The 1991 economic reforms are usually narrated as a story about ideas: a crisis, a conversion at the top of government, and a programme announced to a surprised country. Read as law, the episode is a story about instruments. Almost nothing that happened in July 1991 happened by Act of Parliament. The centrepiece of the reforms, the abolition of industrial licensing, was a statement laid before the two Houses and a set of notifications issued under a power that the licensing statute itself had contained since 1951. The opening to foreign equity was a change in the way an existing permission regime was administered. The statutes came later, in a sequence that stretched across twelve years, and some of them never came at all. A legal account of the reforms therefore has to begin with a distinction that economic accounts can afford to blur: the distinction between what a government decided, what a government was legally able to do on its own authority, and what only Parliament could do.
The sources for such an account are not the sources of political history, and the difference explains why the legal story is so often mistold. A political history can rest on memoirs, speeches and manifestos. A legal history must rest on instruments: the statute book as amended, the Gazette notifications that exempted and dereserved, the circulars of the Reserve Bank, the office memoranda that created and killed the investment board, the letters of intent addressed to the Fund. Read from those sources, the period looks different in a specific way. The celebrated texts, the statement and the budget speech, shrink to their true legal size, which is that of announcements; and the unglamorous texts, the exemption notification and the repeal ordinance, grow to theirs, which is that of the operative law. Anyone who wishes to know what the law of India said about setting up a factory in 1990, in 1992 and in 2000 will get three different answers, and none of them is recoverable from the speeches of any of those years.
That distinction explains the shape of Indian liberalisation better than any account based on intention. The reforms that moved fastest were the ones for which the old control statutes already contained the necessary power, so that a signature on a notification achieved what a Bill would have taken sessions to achieve. The reforms that required repealing or replacing a statute took between one and eighteen years. The reforms that would have required wholly new legislation in fields where the political cost was highest, labour, land and agriculture, were not attempted in 1991 and took between twenty-two and twenty-nine years to arrive in statutory form. The sequence was not a ranking of importance. It was a map of where executive power happened to be sufficient and where it ran out.
This article reconstructs that map. It states the crisis in figures attributed to the records that published them, sets out what was announced on 24 July 1991 and in what legal form, separates the executive strand of the reforms from the legislative strand instrument by instrument, records the intervals between decision and statute, identifies what was left unreformed and for how long, and examines the judicial doctrine within which the executive route operated. The industrial cluster of this series owns the mechanism by which each individual control was dismantled, in the licence raj to liberalisation laws article; the territory here is the moment itself, the choice of instrument at each step, and what that choice determined.
The 1991 economic reforms begin: the crisis as a legal problem
A balance of payments crisis is an economic event, but the form in which it pressed on the Government of India in the first half of 1991 was a set of legal and institutional facts: obligations falling due in foreign currency, reserves held by a statutory central bank running down to a level measured in weeks, and a statutory regime of exchange control that made every external payment a matter of permission. The Economic Survey for 1991-92 recorded the descent of the reserves in figures that have structured every later account. Foreign exchange reserves, which had stood at 3.11 billion dollars at the end of August 1990, fell through the autumn as the Gulf conflict raised the price of oil and interrupted remittances from Indian workers in the Gulf, and the foreign currency assets of the Reserve Bank of India stood at 896 million dollars on 16 January 1991. The Centre’s fiscal deficit, on the same Survey’s figures, had reached 8.4 per cent of gross domestic product in 1990-91, and wholesale price inflation was running in double digits. Credit rating agencies downgraded Indian paper, commercial borrowing dried up, and deposits from non-resident Indians, a prop of the external account through the previous decade, began to flow out rather than in.
By June 1991 the reserve position had deteriorated past the point at which the language of weeks was a courtesy. The Reserve Bank’s published history of the episode records reserves at about 1.2 billion dollars in January 1991, depleted by about half by June, and describes the June position as barely enough for roughly three weeks of essential imports. A state that cannot pay for three weeks of imports and cannot borrow commercially has a short list of moves, and every one of them is an act of public authority rather than of market choice: it can compress imports by administrative order, it can seek official credit, it can pledge or sell assets, and it can change the price of its currency. India did all four within a single quarter, and each is worth stating precisely because each was done without legislation.
Before any of that, the government of the day had already resorted to the oldest instrument in the control armoury, administrative compression of imports. Through the second half of 1990 the Reserve Bank raised cash margin requirements on imports, tightened the licensing of import items, and surcharged the cost of import finance, while the government raised petroleum prices by a surcharge in October 1990 and took further fiscal measures in December of that year aimed at bringing the deficit down. These were the reflexes of a system that managed scarcity by order, and they bought weeks rather than a solution, because compression could reduce the import bill only by reducing the activity that imports fed. Their significance for this article is that they mark the boundary between the emergency and the reform. Everything done between the autumn of 1990 and June 1991 was done to survive the quarter. Everything announced from 24 July 1991 onwards was presented, and is best understood, as an attempt to ensure that the quarter would not recur, by changing the structure that had made the reserves hostage to confidence in the first place.
The recourse to the International Monetary Fund came first and in two stages. In January 1991 the Fund approved a stand-by arrangement for India of 551.9 million special drawing rights, and India drew 716.9 million special drawing rights under the compensatory and contingency financing facility, the Fund’s window for export shortfalls and excess import costs of the kind the Gulf shock had produced. A larger arrangement followed once a new government was in place after the general election of May and June 1991: a stand-by arrangement of 1,656 million special drawing rights, approved on 31 October 1991, supported by a letter of intent and a memorandum of economic policies in which the government set out its programme of stabilisation and structural reform. The legal character of these arrangements matters for everything that follows. A stand-by arrangement is not a treaty that amends domestic law, and its conditions do not enact anything. It is a financing decision of the Fund taken in reliance on the member’s stated policies. Whatever the programme was to contain, its content had to be carried into Indian law, or out of it, by Indian instruments.
The pledging of gold was the most politically searing of the emergency measures, and it too was executive action under existing statutory powers over the reserves. In May 1991 the Reserve Bank shipped twenty tonnes of gold to the Union Bank of Switzerland, against which 200 million dollars …[truncated 1218 chars] The crisis had been a decade in the making, and the legal instruments of the response were shaped by the manner of its making. Through the nineteen eighties the Centre had run large and growing fiscal deficits, financed in part by borrowing abroad on commercial terms and in part by the captive market for government paper that the banking statutes and the Reserve Bank’s directions created at home. External debt climbed steadily, and its composition worsened as concessional borrowing gave way to commercial loans and to deposits from non-resident Indians that could be withdrawn at short notice. The economy grew respectably for much of the decade, but the growth was financed in a way that left the external account dependent on the continued confidence of creditors who had no obligation to remain confident. The trigger, when it came, was external: the invasion of Kuwait in August 1990 raised the price of oil sharply, increased the import bill, and cut the remittances of Indian workers evacuated from the Gulf, while the political instability of the period, with two governments falling in quick succession and a general election fought in the middle of the emergency, deprived the state of the authority to take hard decisions at the moment they were first needed. An interim budget in the spring of 1991 postponed the correction to the far side of the election. The postponement was not merely political convenience; it followed the constitutional convention that a government holding office through an election does not commit the country to a full year’s financial programme, and confines itself to seeking a vote on account for the months until its successor can present a budget of its own. The consequence was a compression of the entire reform calendar into the weeks after the new government was sworn in on 21 June 1991: a budget within five weeks, an industrial policy statement on the same day, Fund negotiations running in parallel, and the reserves falling throughout. By the time a government with a working programme took office in June 1991, the margin for sequencing had disappeared, which is one reason the programme had to be carried by the fastest instruments available. A government formed in quieter circumstances might have led with a Bill and accepted the delay. The arithmetic of the reserves, stated above in the Survey’s and the Reserve Bank’s figures, forbade it.
The emergency measures had a coda that completes the factual record and is often omitted from accounts that stop at the humiliation of the pledge. The loans raised against the gold were repaid as the reserves recovered, and the pledged gold was redeemed; within three years of the June 1991 trough the reserves had been rebuilt from weeks of import cover to months, on the strength of the returning confidence of the same classes of creditor, commercial lenders, depositors and investors, whose withdrawal had precipitated the crisis. The speed of the recovery became an exhibit in the later assessment debate: defenders read it as proof that the programme’s credibility had been established by the measures of July 1991, critics as proof that the emergency had been shorter and shallower than the programme’s scale implied. The legal record is neutral between those readings, but it fixes the sequence on which both depend: pledge and borrowing in the first half of 1991, devaluation and programme in July, Fund arrangement in October, and recovery thereafter.
Stabilisation and structure: the two tracks of the programme
The programme set out in the memorandum of economic policies that accompanied the Fund arrangement had two tracks, and they ran on different legal fuel. The first track was stabilisation: the correction of the fiscal deficit, restraint of monetary expansion, and the restoration of the reserves. Its instruments were the budget, with its Finance Act, and the Reserve Bank’s statutory powers over credit and the exchange rate, and it was expected to work within two years. The second track was structural reform: the dismantling of the control regime over industry, trade and investment, and the rebuilding of the financial sector, whose instruments are the subject of this article and whose timetable ran to decades. Confusing the two tracks is the commonest error in accounts of the period, because it leads the reader to expect the structural changes to have been completed at the speed the emergency demanded of the stabilisation measures. In law they never could have been: a Finance Act can change tax rates in a single session, but a new system of industrial entry cannot be enacted by a government that has decided, for reasons of time, not to introduce the Bill.
The machinery of the administered rate explains why its adjustment lay so completely in executive hands. The rupee’s external value was not discovered in a market; it was announced. The Reserve Bank set the rate against a basket of the currencies of India’s trading partners, exporters surrendered their proceeds at the official rate under the exchange control framework, and importers bought exchange at that rate for licensed purposes. A devaluation was therefore not a legislative event or even, in form, a policy announcement: it was two alterations of the announced rate on two mornings, executed by the central bank in consultation with the government, and described in the official language of the time as a downward adjustment. The two-step form, a first adjustment on 1 July 1991 and a second on 3 July, reflected the technique of the peg: the rate was moved by administrative decision to the level the authorities judged sustainable, and the absence of any market in which the decision could be tested in advance made the judgement, and its secrecy until execution, the whole of the operation.
The exchange rate illustrates the stabilisation track’s technique across its whole course. The two adjustments of July 1991 were followed within weeks by trade policy measures that replaced the replenishment licensing of exporters with tradable import entitlements, the Eximscrips, devised to give exporters a premium without a budgetary subsidy. In the budget for 1992-93 the government announced the Liberalised Exchange Rate Management System, which took effect on 1 March 1992 and divided export proceeds between an official rate, at which forty per cent was surrendered for essential imports, and a market rate for the remaining sixty per cent. A year later, in March 1993, the dual rate was unified into a market-determined rate, and in August 1994 India accepted the obligations of Article VIII of the Fund’s Articles of Agreement, undertaking not to restrict payments on current account. Each step was taken by the Reserve Bank and the government under the exchange control framework and the central bank’s own statute. At no point in the movement from an administered rate to current account convertibility did Parliament enact anything, because the legal power to set and to manage the rate had never been Parliament’s in the first place. The contrast with the structural track, where the same government waited eight years for a foreign exchange statute that matched its policy, is the cleanest demonstration in the whole record that the speed of a reform was a function of where the relevant power already sat.
