The Coalition Era 1989 to 2014 Explained: Renegotiation

The coalition era in India, the twenty-five years from the general election of November 1989 to the general election of May 2014, is conventionally narrated as a corridor between two houses: the Congress dominance that Indira Gandhi built and the Bharatiya Janata Party majority that Narendra Modi won. On the conventional narration, the corridor matters less than the houses at either end. This article defends the opposite claim. The coalition era was twenty-five years of substantive renegotiation across the three permanent tensions of the Indian founding settlement, and its governments, for all their fragility, rebuilt the post-Independence settlement before the 2014 verdict attempted to renegotiate that settlement again.

The coalition era in India from 1989 to 2014 and the renegotiation of the founding settlement - Insight Crunch

The Conventional Reading: An Interlude of Instability Between Two Dominances

The conventional reading begins from arithmetic that nobody disputes. Between December 1989 and May 2014, India held eight general elections: 1989, 1991, 1996, 1998, 1999, 2004, 2009 and 2014. Seven men held the office of Prime Minister across those twenty-five years, against two in the preceding twenty-five. Three governments of the period died before their second birthday. The V. P. Singh government fell in November 1990, eleven months after it took office. The Chandra Shekhar government, installed on Congress sufferance in November 1990, resigned in March 1991. The first Vajpayee government of May 1996 lasted thirteen days. A reader who stops at the arithmetic sees a system that could not settle, a quarter-century of hung parliaments, outside support, withdrawn letters and midnight confidence votes, and the conclusion follows easily: the period was a transition, and transitions are judged by what they connect rather than by what they contain.

Pradeep Chhibber gave this reading its most rigorous scholarly form in Democracy without Associations, published in 1999. Writing a decade into the era, Chhibber argued that the post-1989 party system was fragmenting because the older aggregative structures of Indian politics, the Congress organisation above all, had decayed without successor institutions of comparable reach. Parties multiplied, state-level formations captured national bargaining space, and the national agenda dissolved into the claims of segments. On Chhibber’s account, the instability of the 1990s was not an accident of personalities but the surface of a structural condition: a party system whose parts no longer added up to a whole. The fragmentation reading has the great merit of being true at the level it describes. The parties did fragment. The arithmetic did fail to produce single-party majorities for twenty-five years. What the reading cannot do is tell us what the fragmented system was doing with its power while it lasted.

The instability framing carries a second, quieter assumption: that nothing of constitutional weight happened between 1989 and 2014 because the governments were too weak to intend anything. The record refutes this on its face. The decade of the weakest governments, the 1990s, contains the implementation of the Mandal Commission’s recommendations in August 1990, the demolition of the Babri Masjid on 6 December 1992, the economic liberalisation launched on 24 July 1991, the Supreme Court’s Bommai judgment of 11 March 1994 that fenced the misuse of Article 356, and the nuclear tests of May 1998. The decade of the supposedly captive UPA governments contains the Right to Information Act of 2005, the National Rural Employment Guarantee Act of 2005 and the National Food Security Act of 2013. A period this dense with settlement-altering events is not an interlude. It is an argument that the settlement itself was being reopened, clause by clause, by the very arithmetic that made governments fragile.

How many Prime Ministers did India have between 1989 and 2014?

Seven men held the office between December 1989 and May 2014: V. P. Singh, Chandra Shekhar, P. V. Narasimha Rao, Atal Bihari Vajpayee, H. D. Deve Gowda, I. K. Gujral and Manmohan Singh. Vajpayee held it in three separate spells, and Manmohan Singh held it continuously for ten years, the era’s longest single tenure.

The comparison that powers the transitional reading is the preceding quarter-century. From 1964 to 1989, the office passed between Lal Bahadur Shastri, Indira Gandhi and Rajiv Gandhi, with Indira Gandhi alone occupying it for fifteen of those years, and single-party majorities were the norm in 1966, 1971, 1980 and 1984. Set against that baseline, seven Prime Ministers look like a system in dissolution. But the baseline is the exception in Indian political history, not the rule, and it was produced by one party’s extraordinary organisational reach, a reach whose construction and decay are themselves subjects of this series. When that reach failed, the underlying plurality of Indian politics became visible at the centre. The seven Prime Ministers of the coalition era were not seven failures of a machine that should have produced one. They were the arithmetic of a polity whose parts had acquired the power to bargain, and bargaining is a form of decision, not the absence of one.

This article therefore defends one namable claim: the coalition era was substantive renegotiation, not transitional interlude, and the post-Independence settlement was rebuilt across twenty-five years before the 2014 verdict. The claim is tested against the strongest rival readings in the adjudication section below, and it is organised by the framework this series applies to every period: the three permanent tensions of the founding, between religion and the state, between rights and development, and between the federal units and the Union. The coalition era renegotiated all three, and the renegotiations, not the falls of governments, are its durable content.

The Settlement the Era Inherited: Three Tensions After Four Decades of Congress Rule

Every renegotiation presupposes a text being renegotiated. For the coalition era, the text was the founding settlement of 1950, the bargain struck in the Constituent Assembly across the three founding tensions: how a Hindu-majority society would constitute a state neutral between religions, how a poor republic would reconcile justiciable rights with directed development, and how a subcontinental union would divide power between a strong centre and provinces that spoke different languages and remembered different histories. The Assembly did not resolve these tensions; it institutionalised them, leaving each generation to strike the balance again under new arithmetic. The four decades after 1950 struck the balance through a single instrument, the Congress party, which contained the tensions inside itself by housing the factions that embodied them.

The instrument began to fail at a datable moment. The 1969 split’s structural-cultural rupture broke the Congress habit of absorbing conflict internally and replaced it with a politics of personal command, and the personality-as-institution framework describes what followed: an organisation rebuilt around one office-holder, in which the party’s federal and factional shock absorbers were deliberately removed. The consequences were registered across the 1970s, in the Allahabad High Court judgment of 12 June 1975 that unseated a sitting Prime Minister, in the Emergency of 1975 to 1977, and in the Janata verdict of March 1977 that punished both. When Indira Gandhi returned in January 1980, the post-1980 Congress(I) organisational architecture was leaner and more centralised still: state leaderships nominated from Delhi, inner-party elections abandoned, the party reduced to an electoral machine whose sole moving part was the leader’s appeal. The decade ended with the office-rebuilding’s legible end in October 1984 and a successor, Rajiv Gandhi, who won 414 Lok Sabha seats in December 1984 on sympathy and inherited an organisation that could not reproduce that result by its own strength.

The inheritance of 1989 can therefore be stated precisely. The Congress that entered the election of November 1989 held 197 seats when the votes were counted, its lowest tally since Independence, and it faced two formations that the centralised system had called into being as its own antibodies. The first was the family of parties descended from the Janata experiment and the socialist tradition, speaking for the intermediate and backward castes whose mobilisation the Mandal Commission had documented. The second was the Bharatiya Janata Party, founded in April 1980, which had grown from two Lok Sabha seats in 1984 to 85 in 1989 on the Ram Janmabhoomi campaign. Neither formation could govern alone. Neither could the Congress. The three tensions of the founding had acquired three separate organisational carriers, none strong enough to impose a settlement, each strong enough to veto one. That is the structural definition of the coalition era, and it explains why the era’s characteristic act was the negotiated programme: the National Front’s common programme of 1989, the United Front’s Common Minimum Programme of 1996, the National Democratic Alliance’s National Agenda for Governance of 1998 and the United Progressive Alliance’s National Common Minimum Programme of 27 May 2004. Each document is a partial redraft of the founding bargain, signed by parties who agreed on clauses rather than on creeds.

The renegotiation reading also explains the era’s geography. Because the veto had migrated to the states, the decisive political units of the period were regional: the Telugu Desam in Andhra Pradesh, the Dravida Munnetra Kazhagam and its rival in Tamil Nadu, the Akali Dal in Punjab, the Shiv Sena in Maharashtra, the Samajwadi Party and the Bahujan Samaj Party in Uttar Pradesh, the Rashtriya Janata Dal in Bihar, the Trinamool Congress in West Bengal after 1998. These were not fragments of a broken national politics. They were the institutional form the federal tension took once the Congress container cracked, and their bargaining at the centre, over portfolios, over fiscal transfers, over the dismissal of state governments, was the federal clause of the settlement being rewritten in practice, year by year, cabinet by cabinet.

Religion and the State Renegotiated: Mandal, Mandir and the Courts

The first tension the era reopened was the oldest: the relation between religion, caste and the state. It reopened on a single day that contains the whole mechanism. On 7 August 1990, Prime Minister V. P. Singh announced in the Lok Sabha that his National Front government would implement the Mandal Commission’s central recommendation, reserving 27 percent of posts in the central government and public sector undertakings for the Other Backward Classes. The Mandal Commission, chaired by B. P. Mandal and appointed by the Janata government in 1979, had submitted its report on 31 December 1980, identifying 3,743 castes as socially and educationally backward and estimating them at 52 percent of the population. For a decade the report had sat unimplemented, a loaded clause in the founding bargain: the Constitution had authorised affirmative discrimination without fixing its beneficiaries beyond the Scheduled Castes and Scheduled Tribes, and the Congress governments of the 1980s had declined to extend the list. V. P. Singh, heading a minority coalition dependent on outside support from both the BJP and the Left, pulled the clause into force as an act of coalition survival and social conviction at once, and in doing so he changed the social basis of national politics permanently.

The Mandal Commission’s report of 31 December 1980 recommended 27 percent reservation in central government posts and public sector undertakings for the Other Backward Classes, identifying 3,743 castes at about 52 percent of the population. Added to the 22.5 percent for Scheduled Castes and Scheduled Tribes, the total stopped just below the 50 percent ceiling the courts had indicated.

The counter-mobilisation arrived within weeks and on the same terrain of religion and community. On 25 September 1990, BJP president L. K. Advani set out from Somnath in Gujarat on a rath yatra toward Ayodhya, where the Vishwa Hindu Parishad demanded the construction of a Ram temple on the site of the Babri Masjid. Bihar’s Chief Minister Lalu Prasad Yadav had Advani arrested at Samastipur on 23 October 1990; the BJP withdrew its outside support the same day; and V. P. Singh lost the confidence of the Lok Sabha on 7 November 1990. Mandal and Mandir, the two mobilisations, are usually narrated as opposites, the one fracturing the Hindu electorate by caste and the other uniting it by faith. The renegotiation reading sees them as twin consequences of a single event: the collapse of the Congress container, which had held caste and religious claims in suspension inside one organisation, released both claims into open competition, and no subsequent settlement of the religion-state tension could be written without answering both. The post-Mandal political system, in which caste parties govern the politics of the Hindi belt and a Hindutva party governs the politics of national symbolism, was the era’s joint product, and the two formations spent twenty-five years negotiating its terms.

The judiciary wrote the second draft of this renegotiation. On 16 November 1992, a nine-judge bench of the Supreme Court delivered the Indra Sawhney judgment, upholding the 27 percent OBC quota in central services by a six-to-three majority, excluding the creamy layer of the beneficiary castes, and reaffirming 50 percent as the ordinary ceiling of reservation. The judgment converted a coalition government’s survival measure into durable constitutional doctrine, and its terms governed every subsequent negotiation of the caste clause, including the Tamil Nadu exception and the later litigation over quotas in promotion. Three weeks after the judgment, on 6 December 1992, kar sevaks demolished the Babri Masjid at Ayodhya in the presence of senior BJP leaders, and the government of P. V. Narasimha Rao dismissed four BJP state governments in the aftermath. The demolition and the dismissals posed the religion-state question in its sharpest form since Partition: whether the Indian state would enforce its own constitutional description against a mobilised majority faith. The litigation that followed, the Liberhan Commission appointed on 16 December 1992 and reporting only in 2009, and the Allahabad High Court’s Ayodhya title judgment of 30 September 2010 form a chain of evidence that runs the length of the era; a later article in this series treats the demolition and its consequences in full.

