Caste in Indian Politics

Caste is the oldest organising principle of Indian society and, on every serious account, the most durable variable in Indian electoral politics. The interesting question is not whether caste survives — it plainly does — but what democracy has done to it. Universal adult franchise gave numbers a value they never had in a ritual hierarchy, and in claiming that value caste stopped being only a system of purity and became a vehicle of interest, organisation and bargaining.

What Caste Is: Jati, Varna and a Problem of Translation

• There is no exact translation of the Indian term jati in any European language, and this is not a linguistic accident — European societies produced nothing structurally equivalent to it.
  • Jati is the endogamous, hereditary and occupationally associated local group into which a person is born and within which they marry; there are several thousand, and their boundaries are regional.
  • Varna is the fourfold classificatory scheme of the classical texts — Brahmin, Kshatriya, Vaishya, Shudra — with the untouchables placed outside it altogether as avarna.
  • Varna is a template; jati is the lived unit. No amount of varna theory will tell you who is dominant in a Bihar village, which is why political analysis works with jati while constitutional categories work with lists.

Feature	Varna	Jati
Nature	Textual, pan-Indian classificatory model	Empirical, local, endogamous group
Number	Four, with a fifth excluded category	Several thousand, varying by region
Basis of rank	Ritual function ascribed by scripture	Ritual status, landholding and numbers
Mobility	Effectively closed	Group mobility through Sanskritisation
Political salience	Rhetorical, used in ideological argument	Operative — the unit of mobilisation and voting

• Colonial administrators used the word caste for the Indian practice of jati, and the choice of word did lasting damage to how caste is understood abroad.
  • The term derives from the Portuguese casta — lineage, breed, pure blood — the vocabulary of stock and descent rather than of ritual rank.
  • Because casta carried the sense of purity of blood, caste was pulled towards the Western category of race, and casteism came to be equated with racialism in Western usage.
  • The conflation is analytically wrong. Caste is a social and ritual construct; race is asserted as a biological one, and caste groups share physical descent with the groups they rank above and below.

Caste, Race and the Durban Dispute

• India has consistently refused to have caste discrimination classified as racial discrimination in international forums, and the dispute came to a head at the World Conference Against Racism, Durban, 2001.
  • India made anti-colonialism and anti-racialism founding objectives of its foreign policy, and has resented the reverse charge that it practises a domestic form of apartheid.
  • At Durban the government argued that caste is a social construct and race a biological one, that the two are not interchangeable, and that caste discrimination was an internal matter already addressed by the Constitution.
  • Dalit organisations campaigned for the opposite position, arguing that caste is descent-based discrimination within the meaning of the racial discrimination convention, and describing untouchability as a “hidden apartheid” whose mechanisms — segregation, endogamy, enforced degradation, violence — mirror those of racism.
  • The Indian position prevailed in the intergovernmental text, though caste entered the parallel non-governmental declaration, and the UN Committee on the Elimination of Racial Discrimination subsequently read descent-based discrimination as covering caste.
• The dispute matters because it fixes where the remedy lies. If caste is race, it becomes a subject of international human rights supervision; if it is a domestic social structure, the remedy is Indian constitutional law and Indian politics — and the Indian state has insisted on the second reading throughout.

Caste was translated into a European vocabulary it does not fit, and the mistranslation still shapes both foreign criticism and Indian defensiveness.

The Sociology of Caste

Dumont: Caste as the Indian Model of Stratification

• Louis Dumont described caste as the Indian model of social stratification, standing opposite the Western model built on class — the contrast is the argument, not an incidental comparison.
  • In Homo Hierarchicus he held that caste is not a variety of stratification at all but a distinct form of inequality grounded in religious ideology.
  • The organising opposition is between the pure and the impure. It generates hierarchy, because the pure ranks above the impure, and separation, because the two must be kept apart — endogamy, commensality rules and residential segregation follow from one principle.
  • The Brahmin embodies ritual purity at the apex; the untouchable absorbs permanent impurity at the base; every other group is placed by reference to that axis.
• Dumont’s most contested claim is that the ritual encompasses the material — that in the caste order, status is separated from and superior to power.
  • The Brahmin holds ritual supremacy without necessarily holding land or arms, and the king holds power while conceding ritual precedence.
  • This disjunction of status and power is what Dumont thought made India different: the West subordinates rank to wealth, India subordinates wealth to rank.
  • He read Indian society as holistic — the whole precedes and defines the parts — against a Western individualism in which the individual is the primary value.

Basis of comparison	Caste	Class
Determinant of status	Birth — ascribed, unalterable in a lifetime	Merit and achievement — status is acquired
Criterion of ranking	Purity and pollution, a ritual criterion	Economic position — income, property, occupation
Boundary	Closed — not crossable by an individual	Open — movement between classes is possible
Foundation	Religious ideology encompassing the material	Material relations determining social standing
Consciousness	Group and community consciousness	Class consciousness of shared economic interest
Marriage and food	Regulated by endogamy and commensality rules	No formal restriction, though social distance operates
Political expression	Caste associations and caste-anchored parties	Trade unions and class-based parties

Srinivas’s Challenge and the Mechanics of Mobility

• M.N. Srinivas challenged Dumont on the ground that his approach is too textual — derived from classical literature rather than from what happens in villages.
  • Read from the field rather than from scripture, the difference between Western and Indian tradition is smaller than Dumont allows.
  • Class is as static as caste in practice, and as much determined by birth: the child of a wealthy family in an industrial society inherits its position about as reliably as a Brahmin child inherits ritual rank.
  • The larger point is methodological — caste must be studied through fieldwork in the village and the region, and when it is, the system permits far more movement than the textual model predicts.
• Mobility was permitted, but it was group mobility, not individual mobility. A whole jati could improve its standing over generations; a person could not change caste.

