The Goods and Services Tax that took effect on 1 July 2017 is usually described as a tax reform, and the description understates what the midnight session in Parliament’s Central Hall actually set in motion. GST as cooperative federalism names the larger event: the Union and the states amended the Constitution together, surrendered pieces of their separate taxing powers into a shared pool, and built a permanent joint institution, the Goods and Services Tax Council under Article 279A, to govern what neither side could any longer govern alone. This article reanalyses that settlement as a piece of political architecture. Its claim is namable and testable against the record: the Goods and Services Tax operates as cooperative federalism in institutional form, and the proof lies less in the launch bargain of 2016 and 2017 than in the Council’s working record under strain, above all the compensation crisis of 2020, when the bargain was loaded with a collapsing economy and did not break.

GST as Cooperative Federalism: The Reanalysis and Its Claim

Three framings compete for the Goods and Services Tax. The technical tax policy framing reads GST as an exercise in tax design: a destination-based value added tax, an input credit chain, a revenue neutral rate, a set of slabs. The policy history framing reads it as a long negotiation running from the Empowered Committee of State Finance Ministers in 2000 through the Constitution (One Hundred and First Amendment) Act of 2016 to the launch of 1 July 2017. Both framings capture real content, and both miss the institutional fact that organises the rest. A tax that the Union cannot set alone and no state can set alone, administered by a council in which the Union holds one third of the weighted vote and the states hold two thirds, and in which every decision of consequence for four years was taken by consensus, is a federal institution before it is a tax instrument. The reanalysis offered here follows the approach set out in the economic-policy-as-politics framework, which treats an economic measure as political content: who conceded what, which institution carries the settlement, and how the settlement behaves when interests diverge.

An earlier article in this series read the 2017 launch as a test of cooperative federalism and asked whether the design would hold. The reanalysis has more evidence to work with. By December 2020 the Council had met forty-three times, had pruned its own rate structure in response to political pressure, had voted exactly once, and had fought its deepest fight over the compensation promise that had bought the states’ consent in 2016. The fight matters more than the launch. Institutions reveal their character under load, and the compensation crisis of 2020 loaded the GST settlement with a pandemic, a revenue collapse, an Attorney General’s restrictive reading of the Union’s obligation, and a set of state finance ministers who said in writing that the Union was betraying the compact. The settlement survived because the Council’s architecture forced a negotiated outcome and because the Union, after months of refusal, borrowed Rs 2.69 lakh crore across 2020 and 2021 and passed the money to the states on a back-to-back basis. That sequence, not the midnight speeches, is the strongest evidence for the cooperative federalism reading.

The article proceeds through five features. The first is the Empowered Committee of State Finance Ministers, constituted in 2000 under Asim Dasgupta of West Bengal, which spent a decade and a half building the trust and the technical draft that made a constitutional bargain imaginable. The second is the 101st Amendment itself, carried through the Rajya Sabha on 3 August 2016 by 203 votes to none, which inserted Articles 246A, 269A and 279A into the Constitution. The third is the implementation of 1 July 2017 through the Central, Integrated and Union Territory Goods and Services Tax Acts and the Goods and Services Tax (Compensation to States) Act. The fourth is the Council’s architecture under Article 279A, its weighted voting rule and its consensus practice. The fifth is the trajectory after the launch: rate rationalisation, the compensation guarantee in calm years and in crisis, the back-to-back loan settlement, and the extension of the compensation cess to March 2026 to service that debt after the guarantee itself expired in June 2022. A five-feature map later in the article sets the framework out in citable form.

Two disciplines govern the argument. Positions are attributed to named holders and rendered at their strongest: Arun Jaitley’s essays present the Council as a model federal body, M. Govinda Rao’s Studies in Indian Public Finance (2022) presents the same record as a surrender of state fiscal autonomy, and Vijay Kelkar’s In Service of the Republic (2019), written with Ajay Shah, presents GST as the outstanding achievement of cooperative economic policymaking in the republic’s history. The disagreement among those readers is genuine, and the article adjudicates it on the Council record rather than on preference. Figures are taken from Finance Ministry release statements and answers laid before Parliament, and where official statements differ across years, the figure used is tied to its dated source. The reader should also know what this article is not. It is not a compliance guide, not a rate manual, and not a verdict on whether GST raised or lowered the tax burden on any sector. It is an account of a federal institution and of the politics that institution made possible and necessary.

The Old Architecture: Seventeen Taxes and a Fractured Market

The pre-2017 indirect tax system divided the power to tax consumption between two levels of government that taxed the same goods at different stages and on different bases. The Union levied central excise duty on manufacture, a tax on services, customs duties and cesses. The states levied value added tax on sales within their borders, central sales tax on inter-state sales collected by the state of origin, entry tax and octroi at municipal borders, purchase tax, entertainment tax, luxury tax and taxes on advertisements and betting. The Seventh Schedule’s entries kept the two spheres apart by design: the Union taxed production, the states taxed sale, and neither could tax the other’s stage. The design belonged to the founding fiscal settlement, the same settlement whose democratic frame was set by the first general election of 1951 to 1952, and it hardened into the structure examined in the federalism of the Indian Constitution.

The costs of the fracture were measured repeatedly by official committees. Because the Union’s excise could not be credited against the state’s sales tax, taxes cascaded: a manufacturer paid excise on a price that already carried taxed inputs, and the wholesaler paid sales tax on a price carrying excise, so the final consumer paid tax on tax. The central sales tax, levied by the exporting state, rewarded states for taxing goods consumed elsewhere and penalised the movement of goods across internal borders; check posts at state boundaries inspected trucks and collected entry levies, so the common market that Article 301 promised existed more in text than on highways. The Kelkar task forces of 2002 to 2004 on direct and indirect taxes recommended a comprehensive goods and services tax as the remedy, and the Thirteenth Finance Commission’s task force on GST in 2009, chaired by Vijay Kelkar, worked out a dual GST design levied at a combined rate it put at ten percent. By then the diagnosis was not in dispute. What remained in dispute for another decade was the federal price of the cure: which level of government would give up what, and who would guarantee the loser against loss.

The destination of the old architecture was visible to governments long before it was reachable. The Union budget of 2006-07, presented by P. Chidambaram, announced the intention to move to a goods and services tax from 1 April 2010, a target date that passed with the design still in committee and the constitutional amendment not yet introduced. The 2010 deadline’s quiet burial became part of the states’ brief against haste in the following decade: a tax that had already missed one commencement date proclaimed by one government could not be allowed to commence on a second date without the insurance clauses the states were demanding, because the costs of a failed or loss-making commencement would fall on state budgets first and hardest. The Union’s own record reinforced the caution. Service tax, introduced in 1994 on three services, had grown by accretion into a levy on more than a hundred services without ever acquiring a constitutional home in the state sphere; excise had been narrowed by exemptions and litigation into a tax whose base bore little relation to manufacturing value; and the states’ VAT, for all its success as harmonisation, still stopped at each border, where the central sales tax and entry levies took their cut of internal trade. Every element of the old system worked tolerably in isolation and badly as a whole, which is the classic condition for a constitutional rather than an administrative remedy.

The stakes were asymmetric, and the asymmetry explains the shape of the eventual bargain. The states stood to surrender the taxes that supplied the bulk of their own revenue. The state value added tax alone had furnished roughly three fifths of the states’ aggregate own tax receipts in the years before 2017, and the states guarded rate-setting power over it as the core of their fiscal autonomy, the one lever a state government could pull without asking New Delhi. Manufacturing states such as Maharashtra, Gujarat and Tamil Nadu also collected central sales tax on goods consumed in other states and feared a destination-based tax that moved revenue to consuming states. Consuming states, poorer on average, stood to gain. The Union, for its part, would surrender the exclusive design of excise and service tax and accept a council in which it could be outvoted on paper by a united states’ bloc. Any settlement had to insure the states against transitional loss, which is how compensation entered the design long before it became the crisis of 2020, and it had to give the states a governing share in the new institution, which is how the Council’s two-thirds state weightage entered Article 279A.

Feature One: The Empowered Committee and the Long Apprenticeship

The first feature of the cooperative federalism architecture is an institution that appears in no constitutional text. The Atal Bihari Vajpayee government constituted the Empowered Committee of State Finance Ministers in 2000 and placed Asim Dasgupta, the finance minister of West Bengal in the Left Front government, in the chair. The choice was itself a federal signal: the committee that would design a national tax was to be chaired not by a Union minister but by a state finance minister from an opposition-ruled state, and its working material was the states’ own tax problem, the replacement of the old sales tax with a value added tax. The committee’s first achievement was the state VAT, adopted from 1 April 2005 by twenty-one states and union territories, with Haryana having moved earlier in 2003 and Uttar Pradesh completing the map on 1 January 2008. VAT mattered beyond its economics. It taught state finance departments to run an invoice-credit tax, harmonised floor rates across states through committee negotiation rather than central direction, and built the habit of deciding tax questions in a room where every state sat.

Why did the Empowered Committee take over a decade and a half to deliver GST?

The committee took sixteen years because the object kept changing shape. It first had to harmonise the state value added tax adopted from 2005, then design a dual GST in its First Discussion Paper of November 2009, then survive changes of chair, two changes of government in New Delhi, and resistance from manufacturing states fearing lost origin revenue.

The committee released its First Discussion Paper on GST on 10 November 2009, proposing a dual GST levied concurrently by the Union and the states on a common base, with an inter-state levy administered through a clearing mechanism. Dasgupta chaired the committee until 2011; Sushil Kumar Modi of Bihar, from the Bharatiya Janata Party, succeeded him and held the chair until 2013, so the drafting institution passed across the three main poles of Indian party politics, Congress-led governments at the Centre watching a process chaired in turn by a communist finance minister and a BJP one. The United Progressive Alliance government introduced the Constitution (One Hundred and Fifteenth Amendment) Bill in March 2011 to carry the design into law; the parliamentary standing committee under Yashwant Sinha examined it, BJP-ruled states objected to clauses they read as a Union veto over state rates, and the bill lapsed when the Fifteenth Lok Sabha dissolved in 2014. The apprenticeship looked like failure from the outside. In retrospect its product was the precondition of 2016: a generation of state finance ministers and officials who had negotiated tax harmonisation among themselves for sixteen years, a settled design vocabulary, and a shared understanding that the bargain would carry compensation and a council, because no other terms had ever survived the committee’s own table.