The courts also renegotiated the vocabulary of the tension. On 11 December 1995, a Supreme Court bench led by Justice J. S. Verma delivered the Hindutva judgments, holding in the Manohar Joshi case and its companions that Hindutva, in the electoral context before the court, denoted a way of life rather than a religious appeal within the meaning of the Representation of the People Act. The ruling was read by its defenders, at their strongest, as a refusal to criminalise a civilisational idiom in campaign speech, and by its critics, at their strongest, as a judicial dilution of the secular discipline that the same court enforced elsewhere. Both readings are attributed positions in a live scholarly and legal dispute; what is settled is the judgment’s date, its bench and its place in the era’s sequence, where it stands as the judiciary’s most contested contribution to the religion-state renegotiation. The Gujarat violence of February and March 2002, in which over a thousand people, predominantly Muslims, were killed after the burning of a train coach at Godhra on 27 February 2002, marked the renegotiation’s darkest passage and its sharpest attribution dispute: the state government’s defenders and critics have contested the character of official conduct in that episode ever since, and the dispute is marked here as disputed, its documents examined in the phase narrative below. What the era settled, across Mandal, Ayodhya, Indra Sawhney and the Hindutva rulings, was not a doctrine but a distribution of power: the religion-state clause of the founding would henceforth be negotiated among caste parties, a national Hindu party, and a constitutional court, and no one of the three could write it alone.

Rights and Development Renegotiated: From the 1991 Reform to the Rights Statutes

The second tension, between the rights the Constitution guaranteed and the development the Directive Principles commanded, was renegotiated twice in the era, in opposite directions, and the double movement is the strongest single evidence for the renegotiation reading. The first movement began in insolvency. By June 1991, India’s foreign exchange reserves covered roughly two weeks of imports, gold had been pledged abroad as collateral, and the fiscal deficit of the centre had reached 8.4 percent of GDP in 1990 to 1991. P. V. Narasimha Rao took office on 21 June 1991 at the head of a minority Congress government; his finance minister, Manmohan Singh, presented the budget and the new Industrial Policy Statement on 24 July 1991. The Statement abolished industrial licensing for all but a short list of industries, opened sectors reserved for the public sector to private entry, raised the foreign direct investment ceiling to 51 percent in priority industries, and set the rupee on the path to convertibility. The licence-permit architecture through which the founding settlement had directed development, an architecture four decades old, was dismantled by executive action in a single monsoon session, by a government that lacked a majority and faced a parliament in which the Left and the BJP both, for opposite reasons, distrusted the programme.

Atul Kohli’s Poverty Amid Plenty in the New India, published in 2012, supplies the political economy of what followed and is the third voice in this article’s adjudication. Kohli argued that post-1991 India settled into a growth model in which the state allied itself with organised business, growth rates doubled, and the distribution of the gains narrowed even as poverty headcounts fell. The renegotiation reading absorbs Kohli’s evidence and disputes one inference: the market turn of 1991 did not close the rights-development tension; it relocated it. Because every government after 1991 depended on partners answerable to electorates that experienced liberalisation unevenly, the development clause acquired a permanent opposition bench inside the ruling arrangement itself. The Left parties that supported the United Front and later the UPA from outside made divestment conditional; the BJP’s Swadeshi Jagran Manch constituency slowed the NDA’s privatisation programme between 1998 and 2004; the Telugu Desam extracted fiscal terms for Andhra Pradesh as the price of its support after 1999. Reform proceeded, but it proceeded as negotiated text, and its pace and exclusions bear the coalition’s fingerprints as plainly as its inclusions.

The second movement reversed the polarity. The UPA government’s National Common Minimum Programme of 27 May 2004, drafted with the Left parties whose members sustained the government from outside, committed the coalition to an employment guarantee, to information rights and to food security as statutory entitlements. The Right to Information Act received assent in June 2005 and came into force on 12 October 2005, converting the citizen’s relation to the development state into an enforceable claim on its files. The National Rural Employment Guarantee Act, passed in August 2005 and launched on 2 February 2006, made one hundred days of wage employment a legal right in rural households, the largest single rewriting of the Directive Principles’ promise into justiciable form since land reform. The National Food Security Act of September 2013 extended subsidised grain entitlements to two-thirds of the population. Yogendra Yadav and Suhas Palshikar, in the essays collected in Indian Democracy: Meanings and Practices, read this phase as the democratic deepening of the third electoral system, in which the post-1989 assertion of the subaltern voter compelled the state to answer with rights. The renegotiation reading concurs, with one amendment: the rights statutes were not the negation of the 1991 settlement but its completion under coalition arithmetic, the development clause renegotiated a second time by the partners whom the first renegotiation had empowered.

Set the two movements side by side and the era’s method becomes visible. In 1991 a minority government dismantled the directed-development state because its partners and its creditors left it no alternative and because the crisis silenced the veto. Between 2004 and 2013, coalition governments built a rights-based welfare state because their partners demanded the statutes as the published price of support. Both movements were substantive; both rewired the relation between the Indian citizen and the Indian state; neither could have been enacted by the fragile arithmetic acting as a mere caretaker. The interlude reading, which sees only the falls of governments, cannot explain why the weakest decades produced the strongest statutory output since the 1950s. The renegotiation reading explains it directly: fragmentation distributed the initiative, and each holder of a fragment spent it on the clause of the settlement that mattered to its electorate.

The Material Republic: Highways, Telephones and the Consumer the Era Created

The renegotiation had a physical register that a hub article on the period must record, because the era’s statutes floated on a transformation of the material republic that altered what citizens expected of any settlement. In 1991 India had about five million telephone connections and a waiting list measured in years; the National Telecom Policy of 1994 opened basic services to private operators, the Telecom Regulatory Authority of India was created by statute in 1997, and the New Telecom Policy of 1999 moved operators from fixed licence fees to revenue sharing, the decision that detonated the mobile market. By March 2014 the country counted over nine hundred million telephone subscribers, the world’s second largest network, and the call charges that the state monopoly had set among the world’s highest had fallen among the world’s lowest. The political consequence ran deeper than convenience. The mobile telephone dissolved the information monopolies on which the old settlement’s mediation rested: a voter in 2009 received the campaign, the scandal and the rebuttal on a handset, unfiltered by the district officer, the party office or the state broadcaster, and every renegotiation after about 2004 was conducted before an electorate in continuous session.

The highways programme performed the same office on land. The National Highways Development Project, launched under the Vajpayee government with the Golden Quadrilateral connecting Delhi, Mumbai, Chennai and Kolkata sanctioned in 1999 and substantially built across the following decade, together with the Pradhan Mantri Gram Sadak Yojana of December 2000 for rural roads, bound the federal units into a single logistical space whose maintenance became a standing item of centre-state bargaining. Civil aviation, a state monopoly’s preserve in 1989, was opened to private carriers who by 2014 carried the great majority of domestic passengers; the information technology services industry, negligible in 1991, exported software and services on a scale that made Bengaluru and Hyderabad synonyms for the new economy and gave the states a development competition of their own. Parliament codified the investment frontier in the Special Economic Zones Act of 2005, and the land conflicts that followed, at Singur in West Bengal in 2006 to 2008 and at Nandigram in March 2007, where police firing killed fourteen protesters, demonstrated the era’s method under stress: the Left Front government of thirty years was broken not by an opposition alliance but by its own attempt to acquire farmland for industry, and the Trinamool Congress that replaced it in May 2011 had ridden the resistance to power. The development clause, in the material republic, had acquired a veto-holder the founding never seated: the smallholder on the plot to be acquired, whose assent every coalition after Singur priced into its land policy and whose claims the 2013 land statute attempted, imperfectly, to codify.

The Third Tier: Panchayati Raj and the Amendment Trail of the Era

The era’s most literal rewriting of the founding text is the one its political narratives mention least, and the hub restores it to the centre of the record. The Constitution (Seventy-third Amendment) Act and the Constitution (Seventy-fourth Amendment) Act, passed by the Rao government’s parliament in December 1992 and brought into force on 24 April 1993 and 1 June 1993 respectively, inserted Parts IX and IXA into the Constitution and created a third tier of elected government: panchayats in the villages and municipalities in the towns, with elections made compulsory every five years under state election commissions, seats reserved for Scheduled Castes and Scheduled Tribes in proportion to population, and not less than one-third of all seats and chairpersonships reserved for women. Roughly three million representatives took elected office under the amendments across the era, about a million of them women, the largest single expansion of the republic’s elected class since universal franchise. The amendments were the federal renegotiation conducted downward: the states, jealous of their own autonomy against Delhi, were compelled by constitutional text to share power with the tier beneath them, and the compulsion was the work of a minority Union government legislating the legacy of Rajiv Gandhi’s failed amendment bills of 1989, which had fallen in the Rajya Sabha in October of that year. West Bengal and Karnataka, whose panchayat systems predated the amendments under Left Front and Janata governments respectively, supplied the working models; the era’s contribution was to make the third tier universal, electoral and reserved.

The amendments that followed trace the era’s three tensions in constitutional shorthand. On caste, the Seventy-seventh Amendment of 1995 restored reservation in promotions for Scheduled Castes and Scheduled Tribes against the Indra Sawhney bar, the Eighty-first of 2000 permitted the carry-forward of unfilled reserved vacancies beyond the 50 percent ceiling’s annual operation, and the Eighty-fifth of 2001 added consequential seniority, a sequence that shows the political system amending the constitutional text repeatedly to defend the Mandal-era settlement against its judicial qualifications, each amendment carried by coalition-era parliaments with near-unanimity. On development and rights, the Eighty-sixth Amendment of 2002 inserted Article 21A, obliging the state to provide free and compulsory education to children aged six to fourteen, the promise the Right of Children to Free and Compulsory Education Act of 2009 then operationalised, enrolling the education clause in the same rights-based renegotiation as the information and employment statutes. On the federal clause, the Eighty-fourth Amendment of 2001 extended the freeze on readjustment of Lok Sabha and assembly constituencies until the first census after 2026, entrenching the bargain by which the states that had curbed population growth would not lose parliamentary weight to those that had not, a coalition-era consensus renewed precisely because no bloc could afford to reopen it. Read as a single trail, the amendments of 1992 to 2011 refute the interlude reading from the statute book’s hardest evidence: in twenty-five years of supposedly transitional politics, parliament amended the founding text to create a third tier of democracy, to defend and extend the caste settlement, to constitutionalise a child’s right to schooling and to postpone the republic’s most explosive federal question by common consent. Transitional regimes do not leave amendment trails of this weight; renegotiating ones do.

One renegotiation of the era is routinely omitted from political histories because its theatre was a committee room, and the hub restores it because it is the federal clause’s most successful amendment in practice before the GST. India’s states in 1989 taxed goods through a thicket of sales taxes, octroi and entry levies whose cascading distortions the Kelkar committees and the finance commissions had condemned for decades. The remedy, a state-level value added tax replacing the sales tax, required every state government, of every party, to surrender the rate competition by which states had bid for trade, and to trust a common design administered by the states collectively. The vehicle was the Empowered Committee of State Finance Ministers, constituted in 2000 under the NDA and chaired by Asim Dasgupta, the finance minister of Left Front West Bengal, a communist minister chairing the era’s largest market reform on behalf of a BJP-led Union: the coalition era’s federalism in a single tableau. Haryana adopted VAT in 2003; on 1 April 2005 the great majority of states followed, on a common floor design with compensation assurances from the centre for revenue losses in the transition. The reform held across alternations of government in state after state, it raised the states’ own-revenue buoyancy it had promised, and it built, in the Empowered Committee’s habits of consensus, the precise institution the GST Council later formalised. Beside it ran the fiscal discipline legislation of the same years: the Union’s Fiscal Responsibility and Budget Management Act of 2003, the Twelfth Finance Commission’s debt relief scheme that conditioned assistance on states enacting their own fiscal responsibility laws, which nearly all did by 2010, and the Thirteenth Finance Commission’s award raising the states’ share of central taxes to 32 percent. The public finance of the republic was rewritten, in the era of its weakest governments, more thoroughly than in any decade since the founding, and rewritten by negotiation among the states themselves rather than by imposition from the centre. The growth the fiscal order underwrote, averaging above 8 percent a year in the boom of 2003 to 2011 and carrying the official poverty ratio, on the Tendulkar Committee’s method, from 37.2 percent in 2004 to 2005 down to 21.9 percent in 2011 to 2012, supplied the revenues from which the rights statutes of the same years were paid. The era’s development clause and its federal clause were, in the committee rooms, the same clause.

