The 1998 NDA formation is usually told as an arrival. The Bharatiya Janata Party, founded in 1980 and twice rebuffed at the threshold of national office, assembled a coalition of some two dozen parties, won the general election of February and March 1998, and placed Atal Bihari Vajpayee in the Prime Minister’s office on 19 March 1998. The telling is accurate in every particular and incomplete in the one particular that matters. The alliance that carried the BJP into government did so on a document, the National Agenda for Governance, that set aside the three commitments around which the party had built its identity: the abrogation of Article 370, the enactment of a Uniform Civil Code, and the construction of a Ram temple at Ayodhya. The founding was therefore two events wearing one name. It was Hindutva’s greatest electoral success to that date, and it was Hindutva’s tactical retreat, executed deliberately, in writing, before the first vote of confidence was taken. This article defends that double reading against the simpler one.

The Standard Reading of the 1998 NDA Formation

The standard reading rests on the Election Commission of India’s statistical report for the 1998 general election, and its facts are not in dispute. Polling ran across 16, 22 and 28 February 1998. The BJP won 182 of the 543 Lok Sabha seats on 25.59 percent of the vote. The Indian National Congress won 141 seats on 25.82 percent of the vote, a higher share of votes returning a smaller tally of seats, because the BJP’s vote was concentrated where its allies completed the arithmetic. The BJP-led alliance, the National Democratic Alliance, held 252 seats between the lead party’s 182 and the 70 seats contributed by its partners. The halfway mark in the House stood at 272. The Telugu Desam Party of Andhra Pradesh, which had contested outside the alliance, extended support from outside, and President K. R. Narayanan, after examining the letters of support that Vajpayee carried to Rashtrapati Bhavan, invited him to form the government. Vajpayee took the oath on 19 March 1998, at the head of a council of ministers drawn from parties that had spent the previous decade on opposite sides of nearly every question the republic had argued about.

On those facts the standard reading builds a story of consolidation. A party that had won two seats in 1984 and 161 in 1996 had become the natural centre of a governing formation. The coalition, on this telling, is a container: it held the BJP’s programme the way a larger vessel holds the same liquid. Christophe Jaffrelot’s The Hindu Nationalist Movement and Indian Politics, published in 1996, had mapped the movement’s long march through Indian politics up to the eve of this victory, and the 1998 result appeared to complete the map’s trajectory. The encyclopaedic summaries that most readers meet first treat the National Agenda for Governance as routine coalition furniture, the common programme every alliance drafts and every reader skips.

The document deserves a slower reading. A common programme can do two kinds of work. It can add the partners’ demands to the lead party’s programme, or it can subtract from the lead party’s programme the items the partners refuse to carry. The National Agenda for Governance, released in March 1998 in the days before the swearing-in, did the second kind of work on the three items that defined the BJP to its own cadre. What the BJP’s manifesto promised, the alliance’s agenda deferred, and the deferral was not an accident of drafting. It was the price schedule of the coalition, written down and signed before power changed hands. The paradox this article adjudicates is already visible in that sequence: the party reached office by formally standing down the programme that had built the party, and the standing-down was the condition of the reaching.

The Election That Made the Document Necessary

The 1998 election was the third general election in twenty-two months, called after the Congress withdrew support from the Gujral government in November 1997 and the eleventh Lok Sabha was dissolved in December 1997. The campaign’s organising fact was that no party expected a majority, and the BJP, alone among the major formations, behaved accordingly from the first day. Where the Congress contested as a national party with a national manifesto and a leadership question it declined to settle, its president Sitaram Kesri giving way to Sonia Gandhi’s campaign leadership only as the polling approached, the BJP contested as the core of a pre-poll alliance, negotiating seat adjustments state by state. In Tamil Nadu it fought inside Jayalalithaa’s front alongside the Pattali Makkal Katchi and the Marumalarchi Dravida Munnetra Kazhagam. In Bihar it aligned with the Samata Party and the Janata Dal remnant that became the Janata Dal (United). In Orissa it partnered the newly formed Biju Janata Dal of Naveen Patnaik. In Punjab it renewed its arrangement with the Akali Dal, and in Maharashtra its long-standing alliance with the Shiv Sena held. The Election Commission’s state-level returns record the yield of that method: the BJP’s 182 seats were assembled from sweeps and near-sweeps in the northern and western states, from alliance arithmetic in the south and east where the party’s own organisation was thin, and from almost nothing in Kerala and the northeast.

The verdict’s geography is the deferral map drawn in advance. Every seat the alliance method added beyond the BJP’s heartland total was a seat won in a constituency where the party’s three signature demands were, at best, neutral assets and, at worst, liabilities to be explained away. The AIADMK’s 18 Tamil seats were won in a state where the temple campaign’s idiom had no purchase. The Akali Dal’s 8 Punjab seats were won by a party whose founding purpose was the defence of a minority’s distinct sphere. The Samata Party’s 12 seats rested on Bihar’s socialist inheritance, which treated the movement’s programme as an adversary’s charter. A lead party reading its own results with care could see that its mandate was a composite, and that the composite’s terms of trade had been set by the partners whose additions made the difference between 182 and 252. The National Agenda for Governance was drafted by men who had read the returns in exactly that way. The document was not a retreat from the election’s meaning. It was the election’s meaning, transcribed.

The pre-poll method also changed the deferrals’ legal character inside the alliance. Seat-sharing agreements are contracts about candidacies; the common agenda extended the contract into governance, and partners invoked it as a contract whenever the movement’s organisations pressed the government to move faster. When the Vishwa Hindu Parishad demanded executive action on Ayodhya, the partners did not argue theology. They cited the agenda. When Sangh economists attacked disinvestment, the government could cite the same document’s economic chapters in reply. The National Agenda for Governance thus functioned as the coalition’s constitution in the strict sense: it constituted the terms on which power was shared, it was written before power was exercised, and it bound the strongest partner most, because the strongest partner had the most to lose from the arrangement’s collapse. That a party would accept such a constitution is the puzzle the consolidation reading cannot solve. That a party would design such a constitution, having studied 1996, is the paradox reading’s central exhibit.

The Inheritance: A Decade That Taught the BJP the Price of Standing Alone

The strategic education behind the 1998 deferrals ran through three episodes, each of which the BJP studied as it studied little else. The first was the post-1989 verdict’s structural rupture, when the National Front government of V. P. Singh took office with BJP support from outside and the party learned that proximity to power without possession of it purchased influence at retail and blame at wholesale. The Mandal Commission’s implementation in August 1990 and L. K. Advani’s arrest at Samastipur in October 1990 ended that arrangement, and the BJP spent the early 1990s converting the Ayodhya mobilisation into state governments and a rising national vote, without converting either into a Lok Sabha majority.

The second episode was the 1996 thirteen-day strategic-signal moment. In May 1996 the BJP emerged as the single largest party with 161 seats, and President Shankar Dayal Sharma invited Vajpayee to form a government. No significant party would join, and several parties that might have bargained instead organised against the prospect. Vajpayee resigned on 28 May 1996 before the confidence vote he knew he would lose, using his resignation speech to present the BJP as a party ready for office and wronged by an exclusionary consensus among its rivals. The signal was received. The lesson the party’s strategists drew from the thirteen days was precise: the obstacle was not the BJP’s vote share, which kept rising, but its acceptability. Parties refused the BJP not because it was weak but because its three defining commitments made association expensive for any party answerable to Muslim, Christian or Sikh voters, or to a regional electorate organised around a non-Hindutva identity.

The third episode was the one the BJP watched from the opposition benches: the United Front model the BJP studied between June 1996 and March 1998. The United Front, a coalition of regional and left parties sustained by Congress support from outside, produced two Prime Ministers, H. D. Deve Gowda and I. K. Gujral, a steering committee that functioned as a second cabinet, and a Common Minimum Programme that subordinated every constituent’s manifesto to a negotiated floor. The Front’s governments fell when the Congress withdrew support, first in April 1997 and again in November 1997, yet the model’s architecture impressed the BJP’s planners. A common programme, a coordination mechanism, and a lead party willing to mute its signature demands had kept ideologically unlike parties inside one government for twenty-one months. The BJP’s 1998 formation copied that architecture and improved its foundations, because the BJP, unlike the Janata Dal core of the United Front, brought 182 seats of its own to the table.

Read together, the three episodes explain why the deferrals of 1998 were strategic rather than improvised. The coalition-era renegotiation framework that governed Indian politics between 1989 and 2014 required every aspiring lead party to renegotiate its programme against the vetoes of regional partners, and the BJP in 1998 was the framework’s most disciplined student. The party that had been excluded in 1996 for the content of its programme rewrote the programme’s presentation in 1998, not by abandoning the three commitments in its own manifesto, which retained them, but by agreeing that the government’s agenda would be a different and smaller document.

Why could the BJP not form a government alone in 1998?

The BJP won 182 seats, ninety short of the 272 halfway mark, and its 25.59 percent vote share was marginally below the Congress share. Its strength was regional, concentrated in the north and west, with almost no presence across the south and east. Government required partners, and partners required the deferrals described below.

The arithmetic also explains the identity of the man the arithmetic produced. Vajpayee, rather than Advani, fronted the alliance because Vajpayee was the partner-facing asset: a leader whose parliamentary career stretched back to 1957, whose reputation for personal moderation reassured parties that feared the BJP’s cadre more than its leader, and whose acceptability the 1996 episode had identified as the scarce resource. The choice of face and the choice of document were the same choice. Both told regional parties that association with the BJP in 1998 would cost them less than association had cost in 1996.

Inside the Drafting: How the Agenda Was Written in a Fortnight

The National Agenda for Governance was produced between the counting of votes in the first week of March 1998 and the swearing-in on 19 March, and its drafting circumstances explain its texture. The BJP had campaigned on its own manifesto, released in February 1998, in which the three demands stood in their customary form. The alliance that emerged from the results had no common text, and the fortnight’s negotiations had to manufacture one under three simultaneous pressures: President Narayanan’s requirement that the claim to govern be documented in letters of support, the partners’ requirement that their vetoes be registered before they signed those letters, and the BJP’s requirement that its cadre be able to read the final text without concluding that the movement had been dissolved into its coalition. George Fernandes acted as the alliance’s principal intermediary, a socialist veteran of the anti-Emergency underground whose acceptability to regional leaders exceeded that of any BJP figure except Vajpayee himself. Chandrababu Naidu negotiated from Hyderabad as an outside supporter whose letter the President would weigh, a position that let him press for moderation with the leverage of a partner and the deniability of a non-member.