The politics of the apprenticeship also fixed the cast of the later argument. Asim Dasgupta’s committee embodied the proposition that tax federalism could be negotiated horizontally, state to state, with the Union as a participant rather than a director. Arun Jaitley, who as Union finance minister after 2014 inherited the lapsed bill and reintroduced it in revised form, adopted the committee’s method wholesale: the GST Council he designed is the Empowered Committee written into the Constitution, with the Union inside the room and a voting rule attached. Readers who treat 2014 as a fresh start miss the continuity that Jaitley himself acknowledged in his essays, where the post-2014 negotiation is presented as the completion of a state-led process. The detailed anatomy of the earlier reform era, against which this continuity stands out, is traced in the detailed reading of the 1991 reforms: the 1991 liberalisation was designed inside the Union finance ministry and announced to the states, whereas GST was negotiated with the states for a decade and a half before the Union could announce anything.

The Insurance Precedent: Central Sales Tax and an Unpaid Promise

One memory governed the states’ side of the table throughout the apprenticeship, and no reanalysis of the 2016 bargain is complete without it. Between 2007 and 2010 the Union and the states had agreed to phase out the central sales tax, the origin-based levy on inter-state sales, as a preliminary to GST, cutting its rate in steps from 4 percent to 2 percent against a Union assurance that states would be compensated for the revenue forgone. The phase-out stopped at 2 percent when GST stalled, and the compensation, on the states’ account, stopped earlier and short: state governments carried into the GST negotiation a ledger of unpaid central sales tax compensation claims that they reckoned in the tens of thousands of crores, pressed the claim at meeting after meeting of the Empowered Committee and in correspondence with successive Union finance ministers, and treated the shortfall as the standing proof that a Union promise on tax transition was worth less than its paper. The Union disputed parts of the account and pointed to the sums it had paid, but the political lesson was fixed in the states’ institutional memory and did not depend on the disputed arithmetic.

The lesson surfaced twice in the final design. It explains why the compensation for GST was written into a separate statute with a dedicated cess and a named fund, rather than left as a budgetary assurance of the kind that had failed after 2010: the states demanded a payment machine, not a promise. It also explains the fourteen percent figure. A growth assurance set at the trend rate of the subsumed taxes would have left states exposed to exactly the underperformance they feared, so the premium above trend was the price of the earlier default, charged in 2016 for a debt of trust incurred a decade before. When the states’ finance ministers in 2020 described the compensation clause as the solemn consideration for their surrender of tax powers, they were speaking in this longer ledger, and the Union’s negotiators of 2016, Arun Jaitley among them, had conceded the point in advance by accepting the design they accepted. The reanalysis therefore treats the compensation architecture neither as generosity nor as an accident of drafting, but as a term of trade set by a documented history of non-payment.

Feature Two: The 101st Amendment and the Constitutional Bargain of 2016

The second feature is the constitutional instrument. Finance Minister Arun Jaitley introduced the Constitution (One Hundred and Twenty-second Amendment) Bill in the Lok Sabha on 19 December 2014. The Lok Sabha passed it on 6 May 2015, but the bill stalled in the Rajya Sabha, where the government’s minority status forced the negotiation that improved it. A select committee examined the bill, opposition parties pressed three demands, and on 3 August 2016 the Rajya Sabha passed the bill with the government’s amendments by 203 votes in favour and none against, the AIADMK walking out in protest at the design’s treatment of state powers. The Lok Sabha accepted the amended text on 8 August 2016. State legislatures then ratified the amendment at a speed that itself evidenced the apprenticeship’s work: Assam ratified first on 12 August 2016, sixteen legislatures had ratified within weeks, comfortably past the half of state legislatures the Constitution requires for amendments touching the federal structure, and President Pranab Mukherjee gave assent on 8 September 2016. The bill became the Constitution (One Hundred and First Amendment) Act, 2016.

The Rajya Sabha’s amendments were the federal price, and they repay attention because each one reappears in the later record. The House deleted the proposed one percent additional tax on inter-state supplies, an origin-based levy that manufacturing states had demanded and consuming states and industry had resisted as a reintroduction of cascading. It wrote compensation for revenue loss into the amendment as a five-year assurance rather than a discretionary period. It provided that petroleum crude, high speed diesel, petrol, natural gas and aviation turbine fuel would enter GST from a date the GST Council itself recommends, leaving the most sensitive state revenue lines outside the tax until the joint body decided otherwise. And it strengthened the Council’s dispute mechanism clause, requiring the Council to establish a mechanism to adjudicate disputes arising from its recommendations. Compensation, exclusion and consensus were thus not administrative details added later; they were the terms on which the upper house, speaking for the states in its federal character, consented to the amendment.

The amendment’s operative articles rearranged the tax constitution. Article 246A gave Parliament and every state legislature simultaneous power to legislate on goods and services tax, with Parliament holding exclusive power over inter-state supplies; the old Seventh Schedule boundary between taxing production and taxing sale was dissolved into concurrency. Article 269A provided for the levy and apportionment of the integrated tax on inter-state trade. Article 366(12A) defined the new tax and excluded alcoholic liquor for human consumption from it, leaving that revenue wholly with the states. Article 279A required the President to constitute the GST Council and specified its membership and functions. The amendment sits in the line of constitutional moments examined in the most-consequential-constitutional-moment framework: like the basic structure settlement, it redistributed power between the Union and the states by text, and like that settlement its meaning would be settled less by its wording than by the practice of the institution it created.

Between the Houses: The Select Committee and the Negotiation of 2015 to 2016

The eighteen months between the bill’s passage in the Lok Sabha and its passage in the Rajya Sabha are the best documented stretch of the whole negotiation, and they show the federal bargaining working through Parliament’s own machinery rather than around it. The Rajya Sabha’s select committee on the bill, chaired by Bhupender Yadav of the Bharatiya Janata Party, reported in July 2015. The committee endorsed the bill’s architecture while recording the opposition’s conditions, which the Congress members stated in their notes: a rate ceiling written into the amendment, deletion of the one percent additional inter-state tax, and compensation for states that was full and assured. The committee proposed easing the compensation burden by tapering it across the five years, heavier at the start and lighter as the tax settled, a suggestion the government ultimately set aside. The Congress conditions, by contrast, were substantially met in the government’s own amendments of August 2016, which is why the final vote could be 203 to none: the opposition had been given the substance of its federal objections and could consent without surrendering its separate quarrels with the government of the day.

Behind the parliamentary record ran a second negotiation, in the correspondence and meetings between the Union finance ministry and the states’ negotiators, which the archival trail preserves in the Finance Ministry’s files. By February 2016 the chair of the Empowered Committee had passed to Amit Mitra, the finance minister of West Bengal, so the states’ side of the table in the decisive year was led by a minister of a Trinamool Congress government that would boycott the launch ceremony sixteen months later. The letters of that period worked through the questions the select committee had left open: the compensation formula’s base year and growth rate, the treatment of petroleum, the drafting of the Council’s voting clause, and the assurance on disputes. Jaitley’s method in this exchange, visible in his later essays, was to concede design points that cost the Union discretion while holding the points that cost it the tax itself: the one percent levy went, the five-year compensation assurance hardened, petroleum’s entry was postponed to the Council’s own decision, and in return the states’ negotiators carried the amendment through legislatures their parties controlled at a ratification speed no Union whip could have commanded. The correspondence matters to the reanalysis because it dates the bargain’s true making. The Rajya Sabha’s August 2016 amendments were not improvisations of a single debate; they were the written form of terms settled across the preceding winter between a Union finance minister and a committee chaired by an opposition state.

Feature Three: The Midnight Launch and the Statutes of 1 July 2017

The third feature is the implementation. Parliament passed the Central Goods and Services Tax Act, the Integrated Goods and Services Tax Act, the Union Territory Goods and Services Tax Act and the Goods and Services Tax (Compensation to States) Act in the spring of 2017, with presidential assent on 12 April 2017, and every state legislature passed its own state GST Act in the weeks that followed. At midnight between 30 June and 1 July 2017, in a special session in the Central Hall of Parliament, the tax commenced. President Pranab Mukherjee and Prime Minister Narendra Modi both addressed the session; Modi described GST as a Good and Simple Tax and as the product of a collective federal effort rather than of any single party or government. The Congress, the Trinamool Congress and the Left parties boycotted the midnight session, objecting to a launch staged as spectacle before administrative readiness, a dissent that belongs in the record because it marked the boundary of the consensus: the amendment had passed with near unanimity in 2016, the celebration of its commencement did not command the same house.

What the statutes replaced, in a single night, was the seventeen-levy system described earlier. On the Union side, central excise duty, service tax, countervailing and special additional customs duties and excise-linked cesses were subsumed. On the states’ side, state VAT, central sales tax, entry tax, octroi, purchase tax, entertainment tax, luxury tax, taxes on advertisements and on betting and gambling were subsumed. Jammu and Kashmir, whose special constitutional position required its own legislative sequence, enacted its state GST law on 7 July 2017 and joined the system a week after the national launch. The economy that entered GST in July 2017 was also an economy still absorbing the cash shock administered eight months earlier, a juxtaposition examined in the political-economy framework for the November 2016 demonetisation; traders’ complaints in the first GST months mixed the two disruptions, and the Council’s first year of rate corrections cannot be read apart from that combined discontent.

The Machinery: GSTN, Returns and the E-Way Bill

A pooled tax needs a shared machine, and the machine’s construction was itself a federal negotiation with a paper trail. The Goods and Services Tax Network was incorporated on 28 March 2013, under the previous government, as a not-for-profit company to build and run the common portal on which registration, returns and payments for both levels of government would ride. Its original shareholding placed the Union at 24.5 percent, the states collectively at 24.5 percent, and private financial institutions at 51 percent, an arrangement that drew criticism on data sovereignty grounds from the start: the tax data of every registered business in the country would sit on infrastructure majority-owned outside government. The GST Council took the question up, and in September 2018 the Union Cabinet approved the network’s conversion into a wholly government-owned company, the Union and the states holding half each, a transfer completed the following year. The episode is a small exact instance of the article’s claim: a design inherited from the pre-2016 world was renegotiated inside the new institution until the ownership matched the federal theory.

