Ask what changed when India replaced its disability statute and the answer usually arrives as arithmetic: seven recognised disabilities became twenty-one. The arithmetic is accurate and the answer is wrong. The Persons with Disabilities Act 1995 was not undone by the shortness of its list. It was undone by a sentence fragment that recurs through its chapters, a short qualifying formula which provided that governments would act within the limits of their economic capacity and development. That formula, and the habits of drafting that surrounded it, converted most of the statute from a set of obligations into a set of intentions. The Rights of Persons with Disabilities Act 2016 is best understood as the statute that deleted the formula from every provision where a duty was owed to an identifiable person, and rebuilt the law around the deletion. Counting disabilities measures the change that is easiest to see. The change that mattered was grammatical before it was anything else.

The two statutes reward being read side by side, because the later one is a commentary on the earlier. The 1995 Act, formally the Persons with Disabilities (Equal Opportunities, Protection of Rights and Full Participation) Act, was assented to on 1 January 1996 and is cited as Act 1 of 1996, the first of this subject’s name and year traps: a statute called after 1995 that carries the number of 1996. It came into force on 7 February 1996. The 2016 Act received the assent of the President on 27 December 2016, is Act 49 of 2016, and came into force on 19 April 2017, on which date the older statute stood repealed by its Section 102. Between those two commencement dates lies a complete change in what Indian law thought disability was. The earlier statute treated disability as a characteristic of a person, certified by a medical authority, which attracted provision from the state. The later statute treats disability as what happens when a long term impairment meets a barrier, and it therefore regulates buildings, transport systems, information formats, examinations, employers and attitudes, and only then turns to benefits. This article traces that passage: what the first statute built, why its obligations failed in the particular way they did, what the replacement kept, what it discarded, what it added, and how the repeal itself was drafted so that certificates, appointments and proceedings survived the death of the Act that created them.

A statute built on a proclamation

The 1995 Act began life far from any Indian legislature’s ordinary business. In December 1992 the Economic and Social Commission for Asia and the Pacific met at Beijing and adopted the Proclamation on the Full Participation and Equality of People with Disabilities in the Asian and Pacific Region, launching the Asian and Pacific Decade of Disabled Persons, which ran from 1993 to 2002. India was a signatory to that Proclamation. The Statement of Objects and Reasons prefixed to the Indian Bill recited the Beijing meeting, the Decade and the Proclamation, and stated that the proposed legislation was brought to give effect to the instrument India had signed. The long title of the enacted statute says the same thing in fewer words: it is an Act to give effect to the Proclamation on the Full Participation and Equality of People with Disabilities in the Asian and Pacific Region. A proclamation of a regional commission is a declaration of intent, not a treaty with obligations of result, and the softness of the founding instrument is worth holding in mind, because the statute built on it inherited a corresponding softness of obligation. The connection between the character of the international instrument and the character of the domestic text is the first link in the chain this article follows.

The constitutional route mattered as much as the diplomatic one. Relief of the disabled and unemployable is Entry 9 of the State List in the Seventh Schedule, which means that in the ordinary distribution of legislative power the subject of disability welfare belonged to the states. A central statute on the subject therefore needed a constitutional door, and the door was Article 253, which empowers Parliament to make any law for the whole or any part of India for implementing a treaty, agreement or convention with another country, or a decision made at an international conference, association or other body. Parliament had walked through that door for environmental legislation a decade and a half earlier, when the Air (Prevention and Control of Pollution) Act 1981 was enacted to implement the decisions taken at the Stockholm conference on the human environment, a use of the treaty-implementation power traced in the account of the Air Act 1981. Article 253 did two things for the 1995 Act. It supplied competence over a State List subject, and it shaped the statute’s self-understanding: a law enacted to give effect to an international instrument tends to speak in the register of that instrument, and the Beijing Proclamation spoke the language of participation goals, targets and endeavour rather than the language of entitlements enforceable at the instance of the person concerned.

The result was a statute that sat inside the welfare tradition of Indian legislation, the broad field in which the state announces provision for a class of citizens and organises departments, committees and schemes to deliver it, a field mapped as a whole in the complete guide to India’s welfare laws. Within that tradition the 1995 Act was not an outlier. It was, in form, a respectable example of its kind: earnest in its recitals, organised in its chapters, and equipped with coordinating machinery at the Centre and in the states. What distinguishes it, and what made its replacement a repeal rather than a repair, is the degree to which its operative sentences withdrew what its recitals appeared to grant. To see that, the statute has to be read as its drafters arranged it, chapter by chapter, before the qualifying formula is followed through the text.

Did the Persons with Disabilities Act extend to Jammu and Kashmir?

As enacted, the 1995 Act extended to the whole of India except the State of Jammu and Kashmir, a limitation printed in its first section and common in central legislation of that period. The 2016 Act dropped the exception and extends to the whole of India, a change that passed without debate.

What the Persons with Disabilities Act built

The 1995 Act contained fourteen chapters and seventy-four sections. Its architecture was administrative before it was anything else. At the top sat coordination: a Central Coordination Committee chaired by the Union Minister in charge of welfare, State Coordination Committees in each state, and executive committees to carry their decisions into departments. Chapter II built that machinery first, before a single substantive duty appeared, which tells a reader where the statute located the problem. Disability, on this design, was a problem of departments failing to pull together, and the cure was a committee in which the departments sat. The chapters that followed then distributed the substantive work: prevention and early detection in Chapter IV, education in Chapter V, employment in Chapter VI, affirmative action in Chapter VII, non-discrimination in Chapter VIII, research and manpower development in Chapter IX, recognition of institutions in Chapter X, an institution for persons with severe disabilities in Chapter XI, the Commissioners in Chapter XII, social security in Chapter XIII, and miscellaneous provisions, including the statute’s single offence, in Chapter XIV.

The definitional chapter did the quiet work on which everything else depended. Section 2(i) defined disability as a list of seven conditions: blindness, low vision, leprosy-cured, hearing impairment, locomotor disability, mental retardation, in the language of the period, and mental illness. Section 2(t) then defined a person with disability as a person suffering from not less than forty per cent of any disability as certified by a medical authority. Two features of that pairing deserve attention, because the 2016 Act treated them differently. The list was closed: a condition outside the seven entries was not a disability for the statute at all, however severe its effects, and no power in the Act allowed the list to be extended by notification. The threshold, by contrast, was a measuring rule rather than a boundary of concern, since the forty per cent figure sorted claimants for benefits without purporting to describe what disability was. The replacement kept the threshold, rebuilt the list, and separated the two ideas into different definitions, and much of the practical confusion in the field since has come from readers who treat the two statutes’ definitional chapters as interchangeable when they are not.

The substantive chapters mixed three kinds of provision, and the mixture is the point. The first kind was programmatic: Section 25 directed appropriate Governments and local authorities to undertake surveys, investigations and research concerning the cause of occurrence of disabilities, to promote methods of preventing disabilities, to screen children at least once a year to identify at-risk cases, to train staff at primary health centres, and to run awareness campaigns. The education chapter directed that every child with a disability have access to free education in an appropriate environment until the age of eighteen, contemplated integrated and special schools, and required governments to frame a scheme by notification covering transport, books, uniforms, scholarships, curricular change and the removal of architectural barriers in schools. The employment chapter required governments to identify posts that could be reserved, to maintain Special Employment Exchanges, and to frame schemes covering training, relaxation of age limits, regulation of employment agencies and economic rehabilitation. None of this was worthless. A scaffolding of schemes, exchanges, surveys and committees is the ordinary furniture of Indian social legislation, and some of it was used with effect.

The education chapter repays a closer reading, because it shows the statute’s soft devices working in combination rather than singly. Section 26(a) stated the hard promise: the appropriate Governments and local authorities shall ensure that every child with a disability has access to free education in an appropriate environment till he attains the age of eighteen years. The clauses that followed then retreated into endeavour: governments were to endeavour to promote the integration of students with disabilities in normal schools, and to endeavour to equip special schools with vocational training facilities. Around that core, Sections 27 to 31 arranged the supporting machinery of schemes for non-formal education and functional literacy, research into assistive devices and teaching aids, teacher training institutions, and a redress route for parents aggrieved about the placement of a child. Section 30 required the appropriate Governments to prepare, by notification, a scheme addressing transport facilities for children with disabilities or financial incentives to their parents in the alternative, the removal of architectural barriers from schools and other educational institutions, the supply of books, uniforms and other materials, the grant of scholarships, the restructuring of curriculum, and facilities such as a scribe for students with blindness. Every item in that list is concrete, and a reader can see exactly what a functioning scheme would have delivered. What the chapter never supplied was a date by which the scheme had to exist, a standard against which its adequacy could be judged, or a consequence for a government that framed a thin scheme or none. The chapter’s drafting is the 1995 method in miniature: the goal stated as a duty, the delivery routed through an instrument the duty-bearer drafts for itself, and the soft verb “endeavour” placed at the points where a child’s actual schooling would be decided.