The document that resulted was organised as a programme of governance rather than a treaty of ideology: agriculture, water, education, internal security, economic management, centre-state relations and electoral reform filled its length, and its treatment of the three demands occupied a fraction of its text. That disproportion was itself the message. A reader of the BJP’s manifesto encountered a movement’s charter with a policy annex. A reader of the agenda encountered a policy programme from which the charter had been withdrawn to a set of guarded clauses. The drafting also settled the government’s internal constitution: sensitive subjects were fenced by the consensus formula, the finance and economic portfolios went to figures the partners and the markets could accept, with Yashwant Sinha, a former bureaucrat who had joined the BJP from the Janata tradition, at finance, and the home ministry went to L. K. Advani, the movement’s second figure, whose presence at the cabinet’s centre reassured the cadre that the deferrals were supervised by one of their own. Advani’s installation as Deputy Prime Minister in June 2002 formalised a duumvirate that the 1998 arrangements had implied from the start: Vajpayee faced the partners and the world, Advani faced the movement, and the agenda’s deferrals held in the space between the two faces.

Contemporaries noticed the construction. The phrase that attached itself to the arrangement in the press of the period was that the BJP had agreed to govern on a national agenda rather than a Hindutva agenda, and the party’s spokesmen did not dispute the formulation; they disputed only the inference that the Hindutva agenda had ceased to exist. The distinction they drew, between a party’s programme and a government’s agenda, was real in constitutional terms and unstable in political ones, since the cadre who had won the election had campaigned on the programme and were being asked to defend the agenda. The management of that instability consumed the party’s internal life for six years, through the Govindacharya affair, the Vishwa Hindu Parishad’s campaigns and the periodic restatements by Rashtriya Swayamsevak Sangh chiefs, including K. S. Sudarshan after he succeeded Rajendra Singh in March 2000, that the movement’s patience was a resource with a term. The drafting fortnight of March 1998 therefore deserves its place in the causal chain: the paradox was not discovered in office. It was composed, clause by clause, before office began.

The composition had in fact begun during the campaign itself, in the signals the BJP sent about the government it intended to form. The party projected Vajpayee as its prime ministerial face from the campaign’s opening, an unusual step for a party that had historically contested under collective leadership, and it campaigned on the paired promise of an able leader and a stable government, a slogan set that foregrounded the two commodities the alliance method supplied and the movement’s programme did not. L. K. Advani, the party’s president through the campaign’s early phase before Kushabhau Thakre succeeded him in 1998, publicly entertained the prospect of a common programme with allies during the campaign, preparing the cadre for a document that would not be the manifesto. Each signal lowered the eventual text’s shock. By the time the agenda appeared in March, the party’s voters had been told, in sequence, that the election was about a leader, that the leader was the party’s most acceptable figure, that stability was the prize, and that allies would have terms. The agenda merely wrote down, in programme form, what the campaign had already sold in slogan form, and the ease with which the movement absorbed the document in March 1998, whatever its later discontents, is explained by that preparation. Retreats announced in advance are received as strategy; the 1998 deferrals were announced, in every way except the text itself, well in advance.

Deferral One: Article 370 and the Examination That Replaced Abrogation

The first deferral concerned the constitutional status of Jammu and Kashmir. The demand to abrogate Article 370 was the oldest item in the movement’s programme, older than the BJP itself. Syama Prasad Mukherjee had founded the Bharatiya Jana Sangh in 1951 with the integration of Jammu and Kashmir as a founding grievance, and Mukherjee died in detention in Srinagar in June 1953 during his campaign against the permit system that governed entry into the state. Every Jana Sangh manifesto and every BJP manifesto from 1980 onward carried the demand. The BJP’s 1998 manifesto kept faith with that lineage: it called for the abrogation of Article 370, presenting the provision as a temporary arrangement, which the Constitution’s own placement of the article under a heading of temporary provisions appeared to concede, that had hardened into an obstacle to the state’s full integration.

The National Agenda for Governance replaced the verb. Where the manifesto promised abrogation, the agenda promised examination: the alliance would examine the question of Article 370 rather than act on it. The substitution reads as drafting etiquette until its audience is named. Inside the alliance sat the Jammu and Kashmir National Conference’s rivals and allies in a fluid arrangement, and outside it stood the regional parties whose support the government would need on the floor. More telling still, the National Conference itself, the party of Farooq Abdullah and the institutional heir of Kashmiri autonomy, moved into the alliance’s orbit during the government’s life, an association that would have been unthinkable against a live abrogation programme. The examination formula made that association possible by converting a demand with a destination into a process without one.

Did the Vajpayee government move any measure on Article 370 between 1998 and 2004?

No bill, resolution or constitutional step on Article 370 was introduced in either House across the government’s six years, and no cabinet decision on the subject was recorded. The examination promised in the National Agenda for Governance produced no published report. The demand survived intact in the BJP’s own documents and stayed absent from the government’s.

The deferral also clarifies what Article 370 and the asymmetric-federal design meant for coalition management. The Indian Constitution’s federalism had always carried asymmetric provisions for particular states, and Article 370 was the largest of them. A lead party committed to erasing the asymmetry could govern alone only with a majority the BJP did not possess. A lead party willing to leave the asymmetry unexamined could assemble a majority across parties that disagreed with one another about almost everything except the utility of leaving that question closed. The 1998 formation chose the second course, and in choosing it conceded, in practice, that the asymmetric-federal settlement would be administered rather than contested for as long as coalition arithmetic ruled. The concession held for the government’s full life and beyond it, until a single-party majority, twenty-one years later, made the arithmetic that had imposed the concession disappear.

Within the movement, the deferral was defended as sequencing rather than surrender. The argument, made by BJP spokesmen through 1998 and repeated in the party’s internal literature, ran that a party at 182 seats governed on borrowed numbers and could not spend borrowed numbers on its own programme. Critics inside the Sangh Parivar answered that borrowed numbers spent on everything except the programme raised the question of what the numbers were for. Both arguments are recorded here at their strongest because the article’s adjudication section returns to them; the factual floor beneath both is the textual one. The manifesto said abrogate. The agenda said examine. The government, for six years, did neither the first nor, in any published form, the second.

Deferral Two: The Uniform Civil Code and the Consideration That Replaced Enactment

The second deferral concerned the Uniform Civil Code, the proposal to replace the religion-specific personal laws governing marriage, divorce, inheritance and adoption with a single civil code for all citizens. The Constitution’s Directive Principles had carried the aspiration since 1950, in Article 44, and the founding generation’s arguments over it belonged to the religion-state founding tension that ran through the Constituent Assembly’s treatment of religion and law. The BJP’s claim on the issue was distinctive: the party argued that the existing arrangement, under which the state had reformed Hindu personal law in the 1950s and left Muslim personal law to community leadership, represented an asymmetry that the republic’s secularism could not honestly defend. Its 1998 manifesto called for the enactment of a Uniform Civil Code, and its spokesmen presented the code as a measure of gender justice and equal citizenship rather than of cultural uniformity, a presentation pitched at voters well beyond the party’s core.

The National Agenda for Governance replaced enactment with consideration. The alliance would consider the question of a Uniform Civil Code, with the consideration directed toward consensus. As with Article 370, the operative fact was not the new verb alone but the new decision rule it carried: a matter the BJP’s manifesto assigned to parliamentary majority was reassigned to a consensus that included the very parties and community leaderships whose opposition had kept the code unenacted since 1950. A consensus rule on a contested question is a deferral rule, because it grants the status quo a veto that a majority rule denies it. The Law Commission was not asked to draft a code. No bill was prepared. When the Supreme Court, in judgments of the period including the John Vallamattom decision of 2003 striking down a discriminatory provision of the Indian Succession Act as it applied to Christians, remarked on the absence of a common code, the government expressed sympathy and took no step.

The regional logic of this deferral is examined in the partner section below, but its religious logic belongs here. The code’s opponents inside the alliance included parties answerable to Muslim voters in Andhra Pradesh, Tamil Nadu and Bihar, and parties, including the Shiromani Akali Dal, answerable to a Sikh electorate with its own personal-law sensitivities and a strong institutional memory of resisting central uniformity. The BJP’s partners did not need to defeat the code in Parliament. They needed only to make its introduction a coalition-breaking act, and the National Agenda for Governance, by adopting the consensus formula in advance, recorded that they had succeeded before the government was formed.

For the movement’s own constituency, the code deferral cut in a particular way. The Article 370 demand spoke to national integration; the temple demand spoke to civilisational redress; the code spoke to the everyday legal experience of citizens whom the party claimed the existing order treated unequally, including Hindu families governed by the codified law of the 1950s and Muslim women governed by rules the party described as frozen. Deferring the code therefore deferred the one commitment whose beneficiaries the BJP described in universal rather than communal terms, and the deferral exposed the party to the charge, pressed by its opponents through the period, that its universalism was a flag of convenience flown over a programme it would not enact. The charge and the party’s sequencing defence are both part of the record the adjudication must weigh; the textual floor is again plain. The manifesto said enact. The agenda said consider. Six years produced consideration without a draft.

Deferral Three: Ayodhya and the Consensus That Replaced Construction

The third deferral was the costliest to write, because the Ayodhya commitment was not a policy preference but the emotional capital of the movement’s expansion. The Ram Janmabhoomi campaign had carried the BJP from 85 seats in 1989 to 120 in 1991, through L. K. Advani’s rath yatra of September and October 1990, the kar seva mobilisations, and the demolition of the Babri Masjid on 6 December 1992 by a crowd the movement had assembled and its leadership could not restrain. The BJP’s 1998 manifesto reaffirmed the commitment to facilitate the construction of a Ram temple at the Ayodhya site, a commitment the party carried into the election as a settled article of its identity, defended in constituency after constituency where the movement’s organisations had spent a decade building the claim that the temple was a debt owed by the republic to its Hindu citizens.