The machine’s first year tested the theory harder. The return architecture promised in the design, three monthly returns with invoice matching between buyer and seller, proved beyond the portal’s early capacity and the trade’s bookkeeping habits, and the Council improvised: a summary return, the GSTR-3B, introduced for the opening months became the working return for years, filing dates were extended in rolling relaxations, and invoice matching was deferred and then rebuilt in stages. When the electronic way bill for the movement of goods faltered at its first attempted launch on 1 February 2018 and had to be withdrawn within hours, the Council deferred it and relaunched it nationally on 1 April 2018, after which it became a routine instrument generating movement data in volumes no check-post regime had ever assembled. The Council also constituted, in September 2017, a group of ministers on information technology issues under Sushil Kumar Modi of Bihar to ride the portal’s failures week by week, an echo of the Empowered Committee method: a standing political committee, chaired by a state minister, holding a technical system to account. Critics read the first year as proof the tax had been launched unprepared, and the filing record supports them; the reanalysis notes the corrective route. Every repair was routed through the Council and its committees rather than imposed by the Union’s revenue department, so the machinery’s federal character survived the machinery’s failures.

Feature Four: Article 279A and the Voting Architecture of the Council

The fourth feature is the Council itself. Article 279A places the Union finance minister in the chair, seats the Union minister of state for finance, and seats the finance minister of every state and of every union territory with a legislature. A vice-chairperson is chosen from among the state members. The President constituted the Council in September 2016, and it held its first meeting on 22 and 23 September 2016 in New Delhi, fixing a registration exemption threshold of Rs 20 lakh of turnover (Rs 10 lakh for the special category states) and settling into a rhythm of meeting roughly every month. Article 279A charges the Council with recommending the taxes to be subsumed, the goods and services to be taxed or exempted, model laws, place-of-supply principles, thresholds, and the rates including floor rates with bands; Article 279A(6) directs it to be guided by the need for a harmonised structure of GST and a harmonised national market. The Council is a recommending body in text. In practice its recommendations have been enacted without alteration by Parliament and every state legislature, which is the first sign that its authority is political rather than merely textual.

The voting architecture is the constitutional heart of the reanalysis, and its arithmetic deserves exact statement. Under Article 279A(9), every decision is to be taken by a majority of not less than three fourths of the weighted votes of the members present and voting. The vote of the Union government carries a weightage of one third of the total votes cast; the votes of all the state governments taken together carry two thirds. A quorum is one half of the total membership. Run the arithmetic in both directions. The states, holding two thirds, cannot reach three fourths without the Union: a united states’ bloc still falls short of the threshold by one twelfth. The Union, holding one third, cannot impose anything: it needs a large body of states beside it to assemble three fourths. Each side therefore holds a veto, and neither side holds a majority. The design is a mutual veto dressed as a voting rule, and it converted the bargaining of the Empowered Committee years into a permanent constitutional condition: no rate, no exemption, no structural change could be decided over the sustained objection of either level of government. Arun Jaitley, who steered the design, presented this in his essays and speeches as the Council’s defining virtue, a forum in which the Union and the states pool sovereignty over a tax base and in which he, as chair, understood his task to be the manufacture of consensus. The record of how the body actually decided is the test of that presentation, and it is taken up after the framework is mapped.

Two silences in the architecture complete the picture and both matter for the adjudication to come. Article 279A(11) requires the Council to establish a mechanism to adjudicate disputes between governments arising from its recommendations; through the whole period this article covers, no such mechanism was constituted, and the appellate tribunal contemplated by the 2017 Acts likewise remained unestablished, leaving advance ruling authorities in the states as the working forum for classification disputes and political negotiation as the working forum for intergovernmental ones. The compensation war of 2020 was therefore fought in correspondence and in meetings because the Constitution’s own dispute room had never been built, a gap the dissenting states noted when they threatened recourse to the Supreme Court under Article 131 instead. The second silence runs the other way. For a body whose recommendations bind no one in law, the Council’s word has been enacted everywhere: Parliament and thirty-one legislatures have passed its rate decisions, its threshold changes and its procedural relaxations without substantive alteration, so the distinction between recommendation and command, real in the text, has dissolved in practice. The architecture thus combines a formal supremacy of the legislatures with a working supremacy of the Council, and both sides of the later argument, the autonomy critique and the cooperative defence, draw on one half of that combination.

The five features are mapped in the table below, cited in this article as the Five-Feature Map of GST as Cooperative Federalism. The map compresses the argument into its citable form: each feature names an institution or instrument, its anchor dates, and the federal content it carries.

Feature	Anchor dates	Institution or instrument	Federal content
One	2000 to 2016	Empowered Committee of State Finance Ministers under Asim Dasgupta	Horizontal negotiation among states built the design and the trust
Two	19 December 2014 to 8 September 2016	Constitution (One Hundred and First Amendment) Act	Articles 246A, 269A and 279A pooled taxing power by constitutional text
Three	30 June and 1 July 2017	CGST, IGST and UTGST Acts; Compensation to States Act	A single statute book replaced seventeen Union and state levies
Four	September 2016 onward	GST Council under Article 279A	Union one third, states two thirds, decisions at three fourths: a mutual veto
Five	2017 to 2026	Rate rationalisation, compensation, back-to-back loans, cess to March 2026	The bargain tested by revenue stress and renegotiated inside the institution

The launch also enlarged the tax’s democratic footprint in a way the reanalysis should record. The registered taxpayer base under the pooled tax grew from the roughly 65 lakh businesses carried on the pre-GST indirect tax rolls to well over a crore of active registrations within the first three years, as the input credit chain pulled small suppliers into registration to keep their buyers, and as a single national registration replaced the separate state and Union registrations of the old system. Each of those registrations is a monthly encounter between a business and the joint institution, mediated by the portal and governed by rules the Council sets, so the architecture’s federal character is experienced administratively by more citizens than any other shared institution of the republic except the election machinery itself. State commercial tax departments, which had feared absorption into a Union system, retained assessment and enforcement over the ninety percent of smaller taxpayers assigned to them in the January 2017 division, and staffed the portal’s state side; the administrative federation mirrored the legislative one, deliberately, and the mirroring was one of the apprenticeship’s design inheritances.

The Council at Work: Consensus as a Working Rule

The Council’s first decisions displayed the apprenticeship’s habits in constitutional dress. At its meetings of October and November 2016 it wrestled with administrative control of taxpayers, settling at its meeting of 16 January 2017 on a division under which states would administer ninety percent of taxpayers below a turnover of Rs 1.5 crore, with larger taxpayers shared equally, an outcome neither side had entered the room holding. On 3 November 2016 it agreed the four-tier rate structure of 5, 12, 18 and 28 percent, with a cess on luxury and demerit goods to fund compensation. In May 2017, meeting at Srinagar, it fitted goods into the slabs. The pattern was set early: contested questions were referred to committees of officials or to groups of ministers, drafts circulated between meetings, and the meeting itself ratified what negotiation had assembled. The Council’s secretariat and the fitment committee of revenue officials did the technical work, but the political work was done by the norm that the chair would not put a question the room had not already settled.

How often does the GST Council meet, and who convenes it?

The Council met forty-three times between its first meeting on 22 and 23 September 2016 and the end of December 2020, roughly once a month across four years and two Union finance ministers. The chairperson, the Union finance minister, convenes each meeting, and the Council’s secretariat circulates the agenda, the fitment papers and the draft recommendation summaries.

Consent of this kind is easy to mistake for absence of conflict. The meetings were frequently preceded by public disagreement: state finance ministers briefed against proposals, the Union revenue secretary defended revenue estimates in print, and groups of ministers were constituted precisely because the full Council could not settle a question in one sitting. What the consensus norm changed was the location of the fight. Because a decision announced as consensus bound every government in the room, and because any government could credibly threaten to withhold consensus, concessions were priced into proposals before they reached the table. The states extracted the compensation design, the exclusion of petroleum and alcohol, the deletion of the one percent inter-state levy and the administrative division of small taxpayers through this mechanism. The Union extracted a single national base, a common portal in the Goods and Services Tax Network, and the surrender of the states’ largest own tax. The record of unanimous decisions is therefore not evidence that the bargain cost nobody anything; it is evidence that the bargaining happened upstream of the vote that never needed to be called. The Council’s minutes and recommendation summaries circulate in a mixture of document and spreadsheet formats, and a reader comparing a state’s compensation statement against a Finance Ministry release can open both in a single office file viewer without leaving the browser.

The working machinery beneath the consensus deserves its own notice, because it is where the Council’s federal character is most concrete. Rate proposals reach the table only after passing the fitment committee, a body of Union and state revenue officials that tests each item against the revenue arithmetic and drafts the recommendation in agreed language; political disagreement is thereby converted, meeting by meeting, into questions of classification and yield that officials can narrow before ministers must decide. Where officials cannot narrow a question, the Council delegates upward to a group of ministers, a subset of its own members chaired in most cases by a state minister, charged to return with a text the whole room can accept. Across the period the Council ran groups of ministers on the tax’s information technology, on real estate, on the lottery question, on calamity cess after the Kerala floods of 2018, when a state-specific levy was debated and set aside in favour of national instruments, and on rate rationalisation. The pattern is constant: delegation to a smaller federal room, a negotiated text, ratification by acclamation. A reader tracking only the headline meetings misses the institution’s real tempo, which is that of a permanent conference of finance ministries whose members change with state elections while its methods persist.

