On 7 August 1990, Prime Minister Vishwanath Pratap Singh rose in Parliament and announced that the Union government would implement the central recommendation of the Mandal Commission, reserving 27 percent of vacancies in civil posts and services under the Government of India for the socially and educationally backward classes. The 1990 Mandal protests that followed over the next hundred days are usually told in one of two ways. The first telling is an arithmetic telling, in which the announcement was a move in a coalition game, calculated by a minority government to consolidate the backward caste electorate of the Hindi belt behind the Janata Dal. The second telling is a narrative of the streets, in which the autumn of 1990 belongs to the students who marched against the decision and to Rajiv Goswami, the Delhi University student whose self-immolation on 19 September 1990 became the enduring image of the agitation. This article defends a different claim. The implementation produced three distinct mobilisations in sequence and in interaction: a pro-Mandal consolidation during August and early September, an anti-Mandal counter-mobilisation during September and October, and a Hindutva mobilisation from late September to November that the Ram Rath Yatra carried across north India. The political architecture of India was rearranged by the three phases acting together, and neither the arithmetic telling nor the protest telling, taken alone, can account for the rearrangement.

The 1990 Mandal Protests and the Announcement of 7 August 1990

The announcement itself was brief, and its brevity concealed a decade of deferral. The Union Cabinet took the decision on 6 August 1990. On 7 August, V. P. Singh informed both Houses of Parliament that the government had decided to accept the recommendation of the Second Backward Classes Commission, chaired by Bindeshwari Prasad Mandal, for a 27 percent reservation for the other backward classes in central government employment. On 13 August 1990, the Department of Personnel and Training issued Office Memorandum No. 36012/31/90-Estt. (SCT), the legal instrument that carried the decision into the recruitment rules. The memorandum reserved 27 percent of vacancies in civil posts and services under the Government of India, and in public sector undertakings, for the socially and educationally backward classes, and it declared that the reservation would take effect from 7 August 1990, the date of the parliamentary statement. The figure of 27 percent was not chosen in August 1990. It had been fixed in December 1980, when the Commission calculated that the backward classes formed roughly 52 percent of the population but that the total of all reservations had to remain below the 50 percent ceiling the Supreme Court had indicated in M. R. Balaji v State of Mysore in 1963, leaving 27 percent as the residue beneath the existing 22.5 percent reserved for the Scheduled Castes and the Scheduled Tribes. What the 1980 methodological-substantive report had derived on paper, the memorandum of 13 August 1990 converted into an operative rule of recruitment.

The parliamentary setting of the announcement matters for everything that followed. V. P. Singh governed at the head of the National Front, a coalition whose Janata Dal held 143 seats in a Lok Sabha of 543 and which survived on the outside backing of two mutually hostile forces: the Bharatiya Janata Party, with 85 seats, and the Left Front, led by the Communist Party of India (Marxist). A government resting on such supports had a narrow margin for error on any question, and the Mandal question touched the deepest division in Indian society, the division of caste. Yet the statement of 7 August was received in the House without uproar. Members across parties understood the electoral weight of the backward classes, who by the Commission’s estimate formed a majority of the population, and open opposition to reservation in Parliament carried a price that few parties were willing to pay in the first week. The contest over Mandal was therefore displaced from the floor of the House to the street, the campus, the courtroom and, within eight weeks, to a chariot procession moving from Somnath towards Ayodhya. The three-phase reading begins from that displacement. Because the parliamentary arena absorbed the announcement quietly, the political energy of the question gathered outside Parliament, and it gathered in three successive waves, each with its own actors, its own geography and its own consequences.

The first wave broke in favour of the decision. During August and the first half of September 1990, the political organisations of the backward classes, the state governments led by parties of the Janata Dal family, the Left parties with qualifications, and the Dravidian parties of Tamil Nadu consolidated behind the implementation, and the Janata Dal presented the decision as the fulfilment of a promise carried in the National Front manifesto of 1989. The second wave broke against it. From the middle of September, students of the forward castes in Delhi and across the university towns of north India built a counter-mobilisation that boycotted classes, blocked roads and trains, courted arrest, and, in its most desperate register, produced a series of self-immolation attempts whose toll remains disputed between the counts preserved in contemporary reporting and the higher figures given by later historians. The third wave ran on a parallel track that intersected the second. On 25 September 1990, Lal Krishna Advani, the president of the Bharatiya Janata Party, set out from Somnath in Gujarat on a Ram Rath Yatra towards Ayodhya, and the procession converted the religious mobilisation around the Ram Janmabhoomi into the vehicle that carried the autumn’s accumulated discontent in a different direction. When the Bihar government of Lalu Prasad Yadav arrested Advani at Samastipur on 23 October 1990, the Bharatiya Janata Party withdrew its backing from the National Front government, and on 7 November 1990 the Lok Sabha voted the government out of office by 346 votes to 142. Any account of the autumn that separates these waves into independent stories misses the mechanism. The consolidation of the first phase defined the target of the second; the anxieties assembled by the second phase supplied part of the audience of the third; and the third phase, by destroying the government that had issued the memorandum, determined the institutional conditions under which the reservation survived.

A Decade of Deferral Before the Announcement

The implementation of August 1990 cannot be understood without the ten years during which the Mandal Report waited on the shelves of the Home Ministry. The Commission submitted its report to President Neelam Sanjiva Reddy on 31 December 1980. The Congress government of Indira Gandhi, returned to power in January 1980, neither accepted nor rejected the recommendations. The report was laid before Parliament and discussed in 1982 and again in 1983. During the 1983 discussion, the Home Minister, P. C. Sethi, told the House that although the Commission had been appointed by the predecessor government, the Congress government desired to continue with the Commission and to implement its recommendations. No office memorandum followed. The statement of 1983 is the documentary measure of the deferral: a government that declared its intention to implement a report and then left the report unimplemented for seven further years. The reasons for the deferral were themselves a form of political arithmetic. The Congress of the 1980s rested on a coalition in which forward caste voters of the Hindi belt remained a load-bearing element, and the party’s state units in Uttar Pradesh, Bihar and Madhya Pradesh were led substantially by men drawn from those castes. To implement Mandal was to invite the anger that finally arrived in September 1990; to reject Mandal was to surrender the backward caste electorate to the parties of the Janata tradition. Deferral was the Congress answer to an impossible equation, and it held for a decade.

The deferral also shaped the instrument that ended it. Because two Congress governments had declined to act, the National Front could present implementation in 1990 as the redemption of a pledge rather than an innovation. The National Front manifesto for the 1989 general election had promised to implement the Mandal Commission recommendations, and V. P. Singh returned to that promise in his parliamentary statements of August and September 1990 whenever critics charged him with improvisation. The promise had a genealogy. The Janata Party government of Morarji Desai had appointed the Commission on 1 January 1979 under Article 340 of the Constitution, and the Janata Dal of 1989 understood itself as the heir of the Janata experiment of 1977 to 1979. Behind the party genealogy stood an older current: the politics of backward caste assertion that ran from the state-level reservation regimes of the south, through the Bihar government of Karpoori Thakur, who as chief minister had applied a backward classes reservation formula in Bihar state employment in 1978, to the socialist tradition of Ram Manohar Lohia, whose slogan of the 1960s demanded sixty percent of leadership positions for the backward classes, the Scheduled Castes, the Scheduled Tribes, minorities and women. The 1990 decision was therefore not the sudden discovery of a forgotten file. It was the arrival at the centre of a demand that had already reorganised the politics of Tamil Nadu, Karnataka, Bihar and Uttar Pradesh at the state level, and its arrival at the centre is what gave the autumn its national scale. The first Backward Classes Commission under Kaka Kalelkar had reported in 1955 and been set aside; the second, under Mandal, reported in 1980 and was deferred; the events of 1990 were the third act of a forty-year sequence, and the actors of 1990 on every side knew the earlier acts.

A further fact of the deferral decade conditioned the explosion. During the 1980s the Indian state remained the largest single source of the secure, salaried, pensionable employment that educated families sought for their children. Private sector expansion on the scale that followed the 1991 reforms had not begun. A central government post carried a salary, housing, medical care and a pension, and for the families of the forward castes in the small towns of the Gangetic plain, whose landholdings had fragmented across generations, the government examination was the principal ladder out of agricultural decline. The Mandal memorandum touched that ladder directly. It reallocated more than a quarter of the vacancies at the bottom rung of the central services at the exact moment when the queue for those vacancies was longest and the alternatives were fewest. The student who marched in September 1990 marched, in his own stated account, for his family’s one realistic prospect of a salaried life. The backward caste villager who rallied for the memorandum in August 1990 rallied, in his own stated account, for the first opening of a door that his family had found closed for generations. The three-phase reading holds both accounts in view, because the phases cannot be explained without the material stake that each side believed it was defending.

The State Reservations That Preceded the Centre

The centre arrived late to reservation for the backward classes, and the lateness explains the geography of the explosion. The princely state of Mysore had reserved posts for backward communities under the Miller Committee’s recommendations as early as 1919, and the Madras Presidency had distributed government posts and college seats among communities under the Communal Government Orders issued from 1921 onward by the Justice Party governments. After Independence, the southern states rebuilt these regimes inside the constitutional frame. Tamil Nadu, after the First Amendment of 1951 restored the power of special provision that the Supreme Court’s judgment in State of Madras v Champakam Dorairajan had denied, settled into a regime that M. G. Ramachandran’s government raised in 1980 to 50 percent for the backward classes alone, carrying the state total to 68 percent; the Karunanidhi government of 1989 to 1991 then divided the backward classes quota to carve out a most backward classes share and raised the total to 69 percent. Karnataka, acting on the Havanur Commission report of 1975, operated a backward classes regime from 1977 that classified castes by occupation and educational attainment. In the north, Bihar preceded the centre by a dozen years: the government of Karpoori Thakur, chief minister in the Janata Party years, issued in November 1978 a formula reserving 26 percent of Bihar state posts across the backward classes, the most backward classes, women and the poor of the forward castes, the first application of a Mandal-style scheme by any Hindi belt government.