Sanskritisation

• Srinivas coined Sanskritisation for a lower caste adopting the ritual practices, customs and lifestyle of a higher caste, usually of the twice-born, to claim a higher position locally.
  • The classic move is giving up meat and liquor, adopting vegetarianism and teetotalism and Brahminical rites, thereby reducing the pollution attaching to the group.
  • Other markers followed — claiming a mythical Kshatriya genealogy, adopting a new caste name, restricting widow remarriage, tightening controls on women.
  • There are many instances of Shudra groups elevated to Kshatriya status in this way, across the subcontinent and across centuries.
• The Jatavs of Agra are the standard example among untouchable groups, and their trajectory is instructive.
  • This leather-working community Sanskritised in the early twentieth century, claiming Kshatriya origin and adopting the practices that came with the claim.
  • Once reservation was introduced, they de-Sanskritised. Documented Scheduled Caste status came to be worth more than a claimed higher ritual rank.
  • The reversal is the sharpest single illustration of what democratic politics did to caste: it changed the direction in which it paid to move.
• Sanskritisation has limits Srinivas himself noted. It leaves the hierarchy intact — mobility within the system, not against it; it is available only to groups with an economic base; and it typically imposes new restrictions on its own women.

Westernisation and the Secularisation of Caste

• Westernisation, the companion concept, is mobility outside the ritual system — through English education, professional employment, urban residence, and the adoption of Western dress, science and manners.
  • Its beneficiaries were often the already advantaged upper castes, who had the literacy and the proximity to colonial institutions to take it up first.
  • Ambedkar’s own route — Columbia, the London School of Economics, the bar — was Westernisation, and it made possible a challenge to caste that Sanskritisation could never mount.
• The secularisation of caste is a different process altogether, and confusing it with Sanskritisation is a real analytical error.

	Sanskritisation	Secularisation of caste
What changes	The group’s ritual practice and claimed rank	The group’s function — ritual unit to interest group
Direction	Upward within the ritual hierarchy	Sideways out of the ritual frame, into competition
Reference point	A higher caste whose customs are emulated	The state, its offices, quotas and contracts
Effect on hierarchy	Leaves it intact and validates it	Corrodes it — numbers, not purity, become currency
Instrument	Diet, ritual, genealogy, marriage rules	Caste association, party, alliance, vote
Illustration	Shudra groups claiming Kshatriya descent	Jatavs abandoning that claim for SC entitlement

• Put plainly: Sanskritisation asks a higher caste for recognition; the secularisation of caste asks the state for a share. The first accepts the ladder and climbs it; the second stops caring about the ladder and starts counting heads.

Dominant Caste

• Srinivas held that the term “upper caste” is misleading, because ritual rank and actual power frequently come apart — and it is power that decides political outcomes.
  • A dominant caste holds economic, social and political power in a locality, whatever its position in the varna scheme.
  • He first used the term for the caste in a village that is numerically strong and wields the greatest economic and political power.
  • In most situations the ritually upper castes are not the dominant castes; Brahmins are often numerically tiny and, outside a few regions, not the principal landholders.
• Three characteristics make a caste dominant, with regional variation in how they combine.
  • Landownership — control of the principal productive resource in an agrarian society, and with it control over others’ labour.
  • Numerical strength — a substantial share of the local population, which under universal franchise converts directly into votes.
  • A not-too-low ritual position — enough social standing to be acceptable as a leader.

Region	Dominant caste(s)
Uttar Pradesh and Bihar	Yadavs
Haryana and western Uttar Pradesh	Jats
Maharashtra	Marathas
Andhra Pradesh and Telangana	Reddys, Kammas, Kapus
Karnataka	Lingayats and Vokkaligas
Gujarat	Patidars
Tamil Nadu	Nadars, alongside Thevars, Gounders, Vanniyars
Punjab	Jat Sikhs
Kerala	Nairs, and in numbers the Ezhavas

• The concept explains why the fortunes of caste and of ritual rank moved apart after 1947. Land reform, the Green Revolution and the franchise all rewarded numbers and landholding rather than purity, and the castes holding both became the pivots of state politics.

Ambedkar: Caste as Graded Inequality

• Ambedkar’s sociology is the necessary corrective to any account treating caste as a neutral resource for mobilisation — it is, in origin, a system of graded degradation.
  • He rejected the defence of caste as a division of labour, arguing that it is not a division of labour but a division of labourers, ranked, and assigned before any aptitude can be tested.
  • Caste is endogamy superimposed on an already stratified population; his early formulation was that its mechanism is the enclosure of the group by prohibiting marriage outside it, with sati, enforced widowhood and child marriage as devices for managing the resulting surplus of women.
• Graded inequality is his central structural insight, and it explains why caste has proved so hard to abolish.
  • The system is not a binary of oppressor and oppressed but a ladder in which almost every group has someone below it to look down on.
  • This produces an ascending scale of reverence and a descending scale of contempt — each caste defers upward and despises downward, and each has a stake in the rung it occupies.
  • The political consequence is that there is no single unprivileged class except the one at the very base, and therefore no general mobilisation of the oppressed is spontaneously available. Graded inequality manufactures its own consent.
• Caste is thus not merely a social division but an obstacle to collective action, which is why Ambedkar concluded that political and economic reform without the annihilation of caste would leave the structure standing.

Ghurye: The Features of Caste

• G.S. Ghurye’s enumeration remains the standard structural description, useful precisely because it is descriptive rather than theoretical.
  • Segmental division — closed groups fixed by birth, each with its own council and code of conduct.
  • Hierarchy — the segments are not merely different but ranked, from Brahmins to untouchable groups.
  • Restrictions on feeding and social intercourse — elaborate rules govern from whom food and water may be accepted, and the rules encode rank.
  • Civil and religious disabilities and privileges — differential rights of residence, temple entry, dress, speech and access to public space.
  • Lack of unrestricted choice of occupation — hereditary assignment, with the clean–unclean distinction doing most of the ranking work.
  • Restrictions on marriage — endogamy, and hypergamy within the caste, are the mechanisms that reproduce the structure across generations.
• Ghurye also insisted that the British census and the politics of representation transformed caste from a local ritual order into an all-India political category.