The membership churn itself became a datum. Between 2016 and 2020 the Council absorbed new state governments of every national and regional formation, ministers who had campaigned against aspects of the tax taking their seats and working its machinery within weeks. The institution’s continuity did not rest on party control at either level; it rested on the fact that every state government, whatever its platform, needed the compensation transfers, the input credit chain and the portal to function, and therefore needed the Council to function. That dependence is the mundane foundation under the constitutional poetry of pooled sovereignty, and it helps explain why even the angriest confrontation of the period, the compensation dispute, was conducted by letters addressed to the chair and by speeches inside the chamber rather than by exit threats. No state, at any point in the record this article covers, proposed leaving the system or legislating a divergent rate, although every state retained, in strict constitutional theory, the concurrent power of Article 246A. The exit door existed in text and was never approached in politics, because the economics of the common market had made it unusable, which is the strongest form an institution’s hold can take.

The Council’s handling of pandemic relief rates in 2021 adds a late entry to the working record that both readings must accommodate. When the second wave of COVID-19 struck in the spring of 2021, state governments demanded rate relief on medicines, oxygen and equipment, and the question arrived at the Council as politics before it arrived as tax design: the forty-fourth meeting on 12 June 2021, working from a group of ministers’ report assembled within days, cut the tax on a schedule of COVID relief items, to 5 percent on vaccines’ companion goods, ambulances and oxygen concentrators and to zero or concessional rates on the most urgent supplies, with the concessions time-bound and reviewed at the following meeting. The sequence matters for the adjudication. A body that merely ratified Union preferences would have had no need of the group of ministers, the compressed timetable or the review clauses; a body paralysed by its veto structure would have produced none of them in a fortnight. What the record shows is the consensus machinery running at emergency speed under pressure from member governments of both national parties, several of whose chief ministers had written publicly to the chair demanding the relief the meeting then delivered.

The Single Vote: Lotteries at the Thirty-Eighth Meeting

Exactly one decision in the Council’s record was taken by formal vote, and its singularity is the cleanest proof that the consensus practice was a practice and not a rule of law. The subject was the tax rate on lotteries. State-run lotteries sold within a state carried GST at 12 percent, while state-authorised private lotteries sold outside the organising state carried 28 percent, a dual structure that lottery-running states, Kerala foremost, defended as protection for their own lottery revenue and that other states attacked as an anomaly. A group of ministers under Maharashtra’s Sudhir Mungantiwar examined the question without producing agreement, and the Council took the Attorney General’s opinion. On 18 December 2019, at its thirty-eighth meeting, chaired by Nirmala Sitharaman, who had succeeded Arun Jaitley after his death in August 2019, consensus failed. Kerala’s finance minister T. M. Thomas Isaac pressed for the dual rate and insisted on a division. Twenty-one states voted for a single uniform rate; seven voted against. The Council then settled the rate itself at 28 percent, effective 1 March 2020.

The episode cuts in two directions, and both belong to the reanalysis. Against a romantic reading of the Council, the vote shows that the institution contains a majoritarian mechanism and that the Union could assemble a winning coalition of states when it chose: twenty-one votes with the Union’s weightage comfortably cleared the three-fourths threshold, and the dissenting states, including a state whose own revenue model was at stake, were simply outvoted. Against a cynical reading, the vote shows how exceptional the mechanism’s use was: thirty-seven meetings of rate fixations, exemptions and structural decisions had been carried without a division, and the chair’s own account of the day stressed reluctance. Sitharaman told reporters that every attempt had been made to keep the consensus practice alive, but that the council was reminded that voting was what the rules provided and that practice was not part of the rules. A constitution that is used once, reluctantly, on a question of one state’s lottery revenue, and whose use is then treated by all sides as a warning rather than a precedent, tells the analyst that the operative constitution of the Council is the consensus norm, with Article 279A(9) held in reserve. Later disputes, including the fiercer compensation fight of 2020, were settled without a second vote.

A Stress Test in Miniature: Kerala’s Flood Cess

Between the launch and the compensation crisis, the architecture faced a smaller test that revealed its capacity for state-specific flexibility inside a pooled tax. The Kerala floods of August 2018 destroyed infrastructure and revenue together in one state while the rest of the system ran normally, and the Kerala government asked the Council for permission to levy a cess of up to ten percent on its state GST to fund reconstruction. The request cut against the system’s founding uniformity: a state-specific levy on the shared base was precisely the kind of divergence the common market had been built to end. The Council neither granted the request in the chamber nor refused it. It referred the question to a group of ministers under Sushil Kumar Modi charged with examining disaster funding across the whole design, and on the group’s report the Council decided, at its meeting in January 2019, that Kerala could levy a calamity cess of one percent on the intra-state supply of goods and services for a period of two years. The cess ran from 1 June 2019, the first and only state-specific levy inside the GST base in the period this article covers.

The episode’s significance is out of proportion to its one percent. It demonstrated that the pooled architecture could bend for a named emergency, in a named state, for a named period, by the joint body’s own decision, without reopening the rate structure for everyone and without a divergent state statute that other states would have had to answer. It also demonstrated the discipline of the route: Kerala, whose finance minister would later lead the compensation resistance, obtained its flexibility not by assertion but by submission to the Council’s process, accepting a tenth of the rate it had sought and a two-year sunset it had not proposed. When the pandemic arrived in 2020 and several states floated disaster levies of their own, the Kerala precedent was the text both sides argued from: the states, to show that emergency flexibility inside the tax was an established practice of the institution; the Union, to show that such flexibility had always been bounded, temporary and granted by the collective rather than assumed by the member. The precedent did not decide the 2020 argument, which turned on compensation rather than on cess, but it had already established the principle the 2020 settlement would confirm at a hundred times the scale: inside this architecture, a member’s emergency becomes the institution’s business, processed in the institution’s rooms, on terms the institution sets.

Rates and Rationalisation: The Politics of the Four Slabs

The rate structure the Council agreed on 3 November 2016 departed from the best technical advice it had received, and the departure was political in the precise sense this series gives the word. The committee on the revenue neutral rate, chaired by the Chief Economic Adviser Arvind Subramanian, reported on 4 December 2015. It computed a revenue neutral rate of 15 to 15.5 percent and recommended a structure of few rates: a standard rate of 17 to 18 percent, a lower rate of 12 percent for merit goods and a demerit rate of 40 percent, warning against a proliferation of rates and exemptions that would complicate administration and invite classification disputes. The Council chose instead four rates, 5, 12, 18 and 28 percent, with a large exempt list and a compensation cess over the top slab. The choice priced the politics of the member governments: food and mass consumption had to sit at zero or five percent for state governments facing electorates, revenue had to be found at the top for finance ministries facing the fourteen percent compensation promise, and the middle slabs absorbed the rest. A single standard rate near the revenue neutral figure would have taxed mass consumption visibly and was not available to any government in the room.

When was the Group of Ministers on rate rationalisation appointed, and what was it asked to do?

The Council constituted the Group of Ministers on rate rationalisation in September 2021, days after its forty-fifth meeting at Lucknow, with Basavaraj Bommai, chief minister of Karnataka, as convenor. It was asked to review the four-slab structure, examine slab mergers, correct inverted duty structures and prune exemptions, reporting back to the Council for decision.

Rate politics after the launch ran in one direction: downward, under pressure, in instalments. At its twenty-third meeting at Guwahati on 10 November 2017, four months after the launch and weeks before the Gujarat assembly election, the Council moved 177 items from the 28 percent slab to the 18 percent slab, leaving about fifty lines at the top, and cut restaurant services to 5 percent without input credit. Through 2018 the pruning continued until the top slab held only a few dozen lines, automobiles, cement, air conditioners and demerit goods among them. Each cut was announced as relief and attacked as revenue forgone; state finance ministers who had voted the cuts in the Council complained outside it that compensation claims were rising as collections thinned, a tension the compensation formula converted into arithmetic. The inverted duty problem, in which inputs taxed at 18 percent feed outputs taxed at 5 or 12 percent and trap working capital in refund claims, survived every round and became the stated ground for the rationalisation exercise of 2021. That exercise met its own political limit within months: when the correction of the textile inversion, a rise from 5 to 12 percent notified for 1 January 2022, drew protests from the Gujarat textile industry and from state governments of both national parties, the Council at its forty-sixth meeting on 31 December 2021 deferred the increase and restored the status quo, referring the whole rate structure back to the Bommai group. The episode is small and diagnostic. Rate rationalisation is the Council’s permanent unfinished business because every rationalisation creates a taxed loser with a government to complain to, and the consensus norm gives that government a seat and a voice at the deciding table.

Two further rate episodes fix the pattern in the record. In early 2019 the Council cut the tax on under-construction housing to 5 percent without input credit, and on affordable housing to 1 percent, effective 1 April 2019, against the revenue officials’ preference and on the stated ground that the 12 percent rate with credit was suppressing a sector every state government courted; the decision was taken weeks before the general election, and opposition finance ministers who voted it through in the chamber criticised its timing outside it, without proposing reversal after the election. The second episode is the classification case law that the four-slab structure inevitably generated. Advance ruling authorities in different states answered identical questions differently, and in June 2020 the Gujarat authority’s ruling that a paratha attracts 18 percent while a plain roti attracts 5 percent became a national news story and a standing joke at the tax’s expense. The comedy had a structural moral the Council’s own members drew in debate: every additional slab and every exemption line multiplies classification boundaries, each boundary generates disputes, and the disputes land back on the Council’s table as demands for clarification circulars and rate corrections. Rate rationalisation, on this record, is not a technocrat’s preference for tidiness; it is the Council’s recurring attempt to reduce the surface area of its own politics.

The same pressure explains the fate of the exercise the Bommai group inherited. Its interim recommendations were still being weighed when the Council, at Chandigarh on 28 and 29 June 2022, accepted a first instalment of corrections: rate changes that took effect on 18 July 2022, among them a 5 percent levy on pre-packaged and labelled food items such as curd, lassi and buttermilk that had previously escaped tax in packaged form, and a set of adjustments to services and inverted duty lines. The comprehensive restructuring, the merger of the 12 and 18 percent slabs that every committee since 2015 had circled, was deferred once more, referred back to the group of ministers for further work. Deferral is the rationalisation story’s refrain across seven years: the Subramanian committee’s lean structure in 2015, the Council’s four slabs in 2016, the pruned top slab of 2017 and 2018, the deferred textile correction of December 2021, the partial instalment of July 2022. Each round moved the structure toward the committees’ advice and stopped where a member government’s electorate began. The reanalysis reads this not as failure but as the operating signature of a tax set by thirty-two governments jointly: it converges slowly, in instalments, at the speed of the most exposed member.