These state regimes had two consequences for the autumn of 1990. The first was habituation. In Tamil Nadu and Karnataka, two generations of students had grown up inside reservation regimes whose categories were part of the furniture of public life; the central memorandum of August 1990 altered the prospects of a Tamil or Kannada aspirant at the margin, in central services alone, and the counter-mobilisation in those states was correspondingly slight. The second consequence was contrast. In Uttar Pradesh, Bihar, Madhya Pradesh, Rajasthan and Haryana, state provision for the backward classes in the 1980s was narrower, later and more contested, and the central decision of 27 percent arrived in societies where the forward castes had retained a hold on state employment far beyond their share of population. The Mandal Commission had measured that hold: its report recorded that the other backward classes, estimated at 52 percent of the population, held a small fraction of Class I central posts, a share the Commission’s survey placed in single digits. The students who marched in Allahabad and Delhi in September 1990 therefore defended a distribution of advantage that the students of Madras did not possess in the same measure. The three-phase reading takes this uneven inheritance seriously. Phase One’s consolidation was strongest where the state regimes had prepared a beneficiary coalition with organisational memory; Phase Two’s fury was strongest where no such preparation had softened the centre’s arrival. The map of the protests is the map of the prior settlements, read in reverse.

The state precedents also armed both sides with arguments from experience. Defenders of the implementation pointed to Tamil Nadu, where decades of extensive reservation had coincided with the political integration of the backward classes and, by the measures of literacy and public health that the state’s governments cited, with no collapse of administration. Critics answered from the same state, pointing to the ratchet by which Tamil quotas had climbed from year to year under electoral pressure, and asked what mechanism would hold the central figure at 27 percent when every election rewarded the party that promised more. The argument from the south thus cut both ways, and it was conducted in Parliament and in the press through the autumn with a precision that the purely metropolitan accounts of the protests rarely preserve. What no participant disputed was the premise the state regimes had established: reservation, once enacted and administered for a generation, became a settled feature of the political order that no subsequent government could withdraw. The students of Delhi understood the premise as clearly as the ministers did. It is why the agitation of September had the desperation of a last opportunity. Its participants believed, correctly as events proved, that if the memorandum survived its first year it would survive indefinitely.

Phase One: The Pro-Mandal Consolidation of August and Early September 1990

The first phase of the mobilisation ran from the announcement of 7 August 1990 to roughly the middle of September, and its direction was consolidation. The Janata Dal moved first to frame the decision. Party leaders presented the memorandum as social justice delivered after a decade of Congress evasion, and V. P. Singh, in statements during August 1990, repeatedly grounded the decision in the National Front manifesto and in the constitutional promise of equality of opportunity under Article 16. State governments led by the Janata Dal and its allies carried the consolidation into the provinces. In Bihar, the government of Lalu Prasad Yadav, in office since March 1990, defended the implementation as the entitlement of the majority of the state’s population. In Uttar Pradesh, the government of Mulayam Singh Yadav, a Janata Dal chief minister whose own political base rested on the Yadav cultivating castes, organised the defence of the decision in the state where the counter-mobilisation would soon be fiercest. Across Madhya Pradesh and Rajasthan, associations of the backward classes held meetings and processions in favour of the memorandum during August 1990, and the demand travelled downward into the district towns where caste associations maintained their offices, their hostels and their scholarship funds.

What did the 13 August 1990 office memorandum actually reserve?

The memorandum reserved 27 percent of vacancies in civil posts and services under the Government of India, and in public sector undertakings, for candidates of the socially and educationally backward classes named in the Mandal lists. It took effect from 7 August 1990 and governed direct recruitment to central employment, not admissions to educational institutions.

The consolidation extended beyond the Janata Dal family. The Communist Party of India (Marxist) and the Communist Party of India backed the implementation with stated qualifications, holding that reservation for the backward classes was a necessary instrument against caste hierarchy while pressing for an economic criterion to be applied within it against the better-off strata of the beneficiary castes. The qualification mattered: it marked the distance between the Left, which treated Mandal as one element of a wider programme, and the Janata Dal, which treated it as the centrepiece of a new electoral coalition. In Tamil Nadu, the Dravida Munnetra Kazhagam government of M. Karunanidhi welcomed the central decision from the standpoint of a state that had operated backward classes reservation for decades and had raised its own quota to 69 percent under a state law of 1990 covering the backward classes, the Scheduled Castes and the Scheduled Tribes together. For the Dravidian parties, the Delhi announcement was the centre catching up with a settlement the south had reached a generation earlier, and their backing gave the implementation a federal breadth that the arithmetic telling, fixed on the Hindi belt, tends to miss. Parties that would not back the decision mostly held their fire in August. The Congress, which had deferred the report for ten years, could hardly denounce its implementation, and its spokesmen confined themselves to procedural criticism and to warnings about social peace. The Bharatiya Janata Party, on whose parliamentary backing the government depended, criticised the manner and the timing of the announcement while stopping short of opposing reservation for the backward classes in principle, a hesitation that Phase Three would resolve in an unexpected direction.

The pro-Mandal consolidation was organisational as much as rhetorical. The backward classes of the Hindi belt had never lacked caste associations, but August 1990 gave those associations a single national object. Meetings convened under the banners of Yadav, Kurmi, Koeri, Lodh and other caste sabhas during August and September 1990 passed resolutions in favour of the memorandum and against any dilution of the 27 percent. The All India Yadav Mahasabha and kindred federations, whose ordinary business ran to hostels, marriages and scholarships, became conduits of a political message that reached villages the Janata Dal’s own organisation could not reach. Christophe Jaffrelot, in India’s Silent Revolution of 2003, described the longer process by which the lower castes of north India converted demographic weight into political power, and he treated the Mandal moment as the hinge of that conversion. The consolidation of August 1990 was the hinge beginning to move. Its participants understood the memorandum as the state recognising, at last, a claim their organisations had pressed since the days of the Kalelkar Commission. That understanding explains the speed and the durability of the first phase. A benefit announced by a fragile government might have been discounted as temporary; a recognition framed as historic justice was defended as permanent, and the defence outlasted the government that had announced it.

The Government Behind the Announcement: Devi Lal, the National Front and the Arithmetic Charge

No part of the autumn has attracted more argument than the motive of the announcement, and the argument must be stated precisely before it can be adjudicated. On the night of 1 August 1990, six days before the parliamentary statement, V. P. Singh dismissed his deputy prime minister, Devi Lal, the Jat leader of Haryana whose hold over the farming castes of the northwest made him the most powerful rival inside the Janata Dal. Devi Lal had announced a farmers’ rally in Delhi for 9 August 1990, and the rally was widely expected to become a demonstration of his strength against the prime minister. On 7 August, between the dismissal and the rally, the Mandal announcement was made. Critics of V. P. Singh, in the press and in the opposition parties, read the sequence as a diversion and a counterstroke: the prime minister, they argued, had reached for the Mandal file to divide his opponents, to answer Devi Lal’s agrarian mobilisation with a caste mobilisation of his own, and to bind the backward caste parties to his government ahead of the showdown inside his party. The charge has been repeated so often that it has hardened into the standard account, and it is the core of what this article calls the political-arithmetic framing. The V.P. Singh government’s institutional-political context, a government assembled from the anti-Congress verdict of 1989 and resting on incompatible props, made such readings plausible, because almost every act of that government had to be a survival act before it could be anything else.

The arithmetic framing contains a documented core and an undocumented leap. The documented core is the sequence itself: the dismissal of Devi Lal on 1 August, the Cabinet decision of 6 August, the announcement of 7 August and the farmers’ rally of 9 August, which in the event passed without displacing the prime minister. The timing was available to be read as tactical, and V. P. Singh’s colleagues read it in opposed ways depending on their loyalties. The undocumented leap is the inference that because the timing served the prime minister, the decision had no other ground. Against that inference stand three pieces of evidence. The first is the manifesto: the National Front had promised implementation before the Devi Lal crisis existed, in the election campaign of November 1989, and a promise kept at an opportune moment is not thereby reduced to the moment. The second is the record of V. P. Singh’s own statements across the autumn, in Parliament and outside it, in which he defended the decision on the ground of social justice and accepted, in his final speech of 7 November 1990, that he had saved the Babri Masjid and upheld the principle of reservation at the cost of his office; the defence was consistent whether the audience was friendly or hostile. The third piece of evidence is the behaviour of the decision’s supposed beneficiaries. A purely tactical announcement, designed for one week’s headlines, would have required no defence once the government fell, yet the parties of the backward classes defended the memorandum through the courts until November 1992 and legislated its continuation under a Congress prime minister. The arithmetic framing explains the week of the announcement. It does not explain the following twenty-seven months, and an explanation of the autumn that cannot reach beyond its first week is a partial explanation wearing the dress of a complete one.

The counter-framing, offered by the decision’s defenders and developed by Yogendra Yadav in his writings on the post-1989 democratic upsurge, holds that the implementation belonged to a long democratisation of Indian politics in which the lower orders, having acquired the vote in 1950, acquired in successive waves the party, the ministry and finally the administration itself. On this reading the date of August 1990 matters less than the direction of the half-century. Both framings agree on one thing, and the agreement is the floor on which the three-phase reading stands: the announcement was the most consequential single act of the National Front government, and its consequences ran far beyond the survival calculations of the men who made it. What the framings dispute is the character of those consequences. The three-phase reading answers by following the consequences in their order of appearance.