The National Agenda for Governance replaced construction with exploration. The alliance would explore consensual, legal and constitutional means toward the Ayodhya question, which in the circumstances of 1998 meant that the dispute would proceed where it already lay, in the Allahabad High Court, where the title suits over the site had been pending in one form or another since 1950, and where no party to the coalition could be made answerable for the pace. The formula protected every partner at once. Parties with Muslim voters could tell their electorates that the government had renounced any executive route to a temple. The BJP could tell its cadre that the commitment survived inside the legal process. The Vishwa Hindu Parishad, which was not a party to the document, could be told that nothing in it extinguished the demand. A commitment that every audience can read as preserved is a commitment deferred, and the Ayodhya clause of the agenda was drafted to be read exactly that way by each audience in turn.

The governance record matched the text. The Vajpayee government introduced no legislation on Ayodhya, acquired no land beyond the acreage the Rao government had taken into central custody after the demolition under the Acquisition of Certain Area at Ayodhya Act of 1993, and resisted pressure from the Vishwa Hindu Parishad and the Bajrang Dal to accelerate the court proceedings or to hand the undisputed portion of the acquired land to the Ram Janmabhoomi Nyas. When the Nyas pressed for the performance of a shilanyas ceremony on the undisputed land in March 2002, in the charged weeks after the burning of the Sabarmati Express coach at Godhra on 27 February 2002 and the Gujarat violence that followed, the government’s management of the ceremony, conducted under court-supervised conditions on 15 March 2002, was an exercise in containment. The episode displayed the deferral’s operating cost in its purest form: a government headed by the party of the temple movement deploying its authority to slow the temple movement’s own timetable, in order to hold a coalition whose members would not survive association with that timetable.

The Gujarat violence of February and March 2002 belongs to this section for a further reason. The violence, in which the official figures placed before Parliament recorded more than a thousand deaths, most of them Muslim, strained the coalition precisely along the seam the deferrals had papered over. Chandrababu Naidu of the Telugu Desam Party publicly demanded the removal of the Gujarat Chief Minister, Narendra Modi, in April 2002, and the demand’s failure at the BJP’s national executive meeting in Goa that month marked the limit of what the alliance’s secular-facing partners could extract. Vajpayee’s own remarks at Goa, and the party’s decision to retain Modi, were read by the partners as evidence that the retreat had boundaries, and by the movement’s critics as evidence that the retreat had always been cosmetic. The adjudication below weighs both readings; the chronology itself is settled, and it runs from a written deferral in March 1998 to a coalition surviving a partner’s public ultimatum in April 2002.

The Constitutional Referee: President Narayanan and the Discipline of the Period

The coalition’s discipline had a constitutional enforcer outside the alliance itself. President K. R. Narayanan, who had taken office in July 1997, conducted the government-formation and the crises of the period under a personal doctrine of documentary rigour that raised the cost of every party’s improvisation. In March 1998 he departed from the practice of inviting the single largest party’s leader on a presumption of support and instead required the claim to be evidenced, examining the letters of support that Vajpayee presented before issuing the invitation, and the requirement itself strengthened the National Agenda for Governance, since the letters were pledged against the published document rather than against private assurances. In September 1998, when the cabinet advised the imposition of President’s Rule in Bihar, Narayanan returned the advice for reconsideration, an exercise of the reconsideration proviso that the Constitution’s Article 74 carries, and the cabinet, confronted with a written constitutional objection from a President whose standing the government could not afford to contest, allowed the advice to lapse. The episode taught the government the same lesson its own agenda taught: the period rewarded written texts and punished unwritten assertions of will.

The pattern completed itself in April and May 1999. After the confidence vote was lost, Narayanan explored the alternative formation with the same documentary standard, receiving the Congress president Sonia Gandhi’s claim of support from 272 members and testing it against the lists the claim required, and when the claim could not be substantiated he dissolved the Lok Sabha rather than install a government on an assertion. The dissolution, and the caretaker continuation of the defeated Prime Minister through the ensuing election, set the precedent that coalition claims at the centre would be audited rather than announced. For the argument of this article the presidential record matters because it removed the last informal escape from the 1998 settlement’s terms. A government that might have been tempted to outrun its agenda in a moment of floor strength faced a head of state who had demonstrated, in March 1998, on Article 356 and on the dissolution, that he would ask for the paper. The deferrals held for six years under three disciplines at once: the partners’ vetoes, the movement’s contested patience, and a presidency that governed by the written word.

The Partners Who Priced the Deferrals

The deferrals were not concessions extracted at a single bargaining table. They were the standing requirements of the parties whose seats converted a plurality into a government, and each requirement can be traced to a constituency. The Telugu Desam Party, led by Chandrababu Naidu, the Chief Minister of Andhra Pradesh, won 12 seats in 1998 on the Election Commission’s tally and supported the government from outside; its weight exceeded its tally because Naidu positioned himself as the alliance’s indispensable southern broker, a role that grew when the party won 29 seats in 1999. Andhra Pradesh’s electorate included a large Muslim minority whose political support Naidu actively cultivated, and a state government led by him could not be seen to underwrite an abrogation drive, a civil code drive or a temple drive without paying for the association at home. The Telugu Desam Party’s price for its letters of support was therefore the deferral package itself.

The All India Anna Dravida Munnetra Kazhagam of J. Jayalalithaa won 18 seats in 1998, the largest single bloc inside the alliance after the BJP, on the strength of a Tamil Nadu front she had assembled. Tamil Nadu’s politics carried a Dravidian tradition, descending from the Self-Respect Movement and the Dravida Munnetra Kazhagam’s founding in 1949, in which suspicion of north Indian cultural uniformity was a constitutive sentiment rather than a preference. A Tamil party inside a government committed to a uniform civil code drafted in Delhi, or to a temple campaign conducted in the idiom of the Hindi heartland, would have carried a burden its rivals were waiting to exploit. Jayalalithaa’s support came with the deferrals attached, and her conduct in office, examined below in the government’s fall, showed that she understood the leverage the attachment gave her.

The Shiromani Akali Dal, the party of Punjab’s Sikh electorate and the oldest regional party in the republic, won 8 seats in 1998 and entered the government as the BJP’s most durable partner of the period, an association that stretched back to the two parties’ shared opposition to the Emergency and their Punjab alliances of 1996 and 1997. The Akali Dal’s sensitivities were those of a religious minority party guarding a distinct personal and institutional sphere, including the Sikh community’s own marriage law and the autonomy of its religious institutions. A uniform civil code imposed from the centre sat uneasily with everything the Akali Dal existed to defend, and its presence in the cabinet was itself a standing reminder that the code’s consensus formula could not be outrun. Around these three anchor partners stood the rest of the alliance: the Samata Party of George Fernandes and Nitish Kumar with 12 seats, the Biju Janata Dal with 9, the Trinamool Congress of Mamata Banerjee with 7, the Shiv Sena with 6, and a tail of smaller parties whose combined weight filled the distance between 182 and 252.

What majority mark did the 1998 Lok Sabha arithmetic turn on?

The halfway mark stood at 272 of 543 seats. The alliance’s own tally of 252 left it twenty short, which is why the Telugu Desam Party’s support from outside, and the letters of support President Narayanan examined before issuing his invitation on the eve of the 19 March 1998 swearing-in, carried constitutional weight.

The alliance’s paperwork reflected the pricing. The National Agenda for Governance was negotiated before the government was formed and published as the government’s charter, which meant the deferrals were not mid-term retreats under pressure but founding conditions, agreed at the moment of maximum BJP leverage, when the partners needed the alliance as much as the BJP did. The alliance also maintained a coordination committee, chaired by the Prime Minister, at which partner grievances were heard and the agenda’s boundaries policed; George Fernandes, the Samata Party leader who served as Defence Minister, emerged as the alliance’s convenor and its travelling troubleshooter. The post-March 1998 coordination committee proceedings, preserved in the BJP’s archives and in regional-party collections including the Telugu Desam Party’s records at Hyderabad, document the mechanism in operation: portfolio disputes, state-level frictions between allies who were rivals at home, and recurrent partner demands for public distance from the movement’s sharper edges. Journalists who covered those negotiations for the national press have described the care with which participants guarded their accounts of closed meetings, a professional world in which protecting a source’s identity was a condition of learning anything at all, the daily discipline described in accounts of how journalists use VaultBook to protect sources. The archival record, so far as it has been opened, shows a committee whose recurring achievement was the prevention of votes the government might lose, and whose recurring subject was the boundary the deferrals had drawn.

The Other Side of the Ledger: What the Partners Took Home

A price schedule has two columns, and the partners’ column explains the alliance’s durability as much as the BJP’s restraint explains its existence. The Telugu Desam Party’s gains were structural. G. M. C. Balayogi, its nominee, became Speaker of the Lok Sabha in 1998, the first Dalit and the first leader from a regional party to hold the office in the republic’s history, and the choice announced that the alliance’s honours would be distributed beyond the lead party’s ranks. Naidu himself converted his supporting role into a national standing disproportionate to his party’s numbers, convening opposition conclaves, arbitrating the alliance’s crises and presenting Andhra Pradesh’s reform programme, centred on Hyderabad’s information technology ambitions, to the world through the access that proximity to the Prime Minister’s office supplied. The All India Anna Dravida Munnetra Kazhagam’s column was shorter and more perishable: ministerial berths for its nominees, including senior portfolios for figures such as M. Thambidurai at law and justice in the government’s early composition, and a season of being courted as the alliance’s indispensable southern pillar, which ended when the courtship’s price rose past what the BJP would pay and the withdrawal of April 1999 followed. The Shiromani Akali Dal received the alliance’s most durable settlement: cabinet representation for its nominees, a centre that left Punjab’s internal arrangements and the state’s religious institutions unexamined, and a partnership that its leader Parkash Singh Badal could present to the Sikh electorate as proof that alliance with the BJP carried no threat to the community’s distinct sphere, the very assurance the civil code deferral embodied in text.

The ledger’s smaller entries trace the system’s churn. The Biju Janata Dal converted its 1998 association into the Orissa government of March 2000, with the BJP as its junior partner in the state, a reversal of the national hierarchy that the alliance’s flexibility made ordinary. Mamata Banerjee held the railway ministry from October 1999, resigned in March 2001 amid the Tehelka revelations and a dispute over fuel prices, took the Trinamool Congress out of the alliance for the West Bengal assembly election of that year, and returned to the fold and to the cabinet by 2003 and 2004, a cycle of exit and re-entry that only a formation held together by a written floor, rather than by ideological kinship, could have absorbed without dissolving. The Samata Party’s George Fernandes held defence through the Kargil war, resigned in March 2001 when the Tehelka tapes aired allegations touching his ministry, and returned to the portfolio in October 2001 after an inquiry’s interim course, an episode that tested the government’s reputation for probity and that the alliance weathered because its members’ attachment was transactional in the declared sense: each partner’s continuance was recalculated against the floor document and the spoils ledger, not against a shared creed that scandal could shatter. Read whole, the partners’ ledger shows the deferrals earning interest. The BJP’s restraint was not a gift to its allies. It was the alliance’s currency, and the allies spent it visibly, at home, in the currencies their own electorates valued.

