The land acquisition ordinance of December 2014 is usually remembered in one sentence, and the sentence is a verdict rather than a description: a government tried to take farmers’ land for industry, farmers resisted, and the government was forced to retreat. Every element of that sentence corresponds to something real, and the sentence as a whole is misleading, because it describes only the parliamentary stage of the episode and says nothing about the outcome. The amendment failed in Parliament, completely and publicly. Much of what it proposed is nonetheless law in large parts of India, delivered by state legislatures acting under a proviso of the Constitution and by a Constitution Bench of the Supreme Court reading a single provision of the unamended Act. This article is an account of the whole arc, from an ordinance promulgated on the last day of 2014 through its successive re-promulgations to its lapse in August 2015, and then through the three routes by which its objectives were pursued after its death.

Stated in terms both sides would accept, the disagreement was this. The Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act, 2013 had made acquisition slower and more conditional than at any earlier point in Indian law, by requiring a social impact assessment before acquisition, the consent of a stated proportion of affected families where land was taken for private companies or public private partnerships, and compensation at multiples of market value with a full rehabilitation and resettlement package. The government of the day maintained that these conditions, applied to defence projects, rural roads, electrification, affordable housing and industrial corridors, would obstruct purposes that almost nobody opposed in principle, and that the Act needed a set of exemptions for such projects. Its opponents maintained that the exemptions were drawn widely enough to swallow the rule, that a project list covering most of what governments acquire for would leave the consent and assessment safeguards operating only at the margins, and that in any event a contested change of this size ought not to be carried by ordinance at all. Both of those positions were held by serious people, both were argued in Parliament and outside it, and the episode cannot be understood by adopting either side’s summary of the other.

The subject matter explains the temperature. Land in India is not an asset class among others. Close to half of all workers were engaged in agriculture at the time, holdings were small and shrinking across generations, a family’s land was frequently its only store of wealth, its only collateral and its only claim on the village’s future, and the state’s power to take it compulsorily had been exercised for a century under a statute that asked almost nothing of the taker. A proposal to loosen the conditions on taking was therefore never going to be debated as a drafting exercise, whatever its clauses said, and a proposal to tighten them had carried the 2013 Act to enactment on the strength of two decades of agitation against displacement. Both sides in 2015 knew they were arguing about the terms on which development would be allowed to consume the countryside, and both argued accordingly.

What follows separates the questions that the public argument habitually merges. There is a question about the text: what did the ordinance actually change, provision by provision, and what did it leave untouched. There is a question about merits: whether the exemptions were a reasonable calibration of a cumbersome statute or the removal of its operative safeguards. There is a question about process, which is distinct from both: whether keeping a contested amendment alive by repeated re-promulgation, while a joint committee of the two Houses was examining the very bill that embodied it, was a legitimate use of the ordinance power. And there is a question about consequences: what became of each proposal after the ordinance died. The answers are different in kind. The first is a matter of record. The second is genuinely contested and is presented here through the strongest statements of each side. The third has since been answered by a seven-judge bench in terms that leave little room for argument. The fourth is the most interesting, because it is the part of the story that the standard account omits entirely.

The 2013 settlement the ordinance set out to amend

The ordinance sought to amend a statute that had itself been in force for exactly one year when the amendment was promulgated. The 2013 Act had replaced the Land Acquisition Act, 1894, under which acquisition had proceeded for more than a century on the strength of a declaration of public purpose, a collector’s award and a compensation code that carried no obligation to rehabilitate or resettle anyone. The 2013 settlement changed the structure in four ways that matter for everything that follows. It required, as a preliminary to acquisition, a social impact assessment study under Chapter II of the Act, in which the social costs of the project, the number of affected families and the question whether the land sought was the bare minimum required were examined and placed before an expert group. It imposed, in Chapter III, a special provision to safeguard food security, restricting the acquisition of irrigated multi-cropped land. It required the prior consent of at least eighty per cent of affected families where land was acquired for a private company, and at least seventy per cent where it was acquired for a public private partnership project, as a condition recorded in the second proviso to Section 2(2) of the Act. And it rebuilt the economics of acquisition, with compensation computed on market value multiplied by a factor, a solatium of one hundred per cent, and a rehabilitation and resettlement entitlement for affected families, including families who owned no land but lost their livelihoods.

The statute itself, its full machinery, its compensation arithmetic and its consent procedure, is the subject of the pillar account of the LARR Act 2013, and this article takes the Act’s provisions only as far as the amendments reach into them. Two features of the settlement’s politics matter for 2015. The first is that the 2013 Act had not been a partisan enactment carried against a resisting opposition: it was passed in the last year of the Congress-led government with the support of the Bharatiya Janata Party, then in opposition, whose members spoke for its protections in debate. Throughout the 2015 controversy the Congress returned to that fact, presenting the amendment not as a new government’s legitimate revision of its predecessor’s law but as a party dismantling, within eighteen months, a settlement it had voted for, and the charge of a U-turn did the government’s case real damage because it was documented in the division lists of 2013. The second is that the Act’s compensation code, a market value base multiplied by a factor rising with the rural character of the land, with a solatium of one hundred per cent on the resulting figure, had reset expectations in the countryside very quickly: by the end of 2014 a farming family anywhere in India could state, at least roughly, what the 2013 Act promised it, and a proposal understood to withdraw parts of that promise was always going to be received as a taking from the family rather than as an adjustment between governments. What matters here is the design choice the 2013 Parliament made and the 2014 government proposed partly to reverse. The 2013 Act treated consent and assessment not as preliminaries that a well-intentioned acquiring government could be trusted to compress, but as the substantive protection itself: the mechanism by which a family whose land was wanted could force the acquiring authority to justify the taking, to count the people it displaced, and to negotiate against a statutory floor rather than against its own weakness. Every contested amendment in the ordinance can be located on that design choice. None of them touched the compensation multiples. None of them touched the rehabilitation entitlements. All of the politically explosive ones touched consent, assessment, or the clocks by which the Act measured delay and default.

The machinery deserves a closer look, because the amendments are unintelligible without it. A social impact assessment under Chapter II was not a formality: the study was to be conducted in consultation with the local bodies in the affected area, a public hearing was to be held, the study report was to be appraised by an independent expert group that included social scientists and representatives of the local bodies, and the appropriate government was then to examine the report and record its decision, with the assessment directed to whether the project served a public purpose, whether the land proposed was the bare minimum required, and what the social cost would be and on whom it would fall. Consent, where required, was to be obtained before the preliminary notification, from the affected families as the Act defined them, a definition that reached beyond title holders to families whose primary source of livelihood stood affected, which is why the consent of a village could not be assembled from its landowners alone. It is equally part of the machinery that the 2013 Act had retained, in Section 40, an urgency power under which the government could dispense with the Chapter II assessment where land was needed for the defence of India, national security or relief in a natural calamity. The 2013 Parliament had therefore already accepted that some acquisitions cannot wait for a study; what it had not accepted was that the dispensing judgement should extend beyond urgency to categories of ordinary economic policy. The ordinance’s Section 10A was precisely that extension, which is why the argument about it was an argument about the location of the dispensing power and not about the existence of emergencies.

One feature of the 2013 Act stood apart from the controversy and yet explains its timing. Section 105 of the Act kept thirteen other central enactments, listed in the Fourth Schedule, outside the Act’s operation: the statutes under which land is acquired for national highways, railways, metro railways, atomic energy, electricity, coal bearing areas, petroleum and mineral pipelines, mines, the Damodar Valley Corporation, ancient monuments, tramways, the resettlement of displaced persons and the requisitioning of immovable property. These were not obscure laws. They covered the largest land-taking programmes of the Indian state, and under them an affected family received neither the 2013 compensation scale nor its rehabilitation package. Section 105(3) therefore directed that the Central Government bring the compensation, rehabilitation and resettlement provisions of the thirteen Acts into consonance with the 2013 Act by notification within one year of the Act’s commencement. The Act commenced on 1 January 2014. The year ran out at the end of 31 December 2014. That deadline, and not any parliamentary timetable, is the reason the ordinance bears the date it does, and it is the reason the least controversial item in the ordinance travelled in the same instrument as the most controversial ones, a coupling that shaped the politics of the next eight months.

What the land acquisition ordinance actually changed

The Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement (Amendment) Ordinance, 2014 was promulgated by the President on 31 December 2014. Stripped of characterisation, it did seven things, and precision about the list matters because public argument on both sides persistently described a different instrument from the one that was promulgated: one side describing a repeal of the 2013 Act, which it was not, and the other describing a set of minor drafting corrections, which it was not either.

The central provision inserted a new Section 10A, in a new Chapter IIIA, empowering the appropriate government, by notification, to exempt projects in five categories from the operation of Chapter II, which contains the social impact assessment, and Chapter III, which contains the food security restriction on irrigated multi-cropped land. The five categories were these: projects vital to national security or defence and defence production; rural infrastructure including electrification; affordable housing and housing for the poor; industrial corridors; and infrastructure projects, including projects under public private partnership in which the ownership of the land continued to vest in the government. For the same five categories, the ordinance also removed the consent requirement in Section 2(2), so that land could be acquired for a project within the categories without the prior consent of eighty or seventy per cent of affected families that the 2013 Act would otherwise have demanded. The exemption was a power to be exercised by notification for projects within the categories, not a blanket deletion; but the categories were the operative decision, because once a project fell within one of them, the two preliminary disciplines of the 2013 Act, assessment and consent, could both be switched off for it together.

The second change was the one nobody marched against. The ordinance amended Section 105 so that the compensation, rehabilitation and resettlement provisions of the 2013 Act applied directly to acquisitions under the thirteen Fourth Schedule enactments, achieving by amendment what Section 105(3) had directed to be done by notification within the year. A family whose land was taken for a national highway or a railway line would receive the 2013 Act’s compensation scale and its rehabilitation package, rather than the thinner entitlements of the special statute. This was the fulfilment of a commitment the 2013 Parliament itself had made, and the opposition parties, who had written Section 105(3) when in government, could not and did not oppose it. Its presence inside the same ordinance as the consent exemptions was, from the government’s side, an argument that the instrument as a whole was farmer-facing; from the opponents’ side, it was the wrapping in which a dilution was being carried.