The employment chapter shows the same method with the quantified exceptions standing out in relief. Section 32 required the appropriate Governments to identify posts in establishments which could be reserved for persons with disabilities, and to review the list periodically; identification was the administrative hinge of the whole chapter, since reservation under Section 33 operated in the identified posts, and the pace of identification was set by the governments whose establishments were to be reserved. Section 34 provided for Special Employment Exchanges, through which vacancies and jobseekers were to be matched, and Section 38 contemplated schemes for the employment of persons with disabilities, covering training, the relaxation of upper age limits, the regulation of employment agencies, the collection of statistics and the economic rehabilitation of the class. The affirmative action chapter added Section 42, under which governments were to frame schemes for aids and appliances, and Section 43, under which land was to be allotted preferentially at concessional rates for housing, for setting up businesses and recreation centres, for special schools, for research centres and for factories run by entrepreneurs with disabilities. The architecture was complete on paper: a route into education, a route into work, appliances to make both usable, land to build a livelihood on. What the paper architecture lacked, section after section, was the drafting furniture of compulsion, and the two chapters on research and institutions completed the picture of a statute that organised everything except obedience. Chapter IX directed financial assistance for research into prevention, rehabilitation, assistive devices and employment; Chapter X built a recognition regime for institutions for persons with disabilities, with a competent authority, certificates of recognition and revocation; Chapter XI provided for an institution for persons with severe disabilities. The last provision of the Act, Section 74, reached outside the statute altogether and amended the Legal Services Authorities Act 1987 so that a person with disability qualified for free legal services, a small amendment with a long reach, since it armed the class the statute served with a route to the courts whose absence elsewhere in the text it quietly acknowledged.

The second kind of provision was quantified, and it behaved differently. Section 33 required every appropriate Government to appoint, in every establishment, persons with disability against not less than three per cent of vacancies, one per cent each being reserved for persons with blindness or low vision, persons with hearing impairment, and persons with locomotor disability or cerebral palsy, in posts identified for the purpose. Section 39 required all Government educational institutions and educational institutions receiving aid to reserve not less than three per cent of seats for persons with disabilities. Section 40 required that not less than three per cent be reserved in all poverty alleviation schemes. Section 36 provided for unfilled reserved vacancies to be carried forward. Percentages are enforceable in a way that programmes are not, because a percentage has a numerator and a denominator that a court can count, and the litigation history of the 1995 Act is largely the history of these three sections. The third kind of provision was protective rather than distributive: Section 47 forbade an establishment to dispense with, or reduce in rank, an employee who acquired a disability during service, directing instead that he be shifted to another post with the same pay scale and service benefits. That prohibition, unqualified and addressed to a defined act, was also enforceable, and it too was litigated. The statute, in other words, contained a hard core of justiciable provisions surrounded by a much larger body of programmatic direction. The tragedy of its drafting is that the programmatic body was where most of its promises lived.

Two structural features completed the design. The first was exemption. Section 33’s reservation duty carried a proviso under which the appropriate Government could, having regard to the type of work carried on in an establishment, exempt that establishment from the reservation provision by notification. A duty from which the duty-bearer can release the obligated party by an executive act is a duty with a door in it, and the exemption power was used. The second feature was the statute’s route of complaint. Chapter XII created a Chief Commissioner at the Centre and Commissioners in the states, officers charged among other things with looking into complaints of deprivation of rights. How that office was armed, and how it was not, is examined below, because the Commissioners are the clearest measure of the distance between the two statutes. Before that measure is taken, the qualifying formula itself needs to be followed through the 1995 text, section by section, since the formula is the article’s central exhibit.

How many chapters and sections does each disability statute have?

The 1995 Act contains fourteen chapters and seventy-four sections: coordination committees, programmatic chapters and a short miscellaneous chapter carrying its only offence. The 2016 Act contains seventeen chapters and one hundred and two sections. The growth lies in a rights chapter, a certification chapter, an offences chapter, Special Courts and statutory Funds, machinery with no counterpart in the earlier text.

The qualifying formula and where it sat

The formula reads, in its standard form, “within the limits of their economic capacity and development”. It appears in the Statement of Objects and Reasons of the 1995 Act itself, which is to say it was announced as the design and not smuggled into the text afterwards. The Statement declared that the Bill sought to provide for education and employment, and for the other matters it listed, within the limits of economic capacity and development of the governments concerned. A reader of the Bill therefore knew from its first pages that the obligations to come were conditioned. The enacted text then honoured the announcement with unusual thoroughness. Section 25, on prevention and early detection of disabilities, opens with the formula: the appropriate Governments and local authorities shall, within the limits of their economic capacity and development, undertake the surveys, screening, training and publicity the section lists. Section 41, on incentives to employers who build a workforce of which persons with disabilities form at least five per cent, is governed by the same opening words. The whole non-discrimination chapter is governed by them: Section 44, on transport, Section 45, on roads, and Section 46, on the built environment, each begins by obliging the appropriate Governments to act within the limits of their economic capacity and development. Section 66, on rehabilitation, directs governments within the same limits to undertake or cause to be undertaken the rehabilitation of all persons with disabilities. Section 68, on the unemployment allowance payable to persons with disabilities registered at a Special Employment Exchange for more than two years and not placed in gainful occupation, is drafted within the same frame.

The itinerary matters more than any single instance. Prevention, incentives, transport, roads, buildings, rehabilitation and the unemployment allowance cover most of what a person with a disability in the nineteen nineties would have asked the statute for. A ramp at a railway station fell under Section 44 or Section 46. Aids, appliances and preferential allotment of land sat in the affirmative action chapter, whose operative provisions leaned on scheme-making. The education chapter’s central promise, free education to eighteen, was expressed as a duty, but its content was to be delivered through a scheme that the government notified, and the scheme’s contents were the government’s to settle. The pattern was not that the Act promised nothing. The pattern was that wherever a promise would cost money in a way a budget could not schedule in advance, the promise arrived with its own answer already drafted. A department that had done nothing about accessible transport could reply, in words the statute itself supplied, that it had acted within the limits of its economic capacity and development, and the reply was not a breach of the statute. It was an invocation of it.

The formula had a parent, and identifying it explains both its presence and its effect. Article 41 of the Constitution, among the Directive Principles of State Policy, provides that the State shall, within the limits of its economic capacity and development, make effective provision for securing the right to work, to education and to public assistance in cases of unemployment, old age, sickness and disablement. The words are the statute’s words, and the subject is the statute’s subject: the framers of the 1995 Act lifted the qualifying formula from the constitutional provision that speaks of disablement itself. But a Directive Principle and a statutory duty are different instruments, and the borrowing changed the words’ legal character by moving them. Article 37 declares the Directive Principles fundamental in the governance of the country while making them unenforceable by any court; a fiscal qualification inside a Directive Principle is therefore stating, honestly, the terms on which a non-justiciable aspiration proceeds. Transplanted into the operative sections of a statute, provisions that citizens and Commissioners would naturally read as duties, the same words performed a quieter operation: they carried the non-justiciability of the Directive Principles into a text that wore the dress of enforceable law, without announcing the transfer. Much of the confusion that surrounded the 1995 Act for twenty years, the recurring surprise that a statute so full of the word “shall” could compel so little, is explained by that single act of borrowing. The drafters had written Part IV’s grammar into a statute book where readers expect Part III’s consequences.

Drafting of this kind has a consequence that goes beyond any individual section, and it is the consequence the series’ thesis points at: a statute’s real effect was located in a phrase rather than in a provision. A qualifying formula that the obligated party may invoke for itself is not a standard against which performance is measured. It is a discretion dressed as a standard. No court can find, from the words “within the limits of their economic capacity and development”, what a particular state’s capacity in a particular year required it to build, staff or fund, because the words assign that judgment to the government whose conduct is in question. The judiciary can police the edges of such a formula, striking down a refusal that is mala fide or a policy that ignores the statute altogether, but it cannot convert the formula into a timetable. That is why the enforcement record of the 1995 Act divides so cleanly. Where the text fixed a number, courts enforced it, as the next sections show. Where the text supplied the formula, litigation produced exhortation rather than orders, and the Commissioners, armed with the statute’s own soft instruments, produced recommendations rather than directions.