The Federal Bargain Renegotiated: The Veto Migrates to the States

The third tension, between the Union and its units, underwent the era’s most complete renegotiation, because the coalition arithmetic changed who held power in New Delhi and not merely what that power did. The founding settlement had equipped the centre with formidable instruments over the states, and the federalism of the Indian Constitution as designed in 1950 tilted toward the Union in finance, in legislation and in the emergency power of Article 356. The Congress system had operated that tilt as a party matter: chief ministers were made and unmade by the high command. The coalition era replaced party discipline with arithmetic dependence. From 1996 onward, no national government could be formed or sustained without regional parties, and regional parties price their support in the currency of state autonomy. The federal clause was accordingly rewritten in three registers: judicial, fiscal and governmental.

The judicial register came first and cut deepest. In S. R. Bommai versus Union of India, decided on 11 March 1994, a nine-judge bench of the Supreme Court held that a state government’s majority must be tested on the floor of its assembly and not in the Governor’s assessment, that a proclamation under Article 356 was subject to judicial review, and that secularism formed part of the Constitution’s basic structure. The judgment was delivered against a docket of dismissals that included the BJP state governments removed after Ayodhya, and its effect was immediate and measurable: invocations of President’s Rule, which had numbered near one hundred in the four decades before 1994, fell sharply thereafter, and no coalition government at the centre after 1996 attempted the mass dismissals that Indira Gandhi’s governments had practised in 1977 and 1980. A constitutional clause that had functioned as the centre’s veto over state electorates became, by judicial redraft, a narrow emergency provision, and it became so in the era when the centre was least able to resist the change, because the centre’s own constituents governed the states that the clause had been used against.

The governmental register is documented in an under-cited archive that this hub article flags for the pillar’s later treatments: the correspondence of the coalition negotiations of 1996 to 1998, preserved in the archives of the partner parties, including the Telugu Desam’s records at Hyderabad and the Dravida Munnetra Kazhagam’s records at Chennai. These letters and minutes record the actual mechanisms of the new federalism, the allocation of portfolios to regional parties, the consultation conventions on state subjects, the price lists of support. Students of the period who work through the documentary mass of the era, comparing common minimum programmes against cabinet decisions and state finance records, will find that the files repay assembly into a single research corpus; readers assembling such a corpus from scanned agreements and tabulated programme texts can open the documents and spreadsheets through the Office File Viewer and work across the formats in the browser. What the correspondence shows, and what the published programmes confirm, is bargaining of a kind the founding generation did not anticipate: chief ministers negotiating with prime ministers as principals, not as subordinates, with the survival of the Union government as their security.

The fiscal register completed the movement. The states’ share of central taxes was fixed by successive Finance Commissions whose awards coalition governments accepted without the creative reclassification earlier decades had practised. More consequentially, the era incubated the goods and services tax, the most ambitious fiscal-federal redesign since 1950. The Kelkar Task Force on the implementation of the Fiscal Responsibility and Budget Management Act recommended a national GST in 2004; the Empowered Committee of State Finance Ministers, chaired from 2000 by West Bengal’s finance minister Asim Dasgupta, spent a decade negotiating rates and compensations among the states themselves; and the Constitution (One Hundred and Fifteenth Amendment) Bill to enable the tax was introduced in the Lok Sabha in March 2011 by the UPA government. The tax was enacted after the era closed, and a later article in this series examines the GST Council in full, but its architecture, a council in which the states collectively outvote the centre and decisions proceed by consensus, was designed inside the coalition era’s federal practice and is unintelligible without it. The era also redrew the map the bargain governed: on 1 November 2000 Chhattisgarh was carved from Madhya Pradesh, on 9 November 2000 Uttaranchal from Uttar Pradesh, and on 15 November 2000 Jharkhand from Bihar, each by act of a coalition parliament responding to long-standing regional demands that the Congress system had contained. A polity renegotiating its federal clause does not merely redistribute power over its units; in the end it revises the units themselves.

The Era in the States: Where the Renegotiation Was Governed From

A history of the coalition era written only from New Delhi mistakes the location of its politics. The era’s governments were negotiated in Delhi but they were produced in the states, and several state governments of the period ran longer and governed more consequentially than any Union arrangement. West Bengal’s Left Front, in office from 1977, governed until May 2011, a thirty-four-year tenure that made it the era’s standing refutation of the claim that the period knew only instability; under Jyoti Basu until November 2000 and Buddhadeb Bhattacharjee thereafter, the Front administered the land reforms of Operation Barga, sustained the state-level coalition discipline the centre could not imitate, and in 1996 produced the era’s most famous might-have-been, when the CPI(M)’s central committee declined the United Front’s offer of the prime ministership to Basu, a refusal Basu himself later described, in a much-quoted characterisation, as a historic blunder. The episode belongs in the hub’s record because it shows the era’s bargains being struck inside parties as well as between them: the Left’s doctrine of outside support, which governed its conduct in 1996 and again after 2004, was a state party’s theory of the centre.

Andhra Pradesh supplied the era’s model of the chief minister as national principal. Chandrababu Naidu, Chief Minister from September 1995 to May 2004, converted Hyderabad into an information technology hub, bargained his Telugu Desam’s outside support to the NDA into fiscal and institutional concessions, and convened the chief ministers of the federal opposition into a standing lobby on centre-state questions. Tamil Nadu’s two Dravidian parties, the DMK under M. Karunanidhi and the AIADMK under J. Jayalalithaa, alternated in Chennai and dealt with Delhi as sovereign counterparties: the AIADMK’s withdrawal destroyed a Union government in April 1999, and the DMK’s portfolio claims inside the UPA after 2004 issued, by the prosecution’s case and the CAG’s account, in the spectrum allocations that became the era’s largest scandal, an attribution the DMK contested throughout and which the trial court’s 2017 acquittals left unresolved in law while its political consequences ran on. Bihar traversed the era’s whole arc in one state: the Mandal heartland under Lalu Prasad Yadav from 1990, whose governments made social justice the state’s official idiom and whose administrative record his critics prosecuted as the era’s governance failure and his defenders praised as the dignity of the previously excluded, followed from November 2005 by Nitish Kumar’s JD(U) and BJP coalition, which campaigned on the restoration of administration itself and was returned in 2010 with a large majority. Gujarat under Narendra Modi from October 2001 presented the mirror case, a state government that converted the controversy of 2002 into an electoral fortress and a development claim, contested by critics on the character of the state in the violence and on the inclusiveness of the growth, and carried by its holder, in September 2013, to nomination as the BJP’s prime ministerial candidate.

Uttar Pradesh, the state whose eighty seats had once manufactured Congress majorities, became the era’s purest state arena: the Samajwadi Party and the Bahujan Samaj Party alternated in office through the 1990s, Mulayam Singh Yadav and Mayawati each formed governments with and against the BJP, and in 2007 the BSP won an outright assembly majority, the state’s first in the era, on a coalition of castes its founder Kanshi Ram had spent two decades assembling. Maharashtra’s Shiv Sena and BJP governed from 1995 to 1999, and the Nationalist Congress Party, founded by Sharad Pawar in June 1999 after his expulsion from the Congress, then shared power in Mumbai with its parent party for fifteen years while Pawar held the agriculture ministry in Delhi for a decade. The pattern generalises beyond the examples. The era’s Union was a federation of state verdicts before it was anything else, its prime ministers were made by chief ministers’ arithmetic, and its settlements were drafted by politicians whose careers alternated between state and national office. The pillar’s later treatments of the period’s national episodes presuppose this geography at every step.

The Periphery Renegotiated: Kashmir and the Northeast Under Coalition Arithmetic

The founding settlement’s federal clause had always run thinnest at the periphery, where accession, insurgency and the armed forces’ special powers formed a sub-settlement of their own, and the coalition era renegotiated that sub-settlement by the same method it applied at the centre: negotiation with armed and electoral principals whom the old system had tried to administer. Kashmir consumed the era’s opening. The insurgency that erupted in the Valley in 1989, after the assembly election of 1987 whose manipulation by the National Conference and Congress alliance remains the hinge on which Kashmiri accounts of the descent turn, drove the exodus of Kashmiri Pandits in early 1990 and a decade of militancy and counter-insurgency in which tens of thousands died, their enumeration contested between official and movement counts and marked here as disputed. The coalition-era state rebuilt electoral process in stages: assembly elections in 1996 restored a National Conference government under Farooq Abdullah, the coalition of the People’s Democratic Party and the Congress after 2002 introduced the rotation of the chief ministership between the Valley and Jammu and Mufti Mohammad Sayeed’s healing touch idiom, and the assembly election of 2008 proceeded despite the Amarnath land-transfer agitation that summer, in which the allocation of forest land to a shrine board, its revocation, and the protests and counter-protests across the state’s two regions rehearsed every tension of the founding inside one province. No settlement was reached; the era’s achievement in Kashmir was confined to the restoration of electoral alternation and the containment of the conflict’s spread, and the hub records it as the renegotiation’s clearest limit, the clause where bargaining never found its parties simultaneously at the table.

The Northeast records the method’s fuller application. The coalition governments of the period governed the region through a lattice of ceasefires, accords and autonomous councils inside the Constitution’s Sixth Schedule and its state-specific provisions. The ceasefire with the National Socialist Council of Nagaland (Isak-Muivah), announced in July 1997 under the United Front and maintained by every successor government, suspended one of Asia’s oldest insurgencies pending a political settlement whose negotiation outlasted the era. The Bodo Accord of 10 February 2003, signed under the NDA, created the Bodoland Territorial Council within Assam and converted an armed movement into an elected administration. Tripura’s insurgency was wound down across the 2000s by a combination of policing, development expenditure and the political absorption of tribal formations, and Mizoram’s accord of 1986, the era’s inheritance rather than its product, stood throughout as the template the negotiators cited: a settlement in which former insurgents won and held office through elections. The Armed Forces (Special Powers) Act, whose application to Manipur provoked Irom Sharmila’s fast from November 2000 and whose repeal a government-appointed committee under Justice B. P. Jeevan Reddy recommended in 2005 without effect, records the sub-settlement’s unrepealed core. The periphery thus confirms the hub’s claim in its qualified form: where principals existed who could sign and deliver, the era’s coalition method produced accords that the centralised decades had not; where the principals’ terms remained irreconcilable within the settlement’s text, the era administered the conflict and passed its clause, like the women’s bill and the Lokpal, into the inheritance of the majority age.