Governance as Retreat, First Register: Security, War and the Foreign Policy of a Government Not Pursuing Hindutva

The strongest evidence for the retreat reading is the government’s own record, which across six years was a record of statecraft that any non-Hindutva government might have claimed. On 11 and 13 May 1998, less than two months after taking office, the government conducted five nuclear explosive tests at the Pokhran range in Rajasthan, under the codename Operation Shakti, ending the policy of nuclear ambiguity that every government since 1974 had maintained. The decision belonged to a national-security lineage that ran through the 1974 test under Indira Gandhi and the weaponisation programme of the intervening decades, and its political meaning for the coalition was double. It asserted a muscular nationalism that the movement’s base could celebrate, and it did so on terrain, national security, where no partner’s constituency was threatened. The international sanctions that followed, imposed by the United States under the Glenn Amendment and by Japan and other states, were absorbed without the coalition shedding a single partner, and the diplomatic recovery conducted by Jaswant Singh in his extended dialogue with the American Deputy Secretary of State Strobe Talbott between 1998 and 2000 became a byword for the government’s managerial competence.

The foreign-policy record accumulated in the same register. In February 1999 Vajpayee travelled by bus to Lahore and signed the Lahore Declaration with Pakistan’s Prime Minister Nawaz Sharif on 21 February 1999, an act of outreach that no reading of the movement’s Pakistan doctrine would have predicted from a BJP government in its founding programme. Between May and July 1999 the government fought the Kargil war, after Pakistani soldiers and militants were detected occupying heights across the Line of Control in the Kargil sector, and Operation Vijay cleared the intrusions by 26 July 1999 under a self-imposed restriction, the refusal to cross the Line of Control, that purchased international support at the price of military convenience. The war’s management, and the restraint that accompanied it, consolidated the government’s reputation among exactly the voters and partners the deferrals had been designed to reassure: the state under BJP leadership behaved as a status quo power abroad. The Agra Summit of July 2001 with Pakistan’s President Pervez Musharraf failed to produce a declaration, and the mobilisation that followed the terrorist attack on Parliament on 13 December 2001 kept the two armies facing each other through 2002, yet even these episodes were processed by the government through diplomacy rather than through the idiom of civilisational conflict that the movement’s literature supplied in abundance.

None of this record required the deferrals to be insincere, and the adjudication below does not claim that it was. The claim is narrower and harder to escape. A government elected on a manifesto whose signature commitments were a temple, a code and an abrogation spent its authority instead on nuclear tests, a bus journey to Lahore, a limited war fought under self-denying rules, and a diplomatic courtship of the United States. The substitution was the strategy. Nationalism was to be performed in the register of the state, where coalition partners could applaud it, rather than in the register of the movement, where they could not.

Governance as Retreat, Second Register: The Economic Programme the Manifesto Had Not Headlined

The economic record repeats the pattern in a second register. The governments of 1998 to 2004 pursued a programme of liberalisation and asset construction that the BJP’s older swadeshi inheritance would have contested, and pursued it with the coalition’s regional parties, several of them enthusiastic market reformers, as willing accomplices. A Department of Disinvestment was created in December 1999 and raised to a ministry under Arun Shourie, and the strategic sale of public sector holdings, including the sale of a controlling stake in Bharat Aluminium Company in March 2001 and in Maruti Udyog through 2002, proceeded against the objections of the Swadeshi Jagran Manch, the Sangh Parivar’s own economic platform. The New Telecom Policy of 1999 moved the sector from licence fees to revenue sharing and unlocked the mobile expansion of the following decade. The Insurance Regulatory and Development Authority Act of 1999 opened insurance to private and limited foreign capital. The Fiscal Responsibility and Budget Management Act of 2003 wrote deficit targets into statute. The Securitisation and Reconstruction of Financial Assets Act of 2002 armed banks against defaulting borrowers, and the Electricity Act of 2003 redrew the power sector’s rules.

When did the Vajpayee government launch its national highways programme?

The National Highways Development Project was launched in 1998, in the government’s first year, with the Golden Quadrilateral joining Delhi, Mumbai, Chennai and Kolkata as its flagship component. By 2003 and 2004 the programme had become the government’s most visible construction effort, and the Prime Minister’s Gram Sadak Yojana of 2000 extended the road-building register to rural India.

Two features of this record matter for the paradox. First, its preparatory work reached into the coalition period’s deepest institutional legacy: the Empowered Committee of State Finance Ministers, constituted in July 2000 under the West Bengal finance minister Asim Dasgupta, and the task forces on tax reform chaired by Vijay Kelkar in 2002, designed the value-added tax transition and sketched the goods and services tax architecture that a later government would enact. A goods and services tax is a federal bargain before it is a tax, and the fact that a BJP-led coalition built its foundations, in negotiation with finance ministers of opposition-ruled states, measures how far the government’s working identity had travelled from the centralising idiom of the movement’s programme. Second, the economic record was the partners’ record as much as the BJP’s. Naidu governed Andhra Pradesh as a reforming chief minister and sat inside the alliance’s counsels; the economic programme gave the coalition a positive agenda that its members could carry home, which the deferrals alone, being negative, could not. The retreat on the movement’s three commitments was thus paired with an advance on terrain the partners owned jointly with the lead party. The pairing was the coalition’s stability technology, and it worked for six years, outlasting the United Front and every other formation of the era.

Where the Retreat Did Not Reach: The Cultural Portfolios and Their Limits

Precision about the retreat requires a map of where it did not apply, and the map’s most instructive territory is the cultural and educational portfolios. The deferrals fenced the three commitments that partners had priced as coalition-breaking. They did not fence the wider field of cultural policy, and in that field the government moved with a freedom it denied itself elsewhere. Murli Manohar Joshi, a former BJP president and a figure of the movement’s intellectual wing, held the human resource development portfolio through most of the period, and his tenure produced the era’s sharpest cultural controversies: the revision of the National Curriculum Framework in 2000, the rewriting of history textbooks by the National Council of Educational Research and Training through 2001 and 2002, and appointments to the Indian Council of Historical Research that opposition historians described as an capture of the republic’s memory institutions. The controversy acquired its period label, the saffronisation debate, in the press and in Parliament, and the government’s defence of the revisions consumed political capital that the deferrals were supposed to conserve.

The episode refines the paradox rather than refuting it, for three reasons that the partners’ conduct makes visible. First, the partners tolerated the cultural advance because it occurred inside a ministry’s ordinary discretion and required no legislation they had to vote for, no constitutional step they had to endorse, and no public association with the temple, the code or the abrogation. The fence the agenda built ran around coalition-breaking acts, and textbook revision was not one. Second, the cultural field was where the cadre’s energies were permitted to discharge, and the permission functioned as part of the deferral system’s management: a movement asked to wait on its three founding demands was given ministries in which waiting was not required. Third, the limits showed whenever the cultural agenda approached the fence. The government’s attempts to introduce astrology and related subjects into university grants policy drew judicial and academic resistance and partner unease, and the human resource ministry’s bolder drafts were trimmed in cabinet. The pattern replicates the master pattern in miniature. Advance where the partners were indifferent, halt where they were exposed, and never convert a ministerial discretion into a legislative test of the coalition’s founding document.

A second unreached territory was the states. The deferrals bound the government of India, and BJP governments in the states operated under looser discipline, most consequentially in Gujarat after October 2001. The distinction mattered in April 2002, when the partners’ pressure for the Gujarat Chief Minister’s removal was directed at a state leadership the centre declined to discipline, and Naidu’s failure at Goa marked the boundary of the alliance’s writ. The coalition’s constitution, like the republic’s, was federal in its silences: it specified what the centre would not do, and it left the states to the arithmetic of their own assemblies, with consequences that the period’s history records in the Gujarat violence and its aftermath. The retreat, stated precisely, was a retreat of the Union government from three named commitments, executed inside a federal system whose other floors remained open to the movement’s advance. That precision strengthens the claim this article defends, because it explains the evidence that a cruder version of the retreat thesis, in which the whole movement stood down everywhere, cannot accommodate.

The Fall of the First Government and the Verdict That Vindicated the Design

The deferrals purchased a government, and the government’s first life showed both the purchase’s value and its term limit. J. Jayalalithaa’s demands escalated through 1998 and into 1999: the dismissal of the Dravida Munnetra Kazhagam government in Tamil Nadu, the reversal of decisions touching her legal jeopardy in corruption cases, and a deference in the alliance’s counsels that the BJP was unwilling to formalise. On 14 April 1999 the AIADMK withdrew its support. President Narayanan directed Vajpayee to seek a vote of confidence, and on 17 April 1999 the government lost the motion in the Lok Sabha by 269 votes to 270, the single-vote margin hinging on episodes that entered parliamentary folklore, including the vote of Giridhar Gamang, a Congress member who had become Chief Minister of Orissa and whose eligibility the Speaker, G. M. C. Balayogi, declined to void, and the Bahujan Samaj Party’s last-day reversal of its assurance of support. The opposition’s attempt to assemble an alternative collapsed within days, after the Congress president Sonia Gandhi’s claim of support from 272 members failed to materialise, and the Lok Sabha was dissolved. Vajpayee continued as caretaker Prime Minister into the election of September and October 1999.

The 1999 verdict is the success half of the paradox stated in numbers. The alliance, reconstituted with the Janata Dal (United) inside it and the AIADMK outside, won 303 seats on the brief’s standard tally, with the BJP again returning 182 members on a lower vote share of 23.75 percent. The Kargil war, concluded in July 1999, supplied the campaign’s backdrop, and the alliance campaigned explicitly on stability, on Vajpayee’s person, and on the record of a government that its opponents had removed by a single vote. The second Vajpayee government took office on 13 October 1999 and served the full parliamentary term to 2004, the first non-Congress formation in the republic’s history to do so. The deferrals were carried into the second term essentially unamended: the alliance’s 1999 programme repeated the architecture of 1998, with the three commitments again held in the consensus register, and the BJP’s leadership treated the 1999 result as a ratification of the design rather than as a mandate to retire it.