The counterpoint inside the 1995 Act proves the rule by exception. Section 47, the bar on dispensing with an employee who acquires a disability during service, carries no economic capacity formula and no scheme. It is a prohibition in the imperative mood, and administrations obeyed it in a way they did not obey Section 46, because there was nothing in its text to plead. The reservation sections likewise carried no fiscal qualifier, which is why their drafting battles were fought over identification of posts and the base on which the percentage was computed rather than over whether the duty existed. Even Section 33, though, shows the second softening device at work: its proviso allowed the appropriate Government to exempt an establishment from the duty altogether. The 1995 Act therefore softened its obligations by two distinct techniques, a fiscal formula spread across the programmatic chapters and an exemption power attached to its hardest duty, and any account of the statute that notices only the seven disabilities has missed the machinery that decided what the statute did in the world.

Enforcement without an order

Chapter XII of the 1995 Act, Sections 57 to 65, created the Chief Commissioner for Persons with Disabilities at the Centre and a Commissioner in each state. The functions Section 58 assigned to the Chief Commissioner were to co-ordinate the work of the state Commissioners, to monitor the utilisation of funds disbursed by the Central Government, to take steps to safeguard the rights and facilities made available to persons with disabilities, and to submit reports to the Central Government. Sections 59 and 62 then dealt with complaints. The Commissioner could, of his own motion or on the application of an aggrieved person, look into complaints concerning deprivation of the rights of persons with disabilities, or the non-implementation of laws, rules and guidelines made for their welfare, and take up the matter with the appropriate authorities. Section 63 gave the Commissioner, for the discharge of those functions, the powers of a civil court: summoning and enforcing the attendance of witnesses, requiring the discovery and production of documents, requisitioning public records, receiving evidence on affidavits, and issuing commissions for the examination of witnesses and documents.

Read that machinery slowly and its shape becomes plain. The powers in Section 63 are powers to gather material: to summon, to demand documents, to record evidence. They are the powers a body needs in order to find facts. What follows the finding, in the 1995 design, is that the Commissioner takes up the matter with the appropriate authorities. The statute gave its enforcement officer a courtroom’s tools for inquiry and a correspondent’s tools for remedy. He could establish, on affidavit and on summoned records, that a department had failed to implement the Act, and the legal consequence of that establishment was a representation to the department that had failed. Sections 64 and 65 completed the design at the reporting level. The Chief Commissioner’s annual report was to be laid before Parliament, and the state Commissioner’s before the state legislature, in each case together with a memorandum explaining the action taken on the recommendations and the reasons for the non-acceptance, if any, of any recommendation. A reporting architecture that requires reasons for non-acceptance is candid about what it expects. It expects recommendations to be refused, and it contents itself with making the refusal visible to the legislature. Visibility is not nothing; parliamentary questions and press attention have moved departments on the strength of such reports. But a statute whose enforcement chapter culminates in the laying of a report has told the reader, in its own words, where coercion was meant to stop.

The offence provision confirms the reading from the other end. The 1995 Act contained exactly one penalty. Section 69 punished a person who fraudulently availed, or attempted to avail, any benefit meant for persons with disabilities, with imprisonment for a term which could extend to two years, or a fine which could extend to twenty thousand rupees, or both. The lone criminal provision of the disability statute was thus aimed at the claimant who took a benefit he was not entitled to, not at the government or establishment that failed to perform a duty it owed. There was no penalty for a department that never identified posts, never built the ramp, never framed the scheme, or exempted itself out of the reservation chapter. The compliance asymmetry was total: the statute policed fraud by beneficiaries and left non-performance by the state to reports and recommendations. When later commentary describes the 1995 Act as a welfare statute rather than a rights statute, this is the concrete content of the description: rights are claims that survive the duty-bearer’s unwillingness, and the 1995 Act supplied almost no machinery by which a claim could survive it.

What the courts could enforce, and what they could not

The courts supplied the control experiment. Given a statute in which some duties were numbered and hard and most were qualified and soft, the judiciary enforced the numbered duties with increasing strictness across the Act’s twenty years, and the contrast between the two bodies of case law is the empirical demonstration that the drafting, and not judicial reluctance, was the constraint. The leading line of authority concerns Section 33. In Government of India v. Ravi Prakash Gupta, decided in 2010 and reported at (2010) 7 SCC 626, a candidate with total blindness claimed appointment against the reservation in the civil services. The Union resisted on the ground, among others, that reservation could operate only in posts identified under Section 32. The Supreme Court rejected the sequencing argument: the obligation to reserve under Section 33 was not contingent on the executive first completing identification, and it recorded the stark chronology that although the Act had commenced in 1996, posts in the Indian Administrative Service were identified only in 2005 and reservations were effected from 2006. A decade of administrative delay could not be used to defeat a percentage the statute had fixed. The case is remembered for the candidate’s persistence, but its doctrinal weight lies in the refusal to let an administrative precondition, itself within the defaulting party’s control, dissolve a hard duty.

Section 47 generated the second line of enforcement, in the service jurisdiction of the High Courts and the Supreme Court. An employee who acquired a disability during service and was discharged, reduced in rank or denied adjustment could point to a prohibition in the imperative mood, and courts applied it as one: the establishment had either to retain the employee in the post or to shift him to another post with the same pay scale and service benefits, and an order that ignored the section could not stand. The provision’s litigation history is scattered across service matters rather than gathered in a single leading authority, but its pattern is consistent enough to state. Where the 1995 Act spoke in prohibitions and percentages, the person affected had a case. Where it spoke in programmes and fiscal pleas, he had, at best, a grievance suitable for the Commissioner’s report. Litigants sorted themselves accordingly, and the docket of disability law under the old statute is a map of its drafting: crowded at Sections 33, 39 and 47, empty at Sections 44, 45 and 46. The writ jurisdiction under Article 226 did the rest of the work that the statute’s own machinery would not. Because the Commissioners could not order, the real forum for an aggrieved person with a justiciable complaint was the High Court, and the real sanction for a defaulting department was not the Act’s nonexistent penalty but the court’s power to direct compliance with the hard provisions and to record the default in terms that carried administrative consequences. A statute whose enforcement depends on constitutional litigation for its ordinary operation has outsourced its sanctions to a forum most of its beneficiaries cannot reach, which is another way of stating what the deletion of the qualifiers, and the addition of offences and Special Courts, were intended to repair in 2016.

Three years later a three-judge bench decided Union of India v. National Federation of the Blind, reported at (2013) 10 SCC 772. The dispute concerned the base of the three per cent. The Union’s office memorandum of 29 December 2005 had computed reservation by reference to vacancies in identified posts, which shrank the number of reserved appointments to a fraction of what the percentage suggested. The Court held that the three per cent under Section 33 fell to be computed on the total number of vacancies in the cadre strength of each group of posts, Groups A, B, C and D alike, and that the exercise of identifying posts under Section 32 followed the reservation duty rather than preceding it as a condition. The clauses of the office memorandum that conflicted with that construction were struck down, and the Court directed the governments to compute the vacancies, complete identification within three months, and treat non-observance of its directions as disobedience for which the nodal officers could be proceeded against departmentally. The judgment is the high-water mark of enforcement under the 1995 Act, and its limits are as instructive as its force. Everything the Court did, it did by construing a number. It could fix the denominator of a fraction. It could not have written, from Section 46’s economic capacity formula, a schedule for making government buildings accessible, because the text gave it no fraction, no date and no standard to hold the government to.

The higher education cases run on the same track. In Disabled Rights Group v. Union of India, the Supreme Court required universities and colleges to give effect to the reservation of seats for students with disabilities, treating Section 39’s quantified duty as mandatory rather than aspirational and directing compliance in admissions. Once again the enforceable content of the 1995 Act turned out to be the content that was expressed in figures. Set beside this body of authority, the record on the qualified provisions is thin in a way that proves the drafting thesis: there is no comparable line of judgments ordering governments to complete the non-discrimination chapter, because there was nothing in that chapter a court could order with precision. Litigants and judges together thus mapped the statute’s real boundary. The 1995 Act was enforceable almost exactly to the extent that it had declined to qualify itself, and the lesson was available to the drafters of the replacement in the plainest possible form. Delete the qualifiers, fix the numbers and dates, and attach a penalty, and the same courts would have something to enforce everywhere rather than somewhere.