The remaining changes were narrower and mostly procedural in character, though two of them carried real consequences in litigation. The ordinance substituted the expression private entity for private company throughout the relevant provisions, and defined a private entity as any entity other than a government entity or undertaking, including a proprietorship, a partnership, a company, a corporation, a non-profit organisation or any other entity under any law in force. The 2013 Act’s consent and rehabilitation architecture had been drafted around acquisition for a private company, and organisational forms outside the Companies Act, including the partnership and the trust and the special purpose vehicles through which infrastructure is frequently built, sat awkwardly within it; the amendment regularised their position. It also updated the Act’s reference to the Companies Act, 1956 to the Companies Act, 2013. It removed the exclusion by which the 2013 Act had kept acquisition for private hospitals and private educational institutions outside the definition of public purpose, bringing such acquisition within the Act’s scheme. It amended Section 101, under which land acquired but not utilised must be returned to the original owners or to the land bank, replacing the flat period of five years with the later of five years or any period specified at the time of setting up the project, on the stated reasoning that large projects have gestation periods fixed in advance and longer than five years. It amended the retrospective provision in Section 24(2), under which proceedings under the 1894 Act lapsed where an award had been made five years or more before the 2013 Act commenced and possession had not been taken or compensation had not been paid, by excluding from the computation of that five-year period any period during which the acquisition proceedings were held up by a stay or injunction of a court, any period specified in a tribunal’s award for taking possession, and any period during which possession had been taken but the compensation lay deposited in a court or an account. And it recast Section 87, which in the 2013 Act deemed the head of a government department guilty of an offence committed by the department unless he proved absence of knowledge or due diligence, substituting a requirement that a government official not be prosecuted for an offence under the Act without the prior sanction of the government, on the pattern of Section 197 of the Code of Criminal Procedure, 1973.

Did the exempted categories escape rehabilitation and resettlement as well?

No. The exemption in the new Section 10A reached Chapter II, containing the social impact assessment, Chapter III, containing the food security restriction on irrigated multi-cropped land, and the consent requirement in Section 2(2). The compensation code and the rehabilitation and resettlement entitlements applied to exempted projects exactly as to any other acquisition under the Act.

That negative is worth dwelling on, because a striking amount of the public argument proceeded as though the ordinance had cut the price of land. It had not. The multiplier on market value, the hundred per cent solatium, the rehabilitation package with its housing, annuity and employment elements, all survived the ordinance intact, and for acquisitions under the thirteen scheduled Acts the ordinance raised the entitlement rather than lowering it. What the ordinance withdrew from the five categories was not money but process: the study that counted the displaced and tested whether the land was needed, the food security ceiling on the best agricultural land, and the vote of the affected families. The argument about the ordinance is therefore an argument about whether process protections of that kind are expendable ceremony or are themselves the substance of the 2013 settlement. It is not, on any accurate reading of the text, an argument about compensation.

The Section 24 amendment requires its litigation setting to be understood. The lapse provision had been construed in 2014, in Pune Municipal Corporation v. Harakchand Misirimal Solanki, in the manner most favourable to landowners: the Supreme Court there held that the deposit of compensation in the government treasury did not amount to payment for the purposes of Section 24(2), so that proceedings in which the money had sat in a treasury rather than being paid or deposited in court were liable to lapse, and landowners across the country began invoking the provision to reopen acquisitions that were years or decades old and to claim compensation on the 2013 scale. Acquiring authorities faced a wave of claims founded not on any failure to act but on the form in which the compensation had been held. The ordinance’s exclusions, of stay periods, of tribunal-specified periods and of deposit periods, were drafted directly against that case law and that wave of claims; they were the most technical items in the instrument and, in the government’s presentation, the most obviously corrective. They were also the items whose substance a court could later supply by construction, as the final section of this article records, because the questions they raised were questions about the meaning of a provision already enacted.

The private entity change deserves a similar precision, because it was described in the public argument as the opening of acquisition to any private interest whatever, which it was not. Acquisition under the 2013 Act is acquisition by the state for a public purpose; the Act’s machinery applies where land is acquired for the government’s own purposes, for public sector companies, and for private companies for purposes within the Act’s definition. The expression private company had imported the Companies Act definition, with the consequence that a project executed by a partnership firm, a proprietorship, a society or a non-profit organisation, forms through which a great deal of Indian infrastructure and social infrastructure is in fact built, sat outside or awkwardly across the consent architecture. Substituting private entity, with the definition the ordinance supplied, brought those forms inside the same architecture, consent included, rather than creating a new power to acquire for them. The critics’ objection, fairly stated, was not that the change was a grab in itself but that it widened the class of beneficiaries at the same moment as the categories removed the safeguards for the most consequential projects, and that the two changes had to be read together.

Two limits on the exemption power should also be recorded, because they were real even if they did not satisfy the critics. The five categories were categories of purpose, and the exemption had to be notified for projects within them; the ordinance did not disapply the assessment and consent chapters to acquisition generally. And the categories, however widely drawn, were a closed list. A project that fitted none of them, an ordinary private industrial plant outside a notified corridor, for example, remained subject to the full 2013 discipline, including the consent of eighty per cent of affected families. The critics’ answer to both limits was that the list was closed in form and open in substance, since infrastructure and rural infrastructure are descriptions under which a very large share of all acquisition can be arranged. That dispute, about the real breadth of five phrases, is taken up in the assessment below; the textual point is that everything turned on how those phrases would be read in notification practice.

Why an ordinance, and why December 2014

An ordinance is legislation by the executive for a defined emergency window. Under Article 123 of the Constitution, when Parliament is not in session and the President is satisfied that circumstances exist which render it necessary to take immediate action, the President may promulgate an ordinance having the force of an Act; it must be laid before both Houses and ceases to operate six weeks after Parliament reassembles, or earlier if the Houses disapprove it. The device is designed for the situation in which a legal gap opens while the legislature is away and cannot wait for its return. Its use for a programme of contested amendment, where the gap was a deadline the government had known about for twelve months, was the procedural heart of the controversy from the first day.

The government’s justification had two strands, one strong and one contestable. The strong strand was the Section 105 deadline. The direction in the 2013 Act to bring the thirteen scheduled Acts into consonance within one year expired at the end of December 2014, and the government maintained that allowing it to expire unfulfilled would leave highway and railway landowners, among the largest classes of the displaced, permanently on the inferior compensation scales of the special statutes. An ordinance promulgated on 31 December met the deadline. The contestable strand was the inclusion, in the same instrument and under the same claim of urgency, of the consent and assessment exemptions, for which no deadline existed at all. Nothing in the 2013 Act required the five categories to be created by 31 December 2014 or by any other date. Critics, including the National Alliance of People’s Movements in a statement issued within days, asked what the emergency was: the alliance’s release observed that the new law had been in existence for one year and disputed the claim that investments were stuck because of it. It was also reported in the press that the President, before promulgating the ordinance, sought the government’s reasons for the urgency. The question those reports recorded was the question the opposition pressed for the next eight months, and the government’s answers to it, which rested on the volume of stalled projects and the representations of state governments, are examined in the case for the amendments below.

The land ordinance was also not an isolated use of the device. The government that took office in May 2014 legislated by ordinance across its first months, on subjects that included insurance, coal block allocation and citizenship, in several cases because bills were pending or blocked in the Rajya Sabha and in others because it preferred the speed. The pattern mattered to the land fight because it deprived the government of the argument that land was an exceptional case of urgency: by the time the land bill reached the Rajya Sabha, the ordinance route had become, in the opposition’s presentation, the government’s ordinary method of dealing with an Upper House it did not control, and the President’s reported query about the urgency of this particular ordinance was asked against that background. A government that uses an emergency power occasionally is believed when it pleads an emergency; a government that uses it as a standing supplement to its legislative programme finds the plea examined, which is what happened here.

There was also a parliamentary arithmetic that the ordinance route postponed rather than solved. The Bharatiya Janata Party had won a majority in the Lok Sabha in May 2014, the first single-party majority in that House in three decades, and could carry ordinary legislation there on its own strength. In the Rajya Sabha it was in a minority, and the opposition parties together commanded the numbers to defeat a bill or to amend it against the government’s wishes. An amendment to the 2013 Act therefore had to either attract cross-party support in the Upper House or be kept alive by executive action while that support was sought. The ordinance bought time; whether the time was used to build the support, or merely to defer the admission that it did not exist, is one of the things the sequence of 2015 shows. The opposition’s procedural demand, made from the first sitting of the budget session, was that a change of this character should go through the committee system before either House was asked to vote: the Congress party announced that it would vote against the bill in the Lok Sabha itself unless the government reverted to the 2013 text or sent the bill to a standing committee, and the memorandum later presented to the President repeated the complaint that the standing and select committee institutions were being bypassed. The government declined, relying on its majority to pass the bill in the Lower House first, and a statutory resolution disapproving the ordinance was moved and defeated there. The demand for committee scrutiny was nevertheless met in the end, in the stronger form of a joint committee of both Houses, but by then the bill had already passed one House and the ordinance had already been re-issued once, an order of events that let the opposition argue that scrutiny was being offered only after the decisions it might have influenced had been taken.

Eight months in Parliament and outside it

The replacement bill, the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement (Amendment) Bill, 2015, was introduced in the Lok Sabha on 24 February 2015 by the Minister for Rural Development, Birender Singh, and it reproduced the ordinance. The day before its introduction, on 23 February 2015, Anna Hazare began a two-day protest against the ordinance at Jantar Mantar in Delhi, supported by farmers’ organisations, with Medha Patkar among the participants; the symbolism was deliberate, since Hazare’s name was associated in the public mind with the movement politics that had framed the preceding decade, and his presence announced that the land question would be fought outside Parliament as well as inside it. The government, for its part, did not treat the bill as non-negotiable. On 9 March 2015 it circulated nine official amendments of its own to its own bill, a concession package that remains the clearest evidence of what the government itself, on reflection, thought indefensible or inessential in the December text.

The nine amendments did the following. They removed the change that had brought private hospitals and private educational institutions within public purpose, restoring the 2013 exclusion. They confined the industrial corridor category to corridors set up by the government or government undertakings, and limited acquisition for a corridor to land up to one kilometre on both sides of the designated railway line or road. They deleted social infrastructure from the fifth exempted category, leaving infrastructure projects including public private partnership projects where ownership of the land remained with the government. They added, to the power to exempt a project from assessment and from the multi-cropped land restriction, a condition that the government must first ensure that the extent of land being acquired was in keeping with the minimum required for the project. They required the government to survey its wasteland, including arid land, and to maintain a record of it. They narrowed the rehabilitation employment entitlement so that compulsory employment was assured to at least one member of an affected family of a farm labourer. They required the Land Acquisition, Rehabilitation and Resettlement Authority to hold its hearings in the district in which the acquisition was taking place. They refined the Section 24 exclusion so that the deposit period counted out was a period in which compensation lay in a designated account. And they recast the prosecution provision so that a government employee could be prosecuted where the procedure in Section 197 of the Code of Criminal Procedure, 1973 was followed, that procedure itself requiring the prior sanction of the government. In the debate the government leant on the states and on the character of the exempted purposes. M. Venkaiah Naidu, the Parliamentary Affairs Minister, intervened to insist that land was needed for irrigation, rural housing, rail lines and roads, and that the government would do nothing to hurt the farming community; Birender Singh, replying for the ministry, turned the authorship argument around, saying that the Congress had brought its 2013 bill with an eye on the general election it then lost, and that a party which had lost credibility was attempting to win back a constituency. The exchanges are worth recording because they show the debate’s actual structure: the government argued from purposes, listing works nobody in the chamber opposed, and the opposition argued from coverage, asking what share of acquisition the purposes would in practice absorb, and the two arguments passed each other for two days without ever occupying the same ground. The Lok Sabha passed the bill, with these nine amendments, on 10 March 2015.