Béteille: Caste, Class and Power

• André Béteille’s study of the Tanjore village of Sripuram is the empirical foundation for treating caste, class and power as three distinct dimensions of inequality rather than one.
  • He found that the three used to coincide — Brahmins held ritual rank, land and authority together — and that in the twentieth century they had begun to come apart.
  • Non-Brahmins acquired land and, with adult franchise, political office, while Brahmins kept ritual precedence and migrated into the professions and the towns.
  • He found significant internal differentiation among Brahmins, and class divisions inside single castes, cutting against Dumont’s monolithic ritual apex.
• The analytical yield is the caste–class–power triad: position is fixed by ritual status, economic class and political power, and modern India is characterised by the growing divergence among the three.
  • Where the three coincide, inequality is cumulative and stable; where they diverge it becomes dispersed and contested, and politics becomes the arena in which the mismatch is fought out.
  • This divergence, not the persistence of hierarchy, is what makes caste politically explosive. A group with numbers but no status, or wealth but no rank, has every incentive to organise.

From Ritual Order to Political Resource

Caste in Pre-colonial Politics

• Caste has been present in Indian politics since ancient times, and on a long-standing reading the great religious contests were themselves caste alignments.
  • Classical Brahminical Hinduism expressed an alliance of Brahmins and Kshatriyas — the priest legitimising the king and the king protecting the priest.
  • Buddhism drew on an alliance of Vaishyas and Kshatriyas, the mercantile and ruling groups whose interests were not served by ritual monopoly.
• Pre-colonial kingship worked through and around caste rather than above it. Royal power distributed land and honours, sponsored temples, adjudicated caste disputes and could confirm a group’s rank — which gave Sanskritisation a political as well as a ritual dimension.

The Colonial Census: Classification as an Act of Power

• The British promoted caste politics, and their most consequential instrument was enumeration.
  • Caste was recorded in the decennial census from 1871–72 onward, and compilers had to arrange thousands of local jatis into a single ranked all-India frame.
  • Classification created what it claimed to describe. Groups petitioned to be recorded under a better name or a higher varna; caste associations were founded to lobby census officers; a fluid negotiated order was fixed in writing and given official standing.
  • The last full caste enumeration was the census of 1931, and every reservation calculation since has rested on proportions derived from it or estimated around it.
• Colonial policy then attached representation to the categories it had created — communal electorates, reserved posts, army recruitment by “martial race” — so that being counted became the route to being served.

The Non-Brahmin and Anti-caste Movements

• The anti-caste movements of the nineteenth and twentieth centuries were the first mass politics organised on caste, and they were strongest in the south and west.
  • Jyotiba Phule founded the Satyashodhak Samaj in 1873 in western India — a truth-seekers’ society that attacked Brahminical ritual monopoly, promoted schooling for Shudras, Ati-Shudras and girls, and devised priestless marriage and life-cycle ceremonies to break dependence on Brahmin officiants.
  • Sree Narayana Guru’s SNDP Yogam organised the Ezhavas of Kerala around temple entry, education and the injunction of “one caste, one religion, one god for man”, turning a ritually excluded but numerous community into a political force.
  • The Justice Party, formed in the Madras Presidency in 1916, was the non-Brahmin party of the intermediate castes, and used the reformed councils to press for communal representation in education and public employment.
  • E.V. Ramaswamy Naicker — Periyar — launched the Self-Respect Movement after his attempts to have the Tamil Nadu Congress adopt caste quotas in representation were defeated repeatedly between 1919 and 1925.
• Periyar’s programme was more radical than reform. He treated Hinduism as the instrument of Brahminical domination, ridiculed the Puranas, promoted self-respect marriages without priests, and framed the conflict as Dravidian against Aryan — an ethnic idiom for what was in substance a caste conflict.
  • In 1938 he became president of the Justice Party; in 1944 he reorganised it as the Dravidar Kazhagam, and the Dravidian parties that followed carried non-Brahmin assertion into electoral power.
  • The south’s advantage was demographic. The only significant upper caste was the Brahmin, a very small minority, and the great majority of the population was non-Brahmin; Brahmins could not counter a mobilisation of that size.
• The north followed a different path, and the divergence still shows.
  • The Arya Samaj and the Shuddhi movement strengthened Vedic religion and offered lower castes a route of ritual upgrading inside Hinduism.
  • There was more than one entrenched upper caste to displace, which made confrontation harder to organise.
  • Backward groups therefore chose Sanskritisation over confrontation, Brahminism continued, and autonomous lower-caste parties arrived late — the backward castes remaining inside the Congress long after their southern counterparts had left.

The Communal Award and the Poona Pact

• The Communal Award of 1932 extended separate electorates to the Depressed Classes, bringing the caste question to its constitutional crisis.
  • Separate electorates would have let Dalits elect their own representatives through their own votes alone — the strongest form of political recognition available, and Ambedkar’s demand.
  • Gandhi read it as the vivisection of Hindu society and fasted against it, arguing that separate electorates would permanently divide the Hindu community.
  • The Poona Pact of 1932 replaced separate electorates with reserved seats in joint electorates, and substantially increased the number of reserved seats.
• The design chosen at Poona still governs. A reserved constituency returns a Scheduled Caste member elected by a general electorate, so the winner must be acceptable to non-Dalit voters — presence without independent leverage, and a structural weakness argued over ever since.

The Caste Composition of the National Movement

• The parties of the national movement were, in their social base, caste parties, whatever their stated universalism.
  • The Congress was overwhelmingly led by Brahmins and other upper castes, especially in its early decades and in the presidencies where English education penetrated first.
  • The Justice Party and the Dravidar Kazhagam were the parties of the intermediate and backward castes, built explicitly against Brahmin dominance.
  • The Republican Party of India, descending from Ambedkar’s Scheduled Castes Federation, was the party of the Dalits.
• The nationalist answer to this was that caste would dissolve in the solvent of national unity and economic development — an expectation that shaped the constitutional design and was substantially disappointed.