Phase One, 1989 to 1998: Polarisation, Reform and the Coalition Experiments

The era’s first phase compressed a generation of political change into nine years. The general election of November 1989 returned a Lok Sabha in which the Congress, with 197 seats, remained the largest party yet lost the ability to govern; the Janata Dal won 143 seats, the BJP 85, and the communists and regional parties held the balance. V. P. Singh took office on 2 December 1989 at the head of the National Front, a coalition of the Janata Dal and regional parties sustained from outside by two formations that agreed on nothing else, the BJP and the Left Front. The government lived eleven months and in that span lit both fuses of the decade: Devi Lal’s dismissal and the internal Janata Dal feuds consumed its summer, the Mandal announcement of 7 August 1990 defined its legacy, and the BJP’s withdrawal after Advani’s arrest brought its fall on 7 November 1990. Chandra Shekhar, who split the Janata Dal with sixty-four members, governed from 10 November 1990 to March 1991 on Congress support, which Rajiv Gandhi withdrew over a surveillance complaint in March 1991, sending the country to a mid-term election in the middle of which, on 21 May 1991, Rajiv Gandhi was assassinated at Sriperumbudur.

When was the last single-party majority before 2014?

The last single-party majority before 2014 was won in December 1984, when the Congress under Rajiv Gandhi took 414 of the seats contested in the eighth Lok Sabha, the largest tally any single party had recorded. The verdict followed Indira Gandhi’s assassination by eight weeks, and no party repeated a majority of its own for the following thirty years.

The election of 1991 returned the Congress with 244 seats, short of a majority, and P. V. Narasimha Rao, who had not contested the election, became Prime Minister of a minority government on 21 June 1991. The Rao government is the phase’s standing rebuke to the instability thesis. It survived its full five years, purchasing survival vote by vote, and in those five years it enacted the liberalisation of July 1991, survived the Ayodhya demolition and its aftermath, lost and regained states, and managed a foreign policy reorientation that included full diplomatic relations with Israel in January 1992 and the Look East policy. Its methods of survival, chronicled in the Jharkhand Mukti Morcha bribery case that reached judgment years later, disgraced its arithmetic without diminishing its output. The lesson the political class drew from 1991 to 1996 was precise: a minority government could govern and could legislate the largest economic change since Independence, provided it treated every division as a negotiation. That lesson, learned by the Congress in adversity, became the operating system of every government that followed.

The election of 1996 dissolved even that arrangement. The BJP emerged largest with 161 seats; Atal Bihari Vajpayee formed a government on 16 May 1996 and resigned on 28 May, before the confidence vote he could not win, after thirteen days in office. The United Front, a coalition of thirteen parties anchored by the Janata Dal and the regional formations, took office under Karnataka’s Chief Minister H. D. Deve Gowda on 1 June 1996, sustained from outside by the Congress under its new president Sitaram Kesri. The Front’s Common Minimum Programme, the era’s second great coalition text, codified the federal bargain of the phase: no unilateral dismissal of state governments, a consultative centre, and a go-slow consensus on economic policy that kept divestment and labour reform hostage to the coalition’s Left constituents. When the Congress withdrew support in March 1997, the Front changed its leader rather than fall: I. K. Gujral replaced Deve Gowda on 21 April 1997. When the Congress demanded the removal of DMK ministers after the Jain Commission’s interim report on the Rajiv Gandhi assassination was leaked in November 1997, and Gujral refused, the Congress withdrew support a second time, on 28 November 1997, and the Lok Sabha was dissolved on 4 December 1997. Phase One thus ended with an experiment half-judged. The United Front had demonstrated that regional parties could run the Union for twenty-one months and conduct a respectable foreign policy, the Gujral Doctrine of unilateral accommodation toward smaller neighbours among its products. It had also demonstrated that outside support was a wasting asset, because the supporter bore no cost for withdrawal. The governments of Phase Two were designed, consciously, against that defect.

Phase Two, 1998 to 2004: The Vajpayee NDA and the Discipline of the Pre-Poll Alliance

The general election of February and March 1998 returned the BJP with 182 seats and a pre-poll alliance, the National Democratic Alliance, with the numbers to govern. The institutional innovation of 1998 deserves emphasis, because it is the phase’s contribution to the era’s renegotiation of method. The NDA fought the election on a negotiated text, the National Agenda for Governance, in which the BJP’s three signature commitments, a Ram temple by legislation, the abrogation of Article 370 and a uniform civil code, were expressly deferred for the life of the coalition. The Shiv Sena, the Akali Dal, the Telugu Desam (supporting from outside), the DMK and a dozen smaller formations signed a programme that bound the largest party’s ideology in exchange for its allies’ seats. Critics named the arrangement opportunism; defenders, at their strongest, named it federal maturity, the subordination of one party’s creed to a common text that state-level electorates could read. The attribution matters less than the mechanism: for the first time since 1989, the coalition agreement preceded the election, and the voters, not only the legislators, were its parties.

How many seats did the BJP win in the 1989 general election?

The BJP won 85 seats in the election of November 1989, rising from the two seats it had held after 1984, on a vote share of about 11 percent. The tally made the party a holder of the balance for the first time, able to sustain the National Front from outside and, by withdrawing in October 1990, to end it.

The first NDA government, sworn in on 19 March 1998, tested the new discipline under fire. In May 1998 it conducted the Pokhran-II nuclear tests, renegotiating India’s external posture in the era’s boldest single act of state. Through the year its largest ally, the AIADMK under J. Jayalalithaa, pressed demands the coalition could not meet, and her withdrawal in April 1999 brought the government to a confidence vote on 17 April 1999, which it lost by a single vote, 269 to 270. The caretaker government then fought and won the Kargil war of May to July 1999, evicting Pakistani intrusions from the heights above the Srinagar-Leh highway at a cost of over five hundred Indian soldiers, and went to the electorate with the war as its credential. The election of September and October 1999 returned the BJP again at 182 seats and the NDA, including the DMK, the Trinamool Congress, the JD(U) and the Biju Janata Dal, at a combined strength of about 303. The government formed on 13 October 1999 served until the dissolution of February 2004, the first non-Congress government in Indian history to complete effectively a full term.

The renegotiation content of Phase Two runs through each of the three tensions. On religion and the state, the deferred agenda held: the NDA’s text restrained the BJP’s core commitments for six years, and the restraint broke only at the era’s extremities, in the Gujarat violence of 2002, where the state government’s conduct remains contested between its defenders and its critics and where the Prime Minister’s own public counsel of rajdharma at Ahmedabad in April 2002 entered the record as the coalition’s internal rebuke. On rights and development, the NDA extended the 1991 settlement by negotiation: the Insurance Regulatory and Development Authority Act of 1999 opened insurance, disinvestment proceeded ministry by ministry against Swadeshi resistance inside the alliance’s own family, and the Fiscal Responsibility and Budget Management Act of 2003 wrote the era’s fiscal consensus into statute. On the federal tension, the phase delivered the three new states of November 2000 and governed throughout in dependence on allies who were themselves state governments. When the electorate dismissed the NDA in May 2004, rejecting the India Shining campaign and returning the Congress with 145 seats against the BJP’s 138, the dismissal itself was processed within the new conventions: power passed, on a written programme, to another pre-arranged coalition. The corridor, by 2004, had acquired the furniture of a house.

Phase Three, 2004 to 2009: The UPA Shock and the Rights-Based Redistribution

The verdict of May 2004 was unexpected by every exit poll and by the incumbent alliance, and its processing displayed the era’s mature conventions at their best. The Congress, with 145 seats, assembled the United Progressive Alliance, negotiated the National Common Minimum Programme with the Left parties and the regional allies, and accepted Manmohan Singh as Prime Minister when Sonia Gandhi declined the office on 18 May 2004. The government sworn in on 22 May 2004 rested on a coalition inside the Lok Sabha and on the outside support of the Left parties, whose leverage was written into the programme’s text: no privatisation of profitable public sector units, no labour law dilution, and a development programme weighted toward the rural electorate whose revolt against India Shining had delivered the verdict. Where Phase Two’s innovation was the pre-poll alliance, Phase Three’s was the published price of outside support, and the price, as the rights section of this article records, was a statute book.

Which coalition governed India for the longest continuous stretch of the era?

The United Progressive Alliance under Manmohan Singh governed continuously from 22 May 2004 to 26 May 2014, ten years and two full parliamentary terms. No other arrangement of the coalition era approached that span: the Vajpayee governments totalled about six years across three spells, and every other government of the period lasted under three.

The phase’s crisis tested whether the new conventions could survive the withdrawal of their guarantor. The Indo-United States civil nuclear agreement, initialled in July 2005 and carried through the Hyde Act of December 2006 and the separation plan of 2006, collided with the Left’s opposition to strategic alignment with the United States. On 8 July 2008 the Left parties withdrew their support, reducing the government to a minority, and on 22 July 2008 the government won the ensuing confidence vote by 275 votes to 256, with support assembled from former opponents including the Samajwadi Party. The episode is instructive for the renegotiation reading. The government’s survival no longer depended on sentiment or on the terror of dissolution that had disciplined earlier decades; it depended on the capacity to re-price support issue by issue, and the coalition that emerged from July 2008 was a renegotiated coalition, its programme amended by arithmetic rather than by election. The electorate then renewed it. In the general election of April and May 2009, the Congress rose to 206 seats, the best single-party tally since 1991, and the UPA returned with about 262 seats and a reduced dependence on any single ally. Phase Three closed with the era’s central paradox in full view: the coalition system, designed by its critics’ account to produce drift, had produced two durable governments, a rights statute book and an electorate that rewarded incumbency for the first time since 1984.

Phase Four, 2009 to 2014: Strain, Mobilisation and the 2014 Inflection

The fourth phase is the one the interlude reading cites most confidently, and the renegotiation reading must take its evidence whole. The second UPA government, freed of the Left’s discipline and dependent on allies whose demands ran to portfolios rather than programmes, lost the era’s governing technique. The Commonwealth Games of October 2010 were preceded by corruption findings against their organisers; the Comptroller and Auditor General’s report on the 2G spectrum allocation, tabled in November 2010, estimated a presumptive loss of 1.76 lakh crore rupees from licences issued in January 2008 on a first-come-first-served basis by the DMK’s A. Raja at the telecom ministry, a figure whose method critics of the estimate contested and whose political effect nobody contested. The Supreme Court cancelled 122 licences on 2 February 2012. Anna Hazare’s fasts for a Jan Lokpal, at Jantar Mantar in April 2011 and at Ramlila Maidan in August 2011, mobilised an urban constituency against the political class as such, and the government’s negotiators conceded a parliamentary resolution on the bill’s principles on 27 August 2011. Between the scams and the street, the characterisation of policy paralysis attached itself to the government, pressed by the opposition and by sections of business and the press; the government’s defenders answered with the global slowdown and the coalition’s compulsions, and the dispute over the correct characterisation of 2011 to 2013 is marked here as disputed, its resolution belonging to the detailed treatments of the pillar.

What is not disputed is the phase’s terminal fact. The general election of April and May 2014 returned the BJP with 282 seats, a single-party majority, on 31 percent of the vote, and the Congress with 44 seats, its lowest tally in its history. The renegotiation reading insists on precision about what the verdict did and did not close. It closed the era of arithmetic: for the first time since 1984, a government could be formed without negotiating a programme with regional principals. It did not repeal the era’s renegotiations. The reservation doctrine of Indra Sawhney stood untouched; the rights statutes of 2005 to 2013 remained the law the new government administered; the Bommai discipline on Article 356 bound the new centre as it had bound its predecessors; and the GST architecture, negotiated among the states for a decade inside the era, awaited enactment by the majority the era had made possible. The 2014 verdict is therefore best read, on this article’s claim, as the opening move of a further renegotiation rather than as the restoration of the pre-1989 world. The victorious party itself framed its mandate in those terms, promising to govern a settlement it had inherited from twenty-five years of coalition bargaining and intended to revise, which is precisely the concession the renegotiation reading requires: the era had built something durable enough to be worth revising.