The bill then met the arithmetic. It was not taken up and carried in the Rajya Sabha, where the government lacked the numbers and where the opposition, unusually united, had resolved to defeat it or to send it to a select committee. On 17 March 2015, a week after the Lok Sabha vote, Sonia Gandhi led members of Parliament from fourteen opposition parties in a march from Parliament House to Rashtrapati Bhavan, where they presented a memorandum to President Pranab Mukherjee asking him to intervene and not to allow the government to proceed with the amendments. The memorandum’s complaint joined the merits to the procedure: the amendments did away with the social impact assessment, which the parties described as an essential safeguard against the diversion and excess taking of land, and with the consent requirement, and the bill had been pushed through the Lok Sabha without prior consideration by a standing committee. The Congress party, which had enacted the 2013 Act in the last year of its own government, treated the statute as its legacy and its leader Rahul Gandhi made the contest a national one, describing the government at a farmers’ rally at Ramlila Maidan in April 2015, and again in the Lok Sabha debate on the agrarian situation, as a suit-boot ki sarkar, a government of men in suits and boots, governing for the suited at the expense of the farming majority. The phrase stuck, because it converted a clause-by-clause argument about assessment chapters into a characterisation of the government’s allegiances, which was precisely the conversion the government’s ministers complained the debate had undergone.

Meanwhile the ordinance clock was running. The December ordinance would cease to operate six weeks after Parliament reassembled for the budget session unless replaced by an Act, and no Act was forthcoming. On 3 April 2015 the President promulgated the amendment ordinance afresh, as Ordinance 4 of 2015, this time incorporating the nine amendments the Lok Sabha had accepted, so that the executive law in force matched the text the Lower House had passed. On 13 May 2015 the bill, in the form in which it had passed the Lok Sabha, was referred to a Joint Committee of the two Houses, a committee of thirty members, twenty from the Lok Sabha and ten from the Rajya Sabha, chaired by S. S. Ahluwalia of the Bharatiya Janata Party, and charged with reporting by the first week of the monsoon session. Referral to a joint committee is the parliamentary system’s standard device for a bill on which tempers are high and facts are disputed: it hears the ministries, the states and the affected interests, it proceeds clause by clause, and its report, if it commands support across parties, carries a weight on the floor that a ministry’s assertion does not. Farmers’ organisations appeared before the committee and told it, as the Hindu BusinessLine reported their representations in June 2015, that the amendments were a step backwards towards the regime of the 1894 Act; they also complained that the ordinance had been promulgated a third time while the committee was seized of the bill, which they characterised as contempt of parliamentary norms and as undermining the committee’s authority.

That third promulgation had occurred on 30 May 2015, when the April ordinance was re-promulgated as the Second Ordinance of 2015, Ordinance 5 of 2015, again to prevent the amendments from ceasing to operate. One road was available that the government never took, and its reasons are instructive. Under Article 108, where a bill passed by one House is rejected by the other, or more than six months elapse without the other House passing it, the President may summon a joint sitting, at which the combined strength of the two Houses decides; a government with a large Lok Sabha majority ordinarily prevails in such a sitting, and the device exists for exactly the deadlock the land bill presented. The government did not seek it. The six-month condition could not have matured before the autumn, the bill having passed the Lok Sabha only in March; a joint sitting demanded a confrontation on a single afternoon in which the government’s entire majority would be deployed to override the Upper House on a farmers’ question, weeks from a state election; and the joint committee’s pendency gave the government a standing reason to wait. The choice reveals the government’s own assessment of its position: it preferred the ordinance’s quiet continuation, and then the policy’s migration to the states, to the one parliamentary victory the Constitution offered it, because the price of that victory was a public and counted demonstration that the Rajya Sabha’s refusal counted for nothing. Anyone reconstructing the episode from the primary record, the three successive ordinance texts, the bill as passed by the Lok Sabha, the nine official amendments and the committee’s notices, will handle a stack of gazette prints and committee papers, and a browser-based office file viewer for opening the ordinance, bill and committee documents makes that record readable in sequence without specialist software. The committee did not report by the first week of the monsoon session. Its deadline was extended, then extended again: a short extension into August 2015, then to the winter session, then, on a motion moved by its chairman, to the last day of the first part of the budget session of 2016, by which time it had received its sixth extension and the opposition members on it were asking, with some force, what purpose an examination served of a bill whose companion ordinance the government had already abandoned. The report was never presented. The bill remained formally alive on the parliamentary record and politically dead in fact, and it lapsed when the Sixteenth Lok Sabha was dissolved in 2019.

Contemporary reports through July and August recorded one further possibility that came to nothing: that the government, accepting that the bill in its passed form could not survive, was considering restoring the consent clause in a diluted form as the price of a compromise, and that the joint committee’s report, when it came, might be the vehicle for the softened text. The committee route made such a compromise institutionally available, since a recommendation from a body on which the opposition sat could have been accepted by the government without the appearance of surrender, and opposition members could have claimed the restoration as their work. That the possibility was never tested is part of the episode’s waste: the machinery for a negotiated amendment existed, was seized of the bill, and was allowed to sit while the ordinance expired around it, so that the compromise which might have commanded both Houses was never drafted, and the policy went instead to the states and the courts, where no compromise was required and none was offered.

The monsoon session of 2015, which sat from late July into August, was consumed by other confrontations, including opposition demands arising from the Lalit Modi controversy and the Vyapam scandal in Madhya Pradesh, and the land bill was never brought to a vote in the Rajya Sabha during it. At an all-party meeting on the eve of the session, the Prime Minister had indicated that the government would consider suggestions from the opposition on the bill, and had quoted with approval the Samajwadi Party leader Ram Gopal Yadav’s view that it was time to move forward on the issue incorporating suggestions from all sides; the Rural Development Minister indicated that a fresh bill incorporating various changes could be brought. Nothing came of the signals. Government sources briefed through the session that, with the committee not reporting and consensus absent, the ordinance would have to be promulgated again, for what was described as an unprecedented fourth time, and it was also understood within the government’s own camp that it did not wish to force the question before the Bihar assembly elections in the autumn. The choice presented to the cabinet at the end of August was therefore stark: a fourth promulgation, in the teeth of the committee’s pendency and weeks before a state election in which land was a live grievance, or an abandonment dressed as responsiveness to farmers. It chose the second.

The end of the ordinance came in August 2015, and it came from the government rather than from any vote. Through the monsoon session the expectation in the press and in the government’s own camp was that the ordinance would be promulgated a fourth time when the May ordinance expired. On 30 August 2015, in his radio address Mann Ki Baat, the Prime Minister, Narendra Modi, announced the opposite: the ordinance would lapse the next day and he had decided to let it lapse. He said that a great deal of misinformation had been spread about the ordinance and had created fear and confusion among farmers, that the position would revert to what it had been before his government took office, and that the government remained ready to accept any suggestion that favoured farmers. He added the one substantive salvage from the wreck: the extension of the 2013 Act’s compensation, rehabilitation and resettlement benefits to acquisitions under the thirteen Fourth Schedule Acts, which he described as thirteen points meant to give farmers direct financial benefit, would be carried into the rules with effect from that day, so that farmers would not suffer financial loss from the lapse. The instrument was an order issued two days earlier, on 28 August 2015, under Section 113 of the 2013 Act, the power to remove difficulties: the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement (Removal of Difficulties) Order, 2015, which applied the Act’s provisions on the determination of compensation in accordance with the First Schedule, rehabilitation and resettlement in accordance with the Second, and infrastructural amenities in accordance with the Third, to all cases of land acquisition under the enactments specified in the Fourth Schedule, and was deemed effective from 1 January 2015. The May ordinance duly ceased to operate on 31 August 2015, and with it the five categories, the consent exemption, the private entity substitution and the rest of the December programme ceased to be the law of India at the Centre.

The case for the amendments, at full strength

The government’s case deserves to be stated as its proponents stated it, because it was not a foolish case and the amendments were not explicable only as a favour to industry. It rested on four propositions, each with evidence behind it.

The first proposition concerned time. The 2013 procedures were sequential and cumulative: a social impact assessment study, its appraisal by an expert group, a preliminary notification, the hearing of objections, and, for private and partnership projects, the assembly of consent from seventy or eighty per cent of affected families, a category that the Act defined to include livelihood losers as well as title holders. For a single plant on a single site that sequence was burdensome but navigable. For the linear projects that dominate public acquisition, a road, a canal, a transmission line or a railway crossing hundreds of villages and thousands of holdings, the sequence multiplied by every village it touched, and a defect at any point exposed the whole acquisition to challenge. Ministers argued that the Act had been written with the industrial plant in mind and applied to the railway line by accident. Arun Jaitley, the Finance Minister, made the argument in print repeatedly: that a consent requirement designed as a check on acquisition for private profit operated very differently when the acquirer was the state building an irrigation canal, and that the categories chosen for exemption, defence, rural infrastructure, electrification, affordable housing and corridors, were overwhelmingly the categories of public works rather than of private gain.

The second proposition concerned the states. The demand for amendment did not originate only in Delhi industry. In July 2014 the Rural Development Ministry prepared a report for the Prime Minister’s Office based on inputs received from the states, and the changes the states suggested were substantially the changes the ordinance later made: doing away with the consent clause for public private partnership projects, removing the requirement of a mandatory social impact assessment study, and relaxing the retrospective clause. In the consultations the Rural Development Ministry held with the states in 2014, state governments of more than one party told the Centre that the consent and assessment provisions were proving unworkable in administration, and the ministry’s case was that it was responding to the governments that actually acquire land, most acquisition being executed by state machinery for state projects. The Prime Minister’s August announcement itself recorded that the suggestions to reform the law had come from several quarters including the state governments. On this view the ordinance was a federal request being answered by the only legislature that could amend a central Act, and the subsequent conduct of the states, which is examined later in this article, bears the argument out in at least this sense: when the central route closed, state governments of the ruling party’s persuasion and at least one governed by an opponent proceeded to enact much of the same content for themselves.

Within the first proposition sat a sharper one about irrigation, which Jaitley pressed in his public writing on the bill. Irrigation is a work whose beneficiaries are farmers themselves, frequently the same class of farmers whose consent the Act required before a canal could cross their district; a consent architecture built to protect cultivators against the state and the developer operated, in the canal case, to let a minority of cultivators obstruct a work whose whole purpose was to bring water to the remainder. The government generalised the point into a claim about the Act’s asymmetry: the safeguards had been designed against the paradigm of the factory, in which a private beneficiary profits from the taking, and they malfunctioned across the large class of acquisitions in which the beneficiary is the public of the same countryside, the road’s users, the grid’s villages and the canal’s ayacut. Whether the malfunction justified removing the safeguards for the whole class, rather than repairing the class definition, was the contestable step; but the existence of the malfunction, in the linear and public cases, was a point the opposition’s defence of the Act rarely answered directly, preferring to answer the industrial corridor case, where its argument was strongest.