Read against the founding, the sequence from March 1998 to October 1999 demonstrates the paradox’s mechanism rather than merely asserting it. The party that had stood on its full programme in 1996 won 161 seats and no partners. The party that deferred its programme in 1998 won 182 seats and a government, lost that government to a partner’s withdrawal rather than to any dispute over the deferred programme, and returned in 1999 with a full-term majority coalition behind the same deferrals. Every step of that sequence is explicable by the deferrals’ presence, and none of it is explicable by the programme they deferred.

The Expansion the Retreat Financed

The success half of the paradox has a second ledger, kept in organisational rather than electoral entries. Walter Andersen and Shridhar S. Damle, in The RSS: A View to the Inside, published in 2018, documented the Sangh Parivar’s growth across the coalition years from inside the movement’s own institutional memory: the Rashtriya Swayamsevak Sangh’s shakha network expanded through the period, its affiliate organisations consolidated their presence in education, labour, tribal welfare and student politics, and the BJP’s state units professionalised their machinery in states where the party had been a marginal presence a decade earlier. Government office multiplied the resources available to that expansion. Ministerial office gave the party’s cadres administrative experience at the centre for the first time. Governorships, board memberships and the patronage of incumbency gave the wider movement a stake in institutions it had previously criticised from outside. The party contested state elections through the period from the platform of national incumbency, winning or sharing power in states including Gujarat, Madhya Pradesh, Chhattisgarh, Rajasthan, Orissa and Karnataka across the cycle, and its membership and booth-level organisation deepened in the Hindi heartland and beyond it.

The expansion was not frictionless, and the friction is evidence in its own right. The Sangh’s economic platforms attacked the disinvestment programme in terms the opposition could have drafted. K. N. Govindacharya, a senior BJP general secretary with Rashtriya Swayamsevak Sangh roots, was reported to have described Vajpayee as the party’s mask, a remark that cost him his position in 2000 and that survived because it named the suspicion the deferrals had created: that the face the party showed its partners was a face, and that the cadre were being asked to mistake the mask for the programme. The Vishwa Hindu Parishad’s leadership periodically accused the government of betraying the temple, and the Bajrang Dal’s campaigns ran on a timetable the government repeatedly tried to slow. A movement that had simply been consolidated by electoral success would not have generated this literature of complaint. A movement undergoing a managed retreat, financed by office and contested by its own purists, generated exactly this literature, and its existence corroborates the paradox reading from inside the house.

The State-Level Ledger: Where the Expansion Was Banked

The expansion’s entries can be dated at the state level, where the coalition years converted national incumbency into provincial footholds that outlasted the centre’s own tenure. In Orissa the BJP entered government with the Biju Janata Dal in March 2000, an arrangement that gave the party its first sustained administrative presence in the east. In Gujarat the party held its own government through the period, surviving the replacement of Keshubhai Patel by Narendra Modi in October 2001 and winning the state election of December 2002. The assembly elections of late 2003 delivered the heartland harvest: Madhya Pradesh under Uma Bharti in December 2003, Chhattisgarh under Raman Singh in the same month, and Rajasthan under Vasundhara Raje, three state governments won by a party that a decade earlier had lost those very states in the post-demolition reaction of 1993. Karnataka followed in partial form, the BJP emerging as the single largest party in the assembly election of 2004 and entering its first southern government in coalition two years later. Each of these advances was won by state units campaigning on the record and the respectability of the Vajpayee government at the centre, and each was staffed by cadres whose formation had occurred inside a party that office had made plausible to ambitious entrants.

The Sangh’s affiliate ledger ran in parallel. The Akhil Bharatiya Vidyarthi Parishad consolidated its hold on university politics in the northern states. The Bharatiya Mazdoor Sangh pressed its claim to be the country’s largest labour federation through the period. The Vanvasi Kalyan Ashram’s network in the tribal belts expanded, with consequences that later elections in Chhattisgarh and Jharkhand, the latter created as a state in November 2000 under the Vajpayee government’s reorganisation of Bihar, Madhya Pradesh and Uttar Pradesh, would register. None of these advances required the government to move a single clause of the deferred programme, and all of them were advanced by the government’s existence. The relationship between the two ledgers is the empirical heart of the paradox: the movement’s organisations grew fastest in the years when the movement’s programme was formally stood down, and they grew in the shelter of a government whose founding document had promised its partners that the programme would stay stood down. A reading of the period that counts only the programme’s stagnation misses the growth; a reading that counts only the growth misses the price list pinned to the government’s door.

The Deferrals Under Stress: Episodes That Tested the Boundary

A founding document is tested by the episodes that press against its fences, and three episodes from the government’s middle years show the fences holding under different kinds of load. In January 2001 the Prime Minister’s Office created a dedicated Ayodhya cell, staffed by the officer Shatrughan Singh, to manage the government’s engagement with the dispute and with the movement’s organisations. The cell’s significance lay in its direction of traffic: the government institutionalised a channel through which the Vishwa Hindu Parishad’s demands could be received, delayed and answered without reaching the cabinet’s agenda, and the Parishad’s protests against the arrangement confirmed that it understood the channel as a containment device. When the Parishad announced its intention to begin temple construction activity in March 2002 regardless of the court’s timetable, the cell’s machinery, and the government’s insistence on the Allahabad High Court’s authority over the site, produced the managed ceremony of 15 March 2002 recorded in the Ayodhya section above.

The second episode was literary rather than administrative. At the turn of 2000 and 2001 Vajpayee published his Kumarakom musings, a set of reflections composed during a holiday in Kerala that the press read, correctly, as a state paper in disguise. The musings lamented the quality of governance and the conduct of politics, and they were received inside the movement as a leader’s meditation on the distance between the party’s aspirations and its circumstances. Their relevance here is the reception they met from the Rashtriya Swayamsevak Sangh’s leadership and the party’s own organisation men, who read the reflections as the Prime Minister thinking aloud about the coalition’s constraints and answered with reminders of the movement’s patience. The exchange was courteous and unmistakable. The deferrals were being administered by a Prime Minister who found them wearying and defended by a movement that found them expensive, and the arrangement held because neither side could name a better one that preserved a government.

The third episode ran in the opposite direction. In the state elections of late 2003 the BJP campaigned in Madhya Pradesh, Chhattisgarh and Rajasthan on bijli, sadak and pani, electricity, roads and water, a slogan set that translated the centre’s governance register into state idiom and delivered three victories. The campaign was the deferral strategy operating without a coalition to excuse it, since the party in those states was substantially free to choose its own themes, and it chose the themes of the National Agenda for Governance rather than the themes of its manifesto’s signature chapters. Strategists inside the party drew the conclusion that the Vajpayee model had become the party’s natural idiom. The 2004 defeat interrupted that conclusion’s reign, yet the model’s return in 2014, in a campaign built on governance and addressed past the party’s core to the country’s aspirational voters, suggests that the strategists had identified something durable about the conditions under which their party could win India.

The Complication: Organisation Against Deferral

The strongest counter to the strategic-paradox reading deserves its full force before the adjudication proceeds. The counter holds that the deferrals were epiphenomenal: that the true engine of the BJP’s later capacity was the organisational expansion of 1998 to 2004, which would have proceeded under any BJP-led government, and that the post-2014 reversal of the deferrals was produced by that organisation reaching maturity under a single-party majority, not by any strategic design laid down in 1998. On this reading, defended in its essentials by those who stress the Rashtriya Swayamsevak Sangh’s long institutional patience, the National Agenda for Governance was a routine coalition document, and the paradox is a pattern imposed by hindsight on an unremarkable compromise.

The counter is right about the organisation and wrong about the sequence. Organisational expansion on the scale Andersen and Damle record required inputs that only national office supplied in those years: the visibility of the Prime Minister’s office held by a movement nominee, the flow of capable administrators through the party’s ranks, the credibility with donors, professionals and state-level elites that incumbency confers, and the simple fact that a party in government at the centre is a party that ambitious politicians join. Each of those inputs was conditional on the coalition existing, and the coalition, as the partner section showed, was conditional on the deferrals. The 1996 control case isolates the variable: the same organisation, the same leader, the same rising vote share, and no deferral document, produced thirteen days. The organisation was the engine of everything that followed, and the deferrals were the ignition without which the engine, in 1998, did not start. Suhas Palshikar’s Economic and Political Weekly analyses of the period’s elections reached a cognate conclusion from the electoral side, reading the BJP’s coalition-era conduct as the conduct of a party that had learned to separate the mobilisation of its core from the reassurance of everyone else, and to price the separation correctly. That separation is the strategic-paradox reading in another vocabulary.

The Opposition’s Reading: A Mask, a Façade, or a Genuine Armistice

The government’s opponents read the deferrals from the day the agenda was published, and their reading forms the third strand the adjudication must weigh. The Congress and the Left parties argued that the document changed nothing essential: the movement’s organisations remained committed to the programme, the cadre remained in place, and the agenda was a presentational device, a mask in the word that Govindacharya’s reported remark would later make famous inside the party itself. On this reading, the retreat was cosmetic and the correct response was to treat the government as the movement in office, whatever its charter said. The reading had the virtue of attending to the movement’s continuity of personnel and the vice of attending to nothing else. It could not explain why the partners, who had every incentive to detect cosmetics, priced the document as substance and repeatedly invoked it as a contract. It could not explain why the movement’s own purists attacked the government in terms indistinguishable from betrayal, a strange reaction to a façade they might have been expected to welcome. And it could not explain the six-year legislative silence on the three items, which a merely cosmetic deferral, designed to be quietly outgrown, would have broken at the first favourable moment.