1947: Abolishing Untouchability without Abolishing Caste

• The Constitution abolished untouchability but did not abolish caste; it retained caste as the basis of distributive justice.
  • Article 17 abolishes untouchability and makes its practice in any form an offence — the one categorical prohibition.
  • Articles 15 and 16 forbid discrimination on grounds of caste, while their own provisos permit special provision for socially and educationally backward classes and for the Scheduled Castes and Scheduled Tribes.
  • Articles 341 and 342 require the President to notify lists of Scheduled Castes and Tribes, making caste a category of official record; Articles 330 and 332 reserve legislative seats on that basis.
• The result is a system that prohibits caste as a ground of discrimination and requires it as a ground of remedy, and the tension has never been resolved.
  • Every claim for redress must be framed as a caste claim, which keeps caste identity administratively alive and worth asserting.
  • Linguistic reorganisation, the Green Revolution and the Mandal moment each strengthened rather than weakened caste’s political role — the first by making state politics the arena where dominant castes competed, the second by enriching the middle peasant castes, the third by nationalising backward-class assertion.

The Constitution made caste illegal as a reason to exclude and indispensable as a reason to include; Indian politics has lived inside that contradiction ever since.

The Politicisation of Caste: Kothari’s Argument

• Rajni Kothari’s is the founding argument of the field, and it rejects both positions available to him. Against modernisation theory, caste was not going to disappear; against the alarmists, its persistence did not make democracy impossible.

“It is not politics that gets caste-ridden, it is caste that gets politicised.” — Rajni Kothari

• Kothari credited caste with the success of Indian democracy rather than blaming it for democracy’s difficulties.
  • Caste supplied ready-made structures of mobilisation and integration through which ordinary Indians were attached to a parliamentary system otherwise entirely alien to them — networks of trust, communication, leadership and obligation that no party had to build.
  • The absence of comparable structures in other post-colonial societies is, on his account, part of why democracy failed there; where nothing stood between the individual and the state, mass politics had no intermediary institutions to run through.
  • The influence runs both ways, and it is the second direction — what politics does to caste — that his phrase names.

The Three Faces of Politicisation

• Secularisation — politics changed what caste is for.
  • Caste was formerly significant for ritual purposes; it is now significant for secular benefits — reservation in education and employment, licences, contracts, transfers, tickets and office.
  • Once the currency is material, ritual rank stops determining a caste’s worth and numerical strength starts to. A large low-ranking caste becomes politically more valuable than a small high-ranking one.
  • Caste associations become interest groups — with offices, journals, hostels, scholarships, marriage bureaus, legal funds and a bargaining relationship with parties. The form is traditional; the function is entirely modern.
• Integration — caste is usually described as a disintegrative force, but democracy compels integration.
  • No single caste anywhere in India is a majority. Winning therefore requires alliance, and alliance requires bargaining with groups one’s own ritual code held at a distance.
  • The process produces both fission — a caste splitting along class or sub-caste lines — and fusion, several castes coalescing into a bloc.
  • Caste enters politics and comes out changed: broader, more instrumental, more negotiable.
• Consciousness — entry into politics generates a new self-awareness that then politicises itself.
  • Lower castes discover through electoral arithmetic that their numbers make them important, and the discovery does not stay confined to the polling booth.
  • Dependence gives way to autonomous mobilisation: Dalits were initially led by the Congress, then recognised their own weight and built their own vehicle, the Bahujan Samaj Party.
  • The same sequence runs through backward-class politics — first a vote delivered by patrons, then a vote bargained for, then a party of one’s own.

The Prominent Caste Coalitions

Coalition	Composition	Region and architects
AJGAR, later MAJGAR	Ahirs, Jats, Gujjars, Rajputs, with Muslims added later	Western UP and Haryana; proposed by Sir Chhotu Ram, later pursued by Chaudhary Charan Singh
KHAM	Kshatriyas, Harijans, Adivasis, Muslims	Gujarat; assembled by the Congress to build a majority without the Patidars
Muslim–Yadav	Muslims and Yadavs	UP and Bihar; base of the Samajwadi Party and the Rashtriya Janata Dal
Dalit–Muslim	Scheduled Castes and Muslims	UP; attempted by the BSP, widened by its Brahmin outreach of 2007
Bahujan	All non-elite castes — SCs, STs, OBCs, minorities	The ideological frame of Kanshi Ram’s organising

• These coalitions illustrate the integration thesis exactly. AJGAR yoked a ritually degraded pastoral caste to a Rajput elite; KHAM put Kshatriyas and Harijans in one column. Neither combination has any warrant in the ritual order, and both were rational as electoral arithmetic.
• They also show the fragility of the form. A caste coalition holds only while its members believe they gain more inside it than outside; AJGAR fractured over Jat–Dalit conflict on the ground, and KHAM collapsed under the counter-mobilisation of the excluded Patidars.

Two Stages of the Caste–Politics Interaction

• Kothari analysed the relationship at two levels — how caste affects politics and how politics affects caste — and traced the interaction through two successive stages.
  • What set the process going was that education, government patronage and a slowly expanding franchise penetrated the caste system, and in penetrating it produced new leadership from outside the traditional elite.
  • The involvement of caste did two things at once: it made available to that leadership a structural and ideological basis for mobilisation, and it forced the leadership to make concessions to local opinion and to organised castes for social and economic reasons.
• The first stage is a contest between entrenched and ascendant high castes, driven by an intellectual leadership.
  • The antagonism runs between high entrenched castes — Reddys in Andhra, Patidars in Gujarat, Bhumihars in Bihar — and high ascendant castes such as Kayasthas in Bihar and Jats in Rajasthan.
  • In this stage it is the power structure of caste that operates: mobilisation runs along existing lines of dominance and dependence.
• The second stage is factionalism and fragmentation within the competing castes.
  • Competition inside a dominant caste splits it, and each faction reaches downward for allies, so that multi-caste and multi-factional alignments develop.
  • Lower castes are mobilised to strengthen a faction, and in being mobilised acquire organisation, leaders and expectations of their own.
  • Here it is caste consciousness, not the caste power structure, that operates — which is precisely how the instrument of the dominant becomes the vehicle of the subordinate.