The third proposition concerned the difference between consent as a safeguard and consent as a veto. The 2013 thresholds meant that a minority of affected families, somewhat more than twenty or thirty per cent depending on the project type, could prevent an acquisition that the remainder accepted. The government’s supporters argued that this converted a protective device into a holdout charter, that it rewarded strategic refusal and the political mobilisation of refusal, and that no comparable democracy conditioned a highway or a power line on a supermajority vote of those whose land it crossed, the settled protection elsewhere being generous compensation judicially reviewable rather than a prior vote. The argument that a rural electrification line should not need the consent of a majority of affected families is not obviously wrong, and it was strengthened by the Act’s own internal concession: the 2013 Parliament had itself exempted government projects from the consent requirement altogether, reserving consent for private companies and partnerships, so the principle that consent follows the identity of the beneficiary rather than the fact of displacement was already inside the statute. The ordinance extended that principle to five categories in which the beneficiary was, at least in form, the public.

Running through all four propositions was a claim about the scale of what was at stake. Government speakers put the value of projects stalled for want of land in the region of twenty lakh crore rupees, and presented the amendment as the unlocking of that investment at a moment when the government’s economic programme, of which the Make in India manufacturing initiative was the flagship, depended on factories, corridors and power projects that could not be built without assembled land. The figure was a political figure rather than an audited one, and its weakness is examined in the case against; but the underlying observation, that an acquisition code under which almost no significant acquisition had been completed in the Act’s first year was a problem for a government elected on a development programme, did not depend on the figure being exact. A law can be just in its entitlements and still fail as an instrument if the state cannot in practice acquire under it, and the government was entitled to put that failure, if it could show it, at the centre of its case.

The fourth proposition concerned the balance of the instrument. The ordinance left the compensation code untouched and was the vehicle for the single largest extension of landowners’ monetary rights since the Act itself, the application of the 2013 scales to the thirteen scheduled Acts. The government could and did say that the families displaced by highways and railways, who are far more numerous than the families displaced by industrial corridors, gained more from the December instrument than any family lost by it, since what the five categories lost was a procedure and what the thirteen Acts’ families gained was money and rehabilitation they had never had. Industry associations welcomed the amendments on the narrower ground that land had become the binding constraint on stalled projects, though even sympathetic analysts doubted the drafting: analysts at Ambit Capital wrote that the exempted categories were so broadly drawn that they could cover a majority of the projects for which land is acquired, and asked how rural infrastructure was to be defined and whether a shopping mall could be brought within it. A case can be right in direction and still be badly drawn in its operative phrases, and the breadth objection came from the government’s friends as well as its enemies.

The case against, at full strength

The opposition case also deserves to be stated as its proponents stated it, because it too was not a foolish case, and its central factual claim, that the exemptions were broad enough to consume the safeguard, was a claim about the text rather than about anyone’s motives.

The first proposition was about coverage. The five categories were not five narrow purposes. Infrastructure is the genus of almost everything a modern state builds; rural infrastructure including electrification reaches deep into the countryside where acquisition is most contested; industrial corridors are not lines on a map but development programmes, and the land alongside a designated road or railway line, up to a kilometre on each side in the bill as amended, is among the most commercially valuable land in any district it crosses, destined not merely for the track or the carriageway but for the industry, warehousing and real estate that a corridor is built to attract. Ambit Capital’s question about the shopping mall was not rhetorical mischief. A category system under which the acquiring government classified its own project, by its own notification, and thereby decided for itself whether assessment and consent applied, placed the safeguard at the discretion of the party the safeguard existed to restrain. On the opponents’ account, the honest description of the ordinance was not that it exempted five kinds of project from the 2013 discipline but that it restored official discretion over whether the discipline applied, which was the 1894 settlement in modern dress.

The second proposition was about the pairing. Removing the social impact assessment alone would have left consent; removing consent alone would have left the assessment, with its public counting of the displaced and its expert appraisal of whether the land was the minimum required. The ordinance removed both together for the same categories, and added the removal of the food security restriction on irrigated multi-cropped land, the Act’s specific protection for the most productive agricultural land. What remained for an affected family in an exempted category was compensation alone, at rates the family had no part in negotiating, for a taking whose necessity no independent study had tested and whose social cost no public document had counted. The opposition memorandum to the President put the assessment point precisely: the study was the safeguard against the diversion of acquired land to other uses, against the taking of excess land, and against the acquisition of multi-crop irrigated land except as a demonstrable last resort. Compensation, on this view, answers the question what the displaced are paid; it does not answer the prior questions whether they should be displaced at all, and how many of them there are, and the 2013 Act had been written on the judgment, after a century of displacement without rehabilitation, that the prior questions were the ones the Indian state had systematically refused to ask.

A particular edge of the coverage objection concerned the Scheduled Areas. The 2013 Act had treated acquisition in the Scheduled Areas as a last resort and had required, for such areas, the prior consent of the Gram Sabha or the autonomous district councils, a recognition that tribal land stood under constitutional protections beyond the general law. The five categories were not drawn to stop at the boundaries of the Fifth Schedule: a road, a corridor or a power line through a scheduled district was as capable of notification as one through any other district, and the removal of the assessment, which was the instrument that identified forest and tribal livelihoods dependent on the land, fell hardest where dependence on common and forest land was greatest and recorded title weakest. The point was pressed by tribal rights organisations throughout the controversy, and it exposes the general structure of the opposition’s case in its sharpest form: the broader a category, the more certainly it reaches the people least able to contest their inclusion in it.

The third proposition concerned who bears the cost of the safeguard’s removal. The families within the consent and assessment chapters were not, in the main, title-holding cultivators with lawyers; the 2013 Act’s category of affected families deliberately included agricultural labourers, tenants, sharecroppers and others whose livelihoods depended on the land without a recorded title, and the assessment study was the only instrument in the Act that counted such people before the taking rather than after it. Farmer organisations, including the Bharatiya Kisan Sangh, the farmers’ organisation affiliated to the Rashtriya Swayamsevak Sangh and thus to the ruling party’s own ideological family, objected to the removal of the consent clause, and the ruling alliance’s own partners voiced reservations in public; the government’s difficulty in the Rajya Sabha was not only the opposition’s arithmetic but the visible discomfort of its own camp, in a year when agrarian distress and unseasonal rain had made the politics of land acutely sensitive and with assembly elections in Bihar, a state of small holdings and intense land attachment, due in the autumn. The Congress party’s characterisation of the ordinance as a device to snatch farmers’ land for industrialists flattened separable amendments into a slogan, as slogans do; but the slogan worked politically because the breadth of the categories gave it a textual foothold.

A smaller item drew a criticism out of proportion to its size, and it is worth recording because it shows the opposition reading the instrument as a whole rather than clause by clause. The recasting of Section 87, under which the head of a department had been deemed guilty of the department’s offence unless he proved diligence, into a requirement of prior government sanction before an official could be prosecuted, was presented by the bill’s critics in the parliamentary debate as the ordinance’s most revealing provision: the one place where the instrument’s solicitude ran to the acquiring state rather than to the acquired-from citizen. The 2013 deeming provision had been a deliberate disciplining device aimed at the official who signed a false record or a defective award; to replace it with a sanction gate, the critics argued, was to announce whose defaults the amendment existed to forgive. The government’s answer, that Section 197 of the Code of Criminal Procedure applies the same gate to public servants across the criminal law and that the 2013 provision had exposed heads of department to prosecution for the acts of subordinates they could not have prevented, was a respectable answer; but the exchange fixed an impression, that the ordinance’s sympathies were institutional, which the larger clauses then seemed to confirm.

The fourth proposition was empirical. Rahul Gandhi, in the Lok Sabha debate, cited information obtained from the Finance Ministry under the Right to Information Act for the claim that only eight per cent of stalled projects were held up by problems of land acquisition, the implication being that the government’s central factual premise, that the 2013 Act was the constraint on investment, was weak. The National Alliance of People’s Movements made the matching point about the investment figures in circulation, in terms: that twenty lakh crore rupees of investment were not stuck because of the new land Act, since the law had only been in existence for one year. The defence of the 2013 text in Parliament was led by the party and the minister that had written it. Jairam Ramesh, who as Minister for Rural Development in the previous government had piloted the 2013 Act, argued the case for his own statute against its amendment throughout the controversy, and the Congress position drew on the authority of authorship: the 2013 Act had been five years in the making, through a standing committee and extensive consultation, and the claim that it was unworkable after twelve months of operation, most of them under a government committed to replacing it, was a claim about the government’s intentions as much as about the Act’s text. Medha Patkar, who had been consulted in the making of the 2013 law, led the delegation of the National Alliance of People’s Movements that gave evidence against the bill before the Joint Committee in June 2015, carrying the argument from the street into the committee room in which the bill was then lodged. The government never produced, in the parliamentary debate, a project-by-project account that answered the eight per cent figure, and the absence mattered: a case for urgency that cannot quantify the emergency it invokes is asking the legislature, and the ordinance’s critics in the country, to take the emergency on trust.

The procedural objection: lawmaking by re-promulgation

Distinct from both cases on the merits stands the objection to the method, and it is the objection that time has treated most decisively. The ordinance power exists for action that cannot wait. The December ordinance was promulgated on the last day before a statutory deadline, which supplied a genuine, if partial, justification for acting before Parliament returned. What followed had no such justification of the same order. Parliament reassembled; the bill was introduced, debated and passed by one House; the bill was referred, at the height of the controversy, to a joint committee of both Houses which invited public evidence and heard the affected interests; and while that committee sat, the executive twice re-issued the ordinance so that the contested text remained the law without ever having been enacted. The sequence is the one examined in the account of the ordinance-making power in India, and the land ordinance became its standard modern illustration, because the practice condemned in the analysis of the re-promulgation judgment is precisely the practice this episode exhibited: the conversion of an emergency power into a parallel legislative track that operates for as long as the legislature declines to agree.

The constitutional text does not authorise that conversion. An ordinance ceases to operate six weeks after the legislature reassembles; a government that re-promulgates the same text at the expiry of each window is not responding to successive emergencies but declining to accept the consequence the Constitution attaches to non-enactment. The Supreme Court had said as much a generation earlier in D. C. Wadhwa v. State of Bihar, reported at (1987) 1 SCC 378, condemning the practice by which Bihar had kept ordinances alive for years by successive re-promulgation without placing them before the legislature. Wadhwa was a researcher who had documented the practice as a system: ordinances in Bihar were promulgated and re-promulgated in series, hundreds of them across the period his study covered, some kept continuously in force for well over a decade without ever being enacted by the state legislature, so that a province of India was being governed in material respects by executive decree renewed at intervals as a matter of routine. The Court in 1987 held the practice a fraud on the Constitution in substance, and the description mattered because it identified the wrong precisely: not the making of an ordinance, which the Constitution authorises, but the refusal to submit the ordinance to the only body entitled to make it permanent. The definitive statement came after the land episode, in Krishna Kumar Singh v. State of Bihar, decided on 2 January 2017 by a seven-judge bench of the Supreme Court, reported at (2017) 3 SCC 1. The case concerned a Bihar ordinance of 1989, by which the state had taken over the management of hundreds of private Sanskrit schools, and which had been re-promulgated seven times in succession without being enacted. The majority, in the opinion of Justice D. Y. Chandrachud, held that re-promulgation of ordinances is a fraud on the Constitution and a subversion of the democratic legislative process, the more so where the government persistently avoids placing the ordinance before the legislature; that the obligation to lay an ordinance before the legislature is a constitutional obligation and not a courtesy; and that the satisfaction of the President or the Governor in promulgating an ordinance is not immune from judicial review. The bench was also careful to mark what it was not deciding: it did not hold that an ordinance can never be re-promulgated in any circumstances, since an exigency can recur, but that re-promulgation cannot be a mechanical exercise by which the executive supplies itself with the legislation the legislature has withheld. The land ordinance was not before the Court in that case, and nothing in the judgment turned on it. It did not need to be. The land sequence of 2015 is the most prominent central instance of the practice the seven judges described, and every element of their description, the repeated re-issuance, the pendency of the measure before the legislature throughout, the avoidance of the vote the government expected to lose, is present in it.