Phase Two: The Student Counter-Mobilisation of September and October 1990

The second phase began in the universities. Protests against the memorandum had appeared in parts of Uttar Pradesh, Bihar and Haryana within days of the announcement, but the organised counter-mobilisation was a September phenomenon, and its capital was Delhi. Students of Delhi University, led initially from colleges such as St. Stephen’s and from the Delhi School of Economics, built the agitation through bodies that stood outside the established party student wings, among them the Anti-Mandal Commission Forum, under whose banner some two thousand students marched on the Doordarshan Kendra in Delhi during the early agitation. An Equality Front, drawing in university teachers and employees alongside students, was constituted on 25 August 1990 and called for a Delhi bandh on 27 August. Hundreds of teachers and researchers of Delhi University signed a memorandum to the President and the Prime Minister on 28 August 1990 asking that the implementation be stayed until a national consensus had been reached. On 4 September 1990, students and farmers filled the Boat Club lawns in a joint demonstration of at least ten thousand people, at which supporters of Mahendra Singh Tikait’s Bharatiya Kisan Union stood beside the university students, a conjunction that signalled how far the anxiety had travelled beyond the campus. The demands were consistent across the phase: that reservation be based on economic criteria rather than caste, that the Mandal lists drawn from the census of 1931 be discarded as obsolete, and that merit in public examinations be protected against what the Forum’s literature called a farce in the name of social justice.

When did the anti-Mandal agitation turn from protest to self-immolation?

The turn came on 19 September 1990, when Rajiv Goswami, a nineteen-year-old commerce student of Deshbandhu College, set fire to himself at a student demonstration in Delhi. Goswami survived with severe burns. The photograph of the act carried the agitation across India within days, and further attempts followed in several states.

What the students did after 19 September must be recorded factually, because the grief of the phase is its substance and spectacle would betray it. Goswami’s attempt was followed by a wave of similar attempts across the Gangetic valley and beyond: in Hisar, Sirsa and Ambala in Haryana, in Meerut and Lucknow in Uttar Pradesh, and in towns as distant as Indore, Jabalpur, Hyderabad, Pathankot and Chandigarh. The counts are disputed, and the dispute should be preserved rather than resolved by false precision. Ramachandra Guha, in India After Gandhi of 2007, put the number of attempts at about two hundred, with more than sixty deaths. Other compilations from contemporary reporting give 159 attempts and 63 deaths. Surinder Singh Chauhan, a student of Delhi, is recorded as the first of the protesters to die of his burns. The dead included teenagers of fifteen and sixteen. Police firing on protesting crowds in several towns added further deaths to the toll of the phase, and the state governments of the affected provinces alternated between negotiation, arrest and force. The protesters’ own framing of these acts was sacrifice in defence of merit and of their families’ futures. The framing offered by the implementation’s defenders was that a privileged minority, accustomed to monopolising state employment, had turned its despair into a weapon against the majority’s entitlement. The three-phase reading does not arbitrate between those framings of motive. It records what the acts did politically: they raised the cost of the implementation for every institution that touched it, they held the front pages of the national press for six weeks, and they created the atmosphere of a nation at war with itself in which the third phase could present itself as the answer.

The university archives preserve a record of the phase that the newspaper narrative flattens, and the brief of evidence for this article directs attention to it deliberately. At Delhi University, at Jawaharlal Nehru University, at Banaras Hindu University and at Allahabad University, administrations recorded class boycotts, hostel meetings, examination postponements and the polarisation of campuses into pro-Mandal and anti-Mandal camps that cut across the older divisions of left and right. At Jawaharlal Nehru University, three office-bearers of the students’ union resigned their posts in protest against the implementation, an event unthinkable in the union politics of that campus a year earlier. Students of the backward classes, organised in smaller numbers on most campuses, held counter-demonstrations in favour of the memorandum and faced hostility in hostels where they were suddenly identified as the beneficiaries of their roommates’ ruin. Readers who wish to study the surviving documentation of those weeks, the union resolutions, the administrative circulars, the scanned pamphlets of the Forum and the Front, will find such collections easier to open and compare in a ppt viewer that renders the period’s presentation files and scanned decks without proprietary software. The documents matter because they show the protest’s substantive content: not a riot but an argument, conducted with memoranda, signature campaigns and boycott timetables, by students who believed that the rules of advancement had been rewritten against them in a single week. An argument of that kind does not dissolve when the street empties. It settles into the assumptions of a generation, and the name this article gives to that residue, following the brief, is the anti-Mandal substrate: a durable disposition among the forward caste educated families of north India to experience reservation as dispossession, available for mobilisation by any later politics that spoke to it.

The counter-mobilisation also had a geography, and the geography is evidence. The agitation was fierce in Delhi, Haryana, Uttar Pradesh, Rajasthan, Madhya Pradesh and parts of Bihar; it was weak or absent across most of the south and the east. The difference maps almost exactly onto the prior history of reservation. Tamil Nadu had operated extensive backward classes reservation since the Justice Party era and had normalised it across two generations; Karnataka and Andhra Pradesh had their own state regimes of long standing. In those states the central memorandum changed little and threatened less. In the Hindi belt states, where state-level backward classes reservation was recent, partial or contested, the central decision arrived as a rupture, and the students of Allahabad, Varanasi, Lucknow, Patna and Delhi experienced it with an intensity that their counterparts in Madras or Bangalore did not share. The phase therefore cannot be read as a national student movement. It was a regional insurgency of the examination queue, concentrated where the queue was longest and the prior settlement weakest. Its leaders understood this and framed their demand nationally, as a defence of a single national standard of merit; its opponents understood it too, and framed the agitation as the defence of a regional monopoly. Both framings were arguments about the same map.

The Press, the Photograph and the Making of an Image

The second phase was the first great Indian political agitation of the age of the colour photograph and the satellite-fed newsroom, and its media character shaped its political character. The event of 19 September 1990 was recorded by press photographers at the scene, and the image of Rajiv Goswami in flames, running, was carried by the national newspapers the following morning and by the news magazines of that week to every reading household in the country. India Today and Frontline, the two English-language magazines that set the agenda of the metropolitan reading public, devoted successive covers to the agitation, and the Hindi press of the Gangetic plain, whose circulation had expanded enormously across the 1980s, carried the protests into the district towns where the students’ families lived. Doordarshan, the state broadcaster, remained the only television network, and its restrained coverage was itself a grievance of the protesters, who marched on the Doordarshan Kendra in Delhi in the agitation’s early weeks under the Forum’s banner to demand that their case be shown. The asymmetry is worth recording. A pro-Mandal rally in a Bihar district town generated no metropolitan image and entered the national record largely through the reports of the parties that organised it. A student injured in Delhi entered the national record within hours. The metropolitan bias of the archive is one reason the protest-narrative framing of the autumn came so easily to later writers: the second phase is the best photographed, the best reported and the most intimately documented of the three, and documentation exerts a gravity on narrative that the historian must consciously resist.

The photograph of Goswami also did causal work that no pamphlet could have done. Contemporary accounts and later compilations agree on the sequence: the attempts at self-immolation multiplied after 19 September, in the towns whose names recur through the record, Hisar, Sirsa, Ambala, Meerut, Lucknow, Indore, Jabalpur, Hyderabad, Pathankot and Chandigarh, and the victims were very often younger than the students who had led the marches, in several recorded cases schoolchildren of fifteen and sixteen. The historians who have counted the attempts differ, as this article has noted, between the figure of about two hundred attempts with more than sixty deaths that Ramachandra Guha gives and the 159 attempts and 63 deaths of other compilations drawn from contemporary reporting. The discrepancy itself is instructive. No central registry of the attempts was kept; the numbers were assembled from newspaper reports by writers working to different definitions and different closing dates, and the honest statement is the range with its sources named. What is not in dispute is the mechanism of imitation that the first photograph set running, and the anguish it caused within the agitation itself, whose organisers had built a politics of memoranda and boycotts and found themselves answerable for a sacrificial politics they had not chosen and could not stop. The Forum’s literature continued to argue percentages and criteria while the cameras waited for the next fire. In that divergence between the agitation’s argument and its image lies much of the tragedy of the second phase, and the reason its memory remains raw in the families it touched: the argument was about the rules of a competition, and the image was of children burning.

The first phase and the third phase each understood the media lesson of the second in their own way. The defenders of the implementation learned to speak of Mandal in the register of rights rather than of patronage, presenting the memorandum in interviews and in Parliament as the Constitution’s promise redeemed, a framing addressed to the same national audience the photograph had assembled. The Rath Yatra was, among other things, a media construction of genius: a procession moving at a pace the cameras could follow, through towns whose receptions could be filmed each evening, crowned by a destination and a date that gave the story a climax. The Yatra generated its own archive of images, and the contest of the autumn can be read in the competition of those archives, the burning student against the advancing chariot, each condensing a theory of India into a single frame. A political history that confined itself to the parliamentary record would miss this dimension entirely, which is why the brief of evidence for the autumn rightly includes the newspaper archives and the university documentation alongside the debates of the Lok Sabha. The debates record what the actors said for the record. The images record what the country saw, and the country acted on what it saw.

The Argument Over Merit

Running beneath the street politics of the second phase was an argument about the meaning of merit, and the argument deserves reconstruction at the strength with which each side made it, because the substrate that the phase left behind is composed of it. The protesters’ argument, stated in the Forum’s literature and in the memoranda of the university teachers, held that the public examination was the single institution of Indian life in which the son of a clerk competed on equal terms with the son of a minister, and that marks obtained in that competition were the property of the candidate who earned them. Reservation on caste lines, on this argument, committed three wrongs at once: it punished the present generation of forward caste students for advantages accumulated by their ancestors; it distributed a public good by an inherited identity that the Constitution’s promise of equality was meant to dissolve; and it rested on a classification, the Mandal list of 3,743 castes and communities drawn substantially from the census of 1931, that described the India of two generations earlier. The teachers who signed the memorandum of 28 August 1990 added an institutional argument: that the services would be damaged in quality if entry standards diverged sharply by category, and that a state which doubted its own examinations would soon doubt its own administration.