A subtler opposition reading, developed by commentators who took the document seriously, held that the deferrals were genuine and therefore temporary in a different sense: an armistice that a growing movement would honour only until it no longer needed allies. Jaffrelot’s work on the movement, extended in his Religion, Caste, and Politics in India of 2010, supplied the scholarly version of this reading, treating the coalition period as one phase in a longer strategy of legitimation through which the movement normalised its presence in the republic’s institutions. The armistice reading and the paradox reading agree on the documentary facts and differ on the weight given to intention: the armistice reading treats the retreat as a tactic whose reversal was always scheduled, the paradox reading treats the retreat as a structural condition whose reversal became possible only when the condition, coalition dependence, disappeared. The difference matters for how the post-2014 years are interpreted, and the adjudication returns to it, but the shared ground is the ground this article stands on: in 1998 the BJP governed by standing its programme down, and no serious reading of the period, friendly or hostile, disputes the standing-down itself.

What the Alliance Did to the Party System Around It

The 1998 formation’s effects ran beyond the alliance itself, because it taught every other formation in the system a lesson it then had to practise. The Congress, which had sustained and then destroyed two United Front governments from outside, contested 1999 still organised as a party that expected to rule alone or not at all, and its 114 seats, its lowest tally to that date, completed its education. The United Progressive Alliance of 2004 was the Congress’s answer composed in the BJP’s grammar: a pre-poll and post-poll alliance gathered around a Common Minimum Programme, a lead party moderating its signature positions, in the Congress’s case its economic reform pace and its leadership claims, against the vetoes of allies and of the Left parties that supported the arrangement from outside. The document’s very name conceded the genealogy. What the BJP had named a National Agenda for Governance in 1998, the Congress named a Common Minimum Programme in 2004, and the difference between the titles measures the difference between a lead party at 182 seats dictating the floor and a lead party at 145 seats negotiating it.

Scholarly accounts of the period registered the normalisation. E. Sridharan’s studies of coalition politics in India traced the system’s movement from the Congress-dominant order through the fragmented 1990s into an era in which pre-poll alliances between a national party and regional parties became the standard technology of government formation at the centre. Within that scholarship the 1998 formation occupies a distinctive place, because it was the first occasion on which the technology was operated by a party whose ideology appeared, on its face, least suited to the compromises the technology demanded. The demonstration effect ran in both directions. Regional parties learned that the BJP in office would honour a written floor, which lowered the price of future associations and helps explain the alliance’s rapid reassembly in 1999 after the trauma of April. The BJP learned that the technology rewarded it with a governing durability that ideological assertion had never delivered, which helps explain why the lesson survived the 2004 defeat and reappeared, transformed by majority arithmetic, in 2014. A formation that changes the behaviour of parties outside itself has done more than win an election. It has amended the system’s working constitution, and the amendment of 1998, the written subordination of a lead party’s programme to a negotiated agenda, remained the system’s default until a single-party majority made the device unnecessary for the party that had invented its modern form.

The Findable Artifact: The Three Deferrals in One Table

The argument of this article can be carried away in a single table. Table 1, the three-deferral comparison, sets the BJP’s 1998 manifesto position against the National Agenda for Governance position for each of the three commitments, then records the governance practice of 1998 to 2004 and the post-2014 reversal, so that a reader citing the strategic paradox can point to the exact cells where the retreat and the reversal occurred.

Commitment	BJP manifesto position, 1998	National Agenda for Governance position, March 1998	Governance record, 1998 to 2004	Post-2014 reversal
Article 370	Abrogation of the article as a step toward full integration of Jammu and Kashmir	Examination of the question, with no step announced	No bill, resolution or published examination; the National Conference entered the alliance’s orbit	Presidential orders of 5 and 6 August 2019 and the Jammu and Kashmir Reorganisation Act rendered the article inoperative and bifurcated the state
Uniform Civil Code	Enactment of a common civil code as a measure of equal citizenship	Consideration of the question, directed toward consensus	No draft code commissioned and no bill prepared; judicial remarks on the code answered with sympathy and no step	The 22nd Law Commission reopened consultation in 2023; Uttarakhand enacted a state code in February 2024; a national code remained unenacted
Ayodhya Ram temple	Facilitation of the construction of a temple at the Ayodhya site	Exploration of consensual, legal and constitutional means	The title suits left to the Allahabad High Court; movement pressure for acceleration resisted, including in March 2002	The Supreme Court judgment of 9 November 2019 awarded the site for a temple; the consecration ceremony followed on 22 January 2024

The table’s middle columns are the retreat; its final column is the reversal; and the distance between them, twenty years in the first row’s case, is the measure of what the 1998 design deferred and what it made possible in the interval. The reversal column also disciplines the paradox against exaggeration. Only one of the three commitments had been fully enacted by the reversal period’s own standard: the temple was awarded by a court and consecrated, the article was rendered inoperative by presidential and parliamentary action, and the code remained, at the national level, a consultation. The post-2014 governments reversed the 1998 adaptation at the speed their own majorities allowed, and the order of the reversals, temple and article before code, follows the order of political cost rather than the order of the 1998 text.

The Reversal After 2014

The BJP’s return to office in 2014 with 282 seats, a single-party majority of the kind the party had never held, and its enlargement to 303 seats in 2019, changed the constraint that had written the 1998 document. The party’s 2014 manifesto restored the three commitments to a governing programme that no partner could veto, and its 2019 manifesto repeated the restoration in the confidence of a second majority. On 5 August 2019 the President issued Constitutional Order 272, applying the Constitution in full to Jammu and Kashmir and superseding the 1954 order through which Article 370 had operated, and Parliament passed the Jammu and Kashmir Reorganisation Act, which received assent on 9 August 2019 and divided the state into the Union Territories of Jammu and Kashmir and Ladakh. The Supreme Court upheld the abrogation in its judgment of 11 December 2023. On 9 November 2019 a five-judge bench of the Supreme Court delivered the Ayodhya judgment, awarding the disputed site for the construction of a temple through a trust and directing the allotment of alternate land for a mosque; the temple’s consecration, performed with the Prime Minister at the centre of the ceremony, took place on 22 January 2024. On the code, the reversal remained partial: the 21st Law Commission’s consultation paper of 31 August 2018 had judged a national code neither necessary nor desirable at that stage, the 22nd Law Commission reopened the question in June 2023, and Uttarakhand’s assembly enacted a state-level Uniform Civil Code in February 2024, the first such enactment by a state under the post-2014 dispensation of the party.

Which of the three deferred commitments was reversed first after 2014?

Article 370 was reversed first, by the presidential orders of 5 and 6 August 2019 and the Jammu and Kashmir Reorganisation Act of the same month. The Ayodhya reversal followed through the Supreme Court’s judgment of 9 November 2019. The Uniform Civil Code moved slowest, reaching enactment only at state level in Uttarakhand in February 2024.

The reversal’s authors understood its relationship to 1998 precisely, and said so. The post-2014 leadership framed the coalition decades as a period of compulsion that majority government had ended, and the party’s campaign literature of 2019 treated the fulfilment of the deferred commitments as the dividend of voting the party into office alone. That framing is the paradox reading confirmed by the paradox’s beneficiaries: the deferrals were real, they were known to be real by the movement that made them, and they were reversed at the first moment the arithmetic that had imposed them ceased to bind. A commitment that was never abandoned did not need to be restored; a document that never deferred anything did not need to be superseded. The restoration rhetoric of 2014 and 2019 is unintelligible except against the 1998 text it repudiated.

The Reckoning of 2004 and the Decade of Waiting

The design’s final test inside the period came at the election of April and May 2004, which the government called early on the strength of its economic record and its India Shining campaign, and lost. The BJP fell to 138 seats, the alliance to a tally well short of the Congress-led formation that succeeded it, and the movement’s internal audit of the defeat immediately reopened the 1998 argument. The purist account held that six years of moderation had purchased the cadre’s demoralisation and the core voter’s indifference, and that the party had been defeated as a government that stood for management rather than for the programme. The strategists’ account held that the defeat was assembled from state-level anti-incumbency, drought and agrarian distress in the alliance’s rural base, and the accumulated grievances of partners who had decayed in their own states, and that the same design had won in 1999 with the same document. Both accounts were argued in the party’s post-election reviews through 2004 and 2005, and the friction surrounding L. K. Advani’s remarks in Karachi in June 2005, praising Muhammad Ali Jinnah’s speech of 11 August 1947, showed how raw the underlying question remained: whether the party’s path to national acceptability ran through the moderation of its inheritance or through the inheritance’s vindication. The Rashtriya Swayamsevak Sangh’s public displeasure at the Karachi remarks, and Advani’s resignation as party president under that pressure before its withdrawal and his eventual succession by Rajnath Singh at the end of 2005, demonstrated that the movement’s referee had not retired from the argument the deferrals had started.

The decade that followed supplied the control observations. In opposition from 2004 to 2014 the BJP restored the three demands to the centre of its programme, contested the 2009 election with Advani as its prime ministerial face, and lost ground, falling to 116 seats. The programme undiluted, carried by the leader most identified with it, in a political market unchanged in its regional structure, did not produce a majority. In 2014, under Narendra Modi, the party won 282 seats on a campaign that subordinated the three demands to a development and governance appeal addressed to first-time and aspirational voters, restored them in the manifesto’s text, and held them in reserve through the first term’s early years before enacting the reversals recorded above. The pattern across 1996, 1998, 2009 and 2014 is too regular to be coincidence. The party won when its programme was deferred or subordinated and its acceptability maximised; it lost when its programme led. The 1998 document was the first written instance of that law of the party’s national existence, and its authors, whatever their private intentions, had discovered the law by the most direct method available to them, which was defeat.

The Adjudication: Consolidation, Retreat, and the Record Between Them

Two readings of the 1998 formation stand before the evidence. The consolidation reading, which the standard summaries reproduce, holds that the 1998 result was the BJP’s electoral arrival: the party assembled the largest coalition in the republic’s history to that point, governed, returned with a full term, and thereby became a normal party of government, with the National Agenda for Governance filed as the ordinary paperwork of coalition formation. The paradox reading, defended here, holds that the 1998 founding was an electoral success purchased by a tactical retreat on the programme that defined the party, that the retreat was the coalition’s founding condition rather than its incidental cost, and that the post-2014 reversal of the three deferrals is the paradox’s confirmation.