The twenty years in practice

Between commencement in February 1996 and repeal in April 2017, the 1995 Act had twenty-one years in which to work, and the shape of those years confirms the drafting analysis from the administrative side. The identification of posts, the hinge on which reservation turned, proceeded at the pace the statute’s silence permitted. The Supreme Court’s record in Ravi Prakash Gupta fixes the emblematic dates: the Act in force from 1996, posts in the Indian Administrative Service identified in 2005, reservation in consequence effected from 2006. A decade elapsed between the duty and its precondition because the text supplied no period within which identification had to be completed, and the same pattern, lists drawn late and revised later, repeated across governments and groups of posts. The office memorandum of 29 December 2005, by which the Union computed reservation on identified posts rather than on cadre strength, shows the second administrative habit the drafting invited: where the statute fixed a percentage but left the base to be settled by instruction, the instruction settled it narrowly, and it took a three-judge bench in 2013 to restore the wider base the percentage implied. Neither episode required bad faith to occur. Both required only that administrators, working a text that pleaded fiscal capacity on its own behalf and left its hardest duty dependent on a prior executive exercise, should resolve every ambiguity in the direction of the smaller obligation. Drafting that invites that resolution will receive it, from administrators of ordinary goodwill, as reliably as from hostile ones.

The Commissioners’ offices accumulated the other half of the record. Their annual reports, laid before Parliament and the state legislatures with the memoranda of action taken and the reasons for non-acceptance that Sections 64 and 65 contemplated, documented year after year the gap between the statute’s programme chapters and the departments’ performance: posts unidentified, vacancies unfilled and dereserved, buildings surveyed and left unaltered, schemes announced at scales the reports recorded without being able to enlarge. The reports mattered; they are the principal public archive of the old regime, and researchers of the period work from them. But an archive is not a sanction, and the Commissioners themselves were appointed by, housed within and funded through the governments whose implementation they monitored, a structural dependence the 1995 text did nothing to offset. The office depended in practice on the energy of its holder and the tolerance of the administration of the day, and its influence rose and fell on personalities in a way that rights machinery should not. By the time the Convention had been ratified and the drafting committee appointed, the verdict of the record was available to anyone who read the reports consecutively: the statute’s machinery could describe its own failure in detail, in documents laid before the legislature, without acquiring at any point the power to interrupt it. That is the administrative meaning of the proposition that the qualifiers were the statute. The failure of the 1995 Act was not hidden, episodic or contested. It was chronicled, in the statute’s own prescribed forms, by the statute’s own officers, for twenty years.

The case for the qualifiers, stated fairly

The qualifiers had a defence, and the defence deserves to be put at its strongest before it is answered, because the drafters of 1995 were not careless men and women writing empty words for amusement. The argument runs as follows. A legislature that imposes unqualified duties on a state with limited revenues performs an act of self-deception. Buildings cannot be retrofitted, staff cannot be trained, and transport fleets cannot be replaced by the force of a section number. If the statute demands everything at once, the predictable result is not universal compliance but universal non-compliance, which corrodes the authority of law generally and leaves the intended beneficiary with a paper right and a practical lesson in cynicism. A qualified duty, on this view, is a calibrated duty. It directs the state’s effort, ranks disability provision inside the budget process where it belongs, and allows each government to move at the pace its finances permit, honouring the obligation partially and progressively rather than dishonouring it outright. The international law of economic and social rights supplied a vocabulary for exactly this position. Article 2(1) of the International Covenant on Economic, Social and Cultural Rights obliges each state party to take steps to the maximum of its available resources with a view to achieving progressively the full realisation of the rights the Covenant recognises. Progressive realisation is not an evasion in that tradition. It is the recognised structure of obligations whose fulfilment is resource-dependent, and the 1995 formula can be read as its domestic cousin, written into a statute by drafters who understood that a poor country legislating for a large population had to sequence what it promised.

The defence gains colour from the statute’s origin. The 1995 Act implemented a proclamation, a soft instrument, and its drafters may fairly have thought that fidelity to the Proclamation meant adopting its register of goals and progressive targets. Nor should the counterfactual be romanticised. The India of 1995 had no accessibility standards in force for public buildings, a small cadre of professionals in rehabilitation, and state governments whose finances were under severe strain in the first years of economic reform. An unqualified command to make the built environment accessible, issued into that landscape with a penal clause attached, might have produced prosecutions of the hapless rather than ramps for the waiting. The strongest version of the case for the qualifiers is therefore not that they excused laziness. It is that they matched the promise to the plausible, and that a legislature’s first duty to a disadvantaged class is not to lie to it about what the state can do. The defence can add one more point, the strongest available to it. The formula was not invented by a draftsman looking for an escape. It is the Constitution’s own formula for this very subject, taken from Article 41, where it qualifies the State’s duty to make provision for public assistance in cases of disablement among others. A Parliament legislating in 1995 on a subject the Constitution itself had committed to the register of qualified, resource-dependent provision could plausibly say that it was following the constitutional template rather than evading a constitutional command, and that the courts, which had never treated Article 41 as a source of enforceable individual claims, would have read unqualified statutory duties in the same spirit whatever words were chosen. On that view the qualifiers were candour: the statute declined to promise in the grammar of rights what the constitutional scheme itself assigned to the grammar of policy.

The answer, borne out by the record the statute accumulated, is that the formula as drafted did none of the things the defence claims for a calibration. A calibration has a dial: it specifies the pace, names the milestones, and reserves the judgment of sufficiency to someone other than the party being judged. The 1995 formula had no dial. It set no milestones, required no plan, imposed no reporting duty tied to the qualifier itself, and left the plea of capacity to be raised by the very government whose inaction was in question, in answer to a complaint addressed to a Commissioner who could recommend and not order. What was drafted as progressive realisation functioned as self-judging discretion, and the empirical record shows the difference: the provisions that carried the formula are precisely the provisions for which two decades produced no enforceable jurisprudence, while the unqualified provisions of the same statute, in the same fiscally strained decades, produced Ravi Prakash Gupta and National Federation of the Blind. Fiscal strain did not prevent the enforcement of Section 33 once the Court had a number to enforce. It was the absence of a number, a date and a penalty that prevented enforcement elsewhere. The constitutional template argument fails on the same evidence, for a reason internal to the 1995 Act itself. Parliament was not transcribing Article 41; it was legislating, under Article 253, a statute, and where it wished to legislate in earnest it did so in the same document, in Sections 33, 39, 40 and 47, without the formula and without embarrassment. The drafters therefore knew the difference between the two registers and chose between them section by section. A formula chosen for the programme chapters and withheld from the quantified chapters cannot claim the compulsion of a constitutional template; it can only claim the preference of its authors, and the preference is what twenty years of non-enforcement judged. The verdict the evidence supports is therefore against the formula and not against the concern that motivated it. Resource limits are real, and honest statutes acknowledge them; the acknowledgement belongs in transitional periods, staged deadlines and budgetary machinery, where it can be reviewed, and not in a standing plea embedded in the operative sentence of every duty. The 2016 Act, as the sections below show, adopted exactly that relocation: duties owed to identifiable persons were stated without the fiscal plea, and the resource question was moved to deadlines, rule-making and the design of schemes. That relocation, more than the lengthening of the list, is the reform.

A convention replaces a proclamation

The instrument that displaced the Beijing Proclamation was of a different legal order. The United Nations General Assembly adopted the Convention on the Rights of Persons with Disabilities on 13 December 2006. The Convention is a treaty, and India was among the earliest states to engage with it: India signed the Convention on 30 March 2007 and ratified it on 1 October 2007, so that the obligations it contains became binding on India in international law, subject to the Convention’s own distinction between obligations of immediate application and the progressive realisation it prescribes for economic, social and cultural rights in Article 4(2). The same Convention moment reshaped neighbouring fields of Indian law, including mental health legislation, whose passage from a custodial statute to a rights-based one is traced in the account of the Mental Healthcare Act 2017. Ratification created the situation the 1995 Act could not survive. India had undertaken, in binding form, to guarantee equality and non-discrimination, to recognise the legal capacity of persons with disabilities on an equal basis with others, to ensure accessibility of the physical environment, transport and information, and to understand disability itself in the Convention’s terms, as arising from the interaction between persons with impairments and attitudinal and environmental barriers. Measured against those undertakings, the domestic statute was not merely weak. Its definitional chapter described a different thing, a medical list with a percentage, and its operative chapters qualified away the accessibility and participation duties the Convention stated as rights. Replacement, rather than amendment, followed from the depth of the mismatch: the 1995 Act did not need new clauses so much as a new theory, and a statute’s theory lives in its definitions, its chapter order and its enforcement design, none of which amendment repairs gracefully.