The defenders of the implementation answered each element. To the claim of the level field, they replied with the sociology of the examination itself: the capacity to compete for a central post presupposed English-medium schooling, coaching, books, leisure from labour and family precedent, all of which were distributed by the same caste order that the examination claimed to ignore, so that formal equality of competition ratified a prior inequality of preparation. Marc Galanter, whose Competing Equalities of 1984 remained the standard scholarly treatment of the law of compensatory discrimination in India, had documented how the constitutional scheme from 1950 onward treated preferential treatment not as an exception to equality but as an instrument of it, and the Mandal Commission’s report worked inside that doctrine: caste served the Commission as the practical marker of a class disadvantaged in the aggregate, whatever the variation among its members. To the claim of the outdated list, the defenders answered that the absence of caste enumeration in the census after 1931 was itself a decision of the state, and that a government could not first refuse to count caste and then cite the age of the count against the counted. To the claim of damaged services, they cited Article 335’s own balance, which required the claims of the backward classes to be considered consistently with administrative efficiency, and the record of the southern states, where reserved recruitment had operated for decades without administrative dissolution.

The argument had a biographical edge that sharpened it beyond doctrine. V. P. Singh’s critics observed that the prime minister and most of his cabinet colleagues were themselves products of the open competition they were accused of devaluing, and that the decision transferred opportunity from families like those of the protesters to families unlike those of any minister. The defenders answered in kind: that a Rajput prime minister from an erstwhile princely family of Manda, governing on the votes of the backward classes, was enacting precisely the transfer that the constitutional promise required of those whom the old order had served. The three-phase reading does not award this argument to either side. It notes what the argument did. Conducted at this level of principle, the dispute could not be compromised by adjusting a percentage, because each side experienced concession as the surrender of a theory of the republic. That is a large part of why the autumn escalated from memorandum to self-immolation within six weeks, and why the eventual settlement of November 1992, a judgment of the Supreme Court speaking the language of doctrine rather than the language of the street, could close the legal question while leaving the argument, and the substrate it fed, entirely alive.

The Weeks of the Agitation: A Chronology of September 1990

Because the three-phase reading insists on interaction, the chronology of the central weeks should be set out in one place, drawn from the parliamentary record, the court record and the contemporary press documentation on which this article relies. The Cabinet decision fell on 6 August 1990 and the announcement on 7 August. Through the middle weeks of August the pro-Mandal consolidation organised itself in the states, while in Delhi the first student committees constituted themselves; the Equality Front, joining students with teachers and university employees, was constituted on 25 August and called the Delhi bandh of 27 August. On 28 August, hundreds of Delhi University teachers and researchers addressed their memorandum to the President and the Prime Minister seeking a stay of the implementation until a national consensus existed. On 4 September, the joint demonstration of students and farmers at the Boat Club brought at least ten thousand people onto the lawns by India Gate. On 11 September, the five-judge Constitution Bench of the Supreme Court admitted the writ petitions against the office memorandum and declined, at that stage, to stay it. On 19 September, Rajiv Goswami set fire to himself in Delhi. In the ten days that followed, attempts at self-immolation were reported from Haryana, Uttar Pradesh and beyond, the universities of the north closed or emptied as boycotts took hold, and state governments ordered arrests and, in several towns, police firing. On 25 September, in the middle of that fortnight, L. K. Advani began the Rath Yatra at Somnath. In October the Supreme Court stayed the memorandum and referred the questions to nine judges. On 23 October the Yatra was halted at Samastipur and the Bharatiya Janata Party withdrew its backing. On 30 October and 2 November the Uttar Pradesh police fired on kar sevaks at Ayodhya. On 7 November the Lok Sabha voted 346 to 142 and the government fell.

Read as a list, the chronology conceals what it contains. Read as a sequence of reactions, it shows each actor answering the actor before. The students organised because Parliament had absorbed the announcement without a fight; the teachers petitioned because the students were in the street; the Court admitted the petitions while the memoranda were still polite and stayed the memorandum after the street had turned desperate; the Yatra’s planners kept a date fixed long before, and found their procession received by towns the agitation had heated; the Bihar government arrested Advani because the Yatra, unresisted, would have arrived at Ayodhya with the authority of an accomplished fact; the Bharatiya Janata Party withdrew because the arrest made continued backing a public submission; and the confidence vote assembled every one of these reactions into a single division lobby. The chronology also disciplines two temptations of the literature. Against the arithmetic reading, it shows the announcement preceding the agitation by six weeks of consolidation, a period in which the decision gathered defenders who were not the prime minister’s creatures and would survive him. Against the protest reading, it shows the agitation’s peak, the fortnight after 19 September, coinciding not with the memorandum’s withdrawal but with the opening of the third phase at Somnath, the event that would actually decide the government’s fate. The students believed they were the autumn’s protagonists, and in suffering they were. In causation, the protagonists were plural, and the chronology is the proof.

A word on sources belongs with the chronology. The dates of the parliamentary and court events stand in records that no participant disputes: the Lok Sabha debates of August to November 1990, the orders of the Supreme Court of 11 September and of October 1990, the division figures of 7 November. The dates and scale of the street events rest on the contemporary press and on the university documentation preserved at Delhi University, Jawaharlal Nehru University, Banaras Hindu University and Allahabad University, and where those sources conflict, on the toll of the self-immolations above all, this article has given the conflict rather than a settlement. That practice follows the stance of the series: settled facts stated plainly, disputed points marked as disputed, and no number carried further than its evidence allows.

The Court Steps In: The Supreme Court Stay of October 1990

While the campuses burned in the headlines, a second institutional track opened in the Supreme Court. The constitutional validity of the office memorandum of 13 August 1990 was challenged by writ petition, in the proceedings that became Indra Sawhney v Union of India. A Constitution Bench of five judges admitted the petition on 11 September 1990 and at that stage declined to stay the operation of the memorandum. A month later, in October 1990, with the agitation at its height and the Supreme Court Bar Association among those pressing for intervention, the Court stayed the operation of the memorandum and directed that the questions be heard by a bench of nine judges. The stay froze the implementation without deciding it. Recruitment under the 27 percent quota halted before it had meaningfully begun, and the legal fate of the Mandal decision passed out of the hands of the government that had made it. The stay is a small event in the chronology and a large one in the analysis. It demonstrated that the counter-mobilisation had acquired an institutional instrument: what the students demanded in the street, a pause until the nation had considered the question, the Court supplied in juridical form. It also demonstrated the limit of the street. The final word on Mandal would be spoken neither by the government of V. P. Singh nor by the Forum, but by a nine-judge bench on 16 November 1992, in a judgment that upheld the 27 percent reservation, excluded the creamy layer of the beneficiary castes, and fixed the 50 percent ceiling as constitutional doctrine. Between the stay of October 1990 and the judgment of November 1992, the implementation existed in suspension, and the political system reorganised itself around the suspended question.

Phase Three: The Rath Yatra and the Hindutva Mobilisation of October and November 1990

The third phase opened on 25 September 1990 at Somnath in Gujarat, from where Lal Krishna Advani began the Ram Rath Yatra, a procession in a vehicle refashioned as a mythological chariot, bound for Ayodhya, where the Vishwa Hindu Parishad had announced that volunteers would begin construction of a Ram temple at the disputed site of the Babri Masjid on 30 October 1990. The September-October 1990 Rath Yatra was not created by the Mandal announcement. The Ram Janmabhoomi campaign had its own history running through the opening of the locks of the Babri Masjid in 1986, the foundation-laying ceremony of November 1989 and the Bharatiya Janata Party’s resolution at Palampur in June 1989 committing the party to the temple’s construction. But the Yatra’s timing placed it inside the Mandal autumn, and the intersection was not accidental in its effects, whatever the intention of its planners. The Yatra offered the discontented of the second phase a different banner. Where the student agitation divided Hindu society by caste and said so, the temple campaign summoned Hindu society to unity across caste and said that instead. For a forward caste student who had marched against Mandal in September, the Yatra of October proposed that his grievance be answered not by the abolition of the quota but by the consolidation of a political force in which his caste was one contingent among many. Analysts of the period named the contest Mandal versus Kamandal, the water-pot of the ascetic turned symbol of caste assertion against the water-pot turned symbol of religious mobilisation, and the shorthand survives because it captures the structural rivalry of the autumn: two projects competing to organise the same electorate on incompatible principles.

Why was L. K. Advani arrested at Samastipur on 23 October 1990?

The Bihar government of Lalu Prasad Yadav arrested Advani at Samastipur on 23 October 1990 to halt the Rath Yatra before it entered Uttar Pradesh for the announced Ayodhya mobilisation of 30 October, which both governments held would provoke communal violence. The Bharatiya Janata Party withdrew its backing from the National Front government the same day.

The government in Delhi had watched the Yatra’s approach with an anxiety that the parliamentary record of October 1990 preserves. V. P. Singh had staked his government’s identity on secular governance, and the prime minister’s supporters in the Left parties pressed for the Yatra to be halted; the Bharatiya Janata Party warned that a halt would end its backing. The decision that produced the arrest was taken in Patna. Lalu Prasad Yadav, whose government rested on the backward caste coalition that the Mandal implementation had energised in Bihar, ordered the arrest at Samastipur on 23 October 1990, and Advani was taken into custody and moved to a guest house at Masanjore. The Bharatiya Janata Party withdrew its backing from the National Front government on the same day. Volunteers of the temple campaign nevertheless converged on Ayodhya, and on 30 October and again on 2 November 1990 the Uttar Pradesh police, under orders of the Mulayam Singh Yadav government to prevent the volunteers from reaching the disputed site, opened fire on crowds of kar sevaks in Ayodhya. The number killed in the firings remains disputed between the figures given by the Uttar Pradesh government of the day and the higher figures maintained by the temple campaign’s organisations. The firings completed the third phase’s work. A government in Lucknow drawn from the Mandal coalition had ordered fire on a religious mobilisation; a government in Delhi that depended on the party of that mobilisation had fallen because of an arrest made by another Mandal government in Patna. The three phases were no longer three stories. They were one mechanism, and it had consumed the government that stood at its centre.