Kothari’s Four Conclusions

• A new elite structure emerged — different in caste composition from the old one but common in outlook; the men who displaced the traditional elite adopted its methods and its interest in office.
• New caste-based organisational forms emerged — the caste association, the caste federation, the caste-anchored party — none of which existed in the ritual order.
• The electoral system acquired a new relevance, becoming the arena in which caste claims were registered and settled, and thereby acquiring a social significance formal design alone could not have given it.
• Factional division became a permanent feature of politics, as competition inside castes proved as consequential as competition between them.

Rudolph and Rudolph: The Modernisation of Tradition

• Lloyd and Susanne Rudolph described the politicisation of caste as the “modernisation of tradition” and the “traditionalisation of modernity” — two halves of a single process.
  • Modernisation of tradition: a traditional institution takes on modern functions, so that the caste association becomes a pressure group with a constitution, an office and a bargaining position.
  • Traditionalisation of modernity: modern institutions come to work through traditional identities, so that the party ticket, the ministry and the transfer order are distributed by caste calculation.
  • Neither tradition nor modernity survives the encounter unchanged, which is why arguments about whether India is modernising or regressing generally miss what has happened.

“Caste has helped peasants to represent and rule themselves by attaching them to the idea, processes and institutions of political democracy.” — Lloyd and Susanne Rudolph

• Their formulation is that caste has responded to its changing environment by transforming itself from below and within — hierarchy, privilege and moral parochialism no longer exhaust its secular significance.

Three Types of Caste-Based Mobilisation

Type	Who mobilises	Effect on hierarchy	Their examples
Vertical	“Notables” of the upper or dominant castes lead those beneath them	Leaves the hierarchy largely undisturbed; politics mirrors the old order	The Congress system of the first two decades
Horizontal	A caste association, mobilising one large caste without upper-caste patrons	The unsettling form — turns ritual subordination into electoral strength	The Nadars of Tamil Nadu; the Jats of Rajasthan
Differential	A political party, using electoral arithmetic rather than caste leadership	Reorganises caste itself by fission and fusion	The Mukkulathor fusion of Kallar, Maravar and Agamudayar; the assembly of the OBC category

• Horizontal mobilisation is the master concept for understanding what caste does to a democracy. A caste large enough in numbers can accumulate power in electoral competition without relying on notables at all, and the Rudolphs saw the logic as national wherever the demographic weight existed — Bihar and Uttar Pradesh proving them right.
• Differential mobilisation is the most modern of the three, because the initiative lies with the party. The OBC category is itself a differential construction — several backward castes that had not cooperated before, brought together first by a policy category and then by parties competing for it.

Phases of Caste Politics since 1947

Phase	Character	Markers
Until 1967	Congress worked through locally dominant castes, who delivered blocs of votes for patronage	Upper-caste share of seats far above population; vertical mobilisation; the vote bank
1967–89	The Green Revolution created a prosperous cultivating class that demanded political returns	Jats, Yadavs, Kurmis, Patidars, Marathas, Vokkaligas, Lingayats, Reddys; the Charan Singh phenomenon; Lohia’s preferential-opportunity argument; the Kalelkar Commission of 1953 already shelved
From 1990	Backward-class assertion is nationalised and given an institutional stake	The Samajwadi Party and RJD as OBC-anchored formations; the BSP, founded 1984; counter-mobilisation on religious lines
Since 2014	A national party assembles a cross-caste Hindu bloc while competing on welfare delivery	Cultivation of non-dominant OBCs and non-Jatav Dalits; beneficiaries as a category; caste returning through the caste census demand

• The two idioms of contemporary Indian politics are conventionally described as Mandal and Kamandal — the assertion of backward-caste interest and of religious majority identity — and much of what has happened since 1990 is the contest between them for the same voters.
• Christophe Jaffrelot calls caste-based mobilisation India’s “silent revolution” — silent because it happened without insurrection, revolutionary because it transferred political power from the upper castes to the lower and intermediate castes through the ballot alone.
  • Its measure is that lower-caste leaders now form governments in the largest states, which he reads as genuine democratisation rather than decay.
  • He notes that caste forms the mosaic of Indian politics — the units are many, small and locally distinct, and the picture becomes visible only at a distance.
• Yogendra Yadav describes the interaction of caste and politics as producing two democratic upsurges.
  • The first, in the 1960s, saw OBCs coming out from under the Congress umbrella as an autonomous political force.
  • The second, in the 1990s, saw Dalits leave the Congress and build their own party, alongside a marked rise in turnout among the poor, the rural and the lower castes — the reversal of the Western pattern in which the privileged vote most.
  • His conclusion is that caste mobilisation is what keeps Indian democracy from remaining merely a “topsoil” — a thin institutional layer over an untouched society.
• Satish Deshpande argues that Indian politics must be read with reference to the direction of backward-class politics, which he treats as the heart of the system: the intermediate castes are numerically unavoidable, present in every alliance, and their choices will determine whether caste becomes more prominent or gets diluted.
• On the emancipatory reading of caste politics, caste mobilisation has bridged gaps between sections of society and dislodged entrenched dominance.
  • The standing examples are the RJD, which made the Yadavs a formidable political entity, and the Janata Dal (United) under Nitish Kumar, which united the castes below the Yadavs and gave them a political voice.
  • The second is the more important, because it shows caste politics correcting its own concentration — the beneficiaries of one round of mobilisation becoming the target of the next.

Does Caste Strengthen or Weaken Indian Democracy?

• There is no scholarly consensus here, and the disagreement is genuine rather than terminological. Kothari, Jaffrelot, Yogendra Yadav and Deshpande hold that caste has strengthened Indian democracy; Béteille, Ashutosh Varshney, C.P. Bhambri and Arun Shourie hold that it weakens it and treat it as a long-term threat.