The drafting of the replacement took the better part of a decade, and the delay belongs in the record, because it qualifies any simple story of a rights revolution arriving on ratification. In 2009 the Ministry of Social Justice and Empowerment appointed a committee, chaired by Sudha Kaul, to draft a new legislation in harmony with the Convention, and the committee submitted a draft Bill in 2011 after consulting organisations of persons with disabilities. The draft was debated, criticised from within the disability movement for dilutions, and reworked, and a Bill was eventually introduced in the Rajya Sabha on 7 February 2014 and examined by the departmental Standing Committee on Social Justice and Empowerment, which reported in 2015. Through the Bill’s long passage, organisations of persons with disabilities circulated clause-by-clause comparisons of the proposed text against the 1995 Act, and readers who work through those comparison decks can open them in a free online PowerPoint viewer to follow the substitutions line by line. The Rajya Sabha passed the Bill on 14 December 2016 and the Lok Sabha on 16 December 2016, and the President’s assent followed on 27 December 2016. The finished statute, the Rights of Persons with Disabilities Act 2016, is the subject of its own pillar treatment at the RPwD Act 2016 explained, which states its provisions in full; the concern here is narrower, being what its enactment did to the statute it repealed, provision by provision and habit by habit.

The committee method deserves a sentence of its own, because it marked a change in who drafts Indian disability law. The committee appointed in 2009 worked by consulting organisations of persons with disabilities, and its draft was written against the Convention article by article, so that the resulting text argued with the treaty as it went: legal capacity answered to the Convention’s Article 12, accessibility to its Article 9, the barriers definition to its opening articles. The consultations also set the terms of the arguments that followed the draft through its long passage. Organisations within the disability movement criticised successive versions for diluting the committee’s positions, on the treatment of legal capacity, on the breadth of the Schedule and on the strength of the enforcement machinery, and the Bill that reached Parliament in 2014 was a negotiated text rather than the committee’s draft enacted whole. That history explains features of the finished Act that a reader of the text alone might find puzzling: a rights chapter of genuine strength standing beside a recommendatory enforcement core, and a Schedule of twenty-one conditions that grew during the Bill’s passage as groups whose conditions had been omitted pressed, successfully, for inclusion. The expansion of the list, the change for which the Act is popularly known, was in this way the product of the same consultative politics that produced the deletion of the qualifiers, though the two changes worked on different parts of the machine: one widened the class of persons the statute saw, and the other changed what the statute’s seeing was worth.

When did the RPwD Act come into force?

The Rights of Persons with Disabilities Act 2016 received the President’s assent on 27 December 2016 and was published in the Gazette the following day, but it commenced on 19 April 2017, the date appointed by notification S.O. 1215(E). Commencement is the operative date: the 1995 Act stood repealed from that day, and Section 102 governs everything done before it.

The preamble of the 2016 Act announces the change of foundation in terms. Where the older statute’s long title pointed to a regional proclamation, the new long title states that the Act is one to give effect to the United Nations Convention on the Rights of Persons with Disabilities, and the preamble recites the Convention’s principles: respect for inherent dignity, individual autonomy and freedom of choice, non-discrimination, full and effective participation and inclusion in society, respect for difference and acceptance of persons with disabilities as part of human diversity, equality of opportunity, accessibility, equality between men and women, and respect for the evolving capacities of children with disabilities. A preamble is not an operative provision, and no claim in this article rests on the preamble alone. Its significance is diagnostic. The drafters of 2016 placed the Convention’s principles where the drafters of 1995 had placed a decade-long programme of regional action, and the operative chapters then did what the preamble promised, in the three distinct moves that any repeal and replacement performs: some things were kept, some were discarded, and some were created new. The moves must be separated, because commentary on this transition habitually blurs them, crediting the new Act with novelties it inherited and blaming the old Act for weaknesses the new one retained.

What the 2016 Act kept

Replacement statutes are rarely as new as their preambles, and the 2016 Act kept more of its predecessor than the rhetoric of replacement suggests. The forty per cent threshold survived intact. Section 2(r) of the new Act defines a person with benchmark disability as a person with not less than forty per cent of a specified disability, as certified by the certifying authority, and the benchmark is the gateway to the distributive provisions: free education for children under Section 31, reserved seats in higher education under Section 32, reserved posts under Section 34, and the special schemes under Section 37. The continuity was deliberate and defensible. A threshold for benefits must be administrable, and percentage assessment by medical authorities, for all its crudeness, was machinery that existed in every district. What changed was the threshold’s jurisdiction. Under the 1995 Act the forty per cent line defined who counted as a person with disability for the statute generally. Under the 2016 Act the line governs benefits, while the rights chapters speak to the wider class defined in Section 2(s), persons whose long term impairment in interaction with barriers hinders full and effective participation in society equally with others. The number was kept; its empire was reduced.

The reservation architecture likewise survived, enlarged rather than redesigned. The 2016 Act kept the sequence the courts had litigated under the old law: identification of posts by the appropriate Government, in the new Act expressly through an expert committee with representation of persons with benchmark disabilities and subject to periodic review at intervals not exceeding three years under Section 33; a fixed percentage of vacancies, raised from three to four; allocation of the percentage among disability categories, widened from three groups to five; carry-forward of unfilled vacancies and interchange among categories under Section 34(2), carried over from Section 36 of the old Act; and an exemption proviso, retained in Section 34 itself, under which the appropriate Government may, in consultation with the Chief Commissioner or the State Commissioner and having regard to the type of work, exempt a Government establishment from the section. One survival runs against the current of the reform and should be recorded with the rest. The old Act’s quantified scheme reservations, the three per cent in all poverty alleviation schemes under Section 40 and the preferential allotment of land at concessional rates under Section 43, reappear in Section 37 of the new Act, raised to five per cent and gathered into a single provision: five per cent reservation in the allotment of agricultural land and housing in all relevant schemes and development programmes, five per cent in all poverty alleviation and developmental schemes, and five per cent in the allotment of land at concessional rates for housing, shelter, occupation, business, recreation and production centres, with priority to women with benchmark disabilities in the first two. The percentages rose, but the vehicle changed in the other direction: where the old sections stated reservations directly, Section 37 obliges the appropriate Government and local authorities to make schemes by notification providing for them, routing the entitlement through scheme-making of the kind the old Act used for its softer promises. The drafters evidently regarded land and scheme reservations as programme content even at a fixed percentage, a judgment consistent with their treatment of the fiscal formula in the same territory, and one that leaves Section 37’s five per cent dependent for its bite on the schemes actually notified under it. Continuity, in this corner of the Act, came with a small demotion of form attached to a rise in the figure, and a reader tracing what was kept should keep both halves of that bargain in view.

The Commissioners survived too, reconstituted in Chapter XII with enlarged functions, and the coordination idea survived in altered dress: the Central and State Coordination Committees of 1995, executive bodies charged with reviewing and coordinating departments, gave way to the Central and State Advisory Boards on Disability in Chapter XI, consultative bodies of wider composition, with District-Level Committees beneath them. The Special Employment Exchanges of the old Section 34 reappear in the new Section 36. Even the protection for the employee who acquires a disability during service was kept, and strengthened: Section 20(4) of the new Act repeats the 1995 prohibition on dispensing with or reducing in rank, and adds that an employee who cannot be adjusted against any post may be kept on a supernumerary post until a suitable post is available or superannuation arrives, a second proviso the old Section 47 did not contain. Continuity of this order is not failure of nerve. It is how a replacement statute keeps twenty years of administrative practice, and the precedent built on it, usable the morning after commencement.

What the 2016 Act discarded

Against that continuity, the discards were surgical, and the first of them is the article’s subject. The economic capacity formula was deleted from every provision in which the old Act had used it to condition a duty owed in respect of the built environment, transport and the roads. Sections 44, 45 and 46 of the 1995 Act, the non-discrimination chapter, obliged governments within the limits of their economic capacity and development to take special measures for transport facilities, to make roads accessible, and to adapt the built environment. Chapter VIII of the 2016 Act contains no such plea. In its place stand machinery and dates. Section 40 obliges the Central Government, in consultation with the Chief Commissioner, to formulate rules laying down standards of accessibility for the physical environment, transportation, information and communication, and other facilities and services provided to the public. Section 44 then makes observance of the accessibility rules a condition of building permission and of the certificate that permits occupation, so that a non-compliant new building fails at the gate rather than awaiting a recommendation. Section 45 requires all existing public buildings to be made accessible in accordance with the rules within a period not exceeding five years from the notification of those rules, and Section 46 requires service providers, government or private, to bring their services into conformity within two years from the same notification. A deadline that runs from the notification of rules is not beyond criticism, since the executive controls the starting gun, and the honest assessment below returns to that. But the drafting difference is fundamental: a date can be missed, and missing it is a breach that a court or a Commissioner can name, whereas a fiscal plea cannot be missed at all, because it defines performance down to whatever was done.