The interaction between the second and third phases deserves precision, because it is the point on which the three-phase reading stands or falls. The claim is not that the Rath Yatra was caused by the anti-Mandal agitation; the temple campaign predated the memorandum and followed its own calendar fixed at Palampur and by the Vishwa Hindu Parishad’s announced date of 30 October 1990. The claim is that the Yatra travelled through a country the agitation had prepared. District towns where students had spent September denouncing the Mandal decision received the chariot in October with a receptivity that the campaign’s organisers had not had to manufacture. The anti-Mandal substrate, the settled conviction among forward caste educated families that the state had been turned against them, did not dictate allegiance to the temple campaign, whose social base was broader and whose appeal was devotional before it was political. But the substrate lowered the resistance that a purely religious appeal might have met among the educated urban families whose sons had been on the streets. Pradeep Chhibber, in Democracy without Associations of 1999, argued that India’s party system was being rebuilt in this period around the axis of state intervention and the axis of religious and caste identity, with parties repositioning as the old Congress centre dissolved. The autumn of 1990 is the compressed record of that rebuilding. Within eleven weeks, the electorate of north India was offered a caste-based claim on the state, a merit-based defence against that claim, and a religious claim to supersede both, and the party system that emerged from the following decade was built from the relative strengths the three offers displayed.

The Fall of the Government on 7 November 1990

The withdrawal of the Bharatiya Janata Party’s backing on 23 October 1990 reduced the National Front government to a minority, and the end came by the parliamentary road. Chandra Shekhar, the veteran socialist who had contested the leadership of the Janata Dal against V. P. Singh since 1989, broke from the party with a bloc of members of Parliament and constituted the Samajwadi Janata Party. V. P. Singh sought a vote of confidence in the Lok Sabha. The debate of early November 1990 ranged across all three phases of the autumn: the government defended the Mandal implementation as social justice, defended the halt of the Yatra as the protection of the Babri Masjid and of public order, and charged its opponents with uniting caste privilege and communal mobilisation against a reforming government. The opposition answered from the same three phases in reverse: the implementation as a cynical device, the agitation as the nation’s verdict on it, and the arrest of Advani as proof that the government governed by provocation. On 7 November 1990 the House divided. The government received 142 votes; 346 members voted against. V. P. Singh became the first prime minister of India to be voted out of office on the floor of the Lok Sabha, and he resigned the same day. His final speech asked the House what kind of India it wanted, a question that his supporters and his critics have quoted against each other ever since. Chandra Shekhar was sworn in as prime minister on 10 November 1990 at the head of a minority government backed from outside by the Congress, and Devi Lal, dismissed in August, returned as his deputy prime minister. The Mandal memorandum, stayed by the Supreme Court in October, was not revoked by the successor government. It waited in suspension for the nine-judge bench, outliving the ministry that had issued it and the ministry that replaced it.

The arithmetic of the division of 7 November repays attention, because it locates the causation of the fall where the three-phase reading locates it. The government’s 142 votes comprised the rump of the Janata Dal, the Left parties, and regional allies. The 346 against combined the Congress, the Bharatiya Janata Party and the Chandra Shekhar bloc, three forces that agreed on nothing except the removal of V. P. Singh. The Bharatiya Janata Party voted against the government because of the arrest at Samastipur, an act of the third phase. The Chandra Shekhar bloc voted against the government because of the internal Janata Dal struggle that the Devi Lal dismissal of 1 August had opened, the same struggle out of which the timing of the Mandal announcement had been read. The Congress voted against the government as it voted against every non-Congress government. To say that Mandal caused the fall, as the simplest version of the arithmetic framing says, is to mistake the field for the trigger. The government’s fall required the Yatra, the arrest, the withdrawal, the split and the confidence motion, a chain in which the Mandal decision was the first link and the necessary condition of the political configuration, but in which each link was forged by a different actor pursuing a different end. The three-phase reading preserves that distribution of agency. The protest-narrative framing, by contrast, hands the fall to the students, who in fact never sought the government’s removal as their stated demand, and the arithmetic framing hands it to V. P. Singh’s calculations, which in fact aimed at the government’s preservation.

The Chandra Shekhar Interlude and the Survival of the File

The government that succeeded V. P. Singh deserves a closer look than it usually receives in the Mandal literature, because its composition was the autumn’s three phases rendered as a cabinet. Chandra Shekhar, sworn in on 10 November 1990, led a party constituted three days earlier out of the Janata Dal’s dissidents, and governed with the outside backing of the Congress, the party the National Front had been elected to displace. Devi Lal, whose dismissal on 1 August had opened the sequence, sat as deputy prime minister in the successor government, restored by the fall of the man who had dismissed him. The Bharatiya Janata Party, whose withdrawal had made the fall possible, sat in opposition and husbanded the gains of the third phase. A government so composed had neither the will nor the numbers to legislate on Mandal in either direction, and it did not try. The memorandum rested under the Supreme Court’s stay, the file moved between the Law Ministry and the Department of Personnel and Training as the government’s lawyers prepared the Union’s defence of the scheme in Indra Sawhney, and the state’s position before the nine-judge bench remained the position the fallen government had taken: that the 27 percent was constitutional and would be administered. The continuity is the point. Between November 1990 and the Congress government’s modifications of September 1991, the Indian state’s institutional answer to Mandal never wavered, through a change of prime minister, a change of governing party and the approach of a general election. Whatever the politicians said of one another, the file survived them all, and a file that survives three governments in twenty-five months has ceased to be any government’s instrument. It has become the state’s settled direction, awaiting only the Court’s permission to proceed. The Chandra Shekhar government fell in March 1991 when the Congress withdrew its backing over an unrelated surveillance controversy, and the general election that followed was fought by every major party with the Mandal question in its manifesto and the Ayodhya question at its rallies, the two inheritances of the autumn running in harness into the decade.

Administering the Memorandum: The Machinery of the 27 Percent

A reservation is not self-executing, and part of the autumn’s confusion arose from the distance between the announcement of a percentage and the machinery that turns a percentage into appointments. The office memorandum of 13 August 1990 operated on vacancies as they arose in direct recruitment to civil posts and services under the Government of India and in the public sector undertakings. It did not disturb any serving officer, cancel any pending selection completed before its date, or apply to admissions in educational institutions, which the first memorandum left aside; the extension of backward classes reservation to central educational institutions waited until the amendment of Article 15 by the Ninety-Third Amendment and the Central Educational Institutions Act of 2006. The memorandum’s beneficiaries were the castes and communities in the lists the Mandal Commission had compiled state by state, applied in the first instance from the central list drawn from the Commission’s schedules, and the practical administration of those lists, the issuing of caste certificates by district authorities, the verification of claims, the maintenance of reservation rosters in each ministry and undertaking, became in the months after August 1990 a vast clerical enterprise conducted under political fire. Ministries sought clarifications from the Department of Personnel and Training on the treatment of vacancies already advertised, on the carry-forward of unfilled reserved vacancies, and on the application of the quota to the public sector banks and financial institutions whose recruitment boards worked to their own calendars.

The stay of October 1990 froze this machinery almost before it had engaged a gear. A handful of recruitment cycles had begun under the memorandum; most had not. The freeze had a consequence that the participants noticed at the time and that the analysis should not lose. Because the implementation was stayed before it delivered appointments in any number, the autumn’s argument remained an argument about a prospect rather than about a distribution of actual posts. No backward caste candidate had yet entered the civil services under the 27 percent when the agitation peaked, and no forward caste candidate could point to a named appointment lost. The conflict was therefore conducted entirely in the currency of expectation, which is the most inflammable currency in politics: each side could price the memorandum at the value its fears or hopes assigned, uncontradicted by any settled outcome. Had the implementation run for two years before its challenge, the second phase would have confronted beneficiaries with faces and postings, and the politics of withdrawal would have faced the resistance that entrenched entitlements always face. The Court’s stay, by suspending the fact while the argument raged, preserved the question in its most abstract and most explosive form for twenty-five months. When the judgment of November 1992 finally released the machinery, it released it with the creamy layer excluded and the doctrine settled, into a country whose party system had already been rebuilt around the suspended question. The sequence is a reminder that in the Mandal autumn the institutions did not stand outside the mobilisation as its referees. The Court’s timetable, the ministries’ clarifications, the district officers’ certificates were themselves moves within the three phases, altering what each phase could plausibly demand and what each could plausibly win.

The Memorandum in Suspension: November 1990 to November 1992

The two years between the fall of the government and the judgment of the Supreme Court are the control experiment of the whole episode, the period in which the readings of 1990 can be tested against the behaviour of actors who owed V. P. Singh nothing. The Chandra Shekhar government, which held office from November 1990 to March 1991 on Congress sufferance, did not revoke the memorandum. The election of May and June 1991, interrupted by the assassination of Rajiv Gandhi on 21 May, returned the Congress as the largest party, and the government of P. V. Narasimha Rao took office in June 1991 with every incentive of the arithmetic reading to bury a predecessor’s device. The Rao government instead retained the 27 percent, and by its office memorandum of 25 September 1991 modified the scheme in two respects: within the reserved vacancies, preference would go to the poorer sections among the socially and educationally backward classes, and an additional 10 percent of vacancies would be reserved for other economically backward sections of the people not covered by any existing scheme of reservation. The second modification answered the protesters’ central demand, an economic criterion, in a form the Congress could call its own; the first modification anticipated, in administrative language, the creamy layer doctrine the Court would announce a year later. A Congress government with no debt to the Janata Dal and a fresh memory of the agitation thus amended Mandal rather than annulling it. The explanation available to the three-phase reading is that the first phase’s coalition had become a permanent electoral fact that no governing party could afford to confront. The arithmetic reading, which treats the memorandum as one man’s instrument, has no ready account of why his opponents preserved and extended it.