The phase also bequeathed the era’s unfinished business, and a hub article should inventory it honestly. The Lokpal statute, the era’s answer to its corruption crisis, received assent in January 2014, at the very close of the period, with its institution unbuilt. The land acquisition settlement of 2013 satisfied neither the industry that found it prohibitive nor the movements that found it permissive. The communal violence bill, drafted and redrafted by the National Advisory Council after 2004, was never enacted, and the religion-state tension it addressed remained where Mandal, Ayodhya and Gujarat had left it: distributed among parties, courts and state governments, awaiting the next arithmetic. The inventory supports rather than embarrasses this article’s claim. A settlement renegotiated is not a settlement concluded, and the coalition era’s achievement was to move all three founding tensions into new institutional configurations, not to retire any of them.

Two Organisational Stories: The Congress Decline and the BJP Ascent

The era’s arithmetic was the surface of two organisational trajectories running in opposite directions, and the hub records them because the renegotiation’s parties were these organisations and no others. The Congress trajectory is a descent with one recovery. Its Lok Sabha tallies across the era’s eight elections read 197, 244, 140, 141, 114, 145, 206 and 44. The party that had contained the founding tensions inside itself entered the era without the machinery of containment: state units led by nominees, a working committee that ratified rather than decided, and a leadership whose authority ran downward from a family rather than upward from an organisation. The 244 of 1991 was won under assassination’s shadow and husbanded by Rao’s parliamentary craft; the recoveries of 2004 and 2009 were assembled by Sonia Gandhi, who took the party presidency in April 1998 after Sitaram Kesri’s removal, through pre-poll alliances and the patience of a leader who declined the premiership in 2004 and thereby made her party’s coalition governable. Analysts divide on the family’s role: at their strongest, its defenders treat the Nehru-Gandhi line as the only coordination device a party of India’s scale possessed after its organisation decayed, and its critics treat the dependence on that device as the decay’s cause rather than its remedy. The dispute is marked as disputed. The figures are not: a party reduced to 44 seats in 2014, absent from power in Uttar Pradesh and Bihar for the entire era, had ceased to be the settlement’s container and become one claimant among several, which is the precondition of everything else in this article.

The BJP’s trajectory inverts it: 85, 120, 161, 182, 182, 138, 116 and finally 282. The party grew by absorbing the era’s lessons in sequence. In the 1980s it was a cadre party with a movement’s reach and a legislator’s weakness, two seats in 1984. The Ayodhya mobilisation carried it to 85 in 1989 and 120 in 1991, at the price of confirming every other party’s refusal to sit with it, the untouchability that destroyed its thirteen-day government in 1996. The NDA years were its apprenticeship in governability: the deferred agenda of 1998, the full term of 1999 to 2004, the management of allies from the Dravidian parties to the Akalis, taught the party to convert movement energy into coalition text. Its defeats of 2004 and 2009, at 138 and 116, taught the complementary lesson that coalition text without a national campaign’s momentum could not hold the centre. The organisational inheritance it carried into 2013, a cadre base, a family of affiliates, a corps of full-time workers and a Gujarat government as its exhibit, differed from the Congress inheritance in the one respect the era rewarded: it could reproduce its leadership and its message in every state arena at once. When the party named the Gujarat Chief Minister its prime ministerial candidate on 13 September 2013 and fought the 2014 election as a presidential campaign around a single name, it executed, under majority arithmetic, the personality-centred method that the Indira decades had pioneered and the coalition era had spent twenty-five years constraining. The pillar’s framing article closes with that irony intact: the era ended when one party relearned, at national scale, the trick the era had been invented to prevent, and relearned it inside a settlement the era had rebuilt to withstand it.

The Era’s Unfinished Institutions: Representation, Integrity and the Referee

A hub article owes its readers the inventory of what the era attempted and failed to finish, because the failures map the limits of renegotiation by arithmetic. The largest is the representation of women. A constitutional amendment reserving one-third of seats in the Lok Sabha and the state assemblies for women was first introduced in September 1996 under the United Front, examined by a joint committee chaired by Geeta Mukherjee, and reintroduced by successive governments, each coalition discovering that its own partners held the veto: the parties of the Mandal heartland demanded sub-quotas for backward caste women as their price, a position their spokesmen defended as the refusal of a reform whose benefits would pool among the already privileged and whose opponents dismissed as a wrecking condition. The bill passed the Rajya Sabha on 9 March 2010, in a sitting from which marshals removed protesting members, and was never taken up by the Lok Sabha; it lapsed with the fifteenth Lok Sabha in 2014. The episode is the renegotiation mechanism photographed in negative: a reform commanding an apparent national consensus failed for eighteen years because the era distributed vetoes to precisely the actors the reform’s design did not accommodate.

The integrity institutions trace the same curve. India entered the era without a national ombudsman, the Lokpal bills having failed in every Lok Sabha since 1968, and exited it with the Lokpal and Lokayuktas Act, which received presidential assent on 1 January 2014 after the Hazare mobilisations had made the statute a coalition government’s ransom, and with the institution itself unconstituted. The judiciary’s own settlement shifted inside the era and held: the collegium system of appointments, created by the Supreme Court’s judgment of 6 October 1993 in the Second Judges Case and refined by the advisory opinion of 1998, governed judicial appointments throughout the period, an arrangement every government chafed under and none could assemble the constitutional majority to amend. The Election Commission, by contrast, finished an institutional ascent inside the era: under T. N. Seshan, Chief Election Commissioner from December 1990 to December 1996, the Commission enforced the model code of conduct against state governments, staggered polls across phases for security, and issued the identity card programme, converting a referee the founding had left weak into the era’s most trusted umpire; the first general election conducted entirely on electronic voting machines followed in 2004, and the Supreme Court’s disclosure judgments of 2002 and 2003, in the PUCL and Association for Democratic Reforms litigation, compelled candidates to publish their assets, liabilities and criminal antecedents. The pattern across the three cases disciplines the hub’s claim once more. Where renegotiation required only the assent of sitting governments and courts, the era finished its work; where it required the holders of seats to legislate against the composition of their own houses, the era’s vetoes governed, and the work passed unfinished to the majority age.

The Instruments of Renegotiation: Programmes, Presidents and Confidence Votes

Between the phase narrative and the adjudication, the hub records the era’s machinery, because the renegotiation was effected through instruments the founding generation had not designed and the era standardised by use. The first instrument was the common programme itself, which evolved across the period from the National Front’s brief manifesto of 1989 through the United Front’s Common Minimum Programme of June 1996 and the NDA’s National Agenda for Governance of March 1998 to the UPA’s National Common Minimum Programme of 27 May 2004. Each successive text was longer, more fiscal and more juridical than its predecessor, and each functioned as a coalition constitution: the document against which partners measured defections of policy, the document the Prime Minister cited against his own party’s maximalists, the document the electorate could read. Scholars of the period who compare the four texts possess, in effect, four drafts of the founding settlement’s operative clauses, annotated by the parties who had to live under them.

The second instrument was the presidency of the Republic, an office the coalition era promoted from ceremony to arbitration. President R. Venkataraman managed the government-making of 1989 and 1991 under conventions he largely wrote as he went, inviting the leader best placed to prove a majority and fixing the proof to a date. President Shankar Dayal Sharma invited the BJP, as the largest party, to form the government in May 1996 and accepted its fall thirteen days later. President K. R. Narayanan refined the office’s defensive powers: in October 1997 he returned for reconsideration the Gujral cabinet’s advice to impose President’s Rule in Uttar Pradesh, and in September 1998 he returned the Vajpayee cabinet’s advice on Bihar, and both cabinets let the advice drop. The episodes matter to the federal renegotiation because they show the era’s restraints operating through constitutional offices as well as through arithmetic: a President who could return advice, a court that could void proclamations, and a centre that lacked the numbers to override either. By 2004 and 2009, the conventions were settled enough to be boring, which is the mark of an instrument that has become an institution.

The third instrument was the confidence vote itself, professionalised across the era into scheduled renegotiation. The votes of November 1990, May 1996 (declined), April 1999 and July 2008 mark the era’s hinges, and their management changed character: the 1990 vote was a reckoning, the 1999 vote a one-vote accident that dissolved a house, the 2008 vote a priced transaction conducted in daylight against a published policy. Supporting the votes stood the anti-defection discipline of the Tenth Schedule, tightened by the Constitution (Ninety-first Amendment) Act of 2003, which capped the council of ministers at 15 percent of the strength of the Lok Sabha and abolished the recognition of splits, thereby abolishing at once the era’s two currencies of survival, the jumbo ministry and the engineered defection. A system that begins by falling monthly and ends by outlawing its own methods of falling has not been marking time. It has been legislating its experience into structure, which is what institutions do when a renegotiation succeeds.

Beneath the cabinet arithmetic ran a social transformation that explains why the renegotiations stuck, and the hub article records it because the pillar’s later treatments presuppose it. Yogendra Yadav named the phenomenon in his analysis of the 1996 election and the essays that followed: a second democratic upsurge, succeeding the first upsurge of the 1960s in which the upper castes had led electoral mobilisation, in which the participation rates of the Other Backward Classes, the Dalits, the Adivasis and women rose past the rates of the social groups that had founded the republic’s electorate. The evidence sits in the turnout tables of the Election Commission’s statistical reports across the era: the old gradient, in which the privileged voted more than the plebeian, flattened and in state after state inverted. Politics in the 1990s became, in Yadav’s phrase, plebeian in its personnel as well as its participation. The Lok Sabhas of the era returned rising proportions of OBC members, regional parties carried social coalitions into the Union cabinet that the Congress system had represented only by nomination, and the description of Indian democracy as an elite compact administered downward ceased to fit the observable facts.

The renegotiation reading draws one disciplined inference from the upsurge. The founding settlement’s clauses on caste, on language and on the states had been written for electorates that voted as their patrons directed; once the beneficiaries of those clauses voted in their own names and numbers, the clauses had to be renegotiated with their holders in the room. Mandal was implemented in 1990 and not in 1980 because the arithmetic of 1989 had seated the beneficiaries’ parties at the cabinet’s edge. The federal veto migrated to the states in the 1990s and not earlier because state electorates had acquired parties of their own that no Delhi leadership could dissolve. The rights statutes of the UPA period were priced by allies whose voters measured governments by entitlements delivered. In each case the mechanism is the same, and it is the hub’s central mechanism: social mobilisation changed the identity of the bargainers, the new bargainers reopened the clause that governed their claim, and the reopened clause emerged amended. Chhibber’s fragmentation and Yadav and Palshikar’s consolidation are the two halves of this sequence, the breaking of the old container and the durability of what replaced it, and the adjudication below takes up their disagreement in those terms.

One further social fact belongs to the hub’s record because the pillar returns to it repeatedly: the era normalised the alternation of power at the state level as the system’s school of government. The chief ministers who bargained at the centre in the coalition decades, men and women who had won and lost state elections repeatedly, brought to Union cabinets an experience of opposition and incumbency in alternation that the Congress system’s nominated state leaderships had never possessed. The durability of the era’s renegotiations owes something to this schooling. Settlements lasted because the parties that signed them expected to sit on both sides of the table across a career, and bargains made by actors who expect to inherit their own texts are made, on the whole, in better faith than bargains made by actors who expect permanent possession.