The control regime and the powers it happened to contain
To see why the reforms took the legal forms they did, the reader needs the control regime not as a list of restrictions but as an architecture of powers, because the decisive feature of each control statute turned out to be a different clause: in one, an exemption power; in another, a permission structure; in a third, nothing but the requirement itself, which only Parliament could remove.
Industrial licensing rested on the Industries (Development and Regulation) Act, 1951, the first great statute of the planned economy, which required a licence to establish a new industrial undertaking, to manufacture a new article, or substantially to expand an existing undertaking in any industry listed in its First Schedule. The control was statutory and its breach was an offence. But the same Act contained, in section 29B, a power for the Central Government, where it formed the opinion that applying the Act’s provisions to an undertaking or a class of undertakings would not be in the public interest having regard to matters such as the smallness of the undertaking or the stage of development of the industry, to exempt that undertaking or class by notification in the Official Gazette from all or any of the Act’s provisions. A power drafted for small undertakings and infant industries sat in the statute for forty years as a safety valve. In 1991 it became the hinge on which the central reform turned.
Which provision of the Industries Act allowed delicensing by notification?
Section 29B of the Industries (Development and Regulation) Act, 1951 empowered the Central Government to exempt any industrial undertaking or class of undertakings, or any scheduled industry, from all or any provisions of the Act by notification in the Official Gazette. Delicensing in 1991 was an exercise of that exemption power, not an amendment of the Act.
The control of large business houses rested on the Monopolies and Restrictive Trade Practices Act, 1969, whose Chapter III, sections 20 to 26, subjected undertakings above an asset threshold, set at 100 crore rupees, to prior approval of the Central Government for establishing a new undertaking, expanding, merging, amalgamating or taking over another undertaking, or appointing directors. Unlike licensing, this requirement had no exemption valve of comparable width. The approval requirement was the statute; it could be relaxed administratively at the margin, but it could not be abolished by notification. Its removal therefore waited on Parliament, a fact that determined the two-month gap between the announcement of the change and its legal completion.
Foreign exchange rested on the Foreign Exchange Regulation Act, 1973, a statute of a different temper, which treated foreign exchange as a scarce public resource to be conserved and made a wide range of transactions, including the issue of shares to persons resident outside India, lawful only with the permission of the Reserve Bank. The architecture was prohibitory, but it was prohibitory through permission: the statute did not itself fix the proportion of foreign equity an Indian company might have; it required approval, and approval was an administrative act. A government that controlled the administration of the permissions could therefore change the substance of the regime, raising the ceiling of what would be approved as a matter of course, without touching a line of the Act. That is why the foreign exchange statute, the most feared of the control laws, was the one replaced last among the principal economic statutes, eight years after the policy it administered had been reversed.
The foreign equity ceiling that the 1991 statement raised had its own statutory root, and it repays attention because it shows the permission architecture at its most concrete. Section 29 of the foreign exchange statute restricted companies with non-resident interest above forty per cent from carrying on business in India, or acquiring shares and immovable property, without the permission of the Reserve Bank, and the administrative practice built on that provision produced the working rule of the nineteen seventies and eighties under which foreign shareholding was ordinarily confined to forty per cent and higher participation was an exception to be justified case by case. The forty per cent figure that the 1991 statement displaced was therefore not a policy preference floating free of the law; it was the ceiling that a statutory threshold had hardened into. Yet because the threshold operated through permission, the reform did not need to move it. The statement’s promise of automatic approval up to 51 per cent in the priority industries was implemented by instructing the permission machinery to grant what the statute allowed it to grant, and the forty per cent anchor in section 29 lost its practical force without losing its place in the Act. A lawyer reading only the statute book in 1992 would have found the old regime undisturbed. A businessman reading the approval letters would have found it transformed. The gap between those two readings is precisely the gap this article exists to measure.
It is worth pausing on what the exemption and permission powers had in common, because it is the deepest legal fact in the story. Both had been conferred for the purposes of the control regime itself. Section 29B of the Industries Act existed so that the licensing system could bend where it pinched, for small undertakings and young industries; the Reserve Bank’s permission power existed so that exchange control could admit the deserving case. Neither power had been conferred in order to dismantle the regime it served. Their use in 1991 was lawful, as the courts later confirmed of the exemption power, because the texts were wide enough to bear it: an exemption power expressed to be exercisable in the public interest can be exercised for the public interest as the government of the day understands it. But the fact that the great deregulation of Indian industry was accomplished by turning the control regime’s own safety valves against the regime is not a curiosity of drafting. It is the reason the deregulation was fast, the reason it was legally fragile, and the reason the statute book misdescribed the Indian economy for years afterwards to anyone who read it without the Gazette notifications alongside.
Two further controls complete the map. The issue of capital by companies was governed by the Capital Issues (Control) Act, 1947, under which the Controller of Capital Issues consented to issues and fixed their amount, type and price; the control was statutory, consent-based and administered by an office of government, and ending it meant repealing the Act, though replacing its function with a disclosure-based regulator was a separate and larger legislative task. Imports and exports were governed by the Imports and Exports (Control) Act, 1947, which empowered the Central Government to prohibit, restrict or control trade by order; the actual texture of import licensing lived in the export and import policy and the orders made under the Act, so trade liberalisation could proceed administratively under the old statute while its replacement was drafted. The pattern across the five controls is the article’s spine. Where the statute conferred a power to exempt, to permit or to order, the executive already held the instrument of reform. Where the statute imposed the requirement itself, reform needed a Bill.
24 July 1991: the statement and the budget
The reform programme was announced on a single day in two instruments of entirely different legal weight, and the pairing is itself instructive. The Union Budget for 1991-92 was presented to the Lok Sabha on 24 July 1991. A budget is a legislative event: its taxation proposals take effect through a Finance Bill, and its expenditure through the appropriation process. The statement on industrial policy was laid before Parliament on the same day. A statement of policy is not legislation, is not voted into law, and creates no rights or offences. It announces what the executive intends to do with the powers it already has and what it intends to ask Parliament to change. The programme’s fastest elements travelled in the instrument that required no vote, and its slowest were those that had to wait for the instrument that did.
Was the industrial policy statement of 1991 a law?
No. The statement on industrial policy of 24 July 1991 was a policy announcement laid before Parliament. It was not enacted, not voted on as legislation, and it amended no statute. Its measures took legal effect through notifications and administrative decisions under existing statutes, and through the amendments Parliament passed afterwards.
The statement’s industrial announcements were radical in substance and conservative in form. Industrial licensing was abolished for all projects irrespective of the level of investment, except for a short list of eighteen industries retained on grounds the statement grouped under security, strategic and social considerations. The reservation of industries for the public sector, a feature of industrial policy since the resolution of 1956, was cut from seventeen industries to eight. The requirement of prior approval for large undertakings under the monopolies statute was to go, the statement recording that the provisions on concentration of economic power would be amended. Foreign investment was invited on new terms: automatic approval would be given to foreign equity up to 51 per cent in a list of high priority industries, a rise from the 40 per cent ceiling that had been the working rule under the foreign exchange regime, and proposals beyond the automatic route would be considered by a specially constituted board. The phased manufacturing programmes, under which firms had been required to indigenise production on an official timetable, were abolished. Foreign technology agreements in the priority industries were placed under automatic approval within stated parameters. Each of these announcements can be assigned to its instrument, and the assignment is the analysis: licensing and the phased programmes fell to notification and administrative direction; the foreign equity ceiling moved because the permission practice moved; the monopolies provisions alone required the amendment the statement promised.
The budget carried the programme’s fiscal and trade elements and announced the direction of the rest. It proposed the correction of the fiscal deficit that the Fund arrangement required, setting a target of reducing the deficit to 6.5 per cent of gross domestic product in 1991-92 from the 8.4 per cent recorded for the year before, moved on the structure of customs and excise duties, raised administered prices in fields such as fertiliser where subsidy had absorbed the shock of costs, and set out the intention to reform the financial sector, to restructure public enterprises and to recast trade policy. Some of the administered price increases were partly rolled back within weeks under parliamentary and public pressure, an early sign that the executive strand, for all its speed, operated inside a political legislature whose assent the government needed for the measures that only Parliament could carry, beginning with the Finance Bill that gave the budget its legal force. The relationship between the two instruments of 24 July was therefore one of mutual dependence rather than of rivalry: the statement’s freedom was possible because the budget was going to Parliament anyway, and the budget’s reception was eased by a statement that showed the correction to be part of a design rather than an austerity imposed from outside. The trade measures of the months that followed, the movement of most capital goods and intermediates from licensing to open general licence and the pruning of the negative list of imports, were effected through the export and import policy and orders under the trade control statute rather than through the budget itself, which is further evidence of the same distribution of labour: Parliament was asked for money and for tax law, while the control regime was dismantled in the first instance by the executive acting on powers the control statutes had always contained.
The financial sector: reform by committee and direction
Banking deserves a section of its own because it displays a fourth technique, distinct from notification, permission and repeal: reform by expert committee followed by regulatory direction. The government appointed the Committee on the Financial System, chaired by a former Governor of the Reserve Bank, M. Narasimham, on 14 August 1991, within weeks of the policy statement. The committee reported in November 1991, and its report was laid before Parliament on 17 December 1991. Its central recommendations were that the statutory pre-emptions on bank funds be cut, the statutory liquidity ratio from 38.5 per cent towards 25 per cent and the cash reserve ratio from fifteen per cent towards a range of three to five per cent; that interest rates be deregulated in stages; that banks adopt prudential norms for income recognition, asset classification and provisioning on international lines; that dual control of banking by the ministry and the Reserve Bank be ended in the Reserve Bank’s favour; and that new private banks be admitted.
Almost none of this required an Act, and almost none was given one. The liquidity and reserve ratios are set under the Reserve Bank’s statute and the banking legislation by direction of the central bank; the prudential norms were imposed by the Reserve Bank’s circulars from the financial year 1992-93; interest rate deregulation proceeded by the progressive freeing of rates the Reserve Bank announced; the licensing of new private banks was done by guidelines the Reserve Bank issued in January 1993 under its existing licensing power over banking companies. The public sector banks themselves owed their existence to the acquisition Acts of 1970 and 1980, and those Acts were left in place; what changed was the environment of directions within which the nationalised banks operated, and their access to the capital market, which was opened by permitting them to issue shares to the public within limits the government fixed. The one banking measure of the period that plainly needed a statute received one, and it is the exception that identifies the rule: the recovery of bank debts through dedicated tribunals required the Recovery of Debts Due to Banks and Financial Institutions Act, 1993, Act 51 of 1993, because tribunals displacing the jurisdiction of the civil courts cannot be conjured by circular. Where the reform changed the behaviour of an existing regulator’s supervisees, direction sufficed. Where it created a new forum with coercive jurisdiction, Parliament had to act. The financial sector therefore confirms the article’s rule from a second direction, and adds the committee as the characteristic institution of the era: a body with no legal power at all, appointed by resolution, whose report supplied the programme that the holders of legal power then executed.
