it has produced counter-mobilisation on religious lines, since a politics organised by group numbers invites the largest group to organise as one. The examiner’s line. Both, and the sequence matters. Caste mobilisation was democratising in its first phase because it broke a monopoly; it becomes an obstacle in its later phase when the groups it empowered begin to defend their own dominance. The test to apply is whether mobilisation is expanding access or protecting a share. Kothari’s own later writing recognised this, and the caste-census debate is exactly this question in current form.
Where answers lose marks
- Treating caste politics as a deviation from real politics. Kothari’s point is that in a caste society there is no other politics available, and that participation changes caste.
- Confusing sanskritisation with secularisation of caste. The first is ritual emulation upward; the second is conversion into a political interest group.
- Saying communalism is ancient. Its instruments, the census, separate electorates and competition for state employment, are colonial and modern.
- Omitting ethnicity. The syllabus names it, and answers that cover only caste and religion lose the North-East and Adivasi politics entirely.
- Giving the caste census without the 1931 baseline and the SECC 2011 failure. Those two facts are what make the demand intelligible.
- Presenting reservation politics as identical to caste politics. Reservation is a constitutional remedy; caste mobilisation is an electoral strategy, and they diverge often.
Asked before
- Caste remains a vital axis for political mobilisation in India. How would the caste census address the aspirations of people? Discuss. (2025, Paper I, 15 marks)
- The rise of caste politics is to be attributed to both regional aspirations and electoral manifestations. Comment. (2023, Paper I, 20 marks)
Answer skeleton
Caste remains a vital axis for political mobilisation in India. How would the caste census address the aspirations of people? Discuss. (15 marks, 250 words)
Frame. Accept the first clause with Kothari, then treat the second as a genuine question rather than an endorsement: data addresses aspirations only if it is used in a particular way.
Why caste remains the axis. Democracy gives numbers value; caste supplies ready-made networks of trust and communication; and the state’s own categories, SC, ST, OBC, EWS, make caste the currency of entitlement. Three phases: upper-caste dominance to 1967, the middle peasant castes after, and the Mandal moment from 1990.
The factual position. No comprehensive caste enumeration since 1931; SECC 2011 unusable; Bihar’s survey in 2023; the 2025 decision to enumerate caste in the next census.
How it could address aspirations. Reservation currently rests on 1931 proportions and estimates; Indra Sawhney contemplated periodic revision; the Rohini Commission’s sub-categorisation of OBCs is impossible without it; Nagaraj (2006) requires quantifiable data; and evidence would replace assertion in a debate now conducted on both.
How it could fail to. It may harden identity by giving it official statistical existence; it will generate immediate pressure on the 50 per cent ceiling; caste nomenclature is not uniform across regions; and it risks converting redress for the worst-off into proportional bargaining among mobilised groups.
Conclude. Necessary but not sufficient. Enumeration answers the question of who is where; whether aspirations are met depends on sub-categorisation, creamy-layer refinement and delivery of education and employment, none of which the census itself provides.
Last-mile revision
- Kothari Caste in Indian Politics (1970): caste gets politicised, politics secularises caste. Srinivas: dominant caste, sanskritisation, westernisation, vote bank.
- Phases: upper-caste dominance to 1967; middle peasant castes 1967–89 with Lohia and Charan Singh; Mandal from 1990; BSP from 1984.
- Kalelkar Commission 1953 shelved; Mandal Commission report 1980, implemented 1990; Indra Sawhney (1992) with creamy layer and the 50 per cent rule; Rohini Commission from 2017.
- Caste census: last full enumeration 1931; SECC 2011 unusable; Bihar survey 2023; 2025 decision to enumerate.
- Communalism: modern origins in census, separate electorates from 1909, employment competition; Lahore Resolution 1940; Section 123(3) RPA 1951 and Abhiram Singh (2017).
- Ethnicity: Sixth Schedule; statehood 1963 Nagaland, 1972 Meghalaya, 1987 Mizoram and Arunachal; Assam movement 1979–85 and the Accord of 1985; Mizo Accord 1986; Jharkhand 2000; Forest Rights Act 2006.
Read the rest. This chapter is one of 58 in the complete PSIR Optional Notes, covering Paper I and Paper II in full — free to download.