The Compensation Promise: Fourteen Percent, Five Years, One Cess

Compensation was the bargain’s insurance clause, and its terms were fixed by the Goods and Services Tax (Compensation to States) Act of 2017. Each state’s protected revenue was computed from its 2015-16 receipts from the taxes subsumed in GST, grown at fourteen percent a year, compounded. Where a state’s actual GST revenue, including its share of the integrated tax, fell short of the protected line, the Union would pay the difference from a compensation fund fed by a cess on pan masala, tobacco products, coal, aerated drinks and motor vehicles, goods chosen because their consumption was price-inelastic or demerit-rated. The guarantee ran for five years from the launch, to 30 June 2022. Payments were to be made every two months on a provisional basis and finalised against audited figures. Fourteen percent was a generous line: it exceeded the trend growth of the subsumed taxes in most states, and it was set there because the states, remembering the unmet promises that accompanied the phase-out of central sales tax after 2007, demanded an insurance premium for signing, and the Union, needing ratification and a launch date, paid it.

How was the fourteen percent compensation formula calculated?

The formula took each state’s audited 2015-16 revenue from the subsumed taxes as its base, projected that base forward at fourteen percent growth compounded each year, and treated the gap between the projection and the state’s actual GST collections, including its share of the integrated tax, as the loss payable from the cess-fed fund.

In the calm years the insurance was cheap and the fund ran a surplus. The Finance Ministry’s statement of July 2020 records compensation cess collections of Rs 62,612 crore in 2017-18 against releases of Rs 41,146 crore, and Rs 95,081 crore in 2018-19 against releases of Rs 69,275 crore; Rs 47,271 crore of cess collected across the first two years remained unutilised in the fund at the end of March 2019. The strain appeared from August 2019, as a slowing economy thinned collections while the protected line kept compounding at fourteen percent. Nirmala Sitharaman told the Rajya Sabha in December 2019 that the monthly compensation requirement was running near Rs 14,000 crore against average cess collections of Rs 7,000 to Rs 8,000 crore a month, and the Union released Rs 35,298 crore of accumulated dues days later. For 2019-20 as a whole the Union released Rs 1,65,302 crore of compensation against cess collections of Rs 95,444 crore, meeting the gap from the accumulated fund balance and from a transfer of Rs 33,412 crore representing the year-end balance of integrated tax for 2017-18 apportioned into the fund. The arithmetic of 2019 is the prologue to the crisis of 2020: the guarantee had become a large, structural transfer, running at more than the dedicated tax that funded it, a full year before the pandemic arrived.

The distribution of the 2019-20 releases maps the original geography of the surrender. The Finance Ministry’s state-wise statement for that year records the largest transfers flowing to Maharashtra at Rs 19,233 crore, Karnataka at Rs 18,628 crore, Gujarat at Rs 14,801 crore, Tamil Nadu at Rs 12,305 crore and Punjab at Rs 12,187 crore, with Delhi at Rs 8,424 crore and Kerala at Rs 8,111 crore behind them. The pattern is the apprenticeship’s politics rendered in rupees: the manufacturing and high-consumption states that had given up the most buoyant and most origin-linked revenues drew the largest insurance, and the producing states whose consent the 2016 bargain had most needed were, by 2019, its largest beneficiaries. A transfer of this shape is not a subsidy programme and was never defended as one; it is the premium schedule of the constitutional bargain, paid to the parties whose signatures the bargain required, and its concentration at the producing end of the federation is the clearest single measure of what those signatures had cost.

Standard accounts of the tax, the encyclopaedia entries and the policy histories against which this reanalysis is written, treat the compensation story as a fiscal footnote to a tax reform, a transitional arrangement to be noted and passed over on the way to rates and returns. The reanalysis inverts the emphasis deliberately. On the record assembled here, the compensation mechanism is the load-bearing wall of the whole structure: it was the term that purchased the amendment, the arithmetic that governed the Council’s rate politics from 2017 onward, the subject of the institution’s only existential crisis, and the precedent through which the Union and the states renegotiated the meaning of their compact under stress. A reader who understands the compensation bargain understands why the Council votes the way it does, why the slabs look the way they look, and why the tax survived 2020; a reader who skips it holds a description of a tax and has missed the federation that the tax built.

The Compensation Crisis of 2020: The Promise Meets the Pandemic

The pandemic turned a financing gap into a constitutional argument. India entered a national lockdown on 25 March 2020; gross GST collections for April 2020 fell to Rs 32,294 crore, a fraction of the monthly norm, and state governments simultaneously lost their remaining own revenues from liquor, fuel and stamps, which GST had left outside its base and the lockdown froze anyway. The wider politics of that emergency, the centralisation of decision-making it licensed, is the subject of the COVID-19 governance record. Inside the Council, the question was narrower and harder: who owed the states the difference between the protected line and collections that had collapsed for reasons no tax design caused? The Goods and Services Tax (Compensation to States) Act obliges payment from the compensation fund; the fund held cess, and cess was yielding a fraction of the requirement. The Union’s law officers advised that the obligation was bounded by the fund. Attorney General K. K. Venugopal’s opinion, on which the Union acted, separated the shortfall into two parts: the part attributable to the implementation of GST itself, and the part attributable to the pandemic, which the Union’s presentation described through the finance minister’s phrase, an act of God.

At the forty-first meeting of the Council on 27 August 2020, devoted to compensation, the Union placed two borrowing options before the states, computed on a projected shortfall for 2020-21 of Rs 2,35,000 crore, of which Rs 97,000 crore was reckoned the implementation component. Under the first option, states would borrow Rs 97,000 crore through a special window arranged with the Reserve Bank at a reasonable rate of interest, the debt serviced from the compensation cess whose levy would extend beyond 2022, and the borrowing would not count against the states’ own debt limits, carrying an added unconditional relaxation of half a percent of state domestic product in borrowing space. Under the second option, states would borrow the entire Rs 2,35,000 crore from the market, at their own cost and within revised limits, with less protection. What the Union would not do, in August 2020, was borrow the money itself and transfer it: the options were addressed to the states as borrowers, and the compensation promise, which state governments read as a Union guarantee, was restated as a claim on a fund that happened to be empty.

Which state finance ministers rejected the borrowing options in 2020, and what did they demand instead?

Kerala’s T. M. Thomas Isaac, West Bengal’s Amit Mitra and Punjab’s Manpreet Singh Badal led the rejection after the forty-first meeting, joined by ministers from Chhattisgarh, Jharkhand, Rajasthan, Maharashtra and Delhi. They demanded that the Union borrow the full shortfall itself at the sovereign rate and transfer the proceeds as compensation under the statute.

The rejection was argued at full strength, and the reanalysis must render it so. Isaac held that the two options forced states to borrow for what the statute promised them, that the distinction between an implementation shortfall and an act of God appeared nowhere in the Compensation Act’s text, and that Kerala would test the Union’s reading in court if required. Mitra wrote to the Union finance minister that making the states borrow to honour the Union’s assurance was a betrayal of the compact on which the states had surrendered their taxes, and warned that the precedent would govern every future federal bargain. The opposition-ruled states also made the institutional argument: the Council existed to decide such questions together, yet the options had been framed in the Union finance ministry and presented as a menu of two, with the excluded third option, Union borrowing, being the one a majority of states preferred. The Union’s answer, at its strongest, was fiscal and legal: the statute created a fund, not an open-ended charge on the Consolidated Fund; the Union’s own revenues had collapsed in the same months; and a borrowing programme by the states, protected from their debt ceilings and serviced by an extended cess, delivered the identical cash to state treasuries at a modest spread. At the forty-second meeting on 5 October 2020, twenty states had taken the first option; ten states and union territories held out. The forty-third meeting on 10 October 2020 sat for hours and ended without consensus, the only meeting in the compensation sequence to close in open disagreement, after which the Union finance minister announced that states choosing the first option would be facilitated individually and that no state would be denied the arrangement for want of a collective decision.

The States’ Case in Their Own Words

The dispute’s paper trail is the finest source the reanalysis has, because the principals argued in dated documents rather than in recollection. The states’ case, assembled from the letters and statements of September and October 2020, ran on three rails. The legal rail held that the Compensation Act promised payment of the gap between actual collections and the protected line, that the Act named the fund as the payment route and not as a ceiling on the obligation, and that a shortfall in an earmarked fund could not dissolve a statutory assurance given as the consideration for a constitutional surrender. The historical rail reached back to the central sales tax compensation ledger: the states had been underpaid on a transition promise once, had priced the memory into the fourteen percent premium of 2016, and were being asked, on their account, to accept a second default dressed as a borrowing menu. The institutional rail held that the options had been framed outside the Council, on a law officer’s opinion the states had not sought, and announced as a choice between two forms of the same burden, so that the body the Constitution built for exactly this class of question was being used to ratify a decision taken elsewhere.

Named ministers carried each rail in public. Isaac, in Kerala, combined the legal and historical arguments and added the distributional one: Kerala’s own revenues had collapsed with remittances and consumption, the state faced the Union across a table at which it held a two-thirds collective vote it could not convert into three fourths, and recourse to the Supreme Court under Article 131 remained open to it. Mitra, in West Bengal, had in fact opened the compensation front a year early: at the thirty-eighth meeting in December 2019 he had warned that the Union’s own projections showed the compensation fund running dry within months, a warning the December release of Rs 35,298 crore in accumulated dues partly vindicated, and his letters of 2020 framed the borrowing options as a betrayal of the assurance given in 2016 by the very finance minister, Jaitley, under whom the states had consented, a framing sharpened by the fact that Mitra had himself chaired the Empowered Committee in the amendment’s decisive year. Manpreet Singh Badal in Punjab pressed the fiscal rail: Punjab’s compensation dependence was among the highest of any state relative to its own revenue, and a menu of borrowings was, for Punjab, an instruction to mortgage the revenues the tax had taken. Chhattisgarh’s T. S. Singh Deo wrote in support of Union borrowing serviced by an extended cess, the mechanism that was, in the end, adopted almost exactly as the dissenting states had specified it in their correspondence.