What happens to an ordinance if Parliament does not approve it?

It ceases to operate at the expiry of six weeks from the reassembly of Parliament, or earlier if both Houses pass resolutions disapproving it. An ordinance is not converted into an Act by the passage of time or by repeated re-issuance; only a bill passed by both Houses and assented to by the President makes the change permanent.

Two qualifications keep the procedural verdict honest. The first is that re-promulgation was not invented in 2015, and ministers said so at the time, pointing to ordinances under earlier governments, including governments of the parties then in opposition, that had been promulgated more than once; governments of the Congress-led alliance had also kept ordinances alive by re-issuance. The practice had a pedigree, and the seven-judge bench’s condemnation falls on the practice as such, not on one government. The second qualification runs in the 2015 government’s favour and should be recorded as plainly as the objection: when the political cost became undeniable and the parliamentary route had visibly failed, the government stopped. It did not promulgate the ordinance a fourth time, although its own supporters expected it to and although nothing but convention prevented the attempt. The announcement of the lapse was made by the Prime Minister personally, in terms that attributed the retreat to the fear the measure had created among farmers rather than to any defect in the measure itself, but the retreat was real, and the 2013 Act resumed its unamended operation at the Centre on 1 September 2015. A constitutional culture is measured partly by whether governments accept the outcomes they dislike; this one, after eight months of resisting the outcome, accepted it.

The eight months in which the ordinance was the law

An ordinance is not a proposal. While it operates it is an Act for all purposes, and between 31 December 2014 and 31 August 2015 the amended text was the land acquisition law of India, a fact that the politics of the episode tends to obscure. The consequences were uneven across the programme. The Section 105 extension operated immediately and continuously: from 1 January 2015, acquisitions under the national highways and railways statutes and the other scheduled enactments attracted the 2013 Act’s compensation and rehabilitation provisions, and when the ordinance lapsed the Removal of Difficulties Order, deemed effective from the same 1 January 2015, continued the position without a gap, so that the single most valuable item in the programme never in practice switched off. The exemption power in Section 10A, by contrast, required a notification for projects within the categories before it changed anything on the ground, and its practical footprint in eight months was necessarily smaller than its symbolic one; the categories mattered in 2015 chiefly as a statement of the law the government intended to operate, and as the text from which the states later copied. The Section 24 exclusions affected the computation in lapse disputes that arose during the window. The smallness of the exemption power’s footprint in the window is itself a datum about how such powers work. An exemption by notification is not self-executing across a category: a project authority must identify its project as falling within a category, the appropriate government must issue the notification, and the acquisition then proceeds on the exempted footing, with every step exposed to challenge by an affected family arguing that the project was misclassified. In eight months overshadowed by a parliamentary fight about whether the power should exist at all, acquiring governments had every incentive to proceed under the unamended chapters for projects already in motion and to reserve the new power for projects not yet begun, which are the projects least visible in any eight-month window. The ordinance’s practical legacy was therefore never going to be measured within its own life. It was measured later, in the state statutes that reproduced its categories when the political question had been moved to friendlier legislatures, and in the acquisitions carried out under those statutes in the years after the national argument had moved on. And the whole episode left behind the ordinary puzzle of a lapsed ordinance: the things done under it while it lived were not unwound by its death, but neither did its text survive to govern anything new, so that from 1 September 2015 an acquiring authority in a state without its own amendment worked under the 2013 Act exactly as enacted, whatever had been notified in the spring. Even the salvaged item carried a legal question that outlived the politics. The Removal of Difficulties Order rested on Section 113, a power to make provisions consistent with the Act for removing difficulties in giving effect to it, and it was later argued in litigation, in proceedings before the Bombay High Court among others, that a power of that character cannot do the work of amending a substantive provision such as Section 105(3), whose own mechanism, a notification laid before both Houses within the year, had been allowed to lapse unused. The argument has an obvious force, and its existence marks the difference between the routes in legal security: what the ordinance could not get from Parliament, and what the order supplied from within the Act, remained open to the objection that the instrument exceeded the power it invoked, in a way that a state amendment under Article 254(2), or a parliamentary amendment, never could be.

What the ordinance sought and how each item ended up

Read the December programme item by item, and the episode’s real shape appears. The bill was defeated; the policy, in large part, was not. The routes differed by item, and no single route delivered the whole: the compensation extension was salvaged by executive order within days of the lapse, the exemptions and the consent changes were enacted state by state over the following years, and the recalibration of the lapse provision was achieved in 2020 by a Constitution Bench construing the unamended text. The table below is the article’s findable artifact, and its final column is the point.

Proposed amendment	What the ordinance provided	Enacted by Parliament	Enacted by a state	Achieved by interpretation	Resulting position
Exemption of five project categories from assessment, food security limits and consent	New Section 10A in a new Chapter IIIA, empowering exemption by notification for defence, rural infrastructure, affordable housing, industrial corridors and infrastructure projects including public private partnerships with government ownership of land, with the consent requirement in Section 2(2) also removed for those categories	No	Yes: state amendments modelled on the ordinance, beginning with Tamil Nadu in 2015 and Gujarat in 2016, with Telangana, Maharashtra, Rajasthan and Jharkhand also enacting amendments with Presidential assent	No	The exemptions operate within states that enacted them; the central Act retains consent and assessment in full
Extension of the 2013 compensation and rehabilitation scheme to the thirteen Fourth Schedule Acts	Amendment of Section 105 so that the 2013 provisions applied directly to acquisitions under the scheduled enactments, replacing the notification mechanism of Section 105(3)	No	Not applicable, the scheduled Acts being central enactments	No	Achieved by executive order: the Removal of Difficulties Order of 28 August 2015 under Section 113 applied the First, Second and Third Schedules to the scheduled Acts with effect from 1 January 2015
Substitution of private entity for private company	A defined expression covering a proprietorship, partnership, company, corporation, non-profit organisation or other entity, so that acquisition for such entities stood within the Act’s scheme	No	Yes, in the state amendments that adopted the ordinance text, the Gujarat amendment of 2016 adopting the ordinance’s changes as a whole	No	Operates in the states that adopted it; the central Act continues to speak of private companies
Return of unutilised land	Section 101 amended so that unutilised land is returned after the later of five years or the period specified at the time of setting up the project	No	Yes: Telangana’s amendment adopted the same formula for the return of unused land, and the Gujarat amendment took the ordinance’s scheme	No	State by state; the central Act retains the flat five-year period
Exclusions from the Section 24 lapse computation	Periods of court stay or injunction, periods specified in a tribunal’s award for taking possession, and periods in which compensation lay deposited in a court or an account, excluded from the five years	No	In part, in state amendments adopting the ordinance text	Yes, in substance: the Constitution Bench in 2020 excluded periods of interim orders from the computation and held deposit in the treasury to satisfy payment	For the lapse provision itself, largely achieved by construction of the unamended section
Prior sanction for prosecuting officials	Section 87 recast so that a government official could not be prosecuted for an offence under the Act without the prior sanction of the government, on the pattern of Section 197 of the Code of Criminal Procedure, 1973	No	Included in state amendments modelled on the ordinance	No	State by state; the central Act retains the deeming provision against the head of department
Acquisition for private hospitals and private educational institutions within public purpose	Removal of the 2013 exclusion that had kept such acquisition outside the definition of public purpose	No: abandoned by the government itself through its nine Lok Sabha amendments of March 2015	No	No	The 2013 exclusion continues; the one item in the programme that died by its sponsor’s own hand before any opponent had to defeat it

The distribution of outcomes across the table follows a logic worth naming, because it predicts how later failed amendments will also end. Each item travelled by the cheapest route constitutionally open to it. The compensation extension needed no new legislative authority at all, because the 2013 Act contained both the mandate and, in Section 113, an executive instrument adequate to execute it once the deadline politics were stripped away; it therefore moved first and fastest. The exemption items needed statutory text, but the text did not have to be central: any legislature with competence over a concurrent subject and access to the assent device could supply it for its own territory, and six did. The lapse items needed no text at all in the end, because they turned on the meaning of words already enacted, and meaning is the courts’ province; they took longest, because constitutional litigation moves at its own pace, but arrived with the greatest authority. A reader who wants to know the fate of a defeated amendment should therefore ask three questions of it: whether an executive instrument inside the existing Act can carry it, whether a state legislature can carry it for its own territory, and whether a court can find it in the enacted words. The land programme is the rare case in which all three answers were yes, for different items of the same bill. The table’s last row teaches the complementary lesson. The private hospitals and educational institutions item is the only one for which every route column reads no, and it is also the only item whose abandonment preceded every institutional decision: the sponsor withdrew it in its own amendments, ten weeks after proposing it, without a vote being lost or a court being asked. Items die in legislation for different reasons, and the reasons predict the afterlife. An item abandoned by its sponsor has no constituency to carry it by another route; an item defeated on a vote retains the constituency that voted for it, and that constituency, if it commands any other institution in the system, will look for the route that institution controls. The land programme’s afterlife is exactly the map of its constituencies: the compensation extension had a constituency in every party, the exemptions had constituencies in the state capitals, and the lapse recalibration had a constituency in every acquiring authority in the country, which is why a court was eventually asked to find it.

Three lessons sit inside the table. The first is that the items were separable, and their fates were separated: the item with cross-party support, the compensation extension, was the item the government could save by executive action because Section 113 gave it a lawful instrument; the items that were contested on principle travelled by the slower federal route; the item that even the sponsor could not defend in the Lok Sabha disappeared without trace. The second is that the state route and the judicial route delivered different kinds of victory. A state amendment delivers the text but only within one state’s territory and only until a later legislature changes it; a constitutional construction delivers the proposition everywhere the Act runs but only within the limits of what a court can honestly find in the words enacted. The third is the one the standard account misses: the ordinance’s defeat and the ordinance’s programme are different subjects, and an account of the episode that stops at the lapse in August 2015 has recorded the defeat and missed the programme.