In the Supreme Court, the questions referred to the nine-judge bench were argued across 1991 and 1992 by the leading counsel of the republic, with the Union of India defending the memoranda and the petitioners attacking them on the equality code of Articles 14, 15 and 16. The judgment delivered on 16 November 1992, Indra Sawhney v Union of India, ran to a majority of six against three. The leading majority opinion of Justice B. P. Jeevan Reddy upheld the reservation of 27 percent for the backward classes in central services, held that caste could constitute a class for the purposes of Article 16(4) although caste alone could not exhaust the inquiry, required the exclusion of the creamy layer, the socially advanced persons within the beneficiary castes, forbade reservation in promotions, and reaffirmed that reservations should remain below 50 percent except in extraordinary situations. The Rao government’s additional 10 percent for the economically backward sections was struck down, as a reservation founded on economic criterion alone. The dissenting judges would have gone further in dismantling the scheme. For the history of the autumn of 1990, the operative facts are simpler. The memorandum of the fallen government was sustained in its core by the highest court, with its sharpest edge, the inclusion of the prosperous within the beneficiary castes, removed by judicial doctrine. The second phase’s most concrete institutional demand, that the better-off among the backward classes not take the benefit, entered the law through the judgment, although the street that had demanded it had dispersed two years earlier. The phases, in other words, continued to act after the mobilisation ended: the first phase’s enactment, corrected by the second phase’s critique, administered under governments formed by the third phase’s party arithmetic. That continuing interaction, legible only in retrospect, is the strongest single piece of evidence for reading the autumn as one integrated event rather than three adjacent ones.

Three Readings of the Autumn of 1990

The scholarly argument over 1990 can be adjudicated on the evidence assembled. The three-reading adjudication framework applied across this pillar distinguishes the arithmetic reading, the protest reading and the mobilisation reading defended here. The arithmetic reading, held by many contemporary journalists and by the Congress and Bharatiya Janata Party politicians who fought V. P. Singh, treats the implementation as a move within elite competition: the dismissal of Devi Lal, the farmers’ rally of 9 August, the minority government’s need for a consolidated backward caste base. Its strength is documentary; the sequence of 1 to 7 August is real, and no honest account can dissolve it. Its weakness is reach. It predicts that a decision taken for a week’s advantage would be bargained away when the week ended, and the opposite occurred: the decision was defended in court by successor governments, upheld in November 1992, and administered by every government since. The protest reading, held by the agitation’s participants and by the strand of commentary that treats September 1990 as a generation’s wound, treats the autumn as the story of the students and of Rajiv Goswami. Its strength is moral and documentary attention to the phase that cost the most lives. Its weakness is that it cannot explain its own ending: the agitation faded by November without achieving the withdrawal of the memorandum, which survived the government, the stay and the decade. A reading that makes the protests the whole story must treat their central demand’s failure as an epilogue, when the failure is precisely what requires explanation.

The three-phase mobilisation map, Table 1, sets out the reading defended in this article, with the phases distinguished by their actor configurations, their geographies and their institutional consequences.

Phase	Period in 1990	Principal actors	Signature events	Institutional consequence
Phase One: pro-Mandal consolidation	7 August to mid-September	Janata Dal and allied state governments, caste associations of the backward classes, the Left parties with qualifications, the Dravidian parties	Announcement of 7 August, office memorandum of 13 August, pro-implementation rallies across Bihar, Uttar Pradesh and Madhya Pradesh	The implementation acquired a defending coalition that outlasted the issuing government
Phase Two: anti-Mandal counter-mobilisation	Mid-September to October	Forward caste students of Delhi and the north Indian university towns, the Anti-Mandal Commission Forum, the Equality Front, sympathetic teachers	Class boycotts, the Boat Club demonstration of 4 September, the self-immolation of Rajiv Goswami on 19 September, the Supreme Court stay of October	The anti-Mandal substrate settled into the forward caste educated families of north India
Phase Three: Hindutva mobilisation	25 September to November	The Bharatiya Janata Party, the Vishwa Hindu Parishad, the Bihar and Uttar Pradesh state governments as counter-actors	The Rath Yatra from Somnath, the arrest of Advani at Samastipur on 23 October, the Ayodhya firings of 30 October and 2 November, the confidence vote of 7 November	The National Front government fell, and the party system reorganised around the Mandal and temple axes together

The complication that any honest statement of the three-phase reading must face is overlap. The phases were not sealed compartments. Pro-Mandal rallies continued while the student agitation burned; the Rath Yatra began on 25 September, six days after the Goswami self-immolation, while the second phase was at its height; the Supreme Court’s stay landed in the middle of the third phase’s march. A framework that requires clean temporal separation would fail on this chronology. The framework defended here requires only analytical distinctness, and that the chronology supplies. The three phases drew on different actor configurations: parties and caste associations in the first, students and teachers in the second, a religious-political campaign and its state opponents in the third. They carried different substantive content: a claim on state employment, a defence of examination merit, a demand for a temple. They left different institutional residues: a defended quota, a durable substrate of resentment, a destroyed government and an enlarged party. Phases in political time are like currents in a river at a confluence, distinguishable by their sources and their directions although their waters mix. The mixing is not an objection to the map; the mixing is the mechanism the map exists to display.

A final test of the framework is predictive rather than descriptive. If the autumn had been only an arithmetic episode, its effects should have decayed with the government that authored it; if it had been only a protest episode, its effects should have decayed with the agitation. The three-phase reading predicts instead that each phase would leave an institution behind it, and the record of the following decade satisfies the prediction three times. The first phase left parties: the Samajwadi Party of 1992, the Rashtriya Janata Dal of 1997, and the state regimes in Lucknow and Patna that made backward caste leadership the normal condition of Hindi belt government. The second phase left a doctrine and a disposition: the creamy layer exclusion written into the judgment of November 1992, which answered the protesters’ strongest argument in law, and the substrate, which answered nothing and persisted in the agitations of 2006 and after. The third phase left a party system: a Bharatiya Janata Party enlarged from 85 seats to 120 and then to government in 1998, and an opposition field reorganised for a generation around the question of how to answer it. Institutions outlast the moods that found them. The autumn of 1990 is remembered in the headlines for its fires, but it is present in Indian politics through its institutions, and a reading of the period that cannot find the institutions has mistaken the weather of the season for its climate.

How did the three phases act on each other?

The consolidation of August defined the decision as the property of a caste coalition, giving the September protesters their target. The agitation then assembled a forward caste public persuaded that the state had been captured against it, and that public received the Rath Yatra as a counter-assertion of unity. The Yatra’s halt destroyed the government that had issued the memorandum.

The Consequences After 1990: Parties, Elections and the Surviving Memorandum

The consequences of the autumn arrived in a sequence that the three-phase reading predicts and the rival readings do not. In the general election of May and June 1991, fought in the shadow of the Rajiv Gandhi assassination of 21 May 1991, the Bharatiya Janata Party rose from 85 seats in 1989 to 120 seats, its advance concentrated in Uttar Pradesh, Madhya Pradesh, Rajasthan and Gujarat, the very geography of the second and third phases. The Janata Dal fell from 143 seats to 59, punished in its northern strongholds although its successor formations inherited its social base. The Congress, with 244 seats, constituted the government under P. V. Narasimha Rao, and the Rao government carried the Mandal file forward: an office memorandum of 25 September 1991 modified the stayed memorandum by directing preference within the 27 percent to the poorer sections of the backward classes and by adding a 10 percent reservation for economically weaker sections of the population not covered by any existing scheme. The nine-judge bench of the Supreme Court then delivered its judgment in Indra Sawhney on 16 November 1992, upholding the 27 percent reservation for the backward classes in central services, requiring the exclusion of the creamy layer, rejecting reservation in promotions, and holding the total at or below 50 percent save in extraordinary circumstances. The implementation that the students of 1990 had demanded be withdrawn entered the administrative law of India under a Congress government, through a judgment of the Supreme Court, twenty-seven months after the government that announced it had fallen. No fact of the period is harder for the arithmetic reading to digest.

How many seats did the Bharatiya Janata Party win in 1991?

The Bharatiya Janata Party won 120 seats in the Lok Sabha election of 1991, rising from 85 seats in 1989. Its advance was concentrated in the northern and western states through which the Rath Yatra had passed and in which the anti-Mandal agitation had been strongest, converting the third phase’s mobilisation into parliamentary presence.

The verdict of 1991 also carried a quieter instruction about the limits of each phase. The Bharatiya Janata Party’s 120 seats made it the second party of the Lok Sabha, yet the government was constituted by the Congress, which had stood aside from all three phases and harvested the electorate’s exhaustion with their combined turbulence. Abstention, too, was a strategy the autumn rewarded, and the Congress of P. V. Narasimha Rao converted abstention into five years of office in which it administered the Mandal settlement, liberalised the economy and watched the temple movement destroy the mosque at Ayodhya in December 1992 under a state government of the third phase’s party. The party consequences in the states ran on the Mandal track as faithfully as the national consequences ran on the temple track. Mulayam Singh Yadav founded the Samajwadi Party in October 1992, carrying the Uttar Pradesh base of the Janata Dal into a formation that joined the backward caste claim of the first phase to an alliance with the state’s Muslim electorate, and governed Uttar Pradesh in coalition after the election of 1993. Lalu Prasad Yadav consolidated the Janata Dal’s Bihar inheritance through the 1990s and, after the party’s division, founded the Rashtriya Janata Dal in July 1997, which carried the Bihar version of the Mandal coalition into the next decade. The trajectory from the first phase through these formations to the 1992 Babri demolition trajectory of December 1992 and onward to the post-1996 BJP-led NDA architecture shows the two axes of 1990, caste assertion and temple mobilisation, organising the party system jointly for a generation, until the post-Mandal arithmetic break of 2014 rearranged their relation once more. Yogendra Yadav’s account of a second democratic upsurge, in which the participation of the lower castes deepened through the 1990s, describes the first axis; Jaffrelot’s silent revolution describes the transfer of power the two axes together effected. The autumn of 1990 is where both accounts begin, and the three-phase reading is the account of how they came to begin together.