The Case That Caste Has Strengthened Democracy

• Caste made democracy substantive rather than formal, bringing into legislatures and offices groups that upper-caste-dominated parties would never have selected, within two generations of universal franchise.
• It converted a hierarchy of ritual status into a competition of numbers, in which the numerous have an advantage they never possessed under any previous dispensation.
• It produced a redistributive politics — of reservation, welfare and land — that a purely programmatic politics of the educated middle class would not have generated.
• It deepened participation. Turnout among the poorest and lowest-status groups in India exceeds turnout among the well-off, the inverse of the pattern in most established democracies, and caste mobilisation is the main reason.
• Asghar Ali Engineer treats identity politics as an indicator of a healthy democracy, because it shows vulnerable sections making an active bid for a share in power.

“Democratic process intensifies power struggle, and hence, assertion of various identities assumes vital importance in a democratic set-up. Under authoritarian regime various identities may be suppressed or subordinated but it cannot be done under a democratic regime.” — Asghar Ali Engineer

• Kancha Ilaiah makes the complementary point about how little had changed before mobilisation began.

“Neither in family relations, nor in market relations and nor in political relations has democratisation of civil society taken place so far. The caste system has frozen mobility in social structures and socioeconomic relations.” — Kancha Ilaiah

The Case That Caste Weakens Democracy

• C.P. Bhambri argues that caste politics is corrosive over the long term because it invites religious counter-mobilisation.
  • When one party mobilises on caste, its opponent is pushed to mobilise on religion — the only identity large enough to overcome a caste coalition is the religious majority.
  • The Mandal–Kamandal sequence is the illustration: backward-class assertion and the temple movement rose together, and were not unrelated.
• Ashutosh Varshney argues in Battles Half Won that caste mobilisation has produced no concrete transformation in the distribution of power in society.
  • His distinction is between elections and democracy. as far as elections go Indian democracy is a great success; but democracy is not only elections, and India has done badly between elections — in the ordinary working of police, courts, schools and local administration.
  • Descriptive representation has advanced far ahead of substantive outcomes. A Dalit chief minister does not by itself alter land relations, wages, or the safety of Dalits in a village.
  • The battle must therefore now be for the deepening of democracy rather than the extension of a franchise that is already complete.
• André Béteille objects on both empirical and normative grounds.
  • Caste, class and power are separate dimensions, and a politics conducted wholly in caste terms misdescribes an inequality that is increasingly economic, so that its remedies reach the wrong people.
  • He is sceptical of the reification of caste in public policy — treating a caste as an undifferentiated block ignores the class differentiation inside it that his own fieldwork documented.
• Arun Shourie holds that the use of caste as a political identity threatens the integrity of the country, that it resurfaces perpetually as competitive reservation demands, and that it manufactures social discord where none need exist.
• R.K. Debbarma adds a critique from inside identity politics rather than against it.
  • An identity meant to be a “site of resistance”, but built on an uncritical evaluation of the self, becomes self-defeating.
  • The politics of the oppressed begins to mirror the politics of the oppressor, and turns oppressive towards the very people it was meant to emancipate.
• A further criticism is that identity politics tends to “fetishise” historical injustice, producing a fossilisation of belief in which grievance becomes the identity’s permanent content and its dissolution becomes unthinkable.

Question	Strengthening view	Weakening view
Participation	Mobilises the excluded; raises turnout among the poor	Mobilises them as clients of caste leaders, not as citizens
Representation	Produces descriptive representation unattainable otherwise	Descriptive gains without substantive redistribution
Competition	Forces bargaining and coalition among groups	Crowds out programmatic competition on health and jobs
Social peace	Breaks a monopoly peacefully, through the ballot	Invites religious counter-mobilisation and inter-caste conflict
Where benefits land	Across whole subordinated blocs	On dominant sub-castes and group leaders
Long run	Caste secularises and dilutes as it politicises	Caste is hardened by official recognition and electoral value

Why Caste Operates in Politics at All

• Rajni Kothari — because the society is traditional, and caste- and religion-based mobilisation is therefore natural rather than aberrant. Politics has no other material to work with.
• M.N. Srinivas — because caste is present in the minds of Indians at a subconscious level, operating long after its formal justifications have been abandoned and even among those who deny holding it.
• André Béteille — a two-part answer.
  • The Constitution abolished untouchability without abolishing caste, and simultaneously made caste a basis of public policy.
  • He therefore put a question mark against Nehru’s expectation: how was a caste-free and communalism-free India to be achieved when the Constitution itself entrenched caste as a category of entitlement? The premise of the national leadership’s optimism was contradicted by its own constitutional design.
• Kanchan Chandra — because India is a patronage democracy, in which the state controls a large share of jobs and services and elected officials have discretion over their distribution.
  • Voters lacking reliable information about candidates use caste as a low-cost signal of who will favour them in office; politicians in turn count heads, because ethnic categories are the only ones whose size they can estimate reliably.
  • When political patronage runs on caste, caste-free politics is not on offer. The behaviour is rational on both sides, which is why exhortation does not touch it.

Caste and Electoral Behaviour

• The aphorism that Indians do not cast their vote but vote their caste has organised popular commentary for six decades. It is too neat, but it survives because the underlying correlation is real and unusually stable.
• Voting is the occasion on which a caste converts numbers into influence, and the study of that conversion is among the oldest strands of Indian political science.
  • It was pursued by Kothari and taken up in the comparative voting-behaviour tradition running through Campbell, Key, Miller and, in India, Norman Palmer.
  • The Rudolphs’ verdict is that the caste association gave caste a new vitality with which to operate inside a democracy.

The Vote Bank and Its Decline

• The term vote bank was coined by Srinivas for the bloc of votes a patron could deliver from his dependents — originally the clients of a dominant-caste landholder, later any group whose vote a leader could commit.
  • The arrangement was vertical and transactional: patronage and protection flowed down, votes flowed up.
  • It has decayed steadily, for reasons that are all consequences of the process described above — land reform and the decline of agrarian dependence, the rise of autonomous lower-caste leadership, the spread of literacy and media, and parties appealing to castes over the heads of their notables.
• What replaced it is not caste-blind voting but negotiated caste voting: a caste bargains rather than obeys, and it can and does split.