The Union’s written case, in the finance ministry’s clarifications issued from 29 August 2020 and in the finance secretary’s meetings with state officials that September, answered rail for rail. On the law, the obligation ran to the fund. On the history, the compensation had been paid in full for three years, including Rs 1,65,302 crore for 2019-20 raised well beyond that year’s cess. On the institution, the options had been placed before the Council at its forty-first meeting and debated at the forty-second and forty-third, and nothing in the Attorney General’s opinion prevented the Council from fashioning a different course if a different course could command the room. The last point, offered as a rebuttal, reads in retrospect as the settlement’s own preview: the course the room eventually commanded was the dissenters’ course, executed through the Union’s machinery once the Union accepted that withholding consent had made its menu untenable. The correspondence thus documents the precise sense in which the Council decided the crisis: not by voting the Union down, which its arithmetic forbids, but by refusing, at sufficient scale and for sufficient time, to let the Union’s question be the question.

The Settlement: Back-to-Back Loans and the Cess That Outlived the Guarantee

The settlement, when it came, conceded the substance of the states’ demand while preserving the form of the Union’s options. On 15 October 2020 the Union finance ministry announced that the Union government would itself borrow under the first option’s architecture and pass the proceeds to the states as back-to-back loans, not counted as state debt, serviced from the extended cess. The Union borrowed Rs 1.1 lakh crore in 2020-21, releasing it to the states in tranches from late October 2020, and repeated the arrangement at Rs 1.59 lakh crore for 2021-22, front-loading releases at the states’ request so the cash arrived in the first part of the year: Rs 2.69 lakh crore in all, raised on the Union’s credit and carrying the Union’s lower rate. Jharkhand, the last holdout, accepted the arrangement in December 2020, and every state and union territory with a legislature was inside the settlement before the year turned. The Union’s April 2022 accounting states that Rs 7.35 lakh crore of compensation had been released in aggregate by that date, the loans included; a further release of Rs 86,912 crore on 31 May 2022 cleared the outstanding dues for 2021-22 in full. The guarantee itself expired on schedule on 30 June 2022. The cess did not. The Council decided at its forty-fifth meeting at Lucknow on 17 September 2021 that the compensation regime would end in June 2022 but that the cess levy would continue to March 2026 solely to service the back-to-back debt, and the Goods and Services Tax (Period of Levy and Collection of Cess) Rules, 2022, notified on 25 June 2022, extended the levy from 1 July 2022 to 31 March 2026. The Union paid Rs 7,500 crore of interest on the loans in 2021-22 and budgeted Rs 14,000 crore of interest for 2022-23, with principal repayment running from 2023-24 to March 2026.

Read federally, the sequence vindicates and disciplines both sides’ theories. The states’ reading of the compact prevailed in substance: the compensation promise was honoured in full, in cash, across all five years, at a total the Finance Ministry itself put at Rs 7.35 lakh crore by April 2022, and it was honoured by the Union borrowing on its own balance sheet, the very course the Union had refused in August 2020. The Union’s reading prevailed in form: the statute was never amended, the Attorney General’s construction was never withdrawn, the loans were booked outside the compensation fund’s ordinary course, and the cost of honouring the promise was pushed onto a cess that consumers will pay until March 2026, four years beyond the guarantee it financed. Neither side called a vote. The Council that had been designed as a mutual veto produced, under maximum stress, a negotiated transfer larger than most Union budgets’ capital outlay, and produced it through the chair’s bilateral diplomacy after the formal meeting had deadlocked at the forty-third sitting. The institution held; it held by bargaining, which is what it was built to do.

The settlement’s tail runs to 2026 and belongs to the trajectory the fifth feature names. The compensation cess, designed as a five-year instrument ending with the guarantee, was extended as a debt service instrument for four further years, its collections from July 2022 committed to interest and then to principal on Rs 2.69 lakh crore of loans whose proceeds the states had already spent as compensation. Consumers of tobacco, pan masala, aerated drinks, coal and motor vehicles will pay, through the decade’s middle, for a revenue collapse that occurred in 2020, under a levy that no Finance Commission shares and no state legislature voted as a tax of its own. The arrangement is defensible on its terms, since the debt was incurred for the states’ benefit at the Union’s cheaper rate, and it is nonetheless the settlement’s most quietly centralising artefact: a Union-administered cess, outside the divisible pool, extended by a Council decision, servicing Union borrowing. Both descriptions are true at once, and holding them together is the habit the reanalysis asks of its reader before the measurements are weighed.

Measuring the Fiscal-Federal Balance

The strongest scholarly statement against the cooperative reading is M. Govinda Rao’s. In Studies in Indian Public Finance (2022), Rao reads the GST settlement as a net transfer of fiscal autonomy from the states to a collective body in which the Union’s one-third weightage operates as a standing veto, since no decision can reach three fourths against the Union’s opposition. The states, on this reading, surrendered rate-setting power over the most buoyant part of their own revenue, accepted a compensation formula whose fourteen percent assurance dulled their incentive to administer the tax, and emerged at the end of the guarantee period in June 2022 facing a revenue cliff with collections that had grown well below the protected path in most years. Rao’s measurement discipline is the point his critics must answer: the protected line compounded at fourteen percent, actual state GST revenues grew far more slowly, and the difference, financed first by cess and then by Union borrowing, is the quantified size of the states’ dependence on the settlement. Pinaki Chakraborty’s analyses of compensation dependence across the period add the distribution: the transfer loomed largest, relative to own revenue, in states such as Punjab, Chhattisgarh, Himachal Pradesh, Uttarakhand and Kerala, several of them states whose finance ministers then led the resistance of 2020, a correlation the autonomy reading predicts and the cooperative reading must absorb.

The Fifteenth Finance Commission’s award, tabled on 1 February 2021 and covering 2021-22 to 2025-26, supplies an independent measurement of the same balance. The Commission reduced the states’ share of the divisible pool from the Fourteenth Commission’s 42 percent to 41 percent, attributing the point to the reorganisation of Jammu and Kashmir, whose government the Union funds directly. More consequential for the GST argument, the Commission recorded the growth of cesses and surcharges in the Union’s gross tax revenue, levies that sit outside the divisible pool and therefore outside devolution, as a structural concern. The compensation cess is exactly such a levy, extended to 2026 outside the pool the states share; the federal balance, measured by what flows through the constitutional devolution channel against what flows through earmarked channels the Union administers, tilted toward the earmarked side across the compensation years. Against both measurements stands Vijay Kelkar’s reading in In Service of the Republic (2019). Kelkar, whose 2009 task force had drawn the design, treats the GST as the republic’s outstanding case of cooperative federalism in economic policy precisely because the states consented, through their own committee and their own ministers in the Council, to every element Rao counts as surrender, and consent renewed monthly in a body where states hold two thirds of the vote is a different constitutional fact from imposition. Kelkar’s prescription follows from the reading: converge the slabs toward fewer rates, and bring petroleum, electricity and real estate into the base by the same consensual route.

The adjudication this article defends sits between the measurements and is resolved by the 2020 record. Rao is right about the direction of fiscal dependence and about the veto arithmetic, and wrong, on the evidence of the crisis year, about what the arithmetic produced: when the system’s survival was priced, the veto structure compelled the Union to borrow Rs 2.69 lakh crore rather than compel the states to default on a statutory expectation, because the Union could not govern the tax over the states’ sustained objection. Dependence ran both ways, which is the definition of the cooperative architecture rather than its refutation. The settlement’s genuineness is also its limit: petroleum, natural gas, aviation fuel, electricity and alcohol remain outside the base by the Council’s own postponed decision, so the pooling covers the taxes the states were willing to pool in 2016 and no more.

Two further measurements belong in the balance before the adjudication. The Reserve Bank’s annual studies of state finances, and the assessments produced at the National Institute of Public Finance and Policy by Sacchidananda Mukherjee and others, tracked the effective rate of the tax as the slabs were pruned: the weighted average rate drifted steadily below the 15 to 15.5 percent revenue neutral benchmark of 2015, toward a figure near eleven to twelve percent by the end of the decade’s third year, so that the system’s revenue underperformance against the protected line was partly a chosen consequence of the Council’s own rate decisions, a fact both sides of the autonomy argument can claim, the critics as evidence that the states traded revenue for applause lines, the defenders as evidence that rate policy remained responsive to the taxed economy. The second measurement is the recovery path: gross monthly collections crossed Rs 1 lakh crore again in October 2020, at Rs 1,05,155 crore, for the first time since February of that year, and the compensation requirement for 2021-22 was met through the second borrowing tranche rather than through any reopening of the dispute, which suggests the parties themselves treated the October 2020 settlement as a settled precedent for the guarantee’s remaining life. A bargain that survives its worst year and then administers its own aftermath by routine has passed from negotiation into institution, which is the threshold the reanalysis claims the GST crossed in 2020.

One post-guarantee measurement complicates every position in the debate and belongs in the adjudication before the counter-reading is heard. The revenue cliff that the autonomy critique forecast for July 2022 did not arrive on schedule. Gross GST collections across 2022-23 averaged Rs 1.51 lakh crore a month on the Finance Ministry’s reporting, the strongest sustained run of the tax’s first six years, powered by the compliance machinery, by the withdrawal of pandemic distortions and by rate corrections including the July 2022 instalment; state GST revenues grew at rates that, for that year at least, approached the protected line the guarantee had drawn, without any guarantee standing behind them. The datum does not settle the argument. A single strong year does not restore a surrendered rate-setting power, and the earmarked cess that services the back-to-back loans still runs outside the divisible pool until March 2026. What the strong year establishes is narrower and worth having: the dependence the critics measured was a function of a weak-rate, weak-compliance phase and of a pandemic, not a permanent arithmetic of the design, and the design’s own corrective machinery, the same Council that had pruned the slabs in the applause years, retained the capacity to raise the effective rate when its members judged the moment safe. Federal bargains should be judged across a full cycle rather than at their trough, and the cycle’s evidence, trough and recovery together, is the record this reanalysis has tried to state.