The executive strand: reform under powers already conferred
Delicensing came first and fastest, and its legal biography is short enough to state in a sentence and important enough to dwell on. The Industries Act remained on the statute book with its licensing provisions intact. What changed was the issue of exemption notifications under section 29B which withdrew the licensing requirement from industry after industry, giving effect to the statement’s rule that a licence would survive only for the eighteen listed industries. A manufacturer who had needed permission to establish a plant, to make a new article or to expand capacity needed, after the notifications, no permission at all outside the retained list, not because Parliament had repealed the requirement but because the government had exercised a discretion the requirement’s own statute conferred. The distinction sounds pedantic and is not. A repeal is a legislative act that a later executive cannot undo; an exemption is an executive act that a later executive can withdraw by the same route, and section 29B itself contemplates the cancellation of an exemption notification, with a grace period for undertakings that would then need licences. The legal position after July 1991 was therefore that industrial India had been delicensed at the pleasure of the Central Government. That no government used the route back is a fact of politics. That the route back existed is a fact of law, and it is the fact on which the closing analysis of this article turns.
Delicensing generated its own paperwork, and the paperwork is worth describing because it shows how completely the legal character of industrial entry was inverted. Under the licensing system the entrepreneur’s founding document was the licence application, a request for a discretionary grant, supported by particulars of capacity, location, capital goods imports and foreign collaboration, and decided after inter-ministerial consideration. Under the exemption regime the founding document became the Industrial Entrepreneurs’ Memorandum, filed with the Secretariat for Industrial Assistance in the industry ministry, on which the entrepreneur simply recorded the particulars of the project. The memorandum required no approval and received an acknowledgement rather than a grant; it existed so that the state would know what was being built, not so that the state could decide whether it should be. The distinction between a licence and a memorandum is the distinction between permission and information, and it is one of the cleanest measures of what changed in 1991. The state did not withdraw from knowledge of industrial investment; it withdrew from judgement upon it. For the industries remaining on the retained list, the licence application survived in its old form, and for the small scale sector the registration habits of the district industries centres continued, so that the paperwork of the Indian factory in the nineteen nineties was a palimpsest: memorandum for the delicensed, licence for the reserved, registration for the small. Each layer corresponded to a stratum of the control regime that the reforms had reached, and a project’s file, read carefully, recorded exactly how far liberalisation had travelled in its sector.
The durability of the exemption route was tested in a later decade in a different sectoral setting, when the government delicensed the sugar industry by notification and a disappointed incumbent challenged the exercise. In Bajaj Hindustan Limited v Sir Shadi Lal Enterprises Limited, decided in 2010, the Supreme Court upheld the Central Government’s authority to delicense by notification under the Industries Act and reiterated that economic policy choices of that character are for the executive and the legislature, with the court’s review confined to legality. The case concerned sugar rather than the 1991 list, but its premise is the premise of 1991: the exemption power is a real and continuing power of government, exercisable in both directions.
Foreign equity moved by the second executive instrument, the administered permission. Because the foreign exchange statute made the issue of shares to non-residents a matter of Reserve Bank approval, a policy decision that approvals within stated limits would be automatic changed the law in action while changing nothing in the statute book. The Reserve Bank’s automatic route for equity up to 51 per cent in the specified industries, and the general permission framework that accompanied it, were administrative constructions on an unamended Act. For proposals outside the automatic route, the government constituted the Foreign Investment Promotion Board in 1991, first located in the Prime Minister’s Office, to consider and recommend on foreign investment proposals. The Board is the purest specimen of the executive strand: a body created by executive decision, never given statutory form, moved between departments as administrative convenience dictated, transferred to the Department of Industrial Policy and Promotion in 1996 and to the Department of Economic Affairs in 2003, and abolished by executive decision in 2017, when a Cabinet decision of 24 May 2017 and an office memorandum of 5 June 2017 wound it up and distributed its work among the administrative ministries. A quarter century of the country’s principal investment-screening institution lived and died without Parliament ever being asked to create or to end it.
Why could foreign equity be opened without amending the foreign exchange statute?
The Foreign Exchange Regulation Act, 1973 barred foreign exchange transactions unless the Reserve Bank permitted them, and did not itself fix the level of foreign equity allowed. Because the ceiling operated through approval, granting it automatically up to 51 per cent in priority industries changed the regime in practice while the Act stood unamended until 1993 and unreplaced until 1999.
Foreign investment policy acquired, in the years after 1991, an instrument of its own that deserves naming, because it became the characteristic legal form of the field: the press note. The industry ministry issued the policy on foreign direct investment through a series of press notes, each amending the consolidated policy on sectoral caps, prohibited sectors and the boundary between the automatic and the approval routes, and the Reserve Bank gave the changes legal effect through the permissions and notifications issued under the foreign exchange statute. A sectoral cap could thus be raised, a sector opened, or a condition attached by a document that was neither a statute nor a rule laid before Parliament, but a policy announcement operationalised through a permission regime that the statute had always made available. The arrangement persisted after the statute changed in 1999, because the management Act preserved the architecture of prohibition qualified by permission on which the technique depends. For the student of instruments, the press note is the purest product of the reform era: policy made at the speed of publication, reversible at the same speed, and housed in a statutory shell that successive Parliaments have left undisturbed while the policy inside it has been rewritten many times over. It is the exemption notification’s descendant, adapted to a field where the operative law had always been a practice of approvals.
Trade liberalisation travelled by the third executive instrument, the statutory order. The Imports and Exports (Control) Act, 1947 conferred a broad power to prohibit, restrict or otherwise control imports and exports by order, and the export and import policy announced under it carried the lists of restricted items. Successive policies through the first reform years shrank those lists, moved goods to open general licence, and recast the regime from one in which imports were forbidden unless licensed to one in which they were free unless restricted. The 1947 Act, a wartime control continued into peacetime, was capacious enough to contain its own liberalisation, because what it conferred was not a list but a discretion. The licensing of telecommunications tells the same story in a service sector: entry of private operators proceeded by licence granted by the Department of Telecommunications under the Indian Telegraph Act, 1885, an arrangement that allowed a market to be created in the mid-1990s under a colonial statute, years before the sector acquired either a modern policy statute or, in 1997, a regulator.
The same administrative machinery carried the smaller print of the programme. The phased manufacturing programmes had obliged firms with foreign collaboration to indigenise components on an official schedule as a condition of their approvals; abolishing them required no more than ceasing to impose the condition, since the programmes had always been a creature of approval practice rather than of any Act. Foreign technology agreements moved to automatic approval within ceilings on royalty and lump sum payments that the Reserve Bank administered. The office that ran import and export licensing, headed for four decades by the Chief Controller of Imports and Exports, was reconstituted in the reform period as the Directorate General of Foreign Trade, a change of name and mission made by executive order under the same statutory scheme that the 1992 trade Act then confirmed. In each case the pattern holds: the reform was real, it was documented in notifications, circulars and office orders, and a reader of the statute book alone would have missed most of it. For a researcher reconstructing the period, the working papers of the reform are the Gazette notifications and the Reserve Bank’s circulars, collected and organised by instrument rather than by announcement; a reader assembling those papers for study will find a document viewer suited to the task in the office file viewer for spreadsheet, document and presentation files, which opens the formats in which such records circulate.
Public sector reservation, the fourth item on the statement’s list, illustrates the boundary of the executive strand from the other side. Reservation of industries for the public sector had never been a creature of statute. It was a schedule of the industrial policy resolution of 1956 and its successors, an announcement of the fields in which the state intended to operate exclusively or principally. Reducing the reserved list from seventeen industries to eight therefore required nothing but a new announcement, and the statement made it. What executive action could not do was change the ownership of the enterprises themselves or dissolve the statutory corporations, some of which owed their existence to individual Acts; the programme of disinvestment that followed, the sale of bundles of government shareholding in 1991-92 and of larger stakes thereafter, was accordingly conducted as a series of executive decisions about the management of the state’s assets, advised on by bodies themselves created by resolution, such as the Disinvestment Commission constituted in 1996. No privatisation statute was introduced at any point in the reform era, a silence examined below among the things left undone.
The trade machinery in operation
Trade control deserves a closer look than the other executive fields, because it shows the permission technique operating at the greatest granularity and because the paperwork of trade licensing was, for most Indian businesses, the control regime they actually met. Under the Act of 1947 the government issued periodic export and import policies, and beneath the policies sat the lists: items importable freely under open general licence, items importable against a licence issued on application, items whose import was canalised through a state trading corporation so that the individual importer bought from the state rather than from abroad, and items prohibited altogether. An importer of a restricted item applied for a licence whose grant turned on assessments of essentiality and indigenous availability, the doctrine that an import would not be licensed if an Indian producer could supply the article. The system’s texture was therefore discretionary at every joint, and its reform could proceed joint by joint: an item moved from the restricted list to open general licence by an amendment of the policy, and each such movement was a small act of liberalisation complete in itself. Through the reform years the lists were pruned in successive policies until the architecture inverted, and the five-year export and import policy announced on 31 March 1992 for the period from 1992 to 1997 codified the inversion: imports were free except for a defined negative list, and licensing survived as the exception attaching to listed items rather than as the rule attaching to trade itself. The statute under which all of this occurred, for the first year of it, was the wartime-born Act of 1947, whose broad power to control by order proved as serviceable for dismantling control as it had been for imposing it. When the Foreign Trade (Development and Regulation) Act arrived in 1992, it regularised a transformation that administrative orders had already accomplished, and its most consequential continuity was administrative: the licensing organisation continued, rebadged, administering a regime whose default had been reversed. The lesson generalises. In field after field, the statute of the reform era arrived to find the reform already in force, and its work was to stabilise, to regularise and to house the new arrangements in institutions that could outlast the circulars that had improvised them.
The ordinance bridge
Between the speed of the notification and the speed of the Bill, the Constitution supplies a bridge, and the reform government used it repeatedly in the programme’s first legislative season. The President’s power to promulgate an ordinance when Parliament is not in session produces an instrument with the force of an Act and a life limited to the session that follows. The securities regulator’s statute was deemed in force from 30 January 1992, the date of its ordinance; the repeal of capital issues control was first effected by an ordinance of 29 May 1992 which the Repeal Act then replaced; and the foreign trade statute took effect from 19 June 1992, the date of the ordinance that had first carried it. The pattern is not accidental, and it refines the article’s taxonomy. Where a reform needed legislation but could not wait for the parliamentary calendar, the executive legislated provisionally and Parliament ratified: the ordinance converted a second-speed measure into something close to a first-speed one, at the price of a ratification that still had to be obtained, since an ordinance unratified lapses. The reform programme’s legislative history is accordingly written twice over, once in ordinances and once in Acts, and the dual record is itself evidence of the pressure under which the sequence was assembled. A government confident of its parliamentary position and unconstrained by time legislates once. The government of 1991-92, short of both, legislated in instalments, using every instrument the Constitution and the control statutes put within its reach, in ascending order of permanence as the need for permanence grew.