Caste in Candidate Selection

• Caste operates on the supply side of elections at least as strongly as on the demand side, which is why the charge of caste voting is partly a description of what parties do to voters.
  • Ticket distribution is calibrated to the dominant and pivotal castes of each seat, and the exercise is conducted openly.
  • Ministries and party office are distributed to balance caste blocs, so that the composition of a cabinet is read as a social settlement.
  • Where every serious party has selected its candidate by caste, the voter’s caste-consistent choice may be the only one on offer — which makes “voting one’s caste” an incomplete description of what is happening.

The Empirical Debate: Identity or Development?

• Abhijit Banerjee, with Amory Gethin and Thomas Piketty, produced the strongest evidence for the identity view. Examining the changing structure of party electorates across five decades, they found that caste and religion — not income, education or occupation — structure Indian voting.
  • Upper castes vote disproportionately for the right; Scheduled Castes, Scheduled Tribes and Muslims for centrist and left parties, and these alignments have been remarkably persistent.
  • Class cleavages have effectively disappeared once caste is controlled for — the opposite of the Western pattern, where education and income divide the electorate ever more sharply.
  • Their conclusion is that Indian political conflict has moved towards identity and religious-ethnic conflict rather than class-based redistribution, and has done so without any corresponding shift in social spending.
• Milan Vaishnav’s work supplies the countervailing evidence. The conventional wisdom was that good economics does not make good politics in India and that parochialism, populism and patronage decide elections; his research finds voters challenging that view and rewarding governments for performance.
• Arvind Panagariya likewise argues that economic performance matters, which is why parties have invested so heavily in a development idiom and in the measurable delivery of houses, gas connections, toilets, roads and cash transfers.
  • An economically-defined beneficiary category cuts across caste blocs, and the economically weaker sections quota — a non-caste criterion — was the constitutional expression of the same logic.
• Shreyas Sardesai’s analysis of the 2019 election supplies the third possibility — that caste is being displaced not by development but by religion.
  • The National Election Study evidence showed an unprecedented consolidation of Hindu voters behind the governing alliance, cutting across caste divisions.
  • His formulation is that the party overcame the caste hierarchies among Hindus and systematically constructed a Hindu category of voters, with religious minorities voting the other way.
  • He also found insufficient evidence that this support rested primarily on hostility to minorities — a caution against reading consolidation as simple communal antagonism.
• Jaffrelot, despite his own emphasis on caste, judges that its weight has eroded in recent elections — a striking concession from the scholar of the silent revolution.
• Yogendra Yadav’s reading, drawn from the Lokniti-CSDS National Election Studies, is that Indian voters are looking for identity and dignity as much as for goods, and that recognition and development are not alternatives to one another.

What the Recent Elections Show

• The 2024 general election is the best test case, and it cuts both ways.
  • Lokniti-CSDS post-poll evidence shows the governing alliance losing roughly five percentage points of the Dalit vote against 2019 while holding and extending its position among Adivasi voters, and the opposition alliance taking about two-thirds of the Muslim vote.
  • Fear that constitutional reservation guarantees were at risk was the mobilising theme where the swing was largest — evidence that caste-linked entitlement, not economics alone, moved votes.
  • Sub-caste behaviour diverged sharply from bloc behaviour: fewer than half of Jatavs in Uttar Pradesh voted for the party built on their identity, showing that a caste bloc is not an automatic unit even when its leader is of the caste.
• The intermediate and non-dominant backward castes — Kurmis, Kunbis, Patidars and their equivalents — behaved as the swing groups, which fits Deshpande’s argument about where the pivot of Indian politics now sits.
• The Bihar assembly election of November 2025 pointed the other way.
  • The governing alliance won 202 of 243 seats on a coalition assembled from Extremely Backward Classes, Dalits and Mahadalits.
  • A large cash transfer to women voters shortly before the poll, and a record turnout of about 67 per cent, showed caste arithmetic and welfare delivery working together rather than in opposition.
  • The Muslim–Yadav coalition proved insufficient on its own to carry the opposition alliance.
• The state elections of 2026 confirmed that caste is one variable among several rather than the master variable.
  • A new entrant displaced the established Dravidian duopoly in Tamil Nadu on an appeal built around youth and women rather than caste blocs.

Caste has not stopped predicting the vote; it has stopped being the only thing that predicts it, and that difference is what the caste-versus-development argument is really about.

Caste and Class

• Caste worked for a long time as a serviceable proxy for class, and the correspondence is what made caste-based redistribution defensible in the first place.
  • The ritually excluded were also the landless, the illiterate and the poorest, and the ritually superior held land, literacy and office.
  • This is the premise of constitutional policy: the Scheduled Castes and backward classes were listed because caste position and material position coincided closely enough to substitute for a means test.
• The two have since diverged, and that divergence is the central fact of contemporary caste politics.
  • A substantial middle class has emerged within the Scheduled Castes and the backward classes, largely through reservation, urban employment and education, and its interests are not those of a landless labourer of the same caste.
  • Intra-caste inequality now rivals inter-caste inequality, and a rising middle class combined with widening inequality inside caste groups is opening a class-based faultline in a politics organised on caste for decades.
  • The creamy-layer doctrine and the demand for sub-classification within the Scheduled Castes are the policy expressions of the same divergence — the recognition that a caste is no longer a single economic unit.
• Béteille’s Sripuram findings are the empirical anchor: caste, class and power were once cumulative and are now dispersed, so the analysis of inequality must be three-dimensional.