The Mediated Republic: Television, the Vernacular Press and the Sting

The era’s bargains were struck before a transforming audience, and the transformation belongs to the hub’s record because it altered the price of every renegotiation. In 1989 the Indian citizen received politics through a state broadcaster, a state radio network and newspapers whose combined circulations reached a fraction of the electorate. The satellite invasion of 1991 to 1992, when CNN’s Gulf War coverage and then the Hong Kong-based STAR channels entered Indian homes through cable, broke the state’s broadcasting monopoly within the era’s first phase; by the mid-2000s several hundred television channels, scores of them devoted to news in a dozen languages, competed for the political day, and the vernacular newspaper press, led by Hindi dailies whose circulations dwarfed the English titles, set the agenda the national parties answered to. The consequences for coalition politics were mechanical before they were ideological. A partner’s withdrawal threat, a minister’s indiscretion, a chief minister’s price list became national news within the hour, and the secrecy in which the Congress system had conducted its internal settlements became unavailable to any government after about 1998. The sting operation arrived as the new medium’s weapon of choice: on 13 March 2001 the news portal Tehelka released its Operation West End tapes of defence procurement middlemen and politicians accepting cash, the BJP’s president Bangaru Laxman resigned within days, the Defence Minister George Fernandes resigned and later returned, and the episode demonstrated that a media market could execute the accountability that parliamentary arithmetic, balanced between the alliances, could not. The 2011 anti-corruption mobilisation was unimaginable without the apparatus the era had built: twenty-four-hour channels carried Anna Hazare’s fasts live for days, and a government accustomed to negotiating with parties found itself negotiating with a broadcast. A settlement renegotiated under continuous publicity is renegotiated differently, with the partners performing for their state electorates at every session, and the pillar’s later articles treat the era’s episodes with that condition kept in view.

The Era’s External Settlement: Foreign Policy Under Coalition Arithmetic

The founding settlement had a fourth, unwritten clause, the Nehruvian foreign policy of non-alignment administered by a prime ministerial office that treated external affairs as a preserve. The coalition era renegotiated that clause too, and the record matters here because it tests the renegotiation reading on terrain where instability was predicted to be most disabling. The prediction failed at every checkpoint. The minority Rao government executed the era’s boldest reorientations: full diplomatic relations with Israel in January 1992, the Look East policy toward Southeast Asia, the opening to the United States that the Cold War’s end invited, and the economic diplomacy that the balance of payments crisis compelled. The United Front’s Gujral Doctrine of 1996 to 1997 codified unilateral generosity toward India’s smaller neighbours as the coalition’s considered position, a doctrine produced not despite the Front’s thirteen parties but as their common text. The NDA conducted the Pokhran-II tests of 11 and 13 May 1998, weathered the sanctions that followed, received Prime Minister Vajpayee’s bus journey to Lahore in February 1999, fought the Kargil war under the restraint of the Line of Control, and converted the conflict’s diplomacy into the American embrace formalised by President Clinton’s visit of March 2000. The UPA then negotiated the civil nuclear agreement with the United States between July 2005 and October 2008, the era’s most contested external act, defended it against its own supporting coalition at the cost of that coalition, and carried it into force.

The sequence carries a lesson the pillar will use. External policy under coalition arithmetic proved neither paralysed nor erratic; it proved sequential, each government inheriting its predecessor’s openings and extending them in its own idiom, because the underlying renegotiation, from non-alignment to multi-directional engagement underwritten by economic integration, commanded assent across the era’s rival programmes. The National Agenda for Governance and the National Common Minimum Programme disagreed about much; on the external clause their texts converge to a degree that startled their drafters’ opponents. Where a settlement is renegotiated successfully, the signature is exactly this: successor governments of opposed colour treat the amended clause as the given text and argue about the next amendment instead. By 2014 the external clause had been so amended, and the incoming majority government inherited it as it inherited Bommai, Indra Sawhney and the rights statutes, revising at the margins of a text it did not write.

The Four-Phase Trajectory Map of the Coalition Era

The hub’s findable artifact condenses the argument into citable form. The table below, the Four-Phase Trajectory Map of the coalition era, sets each phase against the named events that define it, the renegotiation outcomes it produced on each of the three founding tensions, and its cumulative effect on the post-Independence settlement. Later articles of this pillar cite the map by name when they locate their episodes within the era’s sequence, and readers testing this article’s claim against the record can check each cell against the phase narratives above.

Phase	Named events	Religion and state outcome	Rights and development outcome	Federal outcome	Cumulative effect on the settlement
Phase One, 1989 to 1998	Mandal implemented 7 August 1990; Babri Masjid demolished 6 December 1992; Indra Sawhney judgment 16 November 1992; liberalisation launched 24 July 1991; Bommai judgment 11 March 1994; United Front governments 1996 to 1998	Reservation doctrine constitutionalised; Ayodhya left the religion-state clause in litigation and mobilisation	Directed development dismantled; reform proceeded by negotiation under minority and coalition governments	Article 356 fenced by floor test and judicial review; regional parties entered Union government	The founding clauses on caste, development and centre-state power all reopened within one decade
Phase Two, 1998 to 2004	National Agenda for Governance 1998; Pokhran-II tests May 1998; confidence vote lost by one vote 17 April 1999; Kargil war 1999; three new states November 2000; Gujarat violence 2002	The BJP’s core commitments deferred by coalition text for six years; Gujarat’s official conduct left contested	Insurance opened 1999; disinvestment negotiated against alliance resistance; fiscal responsibility statute 2003	New states created by a coalition parliament; state-based allies held Union portfolios throughout	Pre-poll alliance and written programme normalised as the era’s governing form
Phase Three, 2004 to 2009	National Common Minimum Programme 27 May 2004; information and employment statutes 2005; Left withdrawal 8 July 2008; confidence vote won 22 July 2008; Congress at 206 seats 2009	Secularism administered as coalition discipline; no majoritarian legislation attempted	Rights-based welfare enacted as the published price of outside support; food security legislated by 2013	Allies from the states held the economic ministries; GST design advanced through the Empowered Committee	The development clause renegotiated a second time, from market opening to statutory entitlement
Phase Four, 2009 to 2014	2G report November 2010; licences cancelled February 2012; Hazare mobilisations 2011; BJP majority of 282 seats May 2014	The religion-state tension carried unresolved into the majority era; no communal violence statute enacted	Rights architecture retained intact by the incoming government; land settlement of 2013 contested by all sides	GST amendment bill introduced 2011; states bargaining as a bloc through the Empowered Committee	The era closed with its renegotiations standing as the inherited text the 2014 verdict proposed to revise

Read down the final column and the hub’s claim restates itself without rhetoric. No phase is transitional in its own column: each closes with the settlement in a different configuration from the one it inherited, and the configurations accumulate rather than cancel. The map also disciplines the claim’s limits. The religion-state column records deferrals and contests alongside settlements; the federal column advances furthest; the development column reverses direction once and holds both directions in statute. A rebuilding with uneven storeys is still a rebuilding, and the map’s unevenness is the honest form of the argument.

Fragmentation, the Third System and Renegotiation: The Scholarly Disagreement Adjudicated

Three readings of the coalition era contend in the literature, and this hub article exists to adjudicate among them on the primary record rather than to catalogue them. The first is Pradeep Chhibber’s fragmentation reading, set out in Democracy without Associations in 1999: the party system broke into segments because the associations that once aggregated interests, the Congress organisation and its allied structures, decayed, leaving a politics of fragments incapable of producing national programmes. The second is the third-system consolidation reading of Yogendra Yadav and Suhas Palshikar, developed across their analyses of the elections of the period and collected in Indian Democracy: Meanings and Practices: the post-1989 order was not a decay but a third electoral system, as coherent in its rules as the Congress system of 1952 to 1977 and the competitive system that followed it, characterised by the federalisation of the party system, the primacy of state-level contests, and the democratic upsurge of the plebeian voter. The third is the reading this article defends: the era as substantive renegotiation, in which the fragmented and federalised system, whatever its pathologies, spent twenty-five years rewriting the operative clauses of the founding settlement across its three permanent tensions.

The adjudication proceeds by asking what each reading predicts and checking the prediction against the record assembled above. The fragmentation reading predicts programmatic incoherence: a system of fragments should legislate randomly, reverse itself at each alternation, and leave no durable text. The record shows the opposite pattern on the clauses that matter. The reservation settlement of 1990 to 1992 survived every subsequent government including the majority of 2014. The economic settlement of 1991 was extended by the United Front’s cautious budgets, by the NDA’s statutes and by the UPA’s programmes, and reversed by none. The federal settlement of Bommai and the coalition conventions bound Congress and BJP governments identically. Where the record does show incoherence, in the fall of governments and the pricing of portfolios, it shows it on the terrain of office, not of settlement. Chhibber’s reading is therefore sustained at the level of party organisation and fails at the level of constitutional output: the fragmentation was real, and it was compatible with a sustained, directional renegotiation that his framework does not predict and cannot easily explain.

The third-system reading predicts durability and rule-bound change, and on the mechanics of elections and party competition it is borne out completely: the era’s contests were fought state by state, its national verdicts were sums of state verdicts, and its alternations followed learnable rules, as the phase narrative shows from 1989 to 2014. Its limit is one of level rather than of truth. To name the system and describe its rules is not yet to say what the system did to the republic’s constitutive bargains, and Yadav and Palshikar’s own evidence, the upsurge, the federalisation, the assertion of state arenas, describes the agents of the renegotiation more exactly than its product. The renegotiation reading accepts the third-system account as its mechanism and adds the output the account leaves unmeasured: a settlement whose caste clause was constitutionalised in a new form, whose development clause was rewritten twice, whose federal clause was transferred in practice to the states and their parties, and whose external clause was converted from non-alignment to negotiated engagement. Synthesis, not victory, is the adjudication’s honest verdict: fragmentation is empirically true at the level of parties, consolidation is true at the level of the electoral system, and renegotiation is the name for what the consolidated, fragmented system did. The claim this article defends is the third, and it stands because it alone accounts for the record’s most stubborn fact, that the weakest governments of the republic’s history left the deepest marks on its settlement.

Atul Kohli’s political economy, engaged above, enters the adjudication as a caution against celebration. If the renegotiation’s development clause was written, as Kohli argues in Poverty Amid Plenty in the New India (2012), under the sign of a state-business alliance whose gains pooled at the top, then the rebuilding this article describes was a rebuilding with a distributional tilt, and the rights statutes of the UPA years were partial corrections negotiated by the coalition’s plebeian partners against that tilt. The renegotiation reading survives the caution because it claims direction, not virtue: the settlement was rebuilt, and the question of for whom it was rebuilt remains open in the scholarship and in the politics that followed 2014. A hub article that hid the caution would fail its own charter; the claim defended here is about what the era did, and Kohli’s work stands in the record as the strongest account of the price at which part of it was done.

Instability and Renegotiation Together: Answering the Strongest Counter

The strongest counter to everything argued so far is not a rival theory but a fact, and the brief for this hub requires it to be met head-on. The instability was real. Seven Prime Ministers in twenty-five years, against two in the preceding twenty-five, is not a framing; it is a count. Between November 1990 and April 1999, India had five changes of government outside the electoral cycle or at its edge, two mid-term dissolutions forced by withdrawn support, and one government defeated by a single vote. The costs were real as well. Budgets were delayed and diluted; the divestment programme of two governments stalled against partner resistance; the insurance and pension reforms of the period crawled; foreign interlocutors priced Indian commitments at a discount in the years when letters of support were being drafted and withdrawn in public. Any reading of the era that treats the falls of governments as surface noise fails the record as surely as the interlude reading fails the statute book.

The renegotiation reading answers the counter by refusing its premise, which is that instability and substantive work are alternatives between which a period must choose. The coalition era shows them operating as mechanism and consequence. The same arithmetic that made governments fragile made renegotiation unavoidable, because no clause of the settlement could be administered, let alone amended, without the assent of parties that earlier decades had excluded from the room. The reservation clause could not be governed in 1990 by the Congress method, because the parties of the beneficiaries held the balance. The federal clause could not be administered after 1996 by the dismissal method, because the states’ parties held the centre’s majority. The development clause could not be rewritten in 1991 by the planning method, because the creditors and the partners together foreclosed it. Fragility was the price of admission of the new bargainers; renegotiation was what the new bargainers did once admitted. Treating the price and the work as rival descriptions of the era mistakes a single process for two.