The Settlement of 1993 in Uttar Pradesh and Bihar

The state elections of November 1993 in Uttar Pradesh supplied the first electoral test of the post-autumn configuration in the state where all three phases had burned hottest, and the result vindicated the three-phase reading in detail. The Bharatiya Janata Party, dismissed from office in December 1992 after the demolition at Ayodhya brought down the four state governments of the party under President’s rule, remained the largest single party in the Uttar Pradesh assembly. But the government was constituted by the coalition the first phase had made possible: the Samajwadi Party of Mulayam Singh Yadav in alliance with the Bahujan Samaj Party of Kanshi Ram and Mayawati, joining the backward caste electorate mobilised by Mandal to the Dalit electorate organised by the Bahujan Samaj Party across the preceding decade. The alliance joined precisely the two constituencies that the constitutional reservation framework had always treated as distinct, the other backward classes of the Mandal lists and the Scheduled Castes of the older settlement, and it governed on the formula that the defence of both against the temple party was the state’s first business. In Bihar, the Janata Dal government of Lalu Prasad Yadav won re-election in 1995 against the combined weight of the Congress decline and the Bharatiya Janata Party’s advance, on the strength of the same identification of the state government with the Mandal coalition that the arrest at Samastipur had sealed in October 1990. The two results together mark the point at which the first phase’s consolidation ceased to be a mobilisation and became a regime: in the two largest states of the Hindi belt, the politics the announcement had called into being held office for most of the following fifteen years, and every national coalition of the period from 1996 to 2014 was negotiated in the shadow of that fact.

The Scholarly Reckoning: Upsurge, Fragmentation and the Silent Revolution

The academic literature that gathered around the autumn divides along lines that reproduce the three readings, and the division is itself evidence for the framework used here. Yogendra Yadav, writing in the mid-1990s and gathering the argument in Making Sense of Indian Democracy in 2020, described a second democratic upsurge: the electoral participation of the lower castes, measured in turnout and in representation, rose through the 1990s until the social profile of India’s voters and legislators more nearly matched the social profile of its population. On Yadav’s reading, Mandal was the moment the upsurge acquired a programme and a state instrument, and the protests of the second phase were the resistance of a departing order to the terms of its departure. Pradeep Chhibber’s Democracy without Associations of 1999 offered the colder reading. The Indian party system, Chhibber argued, was reorganising around two cleavages, the reach of state intervention in the economy and the politics of religious and caste identity, in a polity whose associational life was too weak to discipline either; the autumn of 1990 was the cleavage structure announcing itself, with the Mandal decision and the Rath Yatra as the twin signals that the Congress centre could no longer contain the identities it had once aggregated. Jaffrelot’s India’s Silent Revolution of 2003 supplied the longest arc. The transfer of power from the upper castes to the lower castes of north India, which the south had accomplished decades earlier through the Dravidian movement, proceeded in the north by the ballot rather than by movement politics, and the Mandal implementation was its decisive institutional episode, the point at which the lower caste parties ceased to be pressure groups on other men’s governments and became governments themselves.

The three scholarly readings are usually presented as rivals, and at the level of interpretation they are. Yadav’s upsurge is a story of democratic deepening; Chhibber’s realignment is a story of systemic re-sorting without a hero; Jaffrelot’s revolution is a story of a social transfer of power that happened to use democratic means. But at the level of the autumn’s mechanics, the three accounts describe the three phases. Yadav’s upsurge is Phase One generalised across a decade. Chhibber’s identity cleavage, in its caste register, is Phase Two’s grievance and Phase One’s claim locked in one electorate; in its religious register it is Phase Three. Jaffrelot’s transfer of power is the institutional deposit the three phases left in the state governments of Bihar and Uttar Pradesh. The framework of this article does not pretend to supersede these scholars, on whose work it draws throughout. It claims the more modest achievement of sequence: by following the mobilisation in its phases, it shows how the upsurge, the counter-mobilisation and the religious consolidation could be simultaneously real and mutually conditioning, each supplying the others with actors, occasions and audiences. Accounts that begin from one phase and treat the others as background are not wrong about their phase. They are incomplete about the autumn, and incompleteness, in the historiography of 1990, has repeatedly masqueraded as refutation.

One further scholarly caution belongs in the record, because it qualifies the phase language itself. Sudipta Kaviraj, in his essays on the trajectory of Indian politics, warned against reading the post-1989 decades as a simple succession of novelties, observing that the Congress decline, the rise of caste-based parties and the growth of the Bharatiya Janata Party each had long gestations inside the older order. The caution is accepted here in the form the brief’s complication requires: the phases of 1990 were not creations from nothing. The pro-Mandal coalition had been assembled across the 1980s in the state parties; the merit argument had been rehearsed in every litigation over reservation since Champakam Dorairajan in 1951; the temple campaign had been building since the locks were opened in 1986. What the autumn supplied was not the elements but the compression. Within one hundred days, from the Cabinet decision of 6 August to the confidence vote of 7 November, the elements met, reacted and set in a configuration that the following thirty years inherited. Decisive moments in politics are usually compressions of this kind, and the three-phase map is a device for holding a compression still long enough to be examined.

A closing word belongs to the substrate, because it is the consequence least visible in the institutional record. The families whose sons and daughters attempted self-immolation in 1990, and the very much larger number who marched, boycotted and signed, did not experience the judgment of November 1992 as reconciliation. The creamy layer exclusion answered one of their arguments, the argument from the prosperous beneficiary, and the 50 percent ceiling answered another, the argument from unlimited expansion. The core of their grievance, that a child’s examination marks should not be discounted by the caste of the child’s birth, was the core the judgment declined to adopt, holding instead that caste could serve as a marker of class for the purposes of Article 16. The substrate therefore persisted as an available political sentiment, surfacing in the agitation against the extension of reservation to central educational institutions in 2006, when the mother of Rajiv Goswami addressed the protesters, and in the recurring demands of forward caste groups for inclusion in the beneficiary lists. The phases of 1990 ended in November of that year. The mobilisation they composed did not end. It became the standing structure within which Indian politics has since been conducted, which is the precise sense of the claim with which this article began: the integrated shift produced by the three phases together is the event of 1990, and the announcement, the protests and the Yatra are its parts.

Frequently Asked Questions

Q: What was the 1990 Mandal announcement?

On 7 August 1990, Prime Minister V. P. Singh announced in Parliament that the Union government would implement the central recommendation of the Mandal Commission, reserving 27 percent of vacancies in civil posts and services under the Government of India for the socially and educationally backward classes. The Union Cabinet had taken the decision on 6 August, and the Department of Personnel and Training gave it legal form in an office memorandum dated 13 August 1990, which declared the reservation effective from 7 August. The announcement ended a decade of deferral during which the report, submitted on 31 December 1980, had been discussed in Parliament in 1982 and 1983 without implementation by the Congress governments of the period.

Q: Who was Rajiv Goswami?

Rajiv Goswami was a nineteen-year-old commerce student at Deshbandhu College, Delhi University, who set fire to himself on 19 September 1990 during a student demonstration against the Mandal implementation, becoming the first student to attempt self-immolation in the agitation and its best-known face. Photographs of Goswami ablaze appeared across the national press and converted the student protest into a national crisis. He survived the attempt with burns over most of his body, was later elected president of the Delhi University Students’ Union, and withdrew from active politics as his health failed. He died on 24 February 2004 in a Delhi hospital, of complications that followed from the injuries of September 1990.

Q: Why did students protest Mandal?

The students who protested in September and October 1990 stated their case in the literature of the Anti-Mandal Commission Forum and the Equality Front. A central government post was, in 1990, the principal route by which educated families of modest means secured a salaried life, and the reservation of 27 percent of vacancies was experienced by forward caste students as the closing of that route to all but a shrinking open category. The protesters demanded that any reservation rest on economic criteria rather than caste, attacked the Mandal lists as resting on the census of 1931, and defended examination merit as the one impartial standard available to the poor of every caste. Their opponents answered that the open category had never been impartial in practice.

Q: How did the Rath Yatra connect to Mandal?

The Ram Rath Yatra of L. K. Advani, which set out from Somnath on 25 September 1990 towards Ayodhya, belonged to the Ram Janmabhoomi campaign and followed a calendar fixed independently of the Mandal announcement. Its connection to Mandal was positional and consequential rather than causal in origin. The Yatra travelled through a north India polarised by the Mandal agitation and offered the forward caste public assembled by the anti-Mandal protests a politics of Hindu unity across caste, against the caste-based claim on the state that the implementation represented. The shorthand of the period, Mandal versus Kamandal, named the rivalry. When the Bihar government arrested Advani at Samastipur on 23 October 1990, the Bharatiya Janata Party withdrew its backing from the government that had issued the Mandal memorandum.

Q: Did Mandal cause the V.P. Singh government to fall?

The fall of 7 November 1990 had a chain of causes in which Mandal was the first link rather than the immediate trigger. The government lost its majority when the Bharatiya Janata Party withdrew its backing on 23 October 1990, in response to the arrest of L. K. Advani at Samastipur by the Bihar government of Lalu Prasad Yadav, and it lost the confidence vote because the Chandra Shekhar bloc had also broken from the Janata Dal. The Mandal implementation had, however, created the configuration in which the chain could form: it had bound the government to the backward caste state governments of Bihar and Uttar Pradesh, whose actions against the Rath Yatra and the Ayodhya mobilisation supplied the occasions of the rupture.

Q: What is the post-Mandal anti-Mandal substrate?

The term names the durable disposition that the second phase of the 1990 mobilisation left behind among the forward caste educated families of north India: a settled conviction that reservation in state employment and education operates as dispossession of the meritorious, available for mobilisation by later movements and parties. The substrate outlasted the agitation that produced it. It resurfaced in the protests against the extension of backward classes reservation to central educational institutions in 2006, in which the family of Rajiv Goswami took part, and in later demands by forward caste organisations for inclusion among the beneficiaries of reservation. Unlike a protest, a substrate requires no organisation to persist; it is carried in family memory and in the retelling of the autumn of 1990.

Q: How did the BJP benefit from Mandal?