The second discard was the recommend-only conception of enforcement at the level of principle, though here the replacement was a matter of degree rather than of kind, and candour requires the qualification to be stated in the same breath. The 2016 Commissioners, like their predecessors, inquire and recommend; Section 75 arms the Chief Commissioner to identify provisions of any law or policy inconsistent with the Act and recommend corrective steps, to inquire on his own motion into deprivation of rights, and to take up the matter with the appropriate authorities for corrective action, language that descends from Sections 58 and 59 of the old Act. What the new statute added around that recommendatory core is examined with the additions below: a penal chapter behind the duties, Special Courts to try offences, and, by an amendment effective in January 2018, a duty on an authority that receives a Commissioner’s recommendation on a complaint to act on it and report the action taken within three months, or to convey its reasons for non-acceptance to the Commissioner and to the aggrieved person within the same period. The third discard was conceptual. The 1995 Act’s closed medical list ceased to be the definition of the statute’s subject. The list survived, enlarged, as the Schedule of specified disabilities, but in the new Act it feeds the benchmark definition and the certification chapter; the statute’s general subject is defined by Section 2(s) in the Convention’s interaction terms, and Section 2(c) defines a barrier to include communicational, cultural, economic, environmental, institutional, political, social, attitudinal and structural factors. A definition that names attitudes as a statutory obstacle is doing different work from a list of diagnoses, and the discarding of the old definitional monopoly is what allowed the new Act to regulate examiners, employers, broadcasters and builders as well as hospitals and welfare departments.

What the 2016 Act added

The additions begin with a chapter the old statute never attempted. Chapter II of the 2016 Act states rights as claims of the person: equality and non-discrimination in Section 3, under which no person with disability shall be discriminated against on the ground of disability unless the act or omission is shown to be a proportionate means of achieving a legitimate aim, and discrimination is defined in Section 2(h) to include denial of reasonable accommodation, so that a refusal to accommodate is itself discrimination rather than a mere absence of generosity. The remaining clauses of Section 3 complete the provision’s design: sub-section (4) bars the deprivation of personal liberty only on the ground of disability, and sub-section (5) obliges the appropriate Government to take necessary steps to ensure reasonable accommodation for persons with disabilities, converting the concept from a definition into a governmental duty. Reasonable accommodation itself is defined in Section 2(y) as necessary and appropriate modification and adjustments not imposing a disproportionate or undue burden, where needed in a particular case, to ensure that persons with disabilities enjoy rights equally with others; the burden qualification inside the definition is examined in the assessment below, since it is the one point at which the new Act’s rights chapter carries a limiting formula of its own, tied there to the circumstances of the individual case rather than to the resources of the duty-bearer at large. Sections 4 to 12 add the rights of women and children with disabilities, community life, protection from cruelty, inhuman treatment, abuse, violence and exploitation, protection in situations of risk, home and family, reproductive rights, accessibility in voting, and access to justice. Section 13 then states the provision that most sharply separates the two statutes: the appropriate Government shall ensure that persons with disabilities enjoy legal capacity on an equal basis with others in all aspects of life, with the right to own or inherit property, movable or immovable, to control financial affairs, and to have access to bank loans, mortgages and other forms of financial credit. The 1995 Act contained no legal capacity provision at all. A statute that distributed benefits to persons with disabilities while saying nothing about their capacity to hold property, contract or decide was a statute whose subject was an object of provision; Section 13 is the textual point at which the subject becomes a bearer of rights.

Section 14 carries the same thought into guardianship. Where a district court or a designated authority finds that a person with disability, despite adequate and appropriate support, is unable to take legally binding decisions, a limited guardian may be provided to take such decisions on his behalf in consultation with him. The Explanation defines limited guardianship as a system of joint decision which operates on mutual understanding and trust between the guardian and the person with disability, limited to a specific period and for specific decisions and situations, and operating in accordance with the will of the person with disability. The contrast is with the plenary model, under which a guardian appointed for a person with mental illness or intellectual disability under the older dispensation took decisions in the person’s place and without the statutory qualifications of period, specificity and consultation. The 2016 Act also reached backwards to convert the old appointments: Section 14(2) provides that every guardian appointed under any other law for a person with disability shall, from the commencement of the new Act, be deemed to function as a limited guardian, a transitional device of some elegance that recast existing arrangements without unseating the families who relied on them.

The remaining additions are the machinery of consequence. Chapter XVI creates offences where the old Act had one: Section 89 punishes contravention of any provision of the Act or the rules with a fine up to ten thousand rupees for a first contravention and, for later contraventions, a fine of not less than fifty thousand rupees extending to five lakh rupees; Section 92 punishes atrocities against persons with disabilities, including intentional insult or intimidation meant to humiliate in public view, assault, denial of food or fluids by a person in charge, sexual exploitation, injury to a limb, sense organ or supporting device, and procedures terminating a pregnancy without consent, with imprisonment of not less than six months extending to five years and a fine; Section 91 reprices the old Section 69 fraud offence at imprisonment up to two years or a fine up to one lakh rupees; Section 93 fines failures to furnish information. Section 84 requires a Court of Session in each district to be specified as a Special Court for the speedy trial of offences under the Act, with Special Public Prosecutors under Section 85. Sections 86 and 88 create a National Fund and State Funds for persons with disabilities, the National Fund absorbing by name the two earlier funds of 1983 and 2006, so that money accumulated under the old dispensation was carried into the new. Chapter X rebuilds certification: assessment guidelines under Section 56, designated certifying authorities under Section 57, a procedure under Section 58 with an appeal under Section 59, and a sentence in Section 58(3) of practical importance out of all proportion to its length, providing that a certificate of disability issued under the section shall be valid across the country. How these additions work in the lived fields of hiring and building is the province of the treatment of reservation and accessibility under the RPwD Act, which takes the reservation percentages and the accessibility deadlines into workplaces, campuses and public buildings; the comparison here is complete once it is seen that the 2016 Act did not merely lengthen a list but installed, behind each class of duty, either a date, a gate, a complaint route ending in a court, or a fine.

The additions in the service chapters are quieter than the rights chapter and touch more lives. Chapter III recasts education. Section 16 obliges the appropriate Government and local authorities to endeavour that funded or recognised institutions provide inclusive education, and then specifies what inclusion consists in: admission without discrimination, accessible buildings, campuses and facilities, reasonable accommodation according to individual requirements, individualised support in environments that maximise academic and social development, education of blind and deaf children in appropriate languages and modes of communication, early detection of specific learning disabilities, monitoring of every student’s participation and attainment, and transport for children with disabilities and attendants for those with high support needs. Section 17 adds the supporting measures, including a survey of school going children every five years to identify children with disabilities and the extent to which their needs are met, teacher training institutions, teachers qualified in sign language and Braille and in teaching children with intellectual disability, resource centres, books and assistive devices free of cost to students with benchmark disabilities up to the age of eighteen, and modifications to curriculum and examinations, naming extra time and the facility of a scribe. Chapter IV does the same work for employment beyond the reservation clause: schemes for vocational training and self-employment with loans at concessional rates under Section 19, the non-discrimination code of Section 20, an equal opportunity policy that every establishment must notify and register with the Chief Commissioner or the State Commissioner under Section 21, records under Section 22, and a Grievance Redressal Officer in every Government establishment under Section 23, bound to inquire into a complaint within two weeks of its registration, with an appeal from his handling of it to the District-Level Committee. None of these provisions has the drama of Section 13, and each of them converts a relationship that the 1995 Act left to goodwill (the school and the child, the employer and the employee) into a documented, registered or time-bound transaction that a third party can inspect.

Chapter V and the chapters around it widen the statute’s field in the other direction, into the texture of ordinary life. Section 24(3) settles the contents of social security schemes: community centres, facilities for persons without family or shelter, support during natural and man-made disasters, support to women with disability for livelihood and for the upbringing of their children, access to safe drinking water and sanitation, aids and appliances and corrective surgery for those within an income ceiling, disability pension, unemployment allowance for persons registered with a Special Employment Exchange for more than two years, a care-giver allowance for persons with high support needs, and insurance cover for those outside the employees’ insurance schemes. The unemployment allowance that the 1995 Act had carried in Section 68 thus reappears, folded into the scheme chapter; what changed around it is the proviso in Section 24(1) fixing assistance at least twenty-five per cent above comparable general schemes, discussed with the survivals below. Section 26 obliges governments to make insurance schemes for their own employees with disabilities by notification. Sections 29 and 30 take the statute into culture, recreation and sport, directing measures that include access to television programmes with sign language interpretation or sub-titles and the redesign of sporting infrastructure. Section 38 creates the high support jurisdiction, under which a person with benchmark disability who considers himself in need of high support may apply to a notified authority, whose Assessment Board certifies the need and its nature. Section 39 obliges the appropriate Government to conduct awareness campaigns and sensitisation programmes, and specifies their contents down to the inclusion of the rights of persons with disabilities in the curricula of universities, colleges and schools. Chapter IX carries the registration of institutions for persons with disabilities into the new Act in place of the old recognition regime, and its Section 55, examined below, is one of the four places where the old fiscal formula survived the replacement.