The state route: Article 254(2) and the assent device

Acquisition and requisitioning of property is Entry 42 of the Concurrent List in the Seventh Schedule, which means both Parliament and a state legislature may legislate on it, and the 2013 Act is a central law on a concurrent subject. Where a state law on the same subject is repugnant to the central law, the central law ordinarily prevails. The device is the proviso in Article 254(2) of the Constitution, analysed in the account of Article 254 and repugnancy, under which a state law that conflicts with a central law on a concurrent subject nevertheless prevails within that state if the bill was reserved for the consideration of the President and received the President’s assent. Within days of the lapse, the Finance Minister, Arun Jaitley, set the route out publicly in a written note: acquisition of property is a Concurrent List subject, Article 254(2) plainly permits a state government to legislate in conflict with the central legislation provided Presidential assent is given, and the states were thus fully empowered to amend the 2013 land law and seek that assent. The suggestion had institutional backing: the Centre for Policy Research’s mapping of the state response records that the state amendment route proceeded on the suggestion of NITI Aayog in 2015, the Union government’s own policy institution pointing the states to the proviso within weeks of the central bill’s collapse.

What does Article 254(2) allow a state legislature to do?

It allows a state legislature to enact a law on a Concurrent List subject that conflicts with a central law on the same subject, provided the bill is reserved for the President’s consideration and receives the President’s assent. The state law then prevails within that state over the central law, until Parliament legislates again.

The federal option had in fact been tabled while the ordinance still lived. At the Governing Council of NITI Aayog in July 2015, with the bill lodged in committee and the monsoon session about to open, the Prime Minister told the assembled chief ministers that the deadlock at the Centre need not hold up development in the states, and that states wanting a different acquisition regime were free to legislate for themselves. The suggestion converted the proviso from a textbook curiosity into a programme: a state government no longer had to wait for the Centre to win an argument it was visibly losing, and a state opposition to the exemptions had to be organised in each state capital separately, against a state majority, rather than once in the Rajya Sabha. For the movement against the ordinance, which had won its victory by concentrating pressure on a single national decision, the federal route dissolved the target.

The states moved in sequence. Tamil Nadu acted first and by a different technique: its amendment of 2015, which received the President’s assent on 1 January 2015, placed three Tamil Nadu acquisition statutes, the Tamil Nadu Highways Act, the Tamil Nadu Industrial Purposes Act and the Tamil Nadu Acquisition of Land for Harijan Welfare Schemes Act, in the Fourth Schedule of the central Act with retrospective effect from the Act’s commencement, so that acquisitions under those state statutes stood outside the central scheme altogether. The technique was the mirror image of the ordinance’s: where the Centre had used the Fourth Schedule to extend the 2013 entitlements to thirteen central Acts, Tamil Nadu used it to withdraw three state Acts from the 2013 discipline entirely, and the amendment drew a challenge in the Madras High Court, in proceedings brought by landowners in Ramanathapuram district facing acquisition for a thermal power project, in which the court directed that the status quo be maintained while the state explained how the amendment stood with the one-year scheme of Section 105. The Tamil Nadu route is the broadest of the state devices, since exclusion from the Act removes not only assessment and consent but the compensation code itself, leaving the excluded statute’s own terms to govern; it is recorded in the Centre for Policy Research study as the first of the state enactments and the one that reached furthest beyond the ordinance’s text. Gujarat followed in 2016 with the most complete adoption: the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement (Gujarat Amendment) Act, 2016 reproduced the ordinance’s programme, exempting projects for national security and defence, rural infrastructure, affordable housing for the poor, industrial corridors and other infrastructure including public private partnership projects from the assessment and consent provisions, its statement of objects and reasons declaring the purpose of making what it called a lengthy and difficult process smooth and easy in an industrially progressive state. Telangana enacted its own amendment, adopting among other things the ordinance’s formula for the return of unused land. Maharashtra exempted public private partnership projects from the assessment and consent clauses. Rajasthan and Jharkhand also enacted amendments through the same assent route. The Centre for Policy Research’s study counts at least six state governments that legislated in this way: Tamil Nadu, Gujarat, Maharashtra, Telangana, Rajasthan and Jharkhand. The pattern of the list matters as much as its length. Most of the legislating states were governed by the Bharatiya Janata Party or its allies, implementing through their own assemblies what their party’s central government had failed to carry at the Centre; Telangana was not, which indicates that the appetite for the exemptions was administrative as well as partisan, a state government’s desire to acquire on easier terms rather than a party’s programme alone. Andhra Pradesh also enacted an amendment of its own and stood among the five states whose amendments were later challenged in the Supreme Court, so the migration was not confined to the six states in the Centre for Policy Research count, and its full extent is better described as a continuing state practice than as a closed list: any state legislature may still take the same route, with the same assent, on any day it commands the votes.

Did the state amendments face any legal challenge?

Yes. Medha Patkar petitioned the Supreme Court to strike down the amendments made by Gujarat, Andhra Pradesh, Telangana, Tamil Nadu and Jharkhand, arguing that the removal of consent, assessment and public participation destroyed the essence of the 2013 Act. The Court issued notices to the five states; the bench observed that legislatures were within their competence to amend.

The challenge identifies the genuine constitutional unease in the state route, and it should be stated at its strongest. Article 254(2) was designed as a safety valve for local variation on concurrent subjects, a means by which a state with different conditions could depart from a central norm with the Union’s considered approval. Used once, for one state’s circumstances, it is federalism working as intended. Used as a coordinated programme, encouraged by the Union executive that had just lost the vote in Parliament, with the assent of a President who acts on that same executive’s advice, it begins to look like a method by which a central government that cannot carry an amendment in the Rajya Sabha assembles the same amendment in pieces, state by state, with each piece requiring a majority it already commands and an assent its own advice controls. Nothing in that description alleges illegality: each step is within the constitutional text, and the proviso exists precisely to be used. But the democratic objection that attached to the re-promulgations attaches here in a milder form, since the policy that a national legislature declined to enact became national in effect without a national enactment, and the variation it produced is not the variation of local experiment but a patchwork in which the protection of a farming family against compulsory acquisition depends on which side of a state border the family farms. Whether that is federalism’s promise or its abuse is a value question on which the Constitution’s text, unusually, permits both descriptions; what can be said as fact is that the route was lawful, that it was openly announced by the Finance Minister within days of the lapse, and that it delivered, for the states that used it, the operative core of the December programme.

The judicial route: Section 24 read again in 2020

The second route ran through the courts, and it delivered the one item on the December list that was, in substance, a dispute about the meaning of words already enacted. Section 24(2) of the 2013 Act provided that where an award under the 1894 Act had been made five years or more before the 2013 Act commenced, and the physical possession of the land had not been taken or the compensation had not been paid, the acquisition proceedings were deemed to have lapsed, entitling the landowner to fresh proceedings under the 2013 Act with its higher compensation. The provision generated an enormous volume of litigation, because the difference between compensation under the 1894 Act and compensation under the 2013 Act for the same land is very large. The ordinance had proposed to narrow the provision’s operation by excluding from the five-year computation the periods of court stay, the periods specified in tribunal awards, and the periods in which compensation lay deposited. That amendment died with the ordinance. The questions it addressed did not die, because they were already before the Supreme Court in the form of conflicting decisions on the enacted words.

The road to the Constitution Bench ran through a further round of the same litigation. In February 2018 a three-judge bench, in the Indore Development Authority litigation itself, held that the 2014 decision in Pune Municipal Corporation had been rendered per incuriam and took the contrary view on deposit and lapse; a coordinate bench doubted that course, and the questions were referred to a Constitution Bench of five judges. When the Constitution Bench was constituted under Justice Arun Mishra, who had delivered the 2018 view, landowners sought his recusal on the ground that he had already decided the very questions referred; the bench rejected the plea, holding that a judge’s prior expression of a legal view in a judgment is no ground for recusal and that to hold otherwise would licence the selection of benches by litigants. The conflict was resolved on 6 March 2020, when the Constitution Bench delivered judgment in Indore Development Authority v. Manoharlal. The bench overruled the earlier decision in Pune Municipal Corporation v. Harakchand Misirimal Solanki of 2014, which had taken the landowner-favourable view of the provision. The Constitution Bench held, in the judgment delivered by Justice Arun Mishra for the bench, that the word or in Section 24(2) is to be read as and, so that proceedings lapse only where possession has not been taken and compensation has not been paid, the failure of either condition alone being insufficient; that the obligation to pay compensation is discharged by tendering the amount, and that a deposit of the compensation in the government treasury satisfies the provision, so that non-deposit in court does not produce a lapse; that a landowner who has refused to accept compensation cannot thereafter invoke the lapse; and that the period during which interim orders of a court subsisted must be excluded from the computation of the five years, since the authorities’ inaction during a stay is not the lethargy the provision exists to correct. The detailed working of the compensation, consent and lapse machinery, as distinct from its bearing on the 2015 programme, belongs to the specialist treatment of LARR compensation, consent and lapse; the point here is narrower, and it is the connection that public discussion almost never makes. Item by item, the Constitution Bench’s construction delivered the substance of the ordinance’s Section 24 amendments: the exclusion of stay periods from the clock, which the ordinance would have enacted, the bench found in the provision’s purpose; the protection of deposits, which the ordinance would have enacted for deposits in court or an account, the bench generalised into the proposition that tender and treasury deposit satisfy payment; and the narrowing of lapse to cases of genuine double default, which the ordinance sought by its exclusions, the bench achieved in one move by reading the conjunction into the disjunctive text.

The correspondence between the ordinance’s Section 24 items and the judgment is close but not exact, and the differences are instructive. The ordinance’s exclusion for a period specified in a tribunal’s award for taking possession found no separate counterpart in the Constitution Bench’s conclusions. Conversely, the judgment’s central move, reading the word or in Section 24(2) as and, had no counterpart in the ordinance at all: the bill’s drafters, working within the discipline of amendment, had attacked the provision’s width by carving periods out of its clock, while the Court attacked it at its logical root, holding that a provision aimed at official lethargy could not be triggered where the state had either taken the land or tendered the money. The judicial delivery was in that respect bolder than the legislative proposal it vindicated, and it was also differently founded: the ordinance’s exclusions would have applied because Parliament said so, whereas the Court’s construction claimed only to declare what the 2013 Parliament had meant, a claim the landowners’ side disputed then and disputes in the literature since, on the ground that the enacted text used or in a section whose companion provisions, where Parliament wanted a period excluded, said so expressly. That textual objection, drawn from the contrast with the proviso to Section 19(7) and the explanation to Section 69(2), where the 2013 Act does exclude stay periods in terms, is the strongest argument against the 2020 reading, and the bench answered it by subordinating the contrast to the provision’s purpose of correcting inaction. The answer satisfied the acquiring authorities and the finance of every stalled project in the country; it did not persuade all of the provision’s readers, and the disagreement is the one live doctrinal residue of the whole episode.