The Centre-Centred Counter-Reading, at Its Strongest

The complication the brief for this article insists on must be faced without softening. A substantial body of commentary reads the post-2017 record as centralisation wearing federal dress. The Council’s chair is the Union finance minister; its secretariat sits in the Union finance ministry and is headed by the Union revenue secretary; the agenda, the fitment committee’s papers and the drafting of recommendations are Union-administered; and the wider context of the period, examined in the post-2014 PMO architecture context, is one of executive concentration in the Union government, framed as a puzzle in the integrated three-tensions framework for the post-2014 years. The 2020 options episode is the counter-reading’s best exhibit: the decisive document of the crisis, the two-option menu, was drafted in the Union finance ministry, rested on a Union law officer’s opinion, and was presented to the Council as a choice between two forms of state borrowing, with the states’ preferred course excluded. On this reading, the Council ratifies; it does not decide.

The exhibit is real, and the honest answer is concessive in part. Agenda power in the Council is asymmetric, and 2020 proved that the chair can frame a crisis decision before the room assembles. But the counter-reading must then explain the crisis’s ending. If the Council merely ratified Union preferences, the settlement of October 2020 should have been the August menu executed; instead the Union abandoned its menu’s core feature, borrowed on its own account in two successive years, and paid in full, because the states’ collective refusal, sustained through two meetings and a public correspondence campaign, raised the price of the menu above its value. An institution in which the weaker party can, by withholding consent, reverse the stronger party’s settled position is not a rubber stamp, even if the stronger party holds the pen that drafts the agenda. The same two-level pattern appears in the parallel framework this series applies elsewhere: as the foreign-policy-as-domestic-politics framework shows for external policy, a decision’s domestic institutional processing is where its real allocation of power is revealed. Processed through the Council, the compensation decision allocated the borrowing to the Union. That is the cooperative federalism claim, stated at its most defensible strength: not that the partners are equal, but that the architecture makes their consent necessary and therefore makes their interests count in outcomes.

Verdict: What the Reanalysis Establishes

The five features converge on the namable claim with which the article began. GST operates as cooperative federalism in institutional form: designed horizontally by the states’ own committee between 2000 and 2016, constitutionalised by an amendment that sixteen state legislatures ratified within weeks, launched through mirror statutes at both levels, governed by a council whose voting rule is a mutual veto and whose practice is consensus, and stress-tested by a compensation crisis that ended with the Union borrowing Rs 2.69 lakh crore to keep a promise the states had priced into their consent. The technical framing cannot see this because it reads the Council as a committee of the tax; the policy history framing cannot see it because it stops at the launch. The reanalysis, reading the institution across its whole working record to the cess extension of 2026, sees a federal bargain that has become a permanent organ of the republic, flawed in its rate structure, incomplete in its base, asymmetric in its agenda power, and nonetheless the strongest working example of pooled sovereignty in Indian economic government. The arguments that will shape its next decade, slab convergence, the inclusion of petroleum and electricity, the fiscal settlement after compensation, will be fought in the Council chamber and nowhere else. That, finally, is what 1 July 2017 operationalised.

The open questions the record leaves are themselves evidence for the reading, because every one of them sits inside the institution rather than outside it. The slab structure awaits the convergence that the Bommai group’s deferred reports circle, and the deferral of December 2021 and the partial instalment of July 2022 mark the pace at which thirty-two governments can move a price that every household sees. Petroleum, natural gas and aviation fuel await the recommendation Article 279A(5) reserves to the Council, and the postponement, renewed year after year by the joint body itself, is a federal decision of the same dignity as inclusion would be, taken by the governments that collect the excluded revenues and are not yet willing to pool them. The dispute mechanism of Article 279A(11) awaits construction, and the compensation war showed both the cost of its absence and the system’s capacity to settle its deepest fight without it. A reanalysis is judged by whether it leaves the reader able to locate the next event, and on the cooperative federalism reading the location is never in doubt. Whatever happens to the Goods and Services Tax from here will happen in a room where the Union holds one third of the vote, the states hold two thirds, and nothing passes at less than three fourths, which is to say it will happen the way the tax was made: together, slowly, and under the discipline of a consent neither side can manufacture alone.

Frequently Asked Questions

Q: When was GST implemented in India?

The Goods and Services Tax was implemented at midnight between 30 June and 1 July 2017, at a special session of Parliament held in the Central Hall in New Delhi. President Pranab Mukherjee and Prime Minister Narendra Modi addressed the session, and the Central Goods and Services Tax Act, the Integrated Goods and Services Tax Act, the Union Territory Goods and Services Tax Act and the Goods and Services Tax (Compensation to States) Act, all assented to on 12 April 2017, commenced together with the state GST Acts passed by every state legislature. Jammu and Kashmir enacted its own law on 7 July 2017 and joined the system one week after the national launch.

Q: What is the 101st Constitutional Amendment Act?

The Constitution (One Hundred and First Amendment) Act, 2016 is the constitutional instrument that made the Goods and Services Tax possible. Introduced as the One Hundred and Twenty-second Amendment Bill on 19 December 2014, it was passed by the Rajya Sabha with amendments on 3 August 2016 by 203 votes to none, accepted by the Lok Sabha on 8 August 2016, ratified by more than half of the state legislatures beginning with Assam on 12 August 2016, and assented to by the President on 8 September 2016. The Act inserted Article 246A, giving Parliament and the state legislatures simultaneous power to tax goods and services, Article 269A on the integrated tax for inter-state supplies, and Article 279A creating the GST Council, while defining the new tax in Article 366(12A) and excluding alcoholic liquor for human consumption from it.

Q: What is the GST Council’s voting architecture?

Article 279A(9) of the Constitution gives the GST Council a weighted voting architecture unlike any other Indian institution. The vote of the Union government carries a weightage of one third of the total votes cast at a meeting, and the votes of all the state governments taken together carry a weightage of two thirds. A decision requires a majority of not less than three fourths of the weighted votes of the members present and voting, and one half of the total membership forms the quorum. The arithmetic makes the rule a mutual veto. The states, even voting as a single bloc, hold only two thirds and cannot reach three fourths without the Union, while the Union, holding one third, cannot carry any proposal without a large body of states beside it, so neither level of government can impose a decision on the other.

Q: How is GST a cooperative federalism architecture?

GST is a cooperative federalism architecture because the Union and the states surrendered their separate powers over indirect taxation into a shared pool and govern the pool together. The Constitution (One Hundred and First Amendment) Act dissolved the old boundary under which the Union taxed manufacture and services while the states taxed sales, replacing it with concurrent power under Article 246A. Every rate, exemption and structural rule is set by the GST Council, where the Union holds one third of the weighted vote and the states two thirds, and where decisions require a three-fourths majority that neither side can assemble alone. The compensation guarantee, assuring states fourteen percent annual growth in subsumed revenues for five years, was the insurance that made the surrender acceptable, and the Council’s practice of deciding by consensus across forty-three meetings to December 2020 shows the architecture operating as a standing negotiation between the two levels of government rather than as central direction.

Q: What are the four GST tax slabs?

The GST Council, at its meeting of 3 November 2016, adopted a four-tier rate structure of 5, 12, 18 and 28 percent, alongside a zero rate for exempt goods and services and a compensation cess levied over the top slab on luxury and demerit goods such as tobacco products, pan masala, aerated drinks and motor vehicles. Food grains and many mass consumption items sit at zero or 5 percent, standard goods and services cluster at 12 and 18 percent, and the 28 percent slab was progressively pruned after the launch: the Council moved 177 items from 28 to 18 percent at Guwahati on 10 November 2017, leaving the top slab concentrated on automobiles, cement, air conditioners and demerit goods by 2019. The four slabs departed from the revenue neutral rate committee’s 2015 recommendation of fewer rates, reflecting the member governments’ pricing of electoral and revenue pressures.

Q: Who chairs the GST Council?

The Union finance minister chairs the GST Council. Article 279A seats the Union finance minister as chairperson, the Union minister of state for finance as a member, and the finance minister of every state and of every union territory with a legislature as members, with a vice-chairperson chosen from among the state members. Arun Jaitley, who had steered the constitutional amendment through Parliament, chaired the Council from its constitution in September 2016 and through the launch period, presiding over the rate structure decisions and the first rounds of rate correction. After Jaitley’s death in August 2019, Nirmala Sitharaman chaired the Council, presiding over the lottery vote of December 2019, the compensation crisis meetings of 2020 and the settlement that followed. The chair convenes meetings and, by the Council’s consensus practice, carries responsibility for assembling agreement before questions are put.

Q: How did GST replace the pre-2017 indirect tax architecture?

GST replaced the pre-2017 architecture by subsuming the separate levies of both levels of government into a single destination-based tax on supply, charged concurrently by the Union and the states with a continuous credit for tax paid on inputs. On the Union side it absorbed central excise duty, service tax, countervailing and special additional customs duties and excise-linked cesses. On the states’ side it absorbed state value added tax, central sales tax, entry tax, octroi, purchase tax, entertainment tax, luxury tax and taxes on advertisements and on betting and gambling. Inter-state supplies attract the integrated tax, apportioned between the Union and the destination state, which ended the origin-based central sales tax that had rewarded producing states and taxed goods for crossing internal borders. Taxes on alcoholic liquor for human consumption stayed wholly with the states, and petroleum products remained outside GST pending a Council recommendation, so the replacement, though sweeping, was deliberately incomplete.

Q: Why did the Empowered Committee of State Finance Ministers matter for the design of GST?

The Empowered Committee mattered because it built both the design and the trust that a constitutional bargain required, over sixteen years. Constituted in 2000 by the Vajpayee government and chaired by Asim Dasgupta, the finance minister of West Bengal, it first harmonised the state value added tax adopted from April 2005, then produced the First Discussion Paper on GST in November 2009, proposing a dual GST on a common base. Under Dasgupta and his successor Sushil Kumar Modi of Bihar, the committee accustomed state finance ministers and officials to negotiating tax questions horizontally, state to state, with the Union as participant. When the constitutional amendment finally moved in 2014 and 2016, the compensation principle, the council form and the dual structure were already the committee’s settled vocabulary, and the GST Council itself is best understood as the Empowered Committee written into the Constitution with a voting rule attached.