The counter also proves too much in its comparative form. The quarter-century of stability against which the era is measured, 1964 to 1989, contains the republic’s gravest constitutional injuries: the Emergency’s suspension of the rights the settlement guaranteed, the mass dismissals of state governments in 1977 and 1980, the centralisation that hollowed the Congress organisation until it could no longer reproduce its own majorities. Stability of tenure, that is, did not protect the settlement; on the era’s evidence it licensed the settlement’s abuse, and the fragmented decades that followed repaired, clause by clause, much of the damage the stable decades had done. Bommai answered the dismissals. The rights statutes answered, in part, the Emergency’s demonstration of how thin statutory protection could be. The coalition conventions answered the nominated-chief-minister system. The comparison the interlude reading invites, between a stable before and an unstable during, dissolves when the content of the stable years is admitted to the ledger. What the coalition era lacked in longevity it recovered in repair, and a hub for the pillar is entitled to enter that exchange as the era’s epitaph in advance of the detailed evidence: the era’s governments fell often, and the republic’s settlement rose each time a government fell, because the falling was the sound of a clause being reopened.

The Rebuilt Settlement the 2014 Verdict Inherited

The verdict of May 2014 is the pillar’s hinge, treated in full at the close of this pillar’s sequence, and the hub’s task is confined to specifying the inheritance. The government formed on 26 May 2014 took possession of a settlement that differed from the settlement of 1989 on every clause this article has tracked. On religion and the state, it inherited a constitutionalised reservation order capped at 50 percent with a creamy-layer exclusion, a Hindutva vocabulary admitted to electoral speech by the judgments of 1995, an Ayodhya title dispute in its final appellate stage, and a political field in which caste parties and a national Hindu party divided the founding tension between them. On rights and development, it inherited an opened economy whose commanding heights had been renegotiated toward private capital and statutory welfare at once: information rights, employment guarantee, food security, a fiscal responsibility framework and a land statute, all enacted, all administered by the machinery the majority commanded. On the federal clause, it inherited states that bargained as principals, a judiciary that policed the centre’s emergency power, a GST architecture awaiting its final enactment, and a party system in which its own majority was the exception the preceding twenty-five years had defined. Even the office the new Prime Minister occupied had been rebuilt by the era: the post-1989 premiership, constrained by programmes, partners and presidents into a negotiator’s chair, stood available for re-expansion precisely because the era had normalised its constraints, and the pillar’s subsequent articles trace what the new incumbent made of the vacancy the coalition decades had left in the office’s authority.

The three tensions, this series’ standing instruments, were not resolved by the rebuilding; they were re-sited. The religion-state tension moved from the Constituent Assembly’s abstraction into the arithmetic of caste parties, temple movements and constitutional benches. The rights-development tension moved from the Directive Principles’ exhortation into a statute book where market opening and legal entitlement stood side by side, each the product of a coalition’s published price. The federal tension moved from the Seventh Schedule’s text into the cabinet room, where chief ministers’ parties held portfolios and vetoes. A settlement re-sited is a settlement transformed in everything except its questions, and the questions, religion and the state, rights and development, centre and states, are the ones the founding generation posed. That continuity of questions across a discontinuity of answers is the precise sense of the hub’s claim. The coalition era was substantive renegotiation, not transitional interlude; the post-Independence settlement was rebuilt across twenty-five years before the 2014 verdict; and the articles that follow in this pillar examine the rebuilding phase by phase, government by government and clause by clause, testing the hub’s claim against the episodes on which it rests.

Frequently Asked Questions

Q: What was the coalition era in India?

The coalition era was the period from the general election of November 1989 to the general election of May 2014 in which no single party held a majority in the Lok Sabha and every Union government was a coalition or a minority administration dependent on outside support. Eight general elections were held in the period, seven men served as Prime Minister, and government formation proceeded through negotiated common programmes, from the National Front’s in 1989 to the United Progressive Alliance’s National Common Minimum Programme of 27 May 2004. This article reads the period as twenty-five years of substantive renegotiation of the founding settlement across its three permanent tensions, on religion and the state, rights and development, and the federal division of power, rather than as a transitional interlude between the Congress dominance of the Indira decades and the single-party Bharatiya Janata Party majority of 2014. The era opened with the Congress reduced to 197 seats in 1989 and closed with the same party at 44 seats, its organisational container replaced by a federalised party system.

Q: Was the coalition era a period of instability?

It was unstable in the tenure of governments and stable in the direction of the settlement, and both halves are documented. Three governments of the 1990s fell within eighteen months of taking office, the first Vajpayee government lasted thirteen days in May 1996, and the second fell by a single vote on 17 April 1999. Against that record stand the full or near-full terms of the Rao government from 1991 to 1996, the Vajpayee government of 1999 to 2004, and the ten continuous years of the United Progressive Alliance from 2004 to 2014. The era’s largest measures, the Mandal implementation of 1990, the liberalisation of 1991, the Bommai discipline of 1994 and the rights statutes of 2005 to 2013, were enacted by exactly the fragile governments the instability reading describes. The scholarly dispute is over inference: Pradeep Chhibber’s fragmentation reading, set out in Democracy without Associations in 1999, treats the instability as the system’s essence, and this article treats it as the price of admitting new bargainers who then renegotiated the settlement.

Q: What were the major events of the coalition era?

The events that carry the era’s renegotiations are datable. The Mandal Commission’s recommendations were implemented on 7 August 1990, reserving 27 percent of central government posts for the Other Backward Classes. The economic liberalisation was launched on 24 July 1991 with the budget and the Industrial Policy Statement. The Babri Masjid was demolished on 6 December 1992, three weeks after the Supreme Court’s Indra Sawhney judgment of 16 November 1992 constitutionalised the reservation settlement. The Seventy-third and Seventy-fourth Amendments created the elected third tier of panchayats and municipalities in 1993. The Bommai judgment of 11 March 1994 fenced the dismissal of state governments. The Pokhran-II nuclear tests followed in May 1998 and the Kargil war in 1999. The value added tax unified state indirect taxation from 1 April 2005. The Right to Information Act and the employment guarantee followed in 2005, the civil nuclear agreement survived a confidence vote on 22 July 2008, and the Bharatiya Janata Party’s single-party majority of 282 seats in May 2014 closed the period.

Q: How did the coalition era differ from the Indira period?

The Indira period governed the founding settlement’s tensions inside one party and increasingly through one office. The Congress under Indira Gandhi contained the factions that embodied caste, regional and ideological claims, resolved their disputes by high-command decision, and after 1969 rebuilt the organisation around the Prime Minister’s person, the arrangement this series calls the personality-as-institution framework. The coalition era governed the same tensions between parties rather than within one. Decisions required negotiated texts, the common programmes of 1989, 1996, 1998 and 2004; leadership was constrained by partners holding vetoes; state governments bargained with the centre as principals; and the Prime Minister’s office operated as a negotiator’s chair rather than a command post, its occupant chosen in several cases by the coalition’s arithmetic rather than by a party’s succession. The contrast extends to method: the Indira decades centralised, postponing the Mandal report for a decade and dismissing state governments wholesale in 1977 and 1980, where the coalition decades distributed initiative to caste parties, regional principals, courts and presidents, and legislated under those vetoes.

Q: What is the post-Mandal political system?

The post-Mandal political system is the order that crystallised after the implementation of the Mandal Commission’s recommendations on 7 August 1990, in which caste, organised through parties of the backward classes, Dalits and regional communities, became the central axis of political competition in northern India and a permanent term in national coalition arithmetic. Its elements are the Samajwadi Party, the Bahujan Samaj Party and the Rashtriya Janata Dal in the Hindi belt, the Dravidian parties in Tamil Nadu, and the national parties’ dependence on all of them for parliamentary majorities. Yogendra Yadav and Suhas Palshikar describe the order as India’s third electoral system, succeeding the Congress system, and connect it to the second democratic upsurge, the rising electoral participation of the plebeian castes. It is post-Mandal also in its counterweight: the Bharatiya Janata Party’s Hindutva consolidation was in part an answer to Mandal’s caste arithmetic, offering religious unity against caste division, so the post-Mandal system names a single field in which both mobilisations contend. Its constitutional sediment is the reservation doctrine settled by the Indra Sawhney judgment of 16 November 1992.

Q: Did coalition governments work in India?

They worked unevenly by function, and the functions must be separated before the question can be answered. At renegotiating the settlement, they worked historically well: the period’s fragile governments implemented the Mandal settlement, dismantled the licence raj, fenced Article 356, created the third tier of elected government, unified state taxation through the value added tax and enacted the rights statutes of information, employment and food. At routine administration and at reforms requiring the surrender of rents, they worked badly: divestment crawled against partner resistance, the women’s reservation bill failed for eighteen years against coalition vetoes, and the second United Progressive Alliance government lost control of its own integrity agenda between the spectrum findings of 2010 and the verdict of 2014. The mechanism explains the split. Coalitions distribute vetoes, and vetoes block both predation and paralysis; they admit new bargainers, and the bargainers spend their admission on the clauses that matter to their electorates. Judged as caretakers, the era’s governments underperformed their predecessors. Judged as the only instruments through which the settlement could have been reopened in those decades, they have no rival in the record.

Q: How did the coalition era end in 2014?

The era ended at the general election of April and May 2014, when the Bharatiya Janata Party won 282 of the 543 Lok Sabha seats on about 31 percent of the national vote, the first single-party majority since the Congress victory of 1984, and Narendra Modi, named his party’s prime ministerial candidate on 13 September 2013, took office on 26 May 2014. The immediate causes assembled across the preceding phase: the corruption findings and cancelled spectrum licences that fixed the second United Progressive Alliance government’s reputation, the anti-corruption mobilisations of 2011, slowing growth and price rises, and a Congress organisation unable to project a successor leadership. The structural cause was the era’s own success in nationalising political competition: state-by-state verdicts summed, for once, into a single national wave for a party that had learned the period’s lessons about organisation and message. The ending was arithmetic rather than repudiation. The new majority inherited the era’s renegotiations intact, the reservation doctrine, the rights statutes, the Bommai discipline and the negotiated design of the goods and services tax, and proposed to revise the settlement from inside the text the coalition decades had rebuilt.

Q: How many general elections were held during the coalition era?

Eight general elections were held in the period: those of 1989, 1991, 1996, 1998, 1999, 2004, 2009 and 2014, the last of which closed the era by returning a single-party majority. The density is itself evidence about the period’s first phase, in which four elections fell within a single decade because no arrangement of 1989, 1991 or 1996 could hold a Lok Sabha for its full term, against the second phase’s rhythm, in which the houses of 1999, 2004 and 2009 each ran effectively their course. No election of the eight returned a majority for the incumbent coalition’s largest party except the last, and two of them, those of 2004 and 2014, dismissed incumbent alliances that had expected renewal. Turnout across the eight elections moved within a band of roughly 55 to 66 percent, with the election of 2014 recording about 66 percent, then the highest in the republic’s history, a figure Yogendra Yadav’s second democratic upsurge thesis helps explain, since the rising participation of the plebeian electorate was one of the era’s continuities beneath its governmental churn.

Q: Why did coalition governments become stable after 1999?

The stabilisation had institutional causes that can be enumerated. The first was the pre-poll alliance: the National Democratic Alliance of 1998 and 1999 and the United Progressive Alliance of 2004 fought elections on negotiated programmes, so the coalition’s terms were settled before the arithmetic was known and could not be reopened by the first disappointed partner. The second was the lesson of 17 April 1999, when a government fell by one vote and the electorate punished several of the parties held responsible, teaching partners that withdrawal carried an electoral price. The third was legal: the Constitution (Ninety-first Amendment) Act of 2003 capped ministries at 15 percent of the Lok Sabha and abolished recognition of splits, shrinking the currency in which defections were bought. The fourth was distributive practice: partners received portfolios and policy vetoes inside government rather than promises outside it, so the costs of exit rose with the benefits of membership. Eswaran Sridharan’s studies of the period’s coalition durability document the pattern, and the results stand in the terms served: 1999 to 2004, 2004 to 2009 and 2009 to 2014.