	Where caste and class coincide	Where they diverge
Typical location	Rural, agrarian, land-based society	Urban, salaried, educated India
Politics produced	Caste bloc voting and identity mobilisation	Cross-caste class interests — beneficiaries, aspirants
Policy that follows	Group quotas by caste list	Creamy layer, sub-classification, EWS
Chief risk	Deprivation outside listed groups goes unaddressed	Benefits captured by the already-advanced within a group

• The Marxist reading treats caste as superstructure — a form of consciousness generated by the relations of production, with caste oppression understood as the ideological clothing of class exploitation in a semi-feudal agrarian order.
  • Its strength is that it explains why caste and class coincided so closely in the village economy, and why land reform loosened caste hierarchy where it was carried through.
  • Its weakness is that it cannot explain caste discrimination against the prosperous — the Dalit officer refused a house, the Dalit groom pulled off a horse — which has no economic function at all.
• Ambedkar rejected a purely class analysis, and the rejection is foundational.
  • The Indian working class is itself divided by caste, so the solidarity a class analysis presupposes does not exist and cannot be assumed into being.
  • Graded inequality gives almost every stratum an interest in the system, which means economic reform alone will reproduce caste inside whatever new order it creates.
  • His conclusion reverses the Marxist sequence: social reform must precede economic reform, not follow it.
• The caste–class–power triad is therefore the working framework. Caste supplies status, class supplies resources, politics supplies power; much of Indian politics is the story of groups converting a surplus on one axis into a gain on another — numbers into office, office into land, land into respectability.

Caste Beyond Hinduism

• Caste is not confined to Hinduism. It travelled with converts into Islam, Christianity and Sikhism, because conversion changed the object of worship without changing occupation, endogamy, or the attitude of neighbours.
• Indian Muslim society is conventionally described in three tiers, and the distinctions are social rather than doctrinal.
  • Ashraf — those claiming foreign or elite descent: Sayyid, Sheikh, Mughal, Pathan.
  • Ajlaf — the converted artisanal and service groups, the numerical majority; the Pasmanda movement mobilises around precisely this stratum.
  • Arzal — groups descended from untouchable castes, carrying comparable stigma inside the community.
• Christian and Sikh communities show the same pattern. Converts from Dalit castes remain socially marked, with separate seating, burial grounds and congregations recorded in parts of the country, and Mazhabi Sikhs occupy an analogous position despite Sikhism’s explicit rejection of caste.
• The constitutional question is whether Dalit converts to Christianity and Islam should receive Scheduled Caste status.
  • The Constitution (Scheduled Castes) Order, 1950 restricts SC status to Hindus, extended by amendment to Sikhs in 1956 and Buddhists in 1990, leaving Christians and Muslims outside.
  • The argument for inclusion is that the disability survives the conversion, so a religion test for a social disability is arbitrary; the argument against is that the professed religions do not sanction caste, and that inclusion would enlarge the pool at the expense of existing beneficiaries.
  • The Union government constituted a commission of inquiry under former Chief Justice K.G. Balakrishnan in October 2022 to examine the question. Its term was extended more than once, its report has not been made public, and the litigation before the Supreme Court remains pending.

Caste in the 2020s

• Caste enumeration returns to the census.
  • The Union Cabinet decided in April 2025 to include caste in the forthcoming census — the first full caste count since 1931.
  • Houselisting runs through 2026 and population enumeration in early 2027; a plea to stop the exercise was dismissed by the Supreme Court on 20 May 2026.
  • The machinery and the reservation questions belong to the wider debate on backwardness. What matters here is that the state is about to give caste an official statistical existence it has not had for nearly a century.
• The Supreme Court has begun dismantling caste inside state institutions themselves.
  • In Sukanya Shantha v. Union of India, decided 3 October 2024, it struck down provisions of state prison manuals that segregated barracks by caste, allotted manual labour by caste, and treated members of Denotified Tribes as habitual offenders, holding them contrary to Articles 14, 15, 17, 21 and 23.
  • The directions were unusually specific.
    • States were to revise their manuals within three months, and the Model Prison Manual 2016 and the Model Prisons and Correctional Services Act 2023 were to be amended.
    • Caste references were to be deleted from prison registers, and the Court took suo motu cognizance to supervise compliance — which it has since had to press, states having been slow.
• Manual scavenging remains the sharpest surviving expression of caste as assigned pollution.
  • It is prohibited by the Prohibition of Employment as Manual Scavengers and their Rehabilitation Act, 2013.
  • In Dr Balram Singh v. Union of India the Court issued directions for complete eradication, raised the compensation payable for sewer deaths, and has continued to monitor implementation.
  • Deaths in sewers and septic tanks continue to be reported, and virtually all of the dead are from Scheduled Caste communities.
• Caste has moved online, and the move cuts both ways. Social media has given anti-caste politics a publishing and organising infrastructure independent of mainstream media, and has simultaneously become a vehicle for caste-pride assertion, matrimonial segregation by jati, and targeted abuse.
• Caste has followed the diaspora, and Western jurisdictions have begun to legislate.
  • Seattle in February 2023 became the first jurisdiction in the United States to add caste as a protected class in its anti-discrimination ordinance.
  • California’s legislature passed a comparable bill, SB 403, in 2023; the Governor vetoed it, reasoning that existing categories already covered such discrimination.
  • California’s civil rights agency sued Cisco Systems in 2020 over alleged caste discrimination against a Dalit engineer; the case against the two individual managers was dropped in 2023 while proceedings against the company continued. The episode showed that caste travels into workplaces whose law has no vocabulary for it.
• Inter-caste marriage remains the truest measure of how little the structure has loosened.
  • Survey evidence puts inter-caste marriages at roughly five per cent of all Indian marriages, a figure that has moved very little.
  • Honour killings of inter-caste couples continue, and in Shakti Vahini the Supreme Court directed protective measures against khap panchayat interference in the choice of adult partners.

Conclusion

• Caste is neither dissolving nor standing still; it is reorganising. Its ritual content has thinned — commensality rules have weakened, occupational assignment has largely broken in the cities, untouchability is a crime — while its organisational content has thickened, because democracy rewards the group that can count itself.
• Kothari’s proposition still explains most of what happens: politics has secularised caste more than caste has communalised politics. But the sequence he described has a second half, and Varshney’s and Béteille’s objections belong to it — mobilisation that expands access is democratising, and mobilisation that protects a share is not.
• The test to apply to any caste mobilisation is therefore not whether it invokes caste but what it does with it. The caste census, the demand for sub-classification, and the assertion of the Pasmanda and the Mahadalit are all attempts to put that question to caste politics from within.