The ordinance cluster of 1992 also corrects a lazy assumption about the relationship between the executive strand and the legislative strand, which is that the first was chosen against the second out of impatience with democracy. The ordinances were issued in the winter and summer of 1992, months after the emergency phase had closed, and each was in due course laid before Parliament and replaced by an Act passed through both Houses with debate. The securities ordinance of 30 January 1992 became an Act assented to on 4 April 1992; the capital issues ordinance of 29 May 1992 was replaced by the Repeal Act within the same year; the trade ordinance of 19 June 1992 was folded into an Act assented to on 7 August 1992. The intervals between ordinance and Act, between nine weeks and a single session, show a government using the bridge for its proper purpose, which is to legislate in a recess and submit the legislation for ratification at the next sitting, rather than to govern without Parliament. That the ratifications were not in doubt owed something to the same political fact that had made the direct route attractive: by 1992 the programme had acquired the acquiescence of a Parliament that had watched the alternative in the summer of 1991. The ordinance record is therefore best read as the hinge of the whole sequence. It is where the programme built by notification was progressively converted into enacted law, at the first points where conversion was unavoidable, and it marks the moment when the reforms stopped being merely the acts of a government and began to be the law of the legislature that outlasted it.
The legislative strand: what only Parliament could do
Against the speed of the executive strand, the legislative record of the reforms reads as a calendar of delays that are individually explicable and collectively the point. Each statute that had to change fell into one of three classes: requirements to be deleted, controls to be repealed, and authorities to be created. Deletion was quickest, repeal almost as quick, and creation slowest, because a new authority requires a designed institution, a parliamentary debate about its powers, and usually a committee report to prepare the ground.
The first control statute Parliament amended for the reforms was the monopolies statute. The Monopolies and Restrictive Trade Practices (Amendment) Act, 1991, Act 58 of 1991, omitted sections 20 to 26 of the principal Act, the whole of the chapter on concentration of economic power, with effect from 27 September 1991, and removed with them the apparatus of prior approval for expansion, new undertakings, mergers, amalgamations, takeovers and appointments of directors by large undertakings. The interval between the policy decision of 24 July and the legal completion of 27 September was about two months, the fastest legislative response in the entire programme, and it is no accident that it was also the simplest: a deletion, requiring no new institution and creating no regulator. The remainder of the monopolies Act, its provisions on monopolistic, restrictive and unfair trade practices and its Commission, survived the deletion by eighteen years, a shell whose history the industrial cluster traces. The amending Act did one piece of consequential carpentry that is easy to overlook: the provisions of the monopolies statute that had restricted the acquisition and transfer of shares, housed in a chapter inserted in the nineteen eighties, were transferred by the same amendment into the Companies Act, 1956, where regulation of substantial shareholding was thought to belong. The detail matters to the taxonomy because it shows a deletion that was not a deregulation but a relocation, carried out in the same instrument and at the same speed. Parliament in September 1991 was capable of fine work at speed when the work was subtractive or transferential. What it could not do at speed, then or later, was design the institutions that the programme’s second phase required, and the difference between the carpentry of 1991 and the architecture of 1997 to 2003 is the difference between the two halves of the legislative strand.
What was the first control statute Parliament amended for the reforms?
The Monopolies and Restrictive Trade Practices Act, 1969. The Amendment Act of 1991, Act 58 of 1991, omitted sections 20 to 26, the chapter on concentration of economic power that required large undertakings to obtain prior government approval, with effect from 27 September 1991, about two months after the policy statement announced the change.
Capital issues supplied the second lesson, that repeal and replacement are different operations with different speeds. The control exercised by the Controller of Capital Issues could be ended quickly, and it was: the Capital Issues (Control) Act, 1947, Act 29 of 1947, was repealed by the Capital Issues (Control) Repeal Act, 1992, Act 26 of 1992, deemed to have come into force on 29 May 1992, the date of the ordinance that first did the work. But the repeal was only half of the reform. The function of protecting investors in a market where the state no longer fixed prices had to be re-housed in a regulator, and the Securities and Exchange Board of India had existed since 12 April 1988 as a body constituted by a resolution of the Government of India, able to advise and to persuade but not to compel. The Securities and Exchange Board of India Act, 1992, Act 15 of 1992, converted it into a statutory authority, deemed to have come into force on 30 January 1992, the date of the ordinance that preceded the Act. The pairing of the two 1992 statutes is the clearest instance of the article’s rule: the executive could announce free pricing of capital issues, but a regulator with power to register intermediaries, to call for information, to investigate and to penalise could be created only by Parliament, because a body exercising coercive power over citizens needs a statute to stand on. What replaced the Controller’s consent is itself instructive. The old regime had been one of merit regulation: the state decided whether an issue deserved to reach the market and at what price. The new regime, built by the statutory Board through its disclosure and investor protection guidelines issued in 1992, was one of disclosure regulation: the issuer who made the prescribed disclosures could approach the market at a price the market would test, and the state’s role shifted from approving the issue to policing the truth of what was said about it. The guidelines through which that philosophy operated were themselves subordinate instruments, framed under the new Act, which completes the symmetry of the episode. Parliament was needed to kill the old control and to create the new authority; the authority’s working law was then made by the authority, at the first speed, in the manner the era had normalised. The capital markets strand of the series, in the SEBI Act 1992 explained article, owns the regulator’s powers in detail.
Which reform statutes began as ordinances?
The securities regulator’s statute and the capital issues repeal both began as ordinances. The Securities and Exchange Board of India Act, 1992 was deemed to have come into force on 30 January 1992, when its ordinance was promulgated, and the repeal of capital issues control was first effected by an ordinance of 29 May 1992 before the Repeal Act followed.
Trade received its replacement statute in the same season. The Foreign Trade (Development and Regulation) Act, 1992, Act 22 of 1992, which received assent on 7 August 1992 and was deemed to have come into force on 19 June 1992, repealed the Imports and Exports (Control) Act, 1947 and changed the stated object of the field from control to the development and regulation of foreign trade by facilitating imports and augmenting exports. The change of title records a change of theory: the 1992 Act retained ample powers to prohibit and restrict, and the policy made under it kept a negative list, but its premise was that trade was to be facilitated and its restriction justified item by item, rather than forbidden in general and licensed in particular.
Foreign exchange took longer, and the interval is the article’s standard exhibit for the long statutory tail. The permission architecture of the Foreign Exchange Regulation Act, 1973 had allowed the substance of policy to be reversed administratively in 1991, which removed any urgency from replacing the statute, and the Act survived the decade in a softened form, amended in 1993 to strip out some of its harsher features and to align it in part with the new policy, before the Foreign Exchange Management Act, 1999, Act 42 of 1999, replaced regulation with management and converted a regime built on criminal enforcement into one built on civil contravention, with effect from 1 June 2000. The eight-year interval between the opening of foreign equity and the arrival of the statute that matched it was not drift; it was the predictable consequence of the fact that the executive route worked. Why the replacement took the form it did, and what happened to proceedings pending under the old Act, is the subject of the FERA to FEMA transition article, which owns that ground.
The long twilight of the foreign exchange statute is worth a paragraph of its own, because no other instrument shows so clearly what it means for a law to survive its policy. Through the nineteen nineties the Act of 1973 continued to define the offences, the enforcement directorate continued to police it, and prosecutions and adjudications under it proceeded, even as the approval practice it administered was liberalised year by year and the ceiling that had been its most famous feature ceased to describe the economy. The 1993 amendment pruned provisions whose work was done, but the architecture, criminal in character and conservationist in premise, stood. Business therefore lived in two legal worlds at once: a policy world, announced in press notes and approval letters, in which foreign capital was welcome on stated terms, and a statutory world, in which the same capital moved under a statute drafted for scarcity and enforced, at its edges, by penalties designed for smugglers of currency. The dissonance was a standing argument for the replacement statute, and it was pressed by industry throughout the decade; but it was also, from the government’s side, tolerable, because the permission structure let the practice run ahead of the text. The replacement, when it came in 1999, was therefore less a reform than a ratification, and its most debated feature, the conversion of offences into civil contraventions, concerned the accumulated enforcement of the old world rather than the design of the new one. Statutes that survive their policies do not merely become obsolete. They continue to generate rights, liabilities and proceedings of their own, and the cost of the interval is paid in that currency.
The creation of regulators consumed the rest of the decade and ran into the next. The insurance industry, whose public monopoly dated from the nationalisation statutes, was opened only after a committee process, and the Insurance Regulatory and Development Authority Act, 1999, Act 41 of 1999, created the regulator and amended the insurance and nationalisation Acts to admit private and foreign capital. Telecommunications acquired its regulator through the Telecom Regulatory Authority of India Act, 1997, Act 24 of 1997, six years after the reform programme began. Competition policy waited longest among the market institutions: the Competition Act, 2002, Act 12 of 2003, enacted on the report of a high level committee to replace the philosophy of controlling concentration with a law against anti-competitive conduct, reached assent in January 2003, and the monopolies Act it displaced was repealed only with effect from 1 September 2009. Electricity arrived in the same year, the Electricity Act, 2003, Act 36 of 2003, consolidating and replacing the supply statutes of 1910, 1948 and 1998; and the fiscal strand closed with the Fiscal Responsibility and Budget Management Act, 2003, Act 39 of 2003, which wrote deficit targets into statute and came into force on 5 July 2004. Each of these statutes is a reform of 1991 in the only sense that finally matters, the sense of what the law said when the process finished; each arrived between six and twelve years after the decisions that made it necessary.
The intervals were not empty. They were filled with the committee stage that Indian legislation in contested fields characteristically requires, and the committees are part of the explanation of the timetable. Insurance reform waited on the committee chaired by a former Governor of the Reserve Bank, R. N. Malhotra, appointed in 1993 and reporting in 1994, which recommended a regulator and the admission of private insurers; even with the report in hand, the Bill took a further five years, through changes of government and the lapse of an earlier Bill on the dissolution of the House that had been considering it, so that the measure had to be introduced afresh before the Act of 1999 closed the question. The lapse is a detail of parliamentary mechanics with a large moral: a free-standing Bill carries the mortality of the House that receives it, and every dissolution resets the clock of the second speed to zero, which is one reason governments that can act by notification learn to prefer it. Competition reform waited on the high level committee chaired by S. V. S. Raghavan, appointed in 1999 and reporting in 2000, which advised that the monopolies Act be repealed and a modern competition law enacted; the Competition Act followed in 2002. Electricity followed the longest fuse of all. Private generation had been invited as early as October 1991, when the government announced a policy for private participation in generation and amended the Electricity (Supply) Act, 1948 to make room for generating companies, and the first private projects were negotiated as memoranda of understanding with state governments through the nineteen nineties. A regulator for the sector came in two stages, first through the Electricity Regulatory Commissions Act, 1998, Act 14 of 1998, and then through the consolidating Act of 2003. The sector thus passed through policy invitation, statutory amendment, interim regulator and consolidating statute across twelve years, an extreme case of the general pattern in which the executive moved first and the legislature assembled the permanent structure behind it, in stages, over a decade.