Q: What role did the President play in forming governments during the coalition era?

The presidency became an arbitral office because hung parliaments transferred the choice of government to it. President R. Venkataraman handled the hung verdicts of 1989 and 1991 by inviting the leader he judged best placed to prove a majority and fixing the proof to a date, conventions he recorded in his memoir of the period. President Shankar Dayal Sharma invited the Bharatiya Janata Party, as the largest party, to form the government in May 1996 and accepted its fall thirteen days later when the proof failed. President K. R. Narayanan added the defensive precedent: in October 1997 he returned for reconsideration the United Front cabinet’s advice to impose President’s Rule in Uttar Pradesh, and in September 1998 he returned the Vajpayee cabinet’s advice concerning Bihar, and both cabinets let the advice drop, establishing that the power to return advice, used once, could discipline the centre’s hand over state governments. President A. P. J. Abdul Kalam’s formations of 2004 proceeded under conventions by then settled. The office’s era is a study in authority accumulated by precedent rather than by text, each incumbent’s restraint becoming the next incumbent’s rule.

Q: How did the anti-defection law shape coalition politics?

The anti-defection law, the Tenth Schedule inserted by the Fifty-second Amendment of 1985, disqualified legislators who defied their party’s whip, and its existence structured coalition politics on two levels. For individual members it made the party, not the legislator, the unit of bargaining, so coalition negotiations ran between party leaderships whose blocs could be delivered whole, one of the preconditions of the era’s written programmes and priced support. For parties it long contained a loophole: a split recognised at one-third of a legislature party escaped disqualification, and the era’s smaller formations were assembled and dissolved through that clause. The Constitution (Ninety-first Amendment) Act of 2003, passed under the National Democratic Alliance, deleted the split provision and simultaneously capped the council of ministers at 15 percent of the strength of the house, attacking at once the defection incentive and the jumbo ministry with which earlier governments had purchased survival. The law did not end bargaining; it professionalised it, moving the market from legislators to leaderships and from offices promised after formation to programmes negotiated before elections, which is the era’s institutional signature.

Q: How did regional parties change national policymaking during the coalition era?

Regional parties changed national policy by holding, from 1996 onward, the margin of every Union majority, and converting that margin into policy at three points. They held portfolios through which national decisions ran: railways under ministers from Bihar and West Bengal, telecommunications under a DMK minister after 2004, agriculture under Sharad Pawar for the decade of the United Progressive Alliance. They vetoed measures their state electorates opposed, stalling foreign investment in multi-brand retail in 2011 to 2012, slowing disinvestment under both alliances, and conditioning tax reform on compensation through the Empowered Committee that designed the value added tax. And they redirected the fiscal centre of gravity toward the states, securing Finance Commission awards, debt relief and the consensus procedures from which the goods and services tax emerged. Scholars describe the change as the federalisation of the party system, in Yogendra Yadav’s account a structural feature of the third electoral system rather than a transitional bargaining phase. Its durability is testable: the single-party majority of 2014 governed a Council of States in which regional parties retained their veto, so the era’s federalisation of policy outlasted the arithmetic that produced it.

Q: What was the Bommai judgment and how did it limit the dismissal of state governments?

The Bommai judgment, S. R. Bommai versus Union of India, decided by a nine-judge bench of the Supreme Court on 11 March 1994, is the coalition era’s foundational federal precedent. The case arose from the dismissal of state governments, including the Karnataka government of S. R. Bommai in 1989 and the Bharatiya Janata Party governments dismissed after the Ayodhya demolition of December 1992. The court held that the President’s satisfaction under Article 356 was subject to judicial review, that a state government’s majority must be tested on the floor of its legislative assembly rather than assessed by the Governor, that a proclamation could be struck down and the dismissed government restored, and that secularism formed part of the Constitution’s basic structure. The practical effect was immediate: invocations of President’s Rule, which had approached one hundred in the four decades before 1994, fell sharply, and the mass dismissals of opposition state governments practised in 1977 and 1980 were not repeated by any subsequent centre. The judgment converted the emergency power from an instrument of the ruling party into a justiciable exception, and coalition-era governments of both alliances operated inside its discipline.

Q: How did economic reform continue across coalition governments after 1991?

Reform continued because each successive coalition inherited it as the framework inside which its own programme had to be financed, and because the instruments of continuation were chosen for their compatibility with vetoes. The Rao government liberalised trade, investment and industrial licensing between 1991 and 1996. The United Front, despite its Left constituents, held the fiscal and external course in the budgets of 1996 and 1997, whose finance minister P. Chidambaram’s 1997 budget cut income tax rates and was praised across the aisle. The National Democratic Alliance opened insurance by statute in 1999, pursued disinvestment ministry by ministry, enacted the fiscal responsibility law of 2003 and built the highways programme. The United Progressive Alliance, bound by its Common Minimum Programme against privatisation of profitable public units, nonetheless carried the value added tax to adoption in 2005, sustained the growth boom of 2003 to 2011 and legislated the goods and services tax’s enabling design through the states’ Empowered Committee. Atul Kohli’s Poverty Amid Plenty in the New India of 2012 describes the continuity as a pro-business settlement of the state; on the evidence of the budgets, no government of the era, of any colour, reversed the direction of 1991.

Q: Which new states were created during the coalition era and why?

Three states were created in November 2000 by the National Democratic Alliance government: Chhattisgarh, carved from Madhya Pradesh on 1 November; Uttaranchal, renamed Uttarakhand in 2007, carved from Uttar Pradesh on 9 November; and Jharkhand, carved from Bihar on 15 November. Each statehood demand was older than the era. The Chhattisgarh and Uttarakhand movements pressed the claims of distinct hill and tribal regions governed from distant capitals; the Jharkhand movement, led for decades by the Jharkhand Mukti Morcha and its predecessors, claimed a tribal homeland across the Chota Nagpur plateau. The Congress system had contained all three demands as it contained every reorganisation claim after the 1960s. The coalition parliament enacted them together because the Bharatiya Janata Party had committed itself in its manifestos, because its allies raised no veto, and because smaller states had become administratively respectable in the era’s federal discourse. The creations matter beyond their regions: they demonstrated that the era’s renegotiation of the federal clause extended to the map itself, and they set the procedural template, state assembly consent sought, parliamentary act, appointed day, that the creation of Telangana would follow in June 2014, at the era’s very close.

Q: What happened to the Janata Dal and its successor parties?

The Janata Dal, the party of the National Front victory of 1989, dissolved across the era into a family of state parties that governed much of northern and eastern India. The first split came in November 1990, when Chandra Shekhar carried sixty-four members into the Samajwadi Janata Party to form his government. The Mandal decade then state-ised the inheritance: Lalu Prasad Yadav consolidated Bihar in the Rashtriya Janata Dal, founded in July 1997; Biju Patnaik’s legacy in Odisha became the Biju Janata Dal in December 1997 under his son Naveen Patnaik; the Karnataka parent shrank into the Janata Dal (Secular) of H. D. Deve Gowda in 1999; and the Janata Dal (United), assembled in 1999 around Sharad Yadav, George Fernandes and Nitish Kumar, anchored itself in Bihar as the Bharatiya Janata Party’s ally. Mulayam Singh Yadav’s Samajwadi Party, founded in October 1992, drew on the same socialist reservoir in Uttar Pradesh. The parent party’s fate is the era’s organisational parable: a national party whose social base, the backward castes of the states, was strong enough to rule half of India and too federal in structure to survive as one party. Its successors were coalition principals for the whole period.

Q: How did coalition bargaining shape Union budgets and railway budgets?

Coalition bargaining entered the budget documents directly, because the finance and railway ministries were routinely held by, or negotiated with, the alliance partners whose votes the budget had to survive. The railway budget became the era’s most legible bargaining text. Under Lalu Prasad Yadav, Railway Minister from 2004 to 2009, passenger fares were held while freight earnings and turnaround claims were advertised to Bihar’s electorate and the nation; under Mamata Banerjee in 1999 to 2001 and again after 2009, projects and factories were announced for West Bengal in quantities her critics in other states priced as regional tribute. Union budgets bore the same imprint in fiscal form: the fiscal responsibility targets of 2003 were written with escape margins partners could invoke, the farm loan waiver of 2008, at about 71,000 crore rupees, entered the budget as the United Progressive Alliance’s electoral price, and the petroleum subsidy compromises of every government between 2004 and 2014 registered the partners’ refusal to let administered prices rise. The budget, the founding settlement’s most centralised instrument, had by the era’s middle become a negotiated coalition document whose clauses were initialled in the states.

Q: How did the coalition era change the Congress party as an organisation?

The era converted the Congress from the republic’s governing institution into a party among parties, and the conversion was organisational before it was electoral. The party entered 1989 with state units led by nominees, no internal elections and an authority that ran from the Nehru-Gandhi family downward; it exited in 2014 with the same structure asked to hold together alliances in which it was often the junior or besieged partner. Under Sitaram Kesri from 1996 its decline sharpened; under Sonia Gandhi from April 1998 it learned coalition craft, assembling the alliances of 2004 and 2009 and accepting the discipline of common programmes it had once dictated alone. Yet the learning was tactical rather than structural: the organisation in Uttar Pradesh and Bihar, where its vote had collapsed in 1989 and 1990, was never rebuilt, its tallies there negligible for the entire era, and its 2009 recovery to 206 seats rested on incumbent arithmetic and state-level allies rather than on revived machinery. Zoya Hasan’s study Congress After Indira, published in 2012, documents the hollowing; the 44 seats of 2014, the party’s lowest tally, registered how little of the old container survived the era it had once contained.

Q: Why did the Third Front never produce a durable national government?

The Third Front, the project of a non-Congress, non-BJP alternative built from regional and Left parties, failed for reasons internal to its composition. Its fullest trial, the United Front of 1996 to 1998, governed twenty-one months under two prime ministers and fell twice to the withdrawal of the Congress support on which it depended from outside, because the Front’s constituents would not admit the Congress and could not replace its numbers. The constituents’ state rivalries compounded the structural defect: the Samajwadi Party and the Bahujan Samaj Party fought for the same Uttar Pradesh electorate, the Left and the Trinamool Congress for Bengal’s, and the Janata Dal’s successor parties divided their state inheritances, so the Front’s arithmetic dissolved wherever its members competed. Its programmes, from the Common Minimum Programme of 1996 to the declarations of 2009, when a Third Front was announced before the election and dissolved after it, could specify what the Front opposed more readily than the distribution of office it proposed. The era’s lesson is precise: durability required either a national party’s spine, as the Congress and the BJP supplied to their alliances, or a stated price for outside support, and the Third Front possessed neither.

Q: How did the office of Prime Minister change under coalition government?

The coalition era demoted the premiership from command to negotiation, and the change is measurable in the office’s routines. Prime Ministers were selected by arithmetic as often as by hierarchy: H. D. Deve Gowda and I. K. Gujral were chosen by the United Front’s steering committee in 1996 and 1997, and Manmohan Singh was nominated in 2004 by a party president holding the coalition’s confidence separately. Cabinets were assembled by quota among partners, portfolios were surrendered to allies as the price of support and could not be reclaimed at the Prime Minister’s pleasure, as the United Progressive Alliance’s inability to move its telecommunications minister showed. Policy proceeded by groups of ministers, by consultation with coordination committees and by the texts of common programmes that bound the office’s occupant. Critics, at their strongest, describe an office hollowed below the threshold of accountability; its defenders answer that the era’s premiers, Rao, Vajpayee and Manmohan Singh among them, each governed effectively within the constraints and that the constraints were the federal constitution working as designed. The office the majority of 2014 inherited had been rebuilt as a negotiator’s chair; its re-expansion thereafter is a subject of the next pillar.