The judicial route has limits that the legislative route does not, and honesty requires them to be marked. A court construing Section 24(2) could reach the stay exclusion and the deposit proposition because those were questions about the meaning of a transition provision; it could not have created the five exempted categories, abolished the consent requirement or rewritten Section 101, because those are legislative choices with no foothold in the enacted words. What interpretation delivered was therefore exactly the portion of the December programme that was itself interpretive in character, a recalibration of clocks and defaults, and it delivered it with an authority no ordinance could have given it, since a construction of the Constitution Bench binds every court in India and cannot be displaced except by a larger bench or by the very parliamentary amendment that had proved impossible. Landowners’ advocates criticised the 2020 judgment as reading words into the statute against its protective purpose, and the reading of or as and is, as a matter of ordinary language, a strong construction; but the judgment is the law, and its effect on the stock of pending lapse claims was immediate and large. Between the state amendments and the construction of Section 24, the two live controversies the ordinance had addressed, the breadth of the preliminary safeguards and the width of the lapse provision, were both substantially resolved in the direction the ordinance had proposed, by institutions other than the Parliament that had refused it.

What is settled, what is contested, and what the episode proves

The episode is also misremembered in identifiable ways, and correcting the record is part of settling it. The ordinance is frequently described as having been enacted, or the 2013 Act as having been amended in 2015; neither happened, and the distinction is the whole story, since an enacted amendment would have governed nationally and permanently while the ordinance governed for eight months and the Act emerged unamended. The ordinance and the bill are routinely conflated, so that positions taken on the December text, including the inclusion of private hospitals and educational institutions, are attributed to the bill the Lok Sabha passed, from which the government itself had removed them. The state amendments are omitted from most accounts altogether, although they are the channel through which the programme reached the statute book. And the 2020 judgment is reported, where it is reported at all, as a land compensation case unconnected with the politics of 2015, although it disposed of the very provision the ordinance had tried and failed to amend. An account that makes all four errors will describe a simple defeat for a government and a simple victory for a movement, and will be unable to explain the law that a landowner in Gujarat or Telangana actually faces.

What is settled can be stated shortly. The ordinance was promulgated on 31 December 2014 and issued three times in all, on 31 December 2014, 3 April 2015 and 30 May 2015; the bill passed the Lok Sabha on 10 March 2015 with nine government amendments and was never carried in the Rajya Sabha; it was referred to a joint committee on 13 May 2015, which never reported; the ordinance lapsed on 31 August 2015 on the Prime Minister’s announced decision; the compensation extension to the thirteen scheduled Acts survived through the Removal of Difficulties Order of 28 August 2015; at least six states thereafter enacted amendments through the Article 254(2) route; and in 2020 a Constitution Bench construed Section 24(2) in terms that achieved the substance of the ordinance’s amendments to that provision. None of that is disputed by anyone who has examined the record, and accounts of the episode that omit any element of it, most commonly the state amendments and the 2020 judgment, are incomplete in a way that changes the conclusion.

Was the 2013 Act itself amended by Parliament at any point in this episode?

No. Parliament enacted no amendment to the 2013 Act; the Act stands at the Centre in its enacted form. The changes traceable to the programme exist in two places only: in the amendments of individual states, which operate within those states under Presidential assent, and in the construction placed on Section 24(2) by the Supreme Court.

The resulting patchwork has a concrete meaning for the person the law exists to govern, and it is the practical residue of the whole affair. A family whose land is wanted for an industrial corridor in a state that enacted the ordinance’s scheme faces acquisition without a social impact assessment and without a consent vote, on the 2013 Act’s compensation terms. A family in the identical position across a state border, in a state that enacted nothing, faces the full 2013 discipline, assessment, food security ceiling and consent included. A family whose old acquisition is said to have lapsed faces, everywhere in India, the Constitution Bench’s narrower reading of the lapse provision, whoever its state government is. The three families’ positions were all supposed to have been settled by the single parliamentary decision of 2015. They were settled instead by the family’s state legislature, by the Union executive’s difficulties power and by five judges, in that order of visibility and in reverse order of democratic ceremony. That is not an argument that any of the three settlements was unlawful; each was reached by an institution acting within its competence. It is the accurate description of what the failure of the amendment produced, and it is why the episode belongs in a legislative history rather than in a footnote to one.

What is contested on facts is narrower than the volume of argument suggests. The breadth of the five categories is a question about how the phrases would have operated in notification practice, and honest readers of the phrases differ: the categories had a closed list and a notification discipline, and they were also, on any realistic view of infrastructure and rural infrastructure, capable of covering the great majority of acquisitions by number. The empirical premise of the government’s case, that the 2013 Act was the binding constraint on investment, was asserted rather than demonstrated in the debate, and the eight per cent figure produced by the opposition from the Finance Ministry’s own records was never answered with better numbers; on the evidence actually presented in 2015, the premise should be treated as unproven rather than as false, since the Act had operated for one year and neither side had a settled body of project data. The states’ demand for amendment is better evidenced than either, resting on the consultations of 2014 and on the subsequent conduct of state governments that had no obligation to legislate as they did.

What is contested on values is the residue, and it is genuine. Beneath the clauses, the two sides held different views about what a safeguard is. For the ordinance’s defenders, consent and assessment were procedures, valuable when they disciplined an acquisitive state and indefensible when they multiplied delay for projects whose public character was not in doubt; compensation, generous and judicially enforceable, was the protection. For its opponents, consent and assessment were the protection, because they were the only moments in the entire structure at which the people being displaced appeared as participants rather than as recipients, and a system that pays well for a taking it never has to justify has answered the easier question. No evidence settles that disagreement, because it is about what the state owes the citizen it displaces beyond money, and this article does not pretend to settle it. It records that the 2013 Parliament answered it one way, that the government of 2014 to 2015 proposed answering it another way for five categories of project, and that the second answer governs in several states without ever having been enacted by Parliament.

Each route also charged its own price, and a legislator reading the episode for instruction should count the prices as carefully as the deliveries. The executive route was cheap and immediate but reached only what a difficulties power can lawfully do, which is to execute the parent Act’s own mandate; it could not have carried the exemptions, and an attempt to make it do so would have collapsed in court. The state route was lawful and durable within each state, and its price was uniformity: a national compensation and safeguards code became, in the exempted categories, a schedule of state variations, so that the protection of an affected family came to depend on the accident of which legislature governed its district, and the Union executive that sponsored the migration acquired an interest in defending in court, state by state, amendments it had been unable to defend once in Parliament. The judicial route was the most authoritative and the narrowest, confined to the provision whose words were before the Court, and its price was time and contingency: it depended on a reference being made, a bench being constituted and a construction being found, none of which a government can schedule. The parliamentary route, the only one that could have delivered the whole programme at once and for the whole country, was the only one that failed. That inversion, in which the constitutionally ordinary path proved the least available and the three extraordinary paths divided the spoils, is the episode’s durable lesson about how Indian legislation actually moves when the two Houses disagree.

The episode also educated its sponsor. It was the first major legislative defeat of the government elected in 2014, arriving within its first fifteen months, on the measure it had chosen to make the emblem of its development programme; and the lesson the government drew was visible in its subsequent practice, which was not again to stake a flagship policy on a Rajya Sabha vote it did not command, but to prefer instruments and routes, money bill classification, executive action within parent statutes, and state legislation on concurrent subjects, in which the Upper House’s arithmetic could not be decisive. Later controversies about those routes belong to their own articles; the genealogy begins here, in the discovery that a Lok Sabha majority, however large, is a partial legislature, and that a determined executive with a partial legislature will find the Constitution’s side doors. Whether the side doors are a safety valve or a subversion is the question Krishna Kumar Singh answered for one of them; the land episode is the demonstration that the question arises for all of them.

The namable claim of the episode, and the reason it deserves an article of its own in a series on legislation, is this: defeated in Parliament, delivered elsewhere. The land acquisition amendment is the most significant failed amendment of its decade precisely because its failure was so complete at the point of decision, a bill passed by one House, unexamined to a conclusion by the committee that held it, abandoned by its own sponsor in a radio address, and its principal objectives were nonetheless achieved, piecemeal and over years, through a removal of difficulties order issued under the very Act it sought to amend, through state legislatures using a repugnancy proviso designed for local variation, and through a Constitution Bench reading a conjunction into a disjunctive sentence. In a federal system with judicial review, losing a legislative vote is not the end of a policy; it is the end of one route to it. The record of what happened to a bill after it failed is as much a part of its legislative history as its passage would have been, and only following the policy rather than the bill reveals the outcome. The general account in this article is legal information and analysis, not advice on any particular acquisition; the position in an individual case turns on the state in which the land lies and the statute under which it was acquired, and the reader with a live matter should take advice on that case.

Frequently Asked Questions

Q: What did the 2015 land acquisition ordinance change?

The ordinance of 31 December 2014 amended the Land Acquisition Act, 2013 in seven ways. It inserted a new Section 10A empowering the government to exempt five categories of project from the social impact assessment, from the food security restriction on irrigated multi-cropped land, and from the consent requirement. It extended the 2013 compensation and rehabilitation scheme to acquisitions under the thirteen Fourth Schedule Acts. It substituted private entity for private company. It changed the return of unused land to the later of five years or the period specified for the project. It excluded periods of court stay and deposit from the Section 24 lapse computation. It required prior sanction before prosecuting officials. And it brought private hospitals and educational institutions within public purpose, a change later dropped.

Q: Why was the land ordinance re-promulgated three times?

The ordinance was issued three times in all: on 31 December 2014, on 3 April 2015 and on 30 May 2015. An ordinance ceases to operate six weeks after Parliament reassembles, so each re-issuance was necessary to keep the amendments alive while the replacement bill made no progress. The April re-promulgation incorporated the nine amendments the Lok Sabha had accepted when it passed the bill on 10 March 2015. The May re-promulgation kept the text in force while the bill sat with the Joint Committee to which it had been referred on 13 May 2015. The sequence ended when the Prime Minister announced on 30 August 2015 that the ordinance would be allowed to lapse, which it did the next day.

Q: Which project categories did the land ordinance exempt from consent?

The new Section 10A listed five categories: projects vital to national security or defence and defence production; rural infrastructure including electrification; affordable housing and housing for the poor; industrial corridors; and infrastructure projects, including projects under public private partnership where the ownership of the land continued to vest in the government. For projects in these categories, the ordinance removed the requirement of prior consent from affected families, which the 2013 Act had set at eighty per cent for acquisition for private companies and seventy per cent for public private partnership projects. The same categories could also be exempted by notification from the social impact assessment and from the restriction on acquiring irrigated multi-cropped land. Compensation and rehabilitation entitlements were not exempted and continued to apply in full.

Q: Why did the land acquisition amendment bill fail?

The bill failed on arithmetic and on politics rather than on any single vote. The government commanded the Lok Sabha, which passed the bill on 10 March 2015, but it lacked a majority in the Rajya Sabha, where a united opposition had the numbers to defeat it and had resolved to do so. Referral to a Joint Committee on 13 May 2015 removed the bill from the floor without resolving it, and the committee never reported. Outside Parliament, the bill had been successfully characterised as anti-farmer: fourteen opposition parties marched to the President against it, Anna Hazare protested against the ordinance, and Rahul Gandhi’s description of a suit-boot government fixed its public meaning. Facing that combination, and assembly elections in Bihar, the government let the companion ordinance lapse in August 2015.

Q: What happened after the land ordinance lapsed?