Tax reform supplies a third legislative speed that belongs in the taxonomy. The Tax Reforms Committee chaired by Raja Chelliah, appointed in August 1991, reported in stages through 1992 and 1993 and recommended lower and fewer rates of income tax and customs, a broader base, and the eventual destination of a value added tax. Tax law in India is made annually, in the Finance Act that follows each budget, and the Chelliah programme was accordingly implemented across a succession of Finance Acts through the nineteen nineties, rate by rate and schedule by schedule. This was legislation, and only legislation could do it, since tax is imposed by authority of law under the Constitution; but it was legislation of a routine annual kind that no government finds difficult to move, because the Finance Bill must pass each year for the state to function. The contrast orders the whole field. Measures that ride on an annual must-pass Bill move at the speed of the budget. Measures that need a free-standing Bill move at the speed of the parliamentary timetable and the committee process. Measures that need no Bill at all move at the speed of a signature. The reforms of 1991 distributed themselves across all three speeds, and the distribution, more than the intentions of any participant, is what a legal history of the period has to explain.
Act, notification or nothing: the reform table
The thirteen rows below compress the article into one reference. The interval column, which measures the distance between the policy decision of 1991 and the statutory change where there was one, is the argument in numerical form: the rows with no interval are the executive strand, the rows with intervals of years are the legislative strand, and the rows recording no statutory change at all are the reforms that were announced as absences and stayed absent.
| Reform | Instrument used | Date of the instrument | Interval from the 1991 decision to the statutory change |
|---|---|---|---|
| Industrial licensing | Exemption notifications under section 29B of the Industries (Development and Regulation) Act, 1951 | July 1991 onwards | None; the Act was not amended |
| Monopolies clearance | Amendment Act deleting sections 20 to 26 of the Monopolies and Restrictive Trade Practices Act, 1969 | Act 58 of 1991, in force 27 September 1991 | About two months |
| Foreign equity | Automatic approval route administered under the Foreign Exchange Regulation Act, 1973, with the Foreign Investment Promotion Board created by executive decision | 1991 | Eight years to replacement by the Foreign Exchange Management Act, 1999 |
| Capital issues | Repeal of the Capital Issues (Control) Act, 1947 and enactment of the Securities and Exchange Board of India Act, 1992 | Repeal Act 26 of 1992, in force 29 May 1992; Act 15 of 1992, in force from 30 January 1992 | About ten months |
| Imports and exports | Liberalisation by export and import policy under the 1947 Act, then the Foreign Trade (Development and Regulation) Act, 1992 | Act 22 of 1992, in force from 19 June 1992 | About eleven months |
| Foreign exchange | Foreign Exchange Management Act, 1999 replacing the Act of 1973 | Act 42 of 1999, in force 1 June 2000 | Eight years |
| Insurance | Insurance Regulatory and Development Authority Act, 1999 | Act 41 of 1999 | Eight years |
| Telecommunications | Licensing under the Indian Telegraph Act, 1885, then the Telecom Regulatory Authority of India Act, 1997 | Act 24 of 1997 | Six years |
| Electricity | Electricity Act, 2003 replacing the Acts of 1910, 1948 and 1998 | Act 36 of 2003 | Twelve years |
| Competition | Competition Act, 2002, with repeal of the monopolies Act following later | Act 12 of 2003; monopolies Act repealed from 1 September 2009 | Eleven years to the new Act, eighteen to the repeal |
| Labour | Four labour codes consolidating twenty-nine central labour laws | Code on Wages, 2019 and the three codes of 2020 | Twenty-eight to twenty-nine years |
| Land | Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act, 2013 | Act 30 of 2013, in force 1 January 2014 | Twenty-two years |
| Privatisation | Executive decisions on disinvestment, advised by bodies created by resolution | From 1991-92 onwards | None; no framework statute was enacted |
The licence’s afterlife in the states
Delicensing at the Centre did not end the Indian entrepreneur’s acquaintance with permission, and an honest legal account must follow the queue as it moved. The industrial licence had been the most visible gate because it stood at the entrance to manufacture itself, but behind it stood the gates of the states and of the municipalities: land use conversion, building plan approval, factory licensing under the Factories Act, 1948, boiler and pollution consents under the state boards constituted under the water and air Acts, power and water connections from state utilities, registration under the shops and establishments law of the state, and the profession tax and labour registrations that attended the first employee. None of these was touched by the statement of 1991, most of them because the Centre had no power to touch them, resting as they did on state statutes and municipal law. Through the following decades the states competed for investment partly by legislating about their own gates, enacting single window clearance statutes that promised time-bound approvals and deemed clearances on the expiry of the period, a legislative genre that spread across the states in the two decades after the reforms. The results were uneven, and the queue for many projects reassembled at the state counter much as it had stood at the central one, a fact that qualifies any simple claim that 1991 ended the licence raj. What 1991 ended was the licence raj of the Union list: the permission to exist as a manufacturer. The permissions to build, to employ, to draw power and to discharge effluent remained, distributed across the federation, and their reform belonged to the slower speeds of lawmaking in twenty-eight legislatures rather than to any single statement in Delhi. The industrial cluster’s account of the dismantling is the Centre’s account; the states’ account is longer, less complete, and in considerable part still being written in state legislation.
The controls that survived, and the one that was created
A legal history of the reforms that recorded only dismantling would misstate the period in two directions. The first is that substantial controls survived it intact, because they rested on grounds the programme did not disown. Price control under the Essential Commodities Act, 1955 continued to govern scheduled commodities, and the control of drug prices under that framework continued through and beyond the reforms, revised by successive drug price control orders. The licensing of industries on the retained list, for defence, atomic energy and the other specified fields, continued under the Industries Act as a working system rather than a survival. Control of foreign exchange on the capital account, the heart of the old regime’s suspicion of capital movement, was moderated and codified by the statute of 1999 rather than abolished, and the management framework that replaced the regulation framework retained the distinction between current account transactions, freed in principle, and capital account transactions, permitted by category. Reform, in other words, was a redistribution of control from entry to conduct, from the licensing of capacity to the regulation of markets, and the instruments of the new control were the statutes of the legislative strand: the securities, competition, electricity and insurance Acts are control statutes of a different theory, not the absence of control.
The second misstatement runs the other way, and it is the more interesting one. The reform era also created a control by notification, contemporaneously with the dismantling it is remembered for. In January 1994 the government issued the Environment Impact Assessment notification under the Environment (Protection) Act, 1986, requiring prior environmental clearance for a schedule of industrial and infrastructure projects, and thereby erected a new licensing gate at the very moment when the industrial licence was being retired. The episode is not an anomaly to be explained away; it is the taxonomy working in reverse. A government that had discovered how much law could be made by notification, under a parent Act conferring a general power for a stated purpose, used the technique for the purposes it valued as readily as it used it to retire the purposes it did not. Environmental clearance became, for many projects, the successor permission to the industrial licence, administered by committee and appraisal rather than by the licensing directorates, and its subsequent expansion and contraction proceeded by amendment of the notification, exactly as delicensing had. The legal technique of 1991 was neutral between deregulation and regulation. What the period changed was not the state’s reliance on subordinate legislation but the list of objects to which that reliance was put.
The federal boundary
A final boundary on the executive strand was constitutional, and it explains several of the silences in the 1991 programme that a focus on the Centre can miss. Industrial licensing, foreign trade, foreign exchange, banking, insurance and the securities market were Union subjects, administered from Delhi under central statutes, and a single notification could therefore reform them for the whole country at a stroke. The fields the programme left alone were, in large part, the fields where that stroke was unavailable. Land, public health and agriculture sit in the State List; labour sits in the Concurrent List, where central legislation is possible but the administration, the inspectorates and much of the operative law are the states’. Electricity is a concurrent subject whose delivery, through state electricity boards, was in the hands of state governments whose finances and politics determined the pace of the state-level Acts that followed the central policy invitations of the nineteen nineties. Even the indirect tax reforms that the Chelliah programme contemplated required, for their largest prize, the cooperation of the states: a value added tax on the sale of goods arrived only in 2005, and then state by state under a consensus built by a committee of state finance ministers, because the taxing entry belonged to the states. Electricity shows the federation legislating in parallel rather than in sequence: although the subject is concurrent, the distribution industry belonged to the states, and the first structural reform statute of the sector was a state Act, the Orissa electricity reform legislation of 1995, which unbundled the state electricity board and created a state regulator before the Centre’s consolidating Act existed. Other states followed with reform Acts of their own at the end of the decade. For a decade, therefore, Indian electricity law was a patchwork in which the Centre’s 1991 invitation to private generation, the survival of the central Acts of 1910 and 1948, a handful of state reform statutes and the interim central regulator Act of 1998 coexisted, and the consolidating Act of 2003 rationalised rather than initiated the sector’s reform. The episode refines the federal point: where competence is shared, the statutory strand of a reform runs in several legislatures at once, at several speeds, and no single date or instrument describes it. The reform map of 1991 is therefore also a map of the Seventh Schedule. What the Centre owned outright, it reformed by the fastest instruments it held. What it shared, it reformed slowly or by persuasion. What belonged to the states, it left, and the leaving was not always a choice.
The long tail: statutes that survived their own reform
Two survivals deserve separate statement because they show the interval problem in its extreme form. The Sick Industrial Companies (Special Provisions) Act, 1985 had built an apparatus, the Board for Industrial and Financial Reconstruction, around the premise that industrial sickness was a condition to be nursed by the state. The reform era disowned the premise but left the apparatus standing for a quarter of a century. Parliament enacted the Sick Industrial Companies (Special Provisions) Repeal Act in 2003, and the repeal then waited, unnotified, until it was brought into force on 1 December 2016, when the Board stood dissolved and the corporate insolvency provisions of the Insolvency and Bankruptcy Code, 2016 commenced on the same day. Between the decision that the sickness regime belonged to the old order and its legal removal lie twenty-five years, thirteen of them after Parliament had already passed the repealing Act. The interval contains a further curiosity that this series has had repeated occasion to record: Parliament tried twice more to house the subject elsewhere before the repeal took effect. The Companies (Second Amendment) Act, 2002 inserted a new part into the Companies Act, 1956 providing for the revival and rehabilitation of sick industrial companies, and those provisions were never notified into force. The Companies Act, 2013 carried a corresponding chapter on the same subject, and that chapter was omitted when the Insolvency and Bankruptcy Code, 2016 supplied the operative scheme and the repeal of the sickness Act was finally commenced alongside it. Three successive legislative attempts at the same reform, two of them stillborn, measure the distance between enacting a reform and delivering it more precisely than any commentary could. Enactment is an event; commencement is a decision; and a reform of the second kind can wait indefinitely for a government willing to take it. The monopolies Act’s survival from 1991 to 2009, after its operative chapter had been deleted in the programme’s first autumn, is the same phenomenon at a smaller scale. Indian economic reform repeatedly changed policy at the speed of the executive and law at the speed of the legislature, and the distance between the two speeds is where a large part of the legal history of the period lives.