Q: What did the Rajya Sabha change in the Constitution Amendment Bill in August 2016?

The Rajya Sabha’s amendments of 3 August 2016 wrote the federal price of the amendment into its text. The House deleted the proposed one percent additional tax on inter-state supplies, an origin-based levy demanded by manufacturing states that would have reintroduced cascading and rewarded the state of origin rather than the state of consumption. It converted compensation for state revenue loss into a five-year assurance in place of a more discretionary formulation, so that the insurance on which states were surrendering their taxes stood in the amendment itself. It provided that petroleum crude, high speed diesel, petrol, natural gas and aviation turbine fuel would enter GST only from a date the GST Council recommends, postponing the most sensitive revenue question to the joint body. And it strengthened the requirement that the Council establish a mechanism to adjudicate disputes arising from its recommendations. The Lok Sabha accepted the amended bill on 8 August 2016 without further change.

Q: What was the revenue-neutral rate debate, and why did India adopt four slabs instead of a single rate?

The revenue neutral rate is the single rate at which a new tax yields the same revenue as the taxes it replaces. The committee chaired by Chief Economic Adviser Arvind Subramanian reported on 4 December 2015 that India’s revenue neutral rate was 15 to 15.5 percent, and recommended a lean structure: a standard rate of 17 to 18 percent, a lower rate of 12 percent for merit goods, and a demerit rate of 40 percent, with few exemptions. The GST Council instead adopted four slabs of 5, 12, 18 and 28 percent plus a cess on demerit goods at its meeting of 3 November 2016. The wider structure priced the politics of the member governments. A single rate near 17 percent would have taxed food and mass consumption visibly in every state, while the compensation promise required revenue that only a high top slab could promise, so governments distributed the burden across tiers and accepted classification complexity as the cost of consent.

Q: Which was the only GST Council decision taken by a formal vote, and what did it decide?

The only formal vote in the GST Council’s record was taken at its thirty-eighth meeting on 18 December 2019, on the taxation of lotteries. State-run lotteries sold within a state carried GST at 12 percent, while state-authorised private lotteries carried 28 percent, and a group of ministers under Sudhir Mungantiwar had failed to produce agreement on unifying the rates. Kerala’s finance minister T. M. Thomas Isaac defended the dual rate and insisted on a division. Twenty-one states voted for a single uniform rate and seven voted against, and the Council then fixed the uniform rate at 28 percent, effective 1 March 2020. Chairperson Nirmala Sitharaman recorded that every attempt had been made to preserve the consensus practice before the vote was taken, and no second vote followed in the compensation dispute of 2020, which was settled by negotiation instead.

Q: Why did the GST Council move 177 items out of the 28 percent slab in November 2017?

At its twenty-third meeting at Guwahati on 10 November 2017, the Council moved 177 items from the 28 percent slab to the 18 percent slab, leaving about fifty lines at the top, and reduced restaurant services to 5 percent without input credit. Four months of the tax’s operation had produced a concentrated political problem: the top slab had been fitted with many mass-use goods, from shampoos and shaving products to construction materials, and trader and consumer complaints arrived at every member government while the Gujarat assembly election approached. The Council answered with the largest single rate correction of the launch period. The move also exposed the structure’s standing tension: each reduction in the top slab lowered collections against a compensation promise growing at fourteen percent, so relief voted in the Council reappeared as compensation claims on the Union, a connection the members debated openly in the meetings that followed.

Q: What triggered the GST compensation dispute of 2020?

The dispute was triggered when the COVID-19 pandemic collapsed GST collections against a compensation promise that kept compounding. The national lockdown from 25 March 2020 cut gross GST collections for April 2020 to Rs 32,294 crore, and the projected shortfall for 2020-21 reached Rs 2,35,000 crore while the compensation cess, the only source the Compensation Act names for payment, yielded a fraction of the requirement. The Union’s law officers advised that the obligation was bounded by the fund, and the Attorney General’s reading separated the shortfall into a part caused by GST implementation, reckoned at Rs 97,000 crore, and a part caused by the pandemic, which the Union finance minister described as an act of God. States answered that the Act promised compensation without that distinction. The forty-first Council meeting on 27 August 2020, called to address compensation, turned the gap in the fund into a dispute over the meaning of the 2016 compact.

Q: What were the two borrowing options the Union government offered the states in August 2020?

At the forty-first GST Council meeting on 27 August 2020, the Union offered the states two borrowing options against a projected compensation shortfall of Rs 2,35,000 crore for 2020-21. Under the first option, states would borrow Rs 97,000 crore, the component attributed to GST implementation, through a special window arranged with the Reserve Bank at a reasonable rate of interest; the debt would be serviced from the compensation cess extended beyond 2022, would not count against the states’ debt limits, and carried an unconditional relaxation of half a percent of state domestic product in borrowing space. Under the second option, states would borrow the entire Rs 2,35,000 crore from the market themselves, at their own cost and with less protection on limits and servicing. Both options made the states the borrowers for a shortfall the states read as the Union’s statutory liability, which is why a majority of opposition-ruled states rejected the menu and demanded that the Union borrow instead.

Q: How did the back-to-back loans of 2020 and 2021 settle the compensation dispute?

The settlement changed the borrower. In October 2020, after the forty-third Council meeting ended without consensus, the Union finance ministry announced that the Union government would borrow under the first option’s architecture and pass the proceeds to the states as back-to-back loans, booked outside the states’ debt and serviced from the compensation cess extended to March 2026. The Union borrowed and released Rs 1.1 lakh crore in 2020-21 and Rs 1.59 lakh crore in 2021-22, front-loading the second year’s releases at the states’ request, Rs 2.69 lakh crore in aggregate. Jharkhand, the last holdout among the states, accepted the arrangement in December 2020. The settlement conceded the substance of the states’ demand, Union borrowing at the sovereign rate, while preserving the Union’s legal position that the Compensation Act created no open-ended charge on the Consolidated Fund, and it was reached without a Council vote.

Q: How much compensation did the states receive across the guarantee period?

The Finance Ministry’s accounting of 27 April 2022 states that Rs 7.35 lakh crore of compensation was released to the states in aggregate by that date, including the back-to-back loans of Rs 2.69 lakh crore raised in 2020-21 and 2021-22, and a further release of Rs 86,912 crore on 31 May 2022 cleared the outstanding dues for 2021-22 in full. The early years show why the total grew so large. Releases were Rs 41,146 crore in 2017-18 and Rs 69,275 crore in 2018-19, within cess collections of Rs 62,612 crore and Rs 95,081 crore respectively, then jumped to Rs 1,65,302 crore in 2019-20 against cess collections of Rs 95,444 crore as the protected fourteen percent line outran a slowing economy, before the pandemic years required the borrowing programme. The guarantee expired on 30 June 2022, with the cess continuing to March 2026 to service the loans.

Q: What did M. Govinda Rao argue about GST and the fiscal autonomy of the states?

In Studies in Indian Public Finance (2022), M. Govinda Rao argues that the GST settlement transferred fiscal autonomy from the states to a collective institution in which the Union holds the decisive position. The states surrendered rate-setting power over the most buoyant part of their own tax revenue, accepting uniform rates decided in the GST Council, where the Union’s one-third weightage in the voting architecture functions as a standing veto because no proposal can reach the required three-fourths against Union opposition. Rao measures the consequence through the compensation formula: the assured fourteen percent growth path dulled the states’ incentive to administer the tax and left collections growing well below the protected line in most years, so states emerged from the guarantee period in June 2022 dependent on a settlement whose transfers, first cess and then Union borrowing, quantified their lost autonomy. His reading concedes the Council’s consensual practice while denying that consent restores the surrendered power.

Q: What did the Fifteenth Finance Commission decide on tax devolution during the GST compensation period?

The Fifteenth Finance Commission, chaired by N. K. Singh, recommended in its report for 2021-22 to 2025-26, tabled on 1 February 2021, that the states’ share in the divisible pool of central taxes be set at 41 percent, one point below the 42 percent awarded by the Fourteenth Finance Commission, a reduction the Commission attributed to the reorganisation of Jammu and Kashmir into union territories funded directly by the Union. The report also recorded a structural concern that bears directly on GST compensation: the rising share of cesses and surcharges in the Union’s gross tax revenue, levies that sit outside the divisible pool and are not shared with the states. The compensation cess is such a levy, and its extension to March 2026 to service the back-to-back loans means a growing part of the federal fiscal flow runs through an earmarked channel the Union administers rather than through constitutional devolution.

Q: What was the National Anti-profiteering Authority and why was it created?

The National Anti-profiteering Authority was the enforcement body created under Section 171 of the Central Goods and Services Tax Act to ensure that businesses passed the benefit of GST rate reductions and input tax credits to consumers through lower prices, rather than retaining the gain as margin. Constituted in November 2017, months after the launch, it worked through a screening committee and the Directorate General of Anti-profiteering, which investigated complaints and reported to the Authority for orders. The Authority answered a political need of the launch period: a tax reform sold as reducing prices required a visible guard against profiteering while rates were being cut in rounds, and state governments in the GST Council pressed complaints from their own markets. Its tenure, initially two years, was extended in stages to 30 November 2022, after which its functions passed to the Competition Commission of India.

Q: How are exports and imports treated under GST?

Exports are zero-rated under GST, and imports are taxed as inter-state supplies. Section 16 of the Integrated Goods and Services Tax Act treats exports of goods and services, and supplies to special economic zone units and developers, as zero-rated supplies: the exporter charges no tax on the outward supply and claims refund of the input tax credit accumulated on inputs, or exports under bond without payment and claims refund of the credit, so that no Indian indirect tax is embedded in the price of goods sold abroad. Imports run in the opposite direction. Goods entering India attract basic customs duty plus the integrated tax at the rate of the corresponding domestic supply, collected at the border, and the importer takes credit of the integrated tax against output liability, which preserves the destination principle: tax accrues where consumption happens, and trade across the border neither escapes the tax nor bears it twice.