The Bharatiya Janata Party benefited indirectly and then substantially. It had backed the National Front government from outside and opposed neither reservation in principle nor, at first, the implementation itself, which kept it clear of the anti-Mandal agitation’s caste framing. The Rath Yatra then allowed the party to gather the autumn’s discontented voters under a religious banner that crossed caste lines. After withdrawing its backing on 23 October 1990, the party converted the mobilisation into seats: from 85 seats in the election of 1989 it rose to 120 seats in 1991, with its gains concentrated in the northern and western states where the agitation and the Yatra had both been strongest, and it emerged as the principal opposition party of the 1990s.

Q: When did the Supreme Court stay the Mandal implementation in 1990?

The Supreme Court stayed the operation of the office memorandum in October 1990, roughly a month after a five-judge Constitution Bench, admitting the writ petitions that became Indra Sawhney v Union of India on 11 September 1990, had declined to grant a stay at the admission stage. The October stay halted recruitment under the 27 percent reservation before it had begun in earnest and directed that the constitutional questions be placed before a bench of nine judges. The implementation remained suspended through the fall of the V. P. Singh government and the Chandra Shekhar and Rao governments that followed, until the nine-judge bench delivered its judgment on 16 November 1992, upholding the reservation with the creamy layer excluded.

Q: How did the Congress respond to the August 1990 Mandal announcement?

The Congress responded with procedural criticism and strategic silence on the substance. The party had held the Mandal Report for a decade, discussing it in Parliament in 1982 and 1983, and its Home Minister P. C. Sethi had told the House in 1983 that the government desired to implement the recommendations, so outright opposition to implementation would have indicted the party’s own record. Congress spokesmen attacked the timing of the announcement, warned of threats to social peace as the agitation grew, and accused V. P. Singh of dividing the nation for party advantage, while the party voted against his government in November 1990 on the general question of confidence. In office after June 1991, the Congress government of P. V. Narasimha Rao retained and modified the memorandum rather than revoking it.

Q: What happened to Rajiv Goswami after September 1990?

Goswami survived the self-immolation attempt of 19 September 1990 with burns over most of his body and underwent years of treatment and repeated hospitalisation. He remained a symbol of the anti-Mandal agitation, was elected president of the Delhi University Students’ Union in the agitation’s aftermath, and for a period stayed in public life as a speaker for the cause he had come to represent. Failing health, which he and his family attributed to the injuries and their treatment, obliged him to withdraw from active politics, and he turned to private business. He died on 24 February 2004, aged thirty-two, in the Holy Family Hospital in Delhi. His mother, Nandrani Goswami, addressed the anti-reservation protests of 2006 in his memory.

Q: Why did the anti-Mandal protests fade by November 1990?

Several pressures converged. The Supreme Court stay of October 1990 removed the implementation’s immediate operation, which had been the agitation’s concrete object, and transferred the contest to a courtroom where students could not march. The Rath Yatra and the Ayodhya firings of 30 October and 2 November displaced Mandal from the front pages and absorbed the public attention the protests had monopolised. University administrations moved to recover the academic calendar, examinations approached, and the organisations of the agitation, which had always been looser than the party student wings, could not sustain a boycott without a proximate target. The fall of the V. P. Singh government on 7 November then removed the agitation’s named adversary. The sentiment persisted as substrate; the mobilisation, as a mobilisation, was over.

Q: How did Tamil Nadu respond to the 1990 Mandal implementation?

Tamil Nadu responded from the standpoint of a state in which backward classes reservation was an old settlement rather than a new rupture. The Dravida Munnetra Kazhagam government of M. Karunanidhi welcomed the central implementation, and the state’s parties treated the Delhi decision as the centre adopting a principle that the Dravidian movement had carried into state policy across the twentieth century. Tamil Nadu had legislated a total reservation of 69 percent in 1990 for the backward classes, the Scheduled Castes and the Scheduled Tribes combined, a figure far above the central 27 percent for the backward classes alone. Student protest on the Delhi pattern did not materialise, because the central memorandum altered no Tamil student’s prospects: the state regime already governed admissions and state employment, and its legitimacy across parties was long established.

Q: What did V.P. Singh argue in Parliament in defence of the Mandal decision?

V. P. Singh defended the decision on three grounds across the debates of August to November 1990. He argued fidelity to a mandate, since the National Front manifesto of 1989 had promised implementation of the Mandal recommendations and the promise bound the government. He argued constitutional principle, presenting the reservation as the instrument by which Article 16’s equality of opportunity could be made real for classes that a formal equality had left outside the services for forty years. And he argued democratic weight, observing in substance that the backward classes formed the majority of the population and that a state administration recruited almost entirely without them could not claim to represent the nation it administered. In his final speech of 7 November 1990 he joined the defence of Mandal to the defence of the Babri Masjid as the two grounds on which he accepted the loss of office.

Q: Which parties supported the Mandal implementation in 1990?

The implementation was carried by the Janata Dal and the state governments of its family, prominent among them the Bihar government of Lalu Prasad Yadav and the Uttar Pradesh government of Mulayam Singh Yadav, and it was defended in the states by the caste associations and federations of the backward classes. The Communist Party of India (Marxist) and the Communist Party of India backed the implementation with the stated qualification that an economic criterion should exclude the better-off strata of the beneficiary castes. The Dravida Munnetra Kazhagam in Tamil Nadu welcomed the decision from within its own older reservation settlement. The Congress withheld substantive opposition without endorsing the government, and the Bharatiya Janata Party criticised the announcement’s manner and timing while declining, during August and September, to oppose backward classes reservation in principle.

Q: How did the protests spread across north Indian university towns in September 1990?

The spread ran along the networks of the examination economy. Delhi supplied the model and the images: the class boycott, the march on government offices, the demonstration at the Boat Club, and after 19 September the example of self-immolation. University towns that lived by the same queue for government employment reproduced the repertoire within days: Allahabad, Varanasi, Lucknow, Meerut and Agra in Uttar Pradesh; Patna in Bihar; Jaipur in Rajasthan; Kurukshetra, Hisar, Sirsa and Ambala in Haryana; Indore and Jabalpur in Madhya Pradesh. Local student bodies affiliated loosely to the Delhi organisations or acted under their own banners, administrations closed institutions or postponed examinations, and the state governments answered with a mixture of negotiation, arrest and, in several towns, police firing on crowds that had blocked roads and railway lines.

Q: What was the Anti-Mandal Commission Forum?

The Anti-Mandal Commission Forum was one of the coordinating bodies through which the Delhi student agitation of 1990 organised itself outside the established party student wings. Its banner appeared at the march of some two thousand students on the Doordarshan Kendra in Delhi in the early phase of the protests, and its literature stated the agitation’s platform: that the Mandal Report was, in its phrase, a farce in the name of social justice, resting on the outdated census of 1931 and designed for the electoral advantage of a minority government, and that reservation, where it existed, should be governed by economic criteria. Alongside it stood the Equality Front, constituted on 25 August 1990 by students with university teachers and employees, which called the Delhi bandh of 27 August 1990.

Q: How did Mulayam Singh Yadav’s government respond to the 1990 mobilisations in Uttar Pradesh?

The Uttar Pradesh government of Mulayam Singh Yadav stood on the Mandal side of the autumn and acted on both fronts that the state’s position required. As a Janata Dal government resting on the backward caste electorate, it defended the implementation against the student agitation in the state’s university towns, where protests were met with closures, arrests and, in places, police firing. As the government responsible for Ayodhya, it confronted the temple mobilisation directly: volunteers of the Ram Janmabhoomi campaign were stopped from reaching the disputed site, and on 30 October and again on 2 November 1990 the state police opened fire on crowds of kar sevaks at Ayodhya. The firings fixed Mulayam Singh Yadav’s reputation among the temple campaign’s supporters for a generation and confirmed his standing among the coalition the implementation had consolidated.

Q: When was the Samajwadi Party founded, and what did it inherit from 1990?

Mulayam Singh Yadav founded the Samajwadi Party in October 1992, after the Janata Dal’s fragmentation had made a state vehicle necessary for the Uttar Pradesh base he commanded. The party inherited from the autumn of 1990 a defined social coalition and a defined antagonist. The coalition joined the Yadav and other backward caste voters whom the Mandal implementation had mobilised to the state’s Muslim electorate, which the defence of the Babri Masjid in October and November 1990 had attached to Mulayam Singh Yadav’s government. The antagonist was the Bharatiya Janata Party of the Rath Yatra and Ayodhya. In the Uttar Pradesh election of 1993 the Samajwadi Party, in alliance with the Bahujan Samaj Party, constituted the state government, demonstrating that the first phase’s coalition could win office on its own account.

Q: How many seats did the Janata Dal win in the 1991 general election?

The Janata Dal won 59 seats in the general election of May and June 1991, a fall from the 143 seats it had won in 1989 as the core of the victorious National Front. The decline registered the fragmentation that the autumn of 1990 had completed: the Chandra Shekhar bloc had left the party before the confidence vote of November 1990, the Bharatiya Janata Party contested the election as an adversary rather than a supporter, and the party’s state units faced the Congress and the rising Bharatiya Janata Party without a governing record to defend. The social base the party had mobilised through the Mandal implementation did not disappear with its parliamentary decline. It passed, in Bihar and Uttar Pradesh above all, to the state formations that succeeded the Janata Dal through the 1990s.

Q: How did the Mandal implementation reshape Bihar’s party politics in the 1990s?

In Bihar the implementation fused with a government already in office. Lalu Prasad Yadav had become chief minister in March 1990 at the head of a Janata Dal government resting on the backward caste coalition, and the August announcement made his government the state embodiment of the Mandal claim. His order to arrest L. K. Advani at Samastipur on 23 October 1990 completed the identification: the Bihar government appeared to its supporters as the power that had enacted social justice in Delhi’s file and defended it on Bihar’s roads. Through the 1990s the coalition organised around that identification dominated the state’s politics, surviving the Janata Dal’s national decline, and in July 1997 it acquired a vehicle of its own when Lalu Prasad Yadav founded the Rashtriya Janata Dal after the parent party’s division, carrying the Mandal coalition into the following decade.