What was not reformed
The brief for this article requires the unreformed list to be stated plainly, because it is the least discussed and most consequential part of the record.
Labour law was not reformed in 1991 at all. The statutes governing wages, industrial disputes, trade unions, factories and social security, a body of law reaching back in places to the first half of the twentieth century, were left entirely intact by the reform programme, and the consolidation of twenty-nine central labour laws into four codes waited for the Code on Wages, 2019 and the Industrial Relations Code, the Code on Social Security and the Occupational Safety, Health and Working Conditions Code, all of 2020, whose staggered commencement and rule-making the labour cluster traces in the labour codes implementation delay article. The interval between the reform moment and the statutory reform of the field in which the largest number of Indians work was twenty-eight years.
The emblem of the untouched field was Chapter VB of the Industrial Disputes Act, 1947, inserted in 1976 and widened in 1982, which required establishments employing one hundred or more workmen to obtain the prior permission of the government before laying off or retrenching workers or closing an undertaking. Here was a permission regime of exactly the character that 1991 abolished for industrial entry, applied to industrial exit, and it survived the reform era entire. The reasons were not legal; the provision stood in a statute that Parliament could have amended in any session. They were political, and they were the obverse of the calculus that made delicensing easy: entry licensing had imposed its costs on businesses that wanted to expand, a constituency the reformers courted, while exit control conferred its protection on organised workers, a constituency no government wished to confront in the programme’s fragile first years. State governments, administering the permission regime, occasionally liberalised its thresholds on their own account in later decades, which only confirms the constitutional point made above about the fields the Centre could not move alone. When the consolidation finally came, the threshold question was the centre of the argument about the Industrial Relations Code, 2020, which raised the permission threshold to three hundred workers. The distance between 1991 and that debate, on a provision whose logic the reform programme had repudiated in every neighbouring field, is the single most instructive measurement in the unreformed record.
Land acquisition was likewise untouched. The Land Acquisition Act, 1894, a colonial statute built on eminent domain with minimal process, continued to govern compulsory acquisition through the entire reform era and through two further decades of industrial and infrastructure expansion, until the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act, 2013, Act 30 of 2013, replaced it with effect from 1 January 2014, a subject the LARR Act 2013 article owns in full. The interval was twenty-two years, and it was filled with acquisition-driven conflict that the 1894 Act’s procedures were never designed to absorb.
Agricultural marketing was not reformed because, in the constitutional distribution of subjects, it could not be reformed from the Centre in the manner of the other items. Agriculture is a subject in the State List of the Seventh Schedule, and the marketing of agricultural produce is regulated by state agricultural produce marketing committee statutes, under which the first sale of notified produce was channelled through regulated markets and licensed traders. Reform in this field therefore depended on persuading twenty-eight legislatures rather than one Parliament, and the central reform programme, which had no instrument for it, left the field to a slow state-by-state process and to model legislation that states were free to adopt or ignore. The Centre’s characteristic instrument in the following decade was the model Act: a revised model agricultural produce marketing statute circulated to the states in 2003, whose adoption, in whole, in part or not at all, lay with each state legislature. The contrast with industrial licensing could not be sharper. In industry, the Centre held both the statute and the exemption power, and reform took a signature. In agricultural marketing, the Centre held neither, and reform took a generation of persuasion that remains incomplete in the statute books of the states.
Public ownership was changed in practice and never in framework. Disinvestment proceeded from the budget of 1991-92 as a sequence of decisions to sell minority stakes, and from the turn of the century as strategic sales of controlling stakes, each resting on the government’s powers as shareholder and on the Companies Act, and none resting on a statute that defined when the state should exit an industry, by what process, or subject to what parliamentary oversight. The administrative history underlines the point: for the programme’s first eight years disinvestment was conducted from within the finance ministry without a dedicated department; a Department of Disinvestment was created in 1999, and later elevated and renamed in successive reorganisations, every step taken by executive decision under the rules for the allocation of government business. A field in which the state sold assets worth, across the decades, very large sums was thus administered for its entire life without a governing statute, and the only systematic legal control over it was the general law: the Companies Act for the transactions, the Constitution’s equality guarantee for the process, and the vigilance of the Comptroller and Auditor General and of the courts after the event. When a strategic sale was challenged, the Supreme Court in the BALCO case, Balco Employees’ Union v Union of India, decided in 2002, upheld the disinvestment of a majority stake in Bharat Aluminium Company, holding that decisions about the management and ownership of public enterprises are matters of economic policy for the government, that a change of policy is not open to challenge merely because it reverses the policy of an earlier government, and that the court’s concern is confined to the legality and fairness of the process by which the decision is carried out. The judgment is the judicial counterpart of the legislative silence: a field of the first importance was governed, lawfully, without a governing Act.
Two further absences complete the unreformed list, and both were noticed at the time. The first is the exit policy. A programme that freed entry while leaving the law of failure untouched contained an obvious asymmetry, and the government of 1991 acknowledged it, announcing that an exit policy would be formulated and constituting a National Renewal Fund to finance the retraining and compensation of workers displaced by restructuring. The Fund was created by executive decision and in time wound up; the exit policy as legislation never arrived. What India eventually received, twenty-five years later, was not an exit policy of the kind contemplated in 1991 but a general insolvency code, enacted for reasons of credit discipline and industrial restructuring alike. The second absence is the reservation of products for the small scale sector, the quiet twin of industrial licensing, under which hundreds of items could lawfully be manufactured only by small units. The statement of 1991 left the reservation standing. It was dismantled gradually, by notification, in tranches beginning later in the decade and running well into the next, as item after item was dereserved; the legal framework of the sector was recast only by the Micro, Small and Medium Enterprises Development Act, 2006. Here too the instrument determined the tempo: because reservation rested on notifications under the Industries Act, it could be undone by notifications, but because each item had its constituency, it was undone slowly, over years, in a sequence invisible to anyone reading only the statutes.
What did the 1991 statement do to the industries reserved for the public sector?
It reduced the reserved list from seventeen industries to eight. Reservation had been declared in industrial policy resolutions rather than enacted in any statute, so the reduction required no legislation: the statement re-drew the list, retaining fields such as defence production, atomic energy and railway transport for the public sector.
The courts and the executive route
The executive route operated inside a body of doctrine that the courts had been building before 1991 and continued to build through the reform decades, and the doctrine mattered because it determined what a disappointed investor, licensee or competitor could hope to win by challenging a change of policy. Three propositions, each anchored in a decided case, describe the terrain.
The first is that economic policy is for the government. In Balco Employees’ Union v Union of India the Supreme Court declined to sit in judgment on the wisdom of disinvestment, holding that courts do not evaluate the merits of economic policy and interfere only where the decision or its execution is shown to be unlawful, mala fide or contrary to the statute under which it is taken. The proposition predates the reforms and survived them unchanged; it appears in the sugar delicensing litigation after them, where the Court in the Bajaj Hindustan case treated the choice between licensing and delicensing as a matter of policy committed to the government under the statute.
The second proposition concerns legitimate expectation. In Punjab Communications Limited v Union of India, decided in 1999, the Court considered a change in the terms on which a public enterprise dealt with a supplier, and held that a change of policy does not become vulnerable merely because it disappoints an expectation built on the earlier policy. A legitimate expectation can ground a claim to fair procedure, and in some settings to a hearing before a settled practice is abandoned, but it does not freeze the policy itself; the government must be free to change course in the public interest, and the court’s inquiry is into the fairness and legality of the change rather than into the soundness of the new course.
The third proposition is the boundary, and it is the one that keeps the first two from becoming a charter for unlimited executive movement. A power conferred by statute for a purpose must be exercised for that purpose. In Kasinka Trading v Union of India, decided in 1995, the Court refused to estop the government from withdrawing a customs exemption notification, holding that promissory estoppel does not lie against the exercise of a statutory power in the public interest; but the reasoning of that line of cases preserves the corollary that a notification must stay within the four corners of the power that authorises it. An exemption under section 29B of the Industries Act could abolish licensing for an industry because the section authorised exemption in the public interest; a notification that purported to do something the parent Act did not authorise, or to pursue an object foreign to it, would have been ultra vires and void. The executive route of 1991 was therefore wide but not unbounded: it ran along channels the control statutes had dug, and its legality at each step depended on the fit between the measure and the conferred power. The general position described in this article is legal information drawn from the public record of a historical period, not advice on any individual transaction, and a reader facing a live question under any of these instruments should take advice on the instrument as it stands after later amendment.
Two corollaries of that doctrine deserve statement because they mark the real legal exposure of everyone who built on the reforms. The first concerns the nature of what the executive strand conferred. An exemption notification confers no right to its own continuance, and a permission regime confers no right to the permanence of the policy that administers it. A firm that entered an industry on the strength of delicensing held no legal asset against a future notification re-imposing the licence, beyond the grace the exemption provision itself provides and the general protection against arbitrary state action. Investors understood this and priced it; the long campaign for statutory follow through, pressed by industry bodies through the nineteen nineties, was in part a campaign to convert revocable administrative freedom into enacted law, and the conversion was completed for most fields only when the replacement statutes arrived. The second corollary concerns the discipline on the other side. Subordinate legislation is reviewable. A notification that exceeded the power conferred by its parent Act, or that was issued for an extraneous purpose, or that offended the guarantee of equality in its application, was liable to be struck down, and the knowledge that the Gazette was subject to the courts was one of the constraints that kept the reform instruments carefully fitted to the powers invoked. The sugar litigation that reached the Supreme Court two decades later shows both corollaries in one case: the delicensing stood, because the power supported it, and the disappointed incumbent’s remedy failed, because no right to the continuance of a licensing regime had ever vested in it.
The doctrinal settlement had one further feature, easy to miss because it is an absence. The courts were not asked, in the reform decades, to hold that a control once imposed could not lawfully be removed, and there is no line of authority giving a licensee, an incumbent or a protected industry a constitutional complaint against deregulation as such. Challenges ran the other way, against the survival or the manner of exercise of controls, or against the fairness of a transition. The constitutional law of the period is symmetric in a way that mattered: the discretion that had sustained four decades of licensing was the same discretion, located in the same sections, that sustained its abolition, and the guarantee of equality was available against the arbitrary exercise of either. A legal order that wishes to make reform difficult can entrench the status quo, requiring compensation for deregulation or vesting rights in the holders of licences. The Indian order did none of that. It left the switches where the control statutes had put them, in the hands of the executive, subject to legality review; and that institutional choice, made by the drafters of the control era for their own purposes, is what the reformers of 1991 inherited and used.