What counts as a barrier under the 2016 Act?

Section 2(c) defines a barrier as any factor, including communicational, cultural, economic, environmental, institutional, political, social, attitudinal or structural factors, which hampers the full and effective participation of persons with disabilities in society. A staircase is a barrier, and so are an examination format that admits no scribe and a website that no screen reader can parse.

The four places the formula survived

An honest account of the replacement must also record what the celebration of deletion omits. The economic capacity formula was not struck from the statute book in 2016. It was struck from the rights, and it survives, in the new Act, in four provisions, each of which rewards attention. Section 24(1) provides that the appropriate Government shall, within the limit of its economic capacity and development, formulate necessary schemes and programmes to safeguard and promote the right of persons with disabilities to an adequate standard of living, and its proviso adds that the quantum of assistance under such schemes shall be at least twenty-five per cent higher than that under similar schemes applicable to others. Section 27(1) directs the appropriate Government and local authorities, within their economic capacity and development, to undertake services and programmes of rehabilitation in health, education and employment. Section 35 repeats the old Section 41 almost word for word: the appropriate Government and local authorities shall, within the limit of their economic capacity and development, provide incentives to employers in the private sector to ensure that at least five per cent of their workforce is composed of persons with benchmark disabilities. Section 55 permits the appropriate Government, within the limits of their economic capacity and development, to grant financial assistance to registered institutions. Social security schemes, rehabilitation programmes, employer incentives and institutional assistance: the survivors are all provisions under which the state designs and funds programmes, rather than provisions under which an identifiable person holds a claim against an identifiable duty-bearer.

The pattern of survival is not an oversight, and it maps with some precision onto the Convention’s own architecture. Article 4(2) of the Convention accepts progressive realisation, to the maximum of available resources, for economic, social and cultural rights, while the obligations of non-discrimination, equal recognition before the law and accessibility are not framed as progressive in the same way. The 2016 drafters, consciously or by sound instinct, tracked that division. The fiscal plea was withdrawn from the territory of claims (no section of Chapter II, no accessibility deadline, no reservation percentage is conditioned on capacity) and retained in the territory of programme design, where a government must in fact decide how much a pension scheme, an incentive scheme or a grant programme can carry. Even there, the survival is not verbatim. Section 24’s proviso bolts a comparative floor to the formula, so that a government pleading capacity must still set disability assistance at least a quarter above the comparable general scheme, which converts the plea from an answer into a benchmark against which the plea itself can be tested. Section 55 couples the formula with the permissive “may”, the softest drafting in either statute, and confines it to grants to institutions. The correct statement of the 2016 reform is therefore narrower and stronger than the slogan. The reform was not the abolition of fiscal qualification. It was the confinement of fiscal qualification to the provisions where qualification is a feature of the subject matter, and its expulsion from the provisions where it had functioned as a repeal of the duty by the duty-bearer. Readers who repeat the slogan without the confinement will misdescribe four sections; readers who notice only the four sections will miss that the non-discrimination chapter, the rights chapter and the accessibility deadlines were where the formula had done its damage, and that from all of them it is gone.

The threshold’s strange career

The forty per cent figure is the oldest continuous element in Indian disability law’s modern history, and its career across the two statutes is the cleanest illustration of how a replacement can keep a device and transform it by rehousing. In 1995 the figure sat in Section 2(t) as part of the definition of the statute’s subject, certified by a medical authority whose assessment of a percentage of blindness, locomotor disability or mental illness was necessarily an act of clinical judgment dressed as measurement. The certificate that resulted was the master key of the old regime: without it, the reservation chapters, the scheme chapters and the Commissioners’ solicitude were all, in practice, addressed to somebody else. In 2016 the figure moved to Section 2(r) and changed employers. It there defines the benchmark class for the distributive provisions, while the Act’s rights run to the wider class of Section 2(s), and the certification that supports it was rebuilt as a chapter of its own: assessment guidelines notified by the Central Government under Section 56, designated certifying authorities under Section 57, a prescribed procedure under Section 58 with a duty on the authority either to issue the certificate or to inform the applicant in writing that he has no specified disability, an appeal to a designated appellate authority under Section 59, and the portability rule of Section 58(3). The old Act had no certification chapter, leaving the gateway document of its entire scheme to be produced under administrative arrangements it did not describe; the new Act describes the gateway, times it by procedure, makes its refusal reasoned and appealable, and sends it across state lines. The percentage survived because distributive law needs a gate. What the replacement abolished was the gate’s position at the entrance to the whole statute.

The conceptual tension inside the device should be stated, since the article’s stance requires the strongest version of each side. The objection is real: a statute that defines disability as the interaction between an impairment and a barrier, and then rations its benefits by a percentage of impairment, measures the person by exactly the quantity its own definition says is not the disability. A person assessed at thirty-five per cent whose barriers are total is, on the definition, disabled in the full sense, and on the threshold, entitled to none of the distributive provisions. The defence is also real, and it is administrative rather than conceptual: benefits must be rationed by a criterion that a district authority can apply uniformly and an applicant can verify, and percentage assessment against notified guidelines, for all its crudeness at the margins, is such a criterion, whereas a barriers assessment for every claimant would reproduce, inside the benefits system, the discretion and inconsistency that certification exists to exclude. The 2016 Act’s settlement, rights for the wide class and benefits for the certified class, with the certificate made portable, reasoned and appealable, is the drafters’ answer to that tension, and it is a defensible one; the friction it leaves is not a drafting failure but the price of running a rights statute and a rationing system inside one set of covers, which is the condition of every modern welfare state that has made the same transition.

Repeal and savings: the morning after

Section 102 of the 2016 Act does the whole of the repeal work in two sub-sections, and its economy is worth remarking in a statute of one hundred and two sections. Sub-section (1) provides that the Persons with Disabilities (Equal Opportunity Protection of Rights and Full Participation) Act, 1995 (1 of 1996) is hereby repealed. Sub-section (2) provides that notwithstanding the repeal of the said Act, anything done or any action taken under the said Act shall be deemed to have been done or taken under the corresponding provisions of this Act. The deeming is the operative device. Parliament did not transfer certificates, appointments, admissions, registrations and pending proceedings item by item. It enacted a fiction under which the entire administrative past of the old statute is read as the past of the new one, each act done under a provision of 1995 being treated as done under the corresponding provision of 2016. The general savings doctrine of the General Clauses Act 1897 stands behind the clause for accrued rights, incurred liabilities and pending legal proceedings, but Section 102(2) does the practical work by itself, because most of what the old Act produced was not litigation but paper: certificates, rosters, identified-post lists, registrations of institutions, appointments to reserved posts, admissions to reserved seats, and complaints before Commissioners.

The subordinate legislation crossed by a related route. Section 24 of the General Clauses Act provides that where an enactment is repealed and re-enacted, rules, notifications, orders and appointments made under the repealed enactment continue in force, so far as they are not inconsistent with the re-enacted provisions, and are deemed to have been made under them. The rules framed under the 1995 Act, the notifications identifying posts, and the scheme instruments issued under its chapters therefore did not fall with the Act: they continued, to the extent the 2016 Act’s corresponding provisions could carry them, until replaced by instruments under the new statute. The qualification “so far as not inconsistent” did real work in this transition, because some inherited instruments embodied the old Act’s narrower readings: a roster drawn on the identified-posts computation that National Federation of the Blind had condemned could not survive into a Section 34 regime whose text fixes the cadre strength base, and an exemption notification issued under the old proviso could not govern a section whose own proviso has different conditions. Administrators who worked the transition therefore had to audit their inherited paper against the new text instrument by instrument, retaining what the deeming carried and reissuing what it could not. That unglamorous audit, conducted in departments across the country in the months after April 2017, is where a repeal and replacement is actually executed, and its quality determined, district by district, whether the transition felt to the person holding a certificate or a roster position like a change of statute or like a change of nothing at all.

The deeming has one edge that administrators had to work out without textual help, and it is worth recording because it recurs in every repeal and replacement. Section 102(2) speaks of acts done under the repealed Act being deemed done under the corresponding provisions of the new one, and for most of the old Act’s output a corresponding provision exists: certification answers to Chapter X, reservation to Section 34, registration of institutions to Chapter IX. For some of the output, no true counterpart exists. The Central and State Coordination Committees were executive bodies with functions of review and direction across departments; the Advisory Boards that succeeded them are consultative, differently composed and differently empowered, so a direction issued by a coordination committee in the old Act’s last months could not simply be re-badged as an act of an advisory board that the new text does not arm to issue it. Acts of that character rest on the General Clauses savings for things duly done and on the ordinary continuity of government, rather than on the deeming fiction, and they bind as administrative history rather than as exercises of a continuing statutory power. The distinction mattered for a small residue of files, and it illustrates the general grammar of transitions: a savings clause carries acts into the new regime only as far as the new regime has a shelf of the right shape to receive them, and the draftsman who wants a seamless passage must either build the shelves in the new text, as the 2016 Act did for certificates and guardianships, or accept that the residue travels on general law.