The textual evidence decides the first point. A consolidation that leaves the programme intact does not require a second document. The BJP contested the 1998 election on its own manifesto, which carried the three commitments, and then governed on a different document that deferred all three, and the existence of the substitution, dated March 1998 and published before the swearing-in, is a primary-source fact that the consolidation reading has never explained. The partner evidence decides the second point. The deferrals map, partner by partner and constituency by constituency, onto the vetoes of the parties whose seats made the government, as the pricing section documented; they were not distributed randomly across the manifesto but concentrated with surgical precision on the three items those vetoes targeted, while the programme’s other contents, economic liberalisation included, passed into the government’s agenda untouched or enlarged. The governance evidence decides the third point. A party undergoing consolidation of its programme in office would be expected to advance the programme’s signature items at least incrementally across six years: a draft code, an examination report, a legislative step. The record shows none of these, and shows instead a six-year investment of the government’s authority in security, diplomacy, infrastructure and fiscal architecture, the registers where its partners could follow. The reversal evidence, finally, decides the dispute about intention. The post-2014 party restored and then enacted the deferred commitments under majorities that freed it from the vetoes, and described the coalition period as compulsion in its own campaign voice.

Against all of this, the consolidation reading retains one true element, which the paradox reading absorbs rather than denies. The expansion was real. The party of 1998 to 2004 did become a normal party of government, its organisation did deepen on the scale Andersen and Damle recorded, and its leader did acquire the elder-statesman standing that made him the alliance’s indispensable face. The paradox is not that the success was illusory. The paradox is that success and retreat were the same transaction, each the other’s price, and that Indian politics after 2014 cannot be read without holding both halves of the transaction in view at once: the deferrals explain the coalition, the coalition explains the expansion, the expansion explains the majority, and the majority explains the reversal. The 1998 NDA was Hindutva’s electoral success and tactical retreat simultaneously, and the Modi-era reversal of the 1998 adaptation is the proof that the adaptation, and not the consolidation alone, is what 1998 was.

One further discipline on the claim is owed to the reader. The paradox reading does not require the deferrals to have been cynical, and the evidence does not support cynicism as the whole account. Vajpayee’s own disposition toward the three demands was the disposition of a parliamentarian formed in the Jana Sangh’s legislative decades, who had argued for the movement’s positions in the Lok Sabha since the 1950s and who appears, in the record of his Kumarakom reflections and his conduct of the Lahore and Agra diplomacy, to have valued the government’s achievements in their own right. The party’s strategists, Advani among them, described the agenda as the maximum programme available under the arithmetic, a description that was true and that conceded the paradox in the act of denying it, since a maximum set by partners’ vetoes is a retreat measured from the programme the party had set itself. Intentions divide the historians; the document does not divide. Whoever drafted the clauses of March 1998, and whatever they privately intended for the decades after, they wrote a governing charter in which the movement’s three founding demands were stood down as the price of office, and the office, once held, rearranged Indian politics around the party that paid the price. That rearrangement, and not the intentions behind it, is the evergreen fact of the 1998 formation, and it is why the year remains the hinge on which any account of Hindutva’s passage from movement to state must turn.

The documentary method this article has followed is itself part of the claim’s defence. Electoral history is usually written from results, and results alone make 1998 a consolidation story, because results cannot record what a government declined to attempt. The period’s meaning sits in the space between two published texts, a manifesto and an agenda, issued weeks apart by the same political formation for two different audiences, and in the six-year silence that the second text imposed on the first text’s signature chapters. Historians of the movement, from Jaffrelot’s pre-1998 mapping to Andersen and Damle’s insider account of the Sangh’s growth, supply the context in which that silence becomes audible: an organisation expanding, a cadre complaining, a leadership administering a pause it described as sequencing. Read the two texts side by side, as Table 1 invites the reader to do, and the standard story inverts. The arrival was real, and it was the retreat that made the arrival possible, and the decades after 1998 are the working out of that single transaction.

Frequently Asked Questions

Q: What is the National Democratic Alliance?

The National Democratic Alliance is the BJP-led coalition of national and regional parties first assembled for the 1998 general election. At its formation it gathered some two dozen parties, among them the Samata Party, the Shiromani Akali Dal, the All India Anna Dravida Munnetra Kazhagam, the Biju Janata Dal, the Trinamool Congress and the Shiv Sena, with the Telugu Desam Party supporting from outside. The alliance adopted the National Agenda for Governance as its common programme, formed the government under Atal Bihari Vajpayee in March 1998, fell in April 1999, returned with a larger tally that year, and governed for a full term to 2004. Its membership has changed across the decades since, with parties entering and leaving, but the name and the BJP’s leadership of the alliance have continued, and the alliance returned to office in later elections as the vehicle through which the party organised its coalition support.

Q: When was the NDA formed?

The National Democratic Alliance was formed in the weeks around the 1998 general election, as the BJP converted its pre-poll seat-sharing arrangements with regional parties into a governing coalition after the results of February and March 1998. The alliance’s common programme, the National Agenda for Governance, was released in March 1998, and its government, led by Atal Bihari Vajpayee, was sworn in on 19 March 1998 after President K. R. Narayanan examined the letters of support submitted by its constituent parties and supporting parties. The alliance had a prehistory in the BJP’s search for allies after its thirteen-day government of May 1996 collapsed for want of partners, and in the seat adjustments negotiated state by state before the 1998 polling, but the alliance as a named coalition with a common agenda dates from March 1998.

Q: What is the National Agenda for Governance?

The National Agenda for Governance is the common programme adopted by the National Democratic Alliance in March 1998, before the Vajpayee government took office, as the charter on which the coalition would govern. It set out the alliance’s positions on agriculture, water, education, internal security, economic management, centre-state relations and electoral reform, and it replaced the BJP’s own election manifesto as the government’s working programme. Its most consequential feature was its treatment of the BJP’s three signature demands. Where the manifesto promised the abrogation of Article 370, the agenda promised examination of the question. Where the manifesto promised a Uniform Civil Code, the agenda promised consideration directed toward consensus. Where the manifesto promised a Ram temple at Ayodhya, the agenda promised exploration of consensual, legal and constitutional means. Those substitutions made the document the founding record of the coalition’s terms.

Q: Did Vajpayee abandon Hindutva for coalition?

The textual record supports deferral rather than abandonment. The BJP contested the 1998 election on a manifesto that retained the abrogation of Article 370, the Uniform Civil Code and the Ayodhya temple, and the party never removed those demands from its own programme during the coalition years. What changed was the government’s charter: the National Agenda for Governance deferred all three as the price of the partners’ support, and the governments of 1998 to 2004 took no legislative step on any of them. Inside the movement the deferral was contested, with figures such as K. N. Govindacharya and organisations such as the Vishwa Hindu Parishad charging the leadership with dilution, and after 2014 the party, governing with its own majority, restored and in large part enacted the deferred demands. The sequence is best described as a tactical retreat from Hindutva’s programme in government, executed to hold a coalition, rather than a renunciation of the programme itself.

Q: Why did the BJP soften Hindutva in 1998?

The softening was dictated by arithmetic and by the lesson of 1996. In May 1996 the BJP, as the single largest party, formed a government that lasted thirteen days because no significant party would join a formation identified with the party’s three signature demands. By 1998 the BJP held 182 seats, ninety short of a majority, and every route to the remaining seats ran through regional parties answerable to electorates, including Muslim, Sikh and Christian voters and voters in states with strong regional identities, for whom the abrogation of Article 370, a uniform civil code or an Ayodhya temple campaign were unacceptable governing commitments. The Telugu Desam Party, the All India Anna Dravida Munnetra Kazhagam and the Shiromani Akali Dal each required the deferrals as the condition of their support. The National Agenda for Governance recorded those conditions, converting the party’s demands into examinations, considerations and explorations that the partners could defend at home.

Q: How is the Vajpayee NDA different from the Modi government?

The structural difference is the majority. The Vajpayee governments of 1998 to 2004 rested on a coalition in which the BJP held 182 of the alliance’s seats and governed under a common agenda that deferred the party’s three signature demands, because regional partners could veto them and did. The Modi governments from 2014 rested on a single-party BJP majority, 282 seats in 2014 and 303 in 2019, within an alliance the party dominated, so no partner’s veto governed the programme. The consequences appeared in the record: the Article 370 abrogation of August 2019, the Ayodhya temple’s advance after the Supreme Court judgment of November 2019, and the reopening of the Uniform Civil Code question all belonged to the majority era and had no equivalent in the coalition era. The two periods also differed in style, the Vajpayee governments working through a coordination committee and a consensus idiom that the later period’s centralised decision-making displaced.

Q: What were the three core BJP commitments deferred in 1998?

The three commitments were the abrogation of Article 370, the enactment of a Uniform Civil Code, and the construction of a Ram temple at Ayodhya. All three stood in the BJP’s 1998 election manifesto, descending from the programme the party and its predecessor, the Bharatiya Jana Sangh, had carried for decades: Article 370’s removal as a step toward the full integration of Jammu and Kashmir, a single civil code replacing religion-specific personal laws in marriage, divorce, inheritance and adoption, and a temple at the Ayodhya site claimed as the birthplace of Ram. The National Agenda for Governance deferred each in turn, substituting examination for abrogation, consideration and consensus for enactment, and exploration of legal and consensual means for construction. None of the three was acted on by the governments of 1998 to 2004, and all three returned to the centre of the party’s programme under the majorities won in 2014 and 2019.

Q: How many seats did the NDA win in the 1998 general election?

The National Democratic Alliance won 252 of the 543 Lok Sabha seats in the 1998 general election, on the Election Commission of India’s statistical report. The BJP contributed 182 of those seats, won on 25.59 percent of the national vote, and its allies contributed 70, among them the All India Anna Dravida Munnetra Kazhagam with 18, the Samata Party with 12, the Biju Janata Dal with 9, the Shiromani Akali Dal with 8 and the Trinamool Congress with 7. The Congress won 141 seats on a slightly higher vote share of 25.82 percent. The alliance’s 252 seats left it twenty short of the 272 halfway mark, and the government was formed after the Telugu Desam Party, which had won 12 seats, extended support from outside, carrying the alliance’s documented support past the majority line that President K. R. Narayanan required to see demonstrated.

Q: Which regional parties anchored the 1998 NDA alongside the BJP?