The assessment, stated both ways
The register of this series requires the two assessments of 1991 to be put at full strength, in their proponents’ own terms, and neither adopted.
The favourable assessment, made by the programme’s defenders in government at the time and developed by economists such as Jagdish Bhagwati in the years that followed, is that the reforms averted a default that in June 1991 was weeks away, and that they did so by the only route the calendar allowed. Parliament had been elected in an atmosphere of crisis and political fragmentation; the government was a minority one; a legislative programme of a dozen Bills would have taken years the reserves did not contain. On this view the executive route was not a circumvention of democracy but its rescue, and it was vindicated by what followed: reserves rebuilt, growth sustained at rates the earlier decades had not reached, an enormous expansion of the goods and services available to Indian consumers, and the conversion of a near-default into a durable external position. Defenders add that the statutory follow through did come, across the following twelve years, so that the charge of government by notification describes a sequencing necessity rather than a settled method.
The critical assessment, pressed by the opposition in the parliamentary debates of 1991 and developed by critics such as the economist Prabhat Patnaik, is that the programme was adopted under external financial pressure, on the conditions attached to Fund financing, without an electoral mandate for it and with the decisive measures insulated from legislative scrutiny by the very choice of instruments this article has described. On this view the speed that defenders celebrate is the vice itself: licensing abolished by notification could have been debated as a Bill, and the decision to proceed by exemption and permission meant that the largest reorientation of the Indian state since 1947 was the least legislated. Critics add the distributional record: that the programme addressed first the constraints on capital and on imports, and addressed last, or not at all, the fields in which most Indians earn their living, agriculture and unorganised labour, so that the map of what was reformed is also a map of whose interests the state found it easiest to serve.
Between the two assessments sits a third observation, about method rather than merit, which the political scientist Rob Jenkins developed in his study of the period under the description of reform by stealth. On that account the choice of low-visibility instruments was not merely a response to the emergency of June 1991 but a governing technique that outlasted it: measures taken by notification, by administered permission and by committee-guided direction attracted a fraction of the organised opposition that the same measures would have attracted as Bills, and a government without a majority for its programme could therefore sustain the programme for years by keeping its legal footprint small. The observation cuts both ways, and Jenkins’s own treatment allows both readings. To the programme’s defenders, stealth was the democratic art of the possible, delivering changes that commanded broad assent in substance while sparing a fragile polity a series of confrontations it might not have survived. To its critics, stealth was the mechanism by which the absence of a mandate was concealed, and the statutory lag this article has measured is the documentary residue of the concealment. The legal record cannot settle that argument, but it can and does confirm the premise on which it rests: the reforms were, in their first and decisive phase, made in the Gazette and in the circulars of the Reserve Bank rather than on the floor of Parliament, and the floor was reached, field by field, over the following twelve years and in some fields not for a generation.
One point is separable from both assessments, is a matter of law rather than of judgment, and can be stated definitely. Reform by notification under a power conferred for another purpose leaves the underlying statute intact. The Industries Act of 1951 still contains its licensing provisions; they sleep under exemptions that a notification can cancel. The foreign exchange statute of 1973 governed, in softened form, for eight years after its policy had been disowned. Supporters of the programme read this as unfinished business, a statute book that never caught up with the economy built on top of it. Opponents read the same fact as a preserved option, a legal path back that no majority ever voted to close. Both readings are available because the underlying legal fact is what it is: what was done without Parliament can, in law, be undone without Parliament, and the permanence of the reforms rests on political and economic foundations rather than on statutory ones. Whether that is a defect or a safeguard is the question the period leaves open; that it is the position is not in dispute.
How to date a reform
The record teaches a discipline that students of the period ignore at their peril: a reform does not have a date, it has up to four of them, and the choice among them decides what story gets told. There is the date of decision, when the government announces the change; the date of the instrument, when the notification, ordinance or Act that carries it is made; the date of enactment, when a statute completes its passage; and the date of commencement, when the enacted words are brought into force. For the executive strand the first two dates coincide and the others never arrive. For the monopolies amendment the dates cluster within a single season. For the foreign exchange replacement they spread across nine years, decision in 1991, amendment of the old Act in 1993, replacement enacted in 1999, replacement in force in June 2000. For the sickness repeal they spread across thirteen years between enactment and commencement alone. The Competition Act adds a refinement: a single statute can commence in slices, and its central prohibitions on anti-competitive agreements and abuse of dominance were brought into force only in May 2009, more than six years after assent, so that the competition regime of the reform era existed on paper through most of the decade in which commentators described it as operating. The Electricity Act, by contrast, commenced within months of enactment in 2003, and the trade statute of 1992 was given retrospective effect to the date of its ordinance. No single convention governs. The only sound practice is the one this article has followed: to name the instrument, give its own date, and state the interval to the decision it implemented, resisting the compression by which a decade of legal process is flattened into the single year in which the politics happened. Examination answers and newspaper retrospectives alike date the reforms to 1991. The statute book, read with its commencement notifications, dates them to a period running from 1991 to 2016, and the difference between those two datings is the subject matter of legal history as opposed to political memory.
The tribunal settlement
One institutional legacy of the legislative strand deserves separate record, because it altered the structure of economic adjudication in India more durably than any single deregulation. Almost every regulator the reform statutes created was paired, sooner or later, with a specialised appellate tribunal, and the pairing became the standard constitutional settlement for economic regulation: the expert body decides, the tribunal reviews on law and fact within its statute, and the Supreme Court stands at the apex on questions of law. The Securities Appellate Tribunal was created by the securities statute itself in 1992. The telecom regulator’s appellate arrangements were recast by amendment in 2000, when the Telecom Disputes Settlement and Appellate Tribunal was established and the adjudicatory functions were separated from the regulatory ones. The competition statute acquired its appellate tribunal by amendment in 2007, and the debt recovery tribunals created in 1993 grew an appellate tier of their own. The significance for the reforms is structural. Regulation by statute required not merely regulators but a jurisprudence to discipline them, and the tribunal layer is where that jurisprudence was made: the case law on penalties, on the duty to give reasons, on the separation of investigation and adjudication inside a regulator, accumulated in the tribunals through the decades after 1991. A reform programme that began with the abolition of permissions thus ended by generating one of the largest bodies of administrative law in the country’s history, concerned not with whether business needed the state’s leave but with how the state’s referees must behave. That is a measure of how completely the character of economic law changed between the statement of 1991 and the settlement that followed it.
Three speeds of lawmaking
The taxonomy the record supports can be stated as a rule of Indian legislative behaviour, and it is offered as this article’s framework for reading reform episodes beyond 1991. The first speed is the speed of the conferred power: where a statute already authorises the executive to exempt, permit, license, direct or order, policy can be reversed in the time it takes to print a notification, and the Industries Act exemption, the Reserve Bank’s automatic route, the trade orders under the Act of 1947 and the central bank’s prudential directions are all instances. The price of the first speed is reversibility and a statute book that misdescribes the operative law. The second speed is the speed of the Bill: where a requirement must be deleted, a control repealed or an authority created, the measure joins the queue of parliamentary business, passes through committee, and emerges in months when it is a simple deletion and in years when it designs an institution, as the two months of the monopolies amendment and the eight years of the insurance authority mark the ends of the range. The price of the second speed is delay; its product is durability, because an enacted settlement can be undone only by the same process that made it. The third speed is the speed of the unattempted: where reform requires new legislation in a field of high political cost or divided constitutional competence, the interval is measured in decades, and the field waits for a conjunction of parliamentary arithmetic and political will that 1991 could not supply for labour, land and agricultural marketing. A reader equipped with the three speeds can predict, for any announced reform in any period, roughly when the law will actually change, by asking a single question: under which of the three does the measure fall. The question is more reliable than any inquiry into the sincerity or energy of the government announcing it, because it is a question about the location of legal power rather than about intentions.
The taxonomy also locates the outer limit of reform by ordinary legislation, which a later episode in the same series illustrates by contrast. The goods and services tax could not be carried at any of the three speeds, because the taxing entries of the Seventh Schedule divided the field between the Union and the states in a way no statute could rearrange; it required the Constitution (One Hundred and First Amendment) Act, 2016, with ratification by the states, a quarter of a century after 1991. Constitutional amendment is the fourth and slowest instrument, slower even than the unattempted, because it must assemble special majorities and, for the federal entries, state consent. The reformers of 1991 were fortunate in their subject matter: nothing in the core programme required the Constitution to be touched, and the whole of it could be distributed among notification, ordinance and Act. A reformer whose programme crosses the federal distribution of powers enjoys no such fortune, and the interval structure of Indian reform since 1991, in taxation above all, is the record of that difference.
It also explains the one feature of the record that purely political accounts find puzzling, which is that the reforms were never reversed even by governments elected on platforms critical of them. Reversal, like enactment, obeys the taxonomy. The executive strand could lawfully have been reversed by notification at any time, and was not, because the economic and political costs of restoring licensing had by then been internalised by every governing coalition. The legislative strand could not be reversed without Bills that no government wished to spend its majority on. Permanence in India followed not from the legal form of the reforms, which in their first phase was as fragile as administrative law can be, but from the accumulation of interests organised around them. The law supplied the sequence; politics supplied the ratchet.
What the record shows
The verdict this article defends is the namable claim with which it began. The reforms that happened fastest were the ones no legislature had to approve: delicensing and the opening to foreign equity were effected within weeks by policy statement, exemption notification and administered permission, because the statutes already conferred the powers. The reforms that required repealing or replacing a statute took between one and eighteen years. The reforms that would have required entirely new legislation in labour, land and agriculture were not attempted in 1991 at all and took between twenty-two and twenty-nine years. The shape of Indian liberalisation was therefore determined less by what the reformers of 1991 wanted than by what could lawfully be done without a Bill, and any account of the period that begins with intentions rather than instruments will misread the sequence it sets out to explain.
Frequently Asked Questions
Q: What was the crisis that produced the 1991 reforms?
It was a balance of payments crisis: India could not pay for its imports or service its external obligations from its earnings and reserves. The Economic Survey for 1991-92 recorded foreign currency assets of 896 million dollars on 16 January 1991, down from reserves of 3.11 billion dollars the previous August, after the Gulf conflict raised oil prices and interrupted remittances. The Reserve Bank’s published history describes the June 1991 reserves as covering roughly three weeks of essential imports. Commercial credit had dried up after rating downgrades, and the government pledged sixty-seven tonnes of gold abroad and entered into arrangements with the International Monetary Fund.
Q: What did the industrial policy statement of July 1991 announce?
The statement on industrial policy laid before Parliament on 24 July 1991 abolished industrial licensing for all projects except eighteen listed industries, cut the industries reserved for the public sector from seventeen to eight, promised the removal of prior approval for large undertakings under the monopolies statute, offered automatic approval of foreign equity up to 51 per cent in high priority industries, abolished the phased manufacturing programmes, and placed foreign technology agreements in those industries under automatic approval. It was an announcement of policy, not legislation, and each measure took effect through the instrument appropriate to it: notification, administered permission, or later amendment.