The repeal’s precision about its target is its final lesson. Section 102 repealed the 1995 Act and nothing else, and the cluster of statutes around it stood untouched. The National Trust for the Welfare of Persons with Autism, Cerebral Palsy, Mental Retardation and Multiple Disabilities Act 1999, which had supplied the guardianship machinery for four conditions the 1995 Act recognised, continued in force, its local level committees and its registered organisations undisturbed; the new Act’s Section 60 goes so far as to seat the National Trust’s chairperson on the Central Advisory Board as a member by virtue of office, an arrangement possible only because the Trust and its parent statute survived. The Rehabilitation Council of India Act 1992, governing the training and registration of rehabilitation professionals, likewise continued. What the interaction of the survivors produced was a layered field: the 2016 Act as the general rights and benefits statute, the National Trust Act running its guardianship and welfare machinery for its four conditions in harness with the limited guardianship that Section 14(2) deemed every existing guardian to exercise, and the professional regulator intact. A repeal that had swept the cluster away would have destroyed working machinery for the sake of tidiness. The drafters’ restraint, repealing the one statute whose theory had failed and converting rather than abolishing the arrangements built under its neighbours, is the transition’s least visible and most administrative virtue.

The certificate is the case that matters to the largest number of people, and its answer is clean. A disability certificate issued by a medical authority under the 1995 Act was an action taken under that Act. On the repeal, Section 102(2) deems it to have been taken under the corresponding provisions of the 2016 Act, which are the certification provisions of Chapter X. The certificate therefore did not lapse on 19 April 2017, and its holder was not required to be reassessed merely because the statute under which the certificate issued had been repealed. The same reasoning carries appointments made against the three per cent reservation, admissions taken against reserved seats, and registrations of institutions granted under Chapter X of the old Act: each was an act done under the repealed statute and is deemed done under the corresponding provisions of its successor. Proceedings pending before a Chief Commissioner or a State Commissioner under the old Chapter XII at commencement likewise continue, the offices themselves having been reconstituted by Chapter XII of the new Act with the same names and enlarged functions, so that the complaint file crossed the repeal without a change of forum. Section 14(2), deeming every guardian appointed under any other law to function as a limited guardian, works the same passage from the other direction, recasting a status created elsewhere rather than an act done under the repealed Act itself. This account is legal information about two statutes and their transition, not advice on any individual certificate, appointment or proceeding, and a holder whose certificate is questioned by an authority should put Section 102(2) to that authority in writing, since the deeming answers the objection in terms.

One limit of the savings clause deserves emphasis, because it marks the boundary between what repeal preserves and what it cannot. Savings preserve acts done and rights accrued; they do not preserve the repealed statute’s standards as the measure of future conduct. A building that was lawful under the qualified regime of Section 46 of the old Act did not acquire a vested right to remain inaccessible, because accessibility under Chapter VIII of the new Act is a duty about the future state of the building, with its own period running from the notification of the accessibility rules. An establishment exempted from reservation by notification under the old Section 33 held an exemption under a repealed power; reservation after commencement falls to be administered under Section 34 of the new Act, whose exemption proviso is its own and must be exercised afresh, in consultation with the Commissioner, if it is to operate. The transition, in short, carried over everything the old Act had completed and nothing of its permissions to leave things undone. That asymmetry is the correct reading of a deeming clause attached to a replacement whose purpose was to raise the standard: the past is protected precisely so that the future need not be.

Which ministry administers the two disability statutes?

The Ministry of Social Justice and Empowerment administers both statutes, through the Department of Empowerment of Persons with Disabilities. The department that ran the coordination committees and Commissioners under the 1995 Act framed the rules, assessment guidelines and accessibility standards under the 2016 Act, so the old regime’s institutional memory sits inside the new machinery.

The precedent that crossed over

A repeal kills a statute; it does not kill the case law that construed it, and the interesting question for a replacement is which parts of the old jurisprudence the new text adopts, which it reverses, and which it leaves standing as persuasive authority on provisions that were re-enacted. The 2016 Act answers the first two of those questions inside its own text, which is rarer and better than leaving them to inference. National Federation of the Blind had held, construing Section 33 of the old Act against the executive’s office memorandum, that the three per cent fell to be computed on the total number of vacancies in the cadre strength. Section 34(1) of the new Act writes that construction into the statute: reservation is of not less than four per cent of the total number of vacancies in the cadre strength in each group of posts. The judicial reading of the repealed clause thus became the enacted text of its successor, and the point the Union had litigated and lost was removed from the field of argument altogether. Ravi Prakash Gupta had held that reservation could not be made contingent on the executive first identifying posts, and that identification under Section 32 followed the duty under Section 33 rather than conditioning it. The new Act keeps the sequence in that order as a matter of structure: Section 33 imposes identification, with an expert committee and a three-yearly review, as a duty of the appropriate Government, and nothing in Section 34 makes the four per cent wait upon its completion. The two judgments were delivered under a statute that no longer exists, and their propositions govern its replacement because the replacement was drafted around them.

The principle that organises this crossover is a standard one, and stating it protects readers from two opposite errors. Where a legislature re-enacts words that have received a settled judicial construction, the construction is taken to travel with the words, because Parliament legislates against the background of the courts’ reading. Where the new text changes the words, the old construction falls with them: the three-judge bench’s reading of the old exemption and identification disputes, for instance, binds only as far as the new provisions correspond. Under the 2016 Act the corresponding provisions were drafted to codify the outcomes, so the practical effect is continuity of the reservation jurisprudence with the percentages raised and the categories widened. What did not cross over is equally telling. The old Act’s thin case law on the qualified chapters had nothing to codify, there being no enforceable proposition in it worth re-enacting, and the new Act’s accessibility and non-discrimination provisions therefore begin their judicial life afresh, armed with text the courts of the old regime never had: definitions of discrimination and barrier, deadlines, and a penal clause. The precedent map of the transition is thus the drafting map once more. Where the old statute had numbers, it generated law that survived it. Where it had the formula, it generated silence, and silence is not inherited.

An honest assessment of the replacement

The barriers model carries a cost that its advocates should state as plainly as its merits, because the 2016 Act’s own definitions create the difficulty. Once disability is defined as the product of an impairment and a barrier, and a barrier is defined to include attitudinal and structural factors, the class of duty-bearers explodes. The 1995 Act spoke almost entirely to governments and their establishments. The 2016 Act speaks to every establishment that must notify an equal opportunity policy under Section 21 and register it with the Commissioner, to every service provider, government or private, caught by the two-year conformity period in Section 46, to every builder whose plans must pass the accessibility gate in Section 44, to examination bodies, broadcasters and employers across the private economy. Most of these actors are not hostile to the duties. They are unaware of them, or aware of them in outline and unsure what compliance consists in, and the statute’s enforcement design, which still runs in large part through Commissioners who recommend and through prosecution of offences that requires the machinery of criminal process, was not built to educate several million duty-bearers at once. This is a real implementation problem. It is not an objection to the model, which correctly locates the disadvantage in the environment rather than in the person, and the alternative of returning to a medical list administered through welfare departments would sacrifice the model’s truth to its convenience. But a reader comparing the statutes should register that the 2016 Act traded the old failure mode, duties that dissolved on pleading, for a new one, duties that bind multitudes who have never read them, and that the second failure mode is cured by administration and publicity rather than by drafting, which is slower and less certain work.

A further residue sits inside the accessibility chapter itself, and distinguishing it from the old formula sharpens the article’s claim. Section 41, on access to transport, directs the appropriate Government to provide access to modes of transport conforming to the design standards, including retrofitting old modes of transport, wherever technically feasible and safe for persons with disabilities, economically viable and without entailing major structural changes in design. Those are qualifying words, and a hostile reader might set them beside the deleted formula and declare the reform a relabelling. The distinction is worth the paragraph it takes to state. The fiscal formula of 1995 qualified the duty by reference to the duty-bearer’s own assessment of its capacity, a judgment no outsider could test, and it qualified every duty it touched in the same way. The feasibility words of Section 41 qualify a specific engineering operation by reference to facts about vehicles and structures (whether a retrofit is technically feasible, safe and viable without major structural change) that an expert can examine and a court can test against evidence. A qualifier tied to the physical character of the task is a specification of the duty. A