The ordinance ceased to operate on 31 August 2015, and the 2013 Act resumed its unamended operation at the Centre. Two things survived the lapse. First, the extension of the 2013 compensation, rehabilitation and resettlement benefits to the thirteen Fourth Schedule Acts was preserved by an order of 28 August 2015 issued under Section 113 of the Act, effective from 1 January 2015. Second, the policy migrated. State governments began enacting their own amendments to the central Act under Article 254(2) of the Constitution, obtaining the President’s assent so that the amendments prevailed within their states. The amendment bill itself remained with the Joint Committee, which never presented a report, and the bill lapsed when the Sixteenth Lok Sabha was dissolved in 2019.

Q: Did states amend the land acquisition law themselves?

Yes. Because acquisition of property is a Concurrent List subject, a state legislature may amend the central Act for its own territory if the bill is reserved for the President and receives the President’s assent under Article 254(2). The Centre for Policy Research counts at least six states that legislated in this way: Tamil Nadu, whose amendment of 2015 placed three state acquisition statutes in the Fourth Schedule; Gujarat, whose amendment of 2016 adopted the ordinance’s programme including the exemption of the five categories from assessment and consent; and Telangana, Maharashtra, Rajasthan and Jharkhand, which enacted amendments of their own. Most were governed by the ruling party or its allies, but Telangana was not, which suggests the demand was administrative as well as partisan.

Q: Is re-promulgation of an ordinance constitutional?

The Supreme Court has answered no. In Krishna Kumar Singh v. State of Bihar, decided on 2 January 2017 by a seven-judge bench, the Court held that re-promulgation of ordinances is a fraud on the Constitution and a subversion of the democratic legislative process, particularly where the government persistently avoids placing the ordinance before the legislature. The judgment built on D. C. Wadhwa v. State of Bihar, which had condemned Bihar’s practice of keeping ordinances alive by successive re-promulgation. The Court also held that the President’s or Governor’s satisfaction in promulgating an ordinance is open to judicial review. The ruling postdated the land ordinance, which had been issued three times in 2015 while its replacement bill was before Parliament.

Q: Did the ordinance’s objectives ever take effect?

Substantially, yes, through three other routes. The extension of the 2013 compensation scheme to the thirteen Fourth Schedule Acts took effect through the Removal of Difficulties Order of August 2015 and continues in force. The exemptions from assessment and consent, and several of the ordinance’s other changes, took effect within the states that enacted them under Article 254(2), beginning with Tamil Nadu and Gujarat. And the ordinance’s proposed narrowing of the Section 24 lapse provision was achieved in substance by the Constitution Bench in Indore Development Authority v. Manoharlal in 2020, which excluded periods of court stay from the five-year computation and held that deposit of compensation in the treasury satisfies payment. The central Act itself was never amended by Parliament.

Q: When was the land acquisition ordinance promulgated, and for how long did it operate?

The ordinance was first promulgated by the President on 31 December 2014, the last day of the one-year period within which the 2013 Act required the compensation and rehabilitation provisions to be extended to the thirteen Fourth Schedule Acts. It operated, through successive re-promulgations on 3 April 2015 and 30 May 2015, for eight months in all, ceasing to operate on 31 August 2015 after the Prime Minister announced in his Mann Ki Baat address of 30 August 2015 that he had decided to let it lapse. Because an ordinance has the force of an Act while it operates, acquisitions and notifications during those eight months were governed by the amended text; because it was never replaced by an Act, the amendments left no permanent mark on the central statute.

Q: Why did the government legislate by ordinance in December 2014 rather than wait for Parliament?

The government’s strongest reason was a deadline inside the 2013 Act itself. Section 105(3) directed that the compensation, rehabilitation and resettlement provisions be extended to the thirteen Fourth Schedule Acts within one year of the Act’s commencement on 1 January 2014, and that year expired at the end of December 2014. An ordinance promulgated on 31 December met the deadline and prevented highway and railway landowners from being left on the older, thinner compensation scales. The critics’ answer was that the deadline explained only one item in the ordinance. Nothing required the consent and assessment exemptions to be made by 31 December 2014, and the alliance of people’s movements asked publicly what the emergency was that could not wait for the budget session a few weeks later.

Q: What were the nine amendments the government accepted in the Lok Sabha in March 2015?

The amendments, circulated on 9 March 2015 and passed with the bill the next day, were concessions by the government to its own text. They removed private hospitals and private educational institutions from public purpose again. They confined industrial corridors to those set up by the government or its undertakings, with land limited to one kilometre on either side of the designated road or railway line. They deleted social infrastructure from the fifth exempted category. They required the government to ensure that exempted acquisition took the minimum land required, and to survey and record its wasteland. They assured compulsory employment to a member of an affected family of a farm labourer. They required the Land Acquisition Authority to hear cases in the district of acquisition. And they aligned official prosecutions with Section 197 of the Code of Criminal Procedure.

Q: How did the bill passed by the Lok Sabha differ from the ordinance of December 2014?

The bill as passed on 10 March 2015 was narrower than the December ordinance in every respect in which they differed. The ordinance had brought acquisition for private hospitals and private educational institutions within public purpose; the passed bill dropped that change. The ordinance’s industrial corridor category was unqualified; the passed bill limited it to government corridors and to a strip one kilometre wide on either side of the designated line. The ordinance’s fifth category had included social infrastructure; the passed bill removed it. The passed bill added disciplines the ordinance lacked: a duty to ensure that only the minimum land required was acquired for exempted projects, a survey of government wasteland, district-level hearings before the Land Acquisition Authority, and compulsory employment for a member of an affected farm labourer’s family. Conflating the two texts misstates what the Lok Sabha actually approved.

Q: What role did the Joint Parliamentary Committee play in the fate of the land bill?

The Joint Committee, constituted on 13 May 2015 with thirty members drawn from both Houses and chaired by S. S. Ahluwalia, was where the bill went to stop moving. It heard ministries, state governments and affected interests, including farmers’ organisations that told it the amendments revived the approach of the 1894 Act, and it was charged with reporting by the first week of the monsoon session. Instead its deadline was extended repeatedly, reaching a sixth extension by early 2016, with the pace explained at the time by the failure of many states to answer its questionnaire, including on compensation and on proceedings lapsing under Section 24(2). The committee never presented a report. Its existence allowed every side to say the bill was under examination while the ordinance lived and died elsewhere, and the bill lapsed with the Sixteenth Lok Sabha in 2019.

Q: What was the removal of difficulties order issued in August 2015?

It was the instrument by which the most popular part of the ordinance survived the ordinance’s death. The Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement (Removal of Difficulties) Order, 2015, issued on 28 August 2015 under Section 113 of the 2013 Act, applied the Act’s provisions on determination of compensation under the First Schedule, rehabilitation and resettlement under the Second Schedule and infrastructural amenities under the Third Schedule to all land acquisition under the thirteen enactments in the Fourth Schedule, and was deemed effective from 1 January 2015. Section 113 empowers the Central Government to make provisions, consistent with the Act, for removing difficulties in giving effect to it. Using that power, the government preserved by executive order the one ordinance item the 2013 Parliament had itself mandated.

Q: What did the Prime Minister announce about the ordinance in August 2015?

Speaking in his radio address Mann Ki Baat on 30 August 2015, Narendra Modi announced that the ordinance would lapse the following day and that he had decided to let it lapse rather than promulgate it a fourth time. He said that much misinformation had been spread about the ordinance and had created fear and confusion in the minds of farmers, that the position would revert to what it had been before his government came to office, and that the government was ready to accept any suggestion that favoured farmers. He added that the thirteen points, meaning the extension of the 2013 Act’s benefits to the thirteen Fourth Schedule Acts, were being brought under the rules with effect from that day so that farmers would not suffer financial loss from the lapse.

Q: Why did farmers’ organisations oppose the land acquisition ordinance?

Their objection was to the removal, for the five exempted categories, of the two preliminary safeguards of the 2013 Act. The social impact assessment was the only stage at which the number of affected families, including landless labourers and others without title, was publicly counted and the necessity of the land tested before an expert group; the consent requirement was the only stage at which affected families could refuse. Organisations that appeared before the Joint Committee argued that the amendments restored the approach of the 1894 Act, under which the state took first and counted never. The National Alliance of People’s Movements disputed the government’s premise, arguing that investments could not have been stuck because of a law one year old. The Bharatiya Kisan Sangh, affiliated to the ruling party’s own ideological family, also objected to the removal of consent.

Q: What did the government argue the 2013 Act was obstructing?

The government argued that the Act’s sequential procedures, a social impact assessment appraised by an expert group followed, for private and partnership projects, by the assembly of consent from seventy or eighty per cent of affected families, were workable for a single site but unworkable for the linear projects that dominate public acquisition: roads, canals, transmission lines and railways crossing hundreds of villages, where a defect at any point exposed the whole acquisition to challenge. Ministers including Arun Jaitley argued that consent designed as a check on acquisition for private profit operated differently where the state itself was building irrigation works, rural roads, electrification and affordable housing, and that state governments of more than one party had told the Centre in consultation during 2014 that the provisions were proving unworkable in administration.

Q: Did the ordinance permit acquisition for private hospitals and private educational institutions?

Yes, briefly. The 2013 Act had kept acquisition for private hospitals and private educational institutions outside the definition of public purpose. The ordinance of December 2014 removed that exclusion, so that such acquisition would have proceeded under the Act with its compensation and rehabilitation scheme. The change survived only until the government’s own reconsideration: the nine official amendments circulated on 9 March 2015 removed it from the bill, and the bill as passed by the Lok Sabha restored the 2013 position. The item is a useful corrective to two common errors. The ordinance was not a simple dilution, since this change extended the Act’s protections to a class of acquisition previously outside it; and the ordinance and the passed bill were different texts, since the sponsor itself abandoned the change within ten weeks.

Q: What did the ordinance change about the return of unused land?

Section 101 of the 2013 Act requires land acquired under the Act that remains unutilised for five years to be returned to the original owners or to the state land bank. The ordinance replaced the flat five-year period with the later of five years or any period specified at the time of setting up the project. The reasoning offered was that major projects have gestation periods fixed in advance that run well beyond five years, and that a rule returning the land while the project was still being built would force the state to begin acquisition again for the same purpose. The change died at the Centre with the ordinance, but it travelled: Telangana’s state amendment adopted the same formula for the return of unused land, and the Gujarat amendment took the ordinance’s scheme more broadly.

Q: Which of the ordinance’s proposals did the Supreme Court’s 2020 reading of Section 24 achieve by interpretation?

Most of them, for that provision. The ordinance had proposed excluding from the five-year lapse computation any period during which proceedings were held up by a court stay, any period specified in a tribunal’s award for taking possession, and any period in which possession had been taken but compensation lay deposited. In Indore Development Authority v. Manoharlal, decided on 6 March 2020, the Constitution Bench excluded the period of subsistence of interim orders from the computation, on the ground that inaction under a stay is not the lethargy Section 24(2) corrects. It held that deposit of compensation in the government treasury satisfies the requirement of payment, and it read the word or in the section as and, so that proceedings lapse only where neither possession has been taken nor compensation paid.