Three regional parties anchored the alliance. The All India Anna Dravida Munnetra Kazhagam of J. Jayalalithaa, with 18 seats, was the largest ally inside the alliance and the pivot of its Tamil Nadu front. The Telugu Desam Party of Chandrababu Naidu, with 12 seats, supported the government from outside and supplied weight beyond its tally, including the Speakership for its nominee G. M. C. Balayogi. The Shiromani Akali Dal of Punjab, with 8 seats, entered the cabinet as the alliance’s most durable partner. Around them stood the Samata Party in Bihar, the Biju Janata Dal in Orissa, the Trinamool Congress in West Bengal and the Shiv Sena in Maharashtra, with a tail of smaller parties completing the distance between the BJP’s 182 seats and the alliance’s 252. Each anchor party’s electorate set a condition the common agenda had to respect, and the deferrals in the National Agenda for Governance map onto those conditions.

Q: How did Chandrababu Naidu shape the 1998 NDA from outside the government?

Chandrababu Naidu, the Telugu Desam Party leader and Chief Minister of Andhra Pradesh, was the alliance’s indispensable outside supporter and its southern broker. His party’s 12 seats, pledged from outside the government, carried the alliance’s documented support past the majority mark when President K. R. Narayanan examined the claim in March 1998, and his conditions for that support, which included the deferral of the BJP’s three signature demands, were written into the National Agenda for Governance. Naidu took no cabinet office, which left him free to arbitrate the alliance’s disputes and to criticise its government when his state’s interests required, as his public demand for the Gujarat Chief Minister’s removal in April 2002 demonstrated. His party’s strength rose to 29 seats in 1999, deepening his leverage in the second Vajpayee government, and his use of the supporting role, converting a regional mandate into national arbitration power, became a template for regional leaders in later coalitions.

Q: Why did Jayalalithaa withdraw support from the Vajpayee government in 1999?

J. Jayalalithaa withdrew the All India Anna Dravida Munnetra Kazhagam’s support on 14 April 1999 after a year of escalating demands that the government declined to meet. Her stated conditions included the dismissal of the rival Dravida Munnetra Kazhagam government in Tamil Nadu, action on the Cauvery waters dispute favourable to her state, and the removal or transfer of officials and decisions connected with the corruption prosecutions she faced. The government accepted some demands for deference in consultation and refused the dismissal of an elected state government and interference in her legal cases, which it treated as beyond the coalition’s terms. Her withdrawal letter to President K. R. Narayanan reduced the government to a minority, and Narayanan directed Vajpayee to prove his majority on the floor of the Lok Sabha. The withdrawal thus converted a partner’s grievances into the confidence vote of 17 April 1999 that ended the government’s first life.

Q: What did the BJP’s 1998 manifesto promise on the Ayodhya temple?

The BJP’s manifesto for the 1998 general election reaffirmed the party’s commitment to facilitate the construction of a Ram temple at Ayodhya, at the site the Ram Janmabhoomi movement claimed as the birthplace of Ram and where the Babri Masjid had stood until its demolition on 6 December 1992. The commitment descended from the party’s Palampur resolution tradition and the mobilisation of 1989 to 1992, and the manifesto carried it as a settled element of the party’s identity rather than as a negotiable policy. The significance of the promise in 1998 lies in its fate. Within weeks of the election, the National Agenda for Governance replaced the manifesto as the government’s charter and substituted, for the construction commitment, an undertaking to explore consensual, legal and constitutional means on the Ayodhya question, which left the dispute with the Allahabad High Court and deferred the temple for the coalition’s life.

Q: How did the NDA coordination committee settle disputes among allies?

The National Democratic Alliance maintained a coordination committee, chaired by Prime Minister Atal Bihari Vajpayee, at which the leaders of the constituent parties brought grievances, portfolio disputes and policy objections before they could become floor crises. George Fernandes, the Samata Party leader and Defence Minister, served as the alliance’s convenor and its travelling intermediary, carrying messages between the Prime Minister and partner leaders, including Chandrababu Naidu in Hyderabad and J. Jayalalithaa in Chennai. The committee’s method was accommodation inside the boundary drawn by the National Agenda for Governance: demands that the common programme permitted, such as portfolio adjustments, state-specific concessions and consultation rights, were negotiable, and demands that crossed the programme’s fences, such as Jayalalithaa’s call for the dismissal of the Tamil Nadu government, were resisted until the partner chose between the boundary and exit. The proceedings of the committee after March 1998, preserved in party archives, record a mechanism whose achievement was the prevention of votes the government might lose.

Q: What did the Vajpayee government do on the Uniform Civil Code between 1998 and 2004?

Across its six years the Vajpayee government took no legislative or preparatory step toward a Uniform Civil Code. No bill was introduced in either House, no draft code was commissioned from the Law Commission, and no cabinet decision on the subject was recorded. The National Agenda for Governance had committed the alliance only to consideration of the question directed toward consensus, and the government treated that formula as its instruction, answering parliamentary questions and judicial observations with expressions of sympathy rather than proposals. When the Supreme Court’s judgment in the John Vallamattom case of 2003 remarked on the desirability of a common civil code, the government welcomed the sentiment and announced no measure. The code remained in the BJP’s own programme documents throughout the period, and its absence from the government’s record was the clearest single instance of the distance between the party’s manifesto and the coalition’s charter.

Q: By what margin did the 1998 Vajpayee government lose its confidence vote in April 1999?

The government lost the confidence vote of 17 April 1999 by a single vote, 269 votes in its favour against 270 opposed, the narrowest defeat of a government in the Lok Sabha’s history to that point. The vote followed the All India Anna Dravida Munnetra Kazhagam’s withdrawal of support on 14 April, which removed the government’s majority. The margin turned on individual decisions inside the chamber, including the vote of Giridhar Gamang, a Congress member who had become Chief Minister of Orissa and whose eligibility Speaker G. M. C. Balayogi upheld, and the Bahujan Samaj Party’s reversal, announced by Mayawati during the debate, of its assurance that its members would support the government. After the defeat the opposition failed to assemble an alternative majority, the Congress president Sonia Gandhi’s claim of 272 supporters could not be substantiated, and the Lok Sabha was dissolved for the election of 1999.

Q: What was Pokhran-II and when did the Vajpayee government conduct it?

Pokhran-II was the series of five nuclear explosive tests the Vajpayee government conducted at the Pokhran range in Rajasthan on 11 and 13 May 1998, under the codename Operation Shakti, less than two months after the government took office. The tests ended the policy of nuclear ambiguity maintained since India’s first test of 1974 and declared India’s status as a nuclear weapons state, drawing sanctions from the United States, Japan and other countries and a condemnation in the United Nations Security Council’s Resolution 1172 of June 1998. Within the argument of the period, the tests mattered as evidence of the government’s governing register: a decision of civilisational assertion in the field of national security, where the coalition’s partners raised no objection, taken by a government that was simultaneously deferring its movement’s signature demands in the fields where the partners’ vetoes applied. The diplomatic recovery that followed, conducted by Jaswant Singh in dialogue with the United States, occupied the government through 2000.

Q: How did the BJP’s state units expand during the Vajpayee years?

The BJP’s state organisation deepened across the coalition years on the strength of national incumbency, a growth documented by Walter Andersen and Shridhar S. Damle in The RSS: A View to the Inside of 2018. The party entered the Orissa government with the Biju Janata Dal in March 2000, held Gujarat through a change of chief minister in October 2001 and an election victory in December 2002, and won Madhya Pradesh, Chhattisgarh and Rajasthan in the assembly elections of December 2003, recovering heartland states it had lost a decade earlier. It emerged as the single largest party in Karnataka in 2004. Ministerial office at the centre gave its cadres administrative experience and its state units the credibility of a governing party, drawing entrants from professional and political backgrounds that opposition status had not attracted. The expansion proceeded in the same years in which the party’s national programme stood formally deferred, which is the organisational half of the period’s paradox.

Q: What fiscal and infrastructure programmes defined the Vajpayee government’s record?

The record was defined by disinvestment, fiscal legislation and road building. A Department of Disinvestment, created in December 1999 and raised to a ministry under Arun Shourie, carried through strategic sales including Bharat Aluminium Company in 2001 and Maruti Udyog from 2002, against the opposition of the Sangh Parivar’s own Swadeshi Jagran Manch. The Fiscal Responsibility and Budget Management Act of 2003 wrote deficit discipline into statute. The National Highways Development Project, launched in 1998, joined Delhi, Mumbai, Chennai and Kolkata in the Golden Quadrilateral and became the government’s signature construction programme, extended to rural roads by the Pradhan Mantri Gram Sadak Yojana of 2000. Around those anchors stood the New Telecom Policy of 1999, the opening of insurance under the Insurance Regulatory and Development Authority Act of 1999, the tax reform work of the Kelkar task forces of 2002, and the state finance ministers’ empowered committee of 2000 that prepared the ground for a goods and services tax.

Q: What did the RSS expect from the BJP during the coalition years?

The Rashtriya Swayamsevak Sangh expected the BJP’s governments to advance the movement’s programme as circumstances allowed, and its conduct through the coalition years combined public patience with periodic pressure. Its affiliate, the Vishwa Hindu Parishad, pressed the government on the Ayodhya temple timetable and accused it of delay, most sharply around the proposed ceremony of March 2002, which the government contained under court supervision. Its economic platform, the Swadeshi Jagran Manch, campaigned against the disinvestment programme. Inside the party, the reported description of Vajpayee as the BJP’s mask by the general secretary K. N. Govindacharya, which cost him his position in 2000, voiced the suspicion that the government’s face and the movement’s programme had parted company. The Sangh’s chiefs, Rajendra Singh and, from March 2000, K. S. Sudarshan, restated the movement’s patience as finite without setting a public deadline. The expectations were therefore real, voiced and unmet on the three deferred demands for the coalition’s duration.

Q: How did the 1998 NDA differ from the United Front it replaced?

The two formations differed in their cores and in their documents. The United Front of 1996 to 1998 was a coalition of regional and left parties without a dominant member, sustained by Congress support from outside, which produced two Prime Ministers, H. D. Deve Gowda and I. K. Gujral, and governed under a Common Minimum Programme negotiated among near-equals. The National Democratic Alliance was built around a single lead party holding 182 seats, more than the next several partners combined, and governed under a National Agenda for Governance whose central function was to subordinate that lead party’s own manifesto to the coalition’s terms. The United Front’s weakness was its core’s smallness, which left it dependent on an outside supporter that twice withdrew. The Alliance’s strength was its core’s size, which let it survive the loss of its largest ally in 1999, dissolve on its own terms, and return with a full-term tally. The BJP had studied the United Front’s architecture and corrected its foundations.