Q: Which reforms required legislation and which did not?
Delicensing required no Act, because the Industries Act of 1951 itself conferred an exemption power exercisable by notification. Opening foreign equity required no Act, because the foreign exchange statute operated through Reserve Bank permission, which could be liberalised administratively. Removing monopolies clearance required an amendment, because the approval requirement stood in the statute; abolishing capital issues control required a repeal; and creating regulators required new Acts, because coercive authority over citizens cannot be conferred by resolution. The organising rule is that executive action sufficed wherever the old statute already conferred the necessary power, and legislation was needed wherever a statutory requirement had to be removed or an authority created.
Q: Which new regulators were created by statute?
The Securities and Exchange Board of India, constituted by resolution in 1988, was made a statutory regulator by the Securities and Exchange Board of India Act, 1992. The Telecom Regulatory Authority of India was created by Act 24 of 1997. The Insurance Regulatory and Development Authority was created by Act 41 of 1999, which also opened insurance to private and foreign capital. Electricity commissions were established under the Electricity Regulatory Commissions Act, 1998 and continued under the Electricity Act, 2003. The Competition Commission followed under the Competition Act, 2002. Each required an Act, because a body with power to bind, investigate and penalise cannot be created by executive resolution.
Q: What did the economic reforms leave untouched?
Labour law was left entirely intact in 1991; the consolidation of twenty-nine central labour laws into four codes arrived only with the Code on Wages, 2019 and the three codes of 2020. Land acquisition continued under the colonial Act of 1894 until the statute of 2013 replaced it. Agricultural marketing, a subject in the State List, was left to the states and to model legislation. Public ownership was changed by executive decisions on disinvestment but never given a framework statute. The sick industries apparatus also survived: the repeal enacted in 2003 was brought into force only in December 2016.
Q: How long did the statutory follow through to the economic reforms take?
It depended entirely on the class of change. The deletion of the monopolies clearance provisions took about two months, completing in September 1991. The capital issues repeal and the securities regulator’s statute took about ten months. The foreign trade statute took about eleven months. The regulators for telecommunications and insurance took six and eight years respectively. The foreign exchange replacement took eight years, competition and electricity eleven and twelve years, land twenty-two years, and labour twenty-eight to twenty-nine years. The monopolies Act, whose operative chapter was deleted in 1991, was itself repealed only in 2009, eighteen years after the decision that hollowed it out.
Q: How are the economic reforms assessed?
Defenders, including the programme’s authors in government and economists such as Jagdish Bhagwati, argue that the reforms averted an imminent default, that the executive route was the only one the reserve position allowed, and that the growth, reserves and consumer choice that followed vindicated the choice of speed. Critics, including the parliamentary opposition of 1991 and economists such as Prabhat Patnaik, argue that the programme was adopted under external financial pressure without an electoral mandate, that proceeding by notification insulated it from legislative scrutiny, and that its benefits were unevenly distributed while agriculture and unorganised labour were addressed last or not at all. This article adopts neither assessment.
Q: What do the economic reforms show about how India legislates?
They show that the Indian state legislates at two speeds. Where a statute confers a power to exempt, permit or order, policy can change in weeks, and the change remains reversible by the same route because the statute underneath is untouched. Where a requirement stands in the statute, or a new authority must be created, change moves at the pace of Parliament, measured in years. The reforms also show that a repealing Act is not itself a repeal until it is brought into force, as the thirteen-year wait of the sick industries repeal demonstrates. The durability of a reform in India therefore depends on which instrument carried it.
Q: Why did the government pledge gold in 1991 instead of relying only on borrowing?
Because commercial borrowing had ceased to be available on the strength of India’s credit alone. Rating downgrades in late 1990 and early 1991 closed the international capital markets to Indian borrowers, and non-resident deposits were flowing out. Gold held by the Reserve Bank was an asset against which funds could actually be raised at speed: twenty tonnes shipped to the Union Bank of Switzerland in May 1991 raised 200 million dollars, and forty-seven tonnes shipped to the Bank of England raised 405 million dollars. The pledge was a bridge to official financing, not a substitute for it; the Fund arrangements of 1991 supplied the larger sums, against the security of a stated reform programme rather than of metal.
Q: What arrangements did India enter into with the International Monetary Fund in 1991?
In January 1991 the Fund approved a stand-by arrangement of 551.9 million special drawing rights, and India drew 716.9 million special drawing rights under the compensatory and contingency financing facility, the window for shocks such as the Gulf oil price rise. After the general election, a larger stand-by arrangement of 1,656 million special drawing rights was approved on 31 October 1991, supported by a letter of intent and a memorandum of economic policies setting out the government’s stabilisation and structural reform programme. A stand-by arrangement is a financing decision of the Fund, not a treaty; it enacted nothing in Indian law, and every element of the programme still required an Indian instrument to carry it.
Q: Was the rupee devalued as part of the 1991 reforms, and by what route?
Yes. The Reserve Bank adjusted the value of the rupee downward in two steps, on 1 and 3 July 1991, by a cumulative margin of roughly eighteen per cent against the major currencies. The route was administrative: the exchange rate was at that time an administered price set under the exchange control regime, and its adjustment required no legislation and no amendment to the foreign exchange statute. The devaluation preceded the industrial policy statement by three weeks and belonged to the same class of measures as the gold pledge and the Fund arrangements: emergency stabilisation carried out with the instruments the executive already held, before the structural programme was announced.
Q: Did the opening to foreign equity in 1991 amend the Foreign Exchange Regulation Act?
No. The Foreign Exchange Regulation Act, 1973 remained in force, in the form enacted, while the automatic approval route for foreign equity up to 51 per cent in listed industries was administered under it. The Act prohibited the relevant transactions without Reserve Bank permission but did not itself fix the equity ceiling, so raising the ceiling changed the practice of permission rather than the text of the law. The Act was amended in 1993 to remove some of its harsher features, and it was replaced only by the Foreign Exchange Management Act, 1999, which took effect on 1 June 2000, eight years after the policy it administered had been reversed.
Q: What was the Foreign Investment Promotion Board and why did it need no statute?
The Foreign Investment Promotion Board was the inter-ministerial body constituted in 1991, initially in the Prime Minister’s Office, to consider foreign investment proposals that fell outside the automatic approval route and to recommend on them to the government. It needed no statute because it wielded no independent legal power: it advised, and the legal acts of approval or refusal were taken under the foreign exchange law and the industrial licensing framework by the authorities those laws named. Being a creature of executive decision, it could be and was moved between departments, to the industry ministry’s promotion department in 1996 and to the Department of Economic Affairs in 2003, and abolished by a Cabinet decision and office memorandum in 2017.
Q: What is the difference between a regulator created by resolution and one created by statute?
A body created by resolution, as the Securities and Exchange Board of India was in 1988, exists at the pleasure of the government that created it. It can advise, study and persuade, but it cannot compel anyone: it has no statutory power to demand information, to register or debar intermediaries, to investigate, or to impose penalties, because no law has conferred those powers. A body created by statute holds the powers its Act confers, exercised in its own name and enforceable in law, and it can be restructured or abolished only by Parliament. The conversion of the securities board by the Act of 1992 is the standard illustration: the institution was four years old before it could bind anyone.
Q: Which statute governed imports and exports before 1992 and what replaced it?
Imports and exports were governed by the Imports and Exports (Control) Act, 1947, which empowered the Central Government to prohibit, restrict or control trade by order, and under which the export and import policy and its licensing lists were issued. Trade liberalisation in 1991 was carried out under that Act, by shrinking the restricted lists. The Foreign Trade (Development and Regulation) Act, 1992, Act 22 of 1992, replaced it, taking effect from 19 June 1992, and changed the declared object of the field from control to the development and regulation of foreign trade by facilitating imports and augmenting exports, while retaining powers of prohibition and restriction exercisable item by item.
Q: Did the reforms change the ownership of public sector enterprises by statute?
No. No privatisation or disinvestment statute was enacted at any point in the reform era. Disinvestment proceeded as a sequence of executive decisions: sales of minority shareholding bundles from the budget of 1991-92, advised on by bodies created by resolution such as the Disinvestment Commission of 1996, and strategic sales of controlling stakes from the turn of the century, all resting on the government’s position as shareholder under the Companies Act. When a strategic sale was challenged, the Supreme Court upheld it in the BALCO case in 2002, holding that ownership decisions about public enterprises are matters of economic policy, reviewable for legality and fairness of process rather than for wisdom.
Q: What replaced the industrial licence application for delicensed industries?
The Industrial Entrepreneurs’ Memorandum. Under the licensing system, the entrepreneur’s founding document was a licence application: a request for a discretionary grant, decided after inter-ministerial consideration of capacity, location and imports. Under the exemption regime that followed the 1991 statement, a project in a delicensed industry was begun by filing a memorandum with the Secretariat for Industrial Assistance in the industry ministry, recording the particulars of the project. The memorandum required no approval and received an acknowledgement rather than a grant. Its purpose was information, so that the state would know what was being built, not permission, and the change from application to memorandum is the clearest single measure of the inversion of industrial entry in 1991.
Q: Why did a competition statute arrive only in 2002 when monopolies clearance ended in 1991?
Because the two measures did different work. Ending monopolies clearance in 1991 was a deletion: the removal of prior approval for large undertakings, achieved by omitting sections 20 to 26 of the monopolies Act. A competition statute was a creation: it required a new theory of the subject, against anti-competitive agreements and abuse of dominance rather than against size itself, a committee to design it, and a new commission to administer it. The Competition Act, 2002, Act 12 of 2003, followed the report of the high level committee appointed at the end of the decade. The old monopolies Act, reduced to its trade practices provisions, survived alongside until its repeal took effect on 1 September 2009.
Q: What role did the Union Budget of July 1991 play alongside the industrial policy statement?
The budget of 24 July 1991 was the legislative half of the announcement day. It carried the fiscal correction the Fund programme required, proposed the tax changes that only a Finance Act could enact, set the direction for customs and excise restructuring, and announced the intended restructuring of public enterprises and the financial sector. The industrial policy statement, laid before Parliament the same day, carried the deregulatory half, none of which needed a vote. The pairing displays the article’s central distinction in a single date: the measures that required Parliament’s authority travelled in the budget and its Finance Bill, and the measures within executive power travelled in a statement that Parliament was informed of but never asked to enact.
Q: How did the courts respond when a change of economic policy was challenged?
With restraint bounded by legality. In the BALCO case in 2002 the Supreme Court refused to evaluate the merits of disinvestment, holding that economic policy is for the government and that review is confined to unlawfulness, mala fides and fairness of process. In Punjab Communications Limited v Union of India in 1999 the Court held that a legitimate expectation founded on an earlier policy does not prevent a change of policy in the public interest, though it may ground a claim to fair treatment. In Kasinka Trading in 1995 the Court refused to estop the government from withdrawing an exemption notification. The consistent limit is that a power must be exercised within the statute that confers it and for the purposes for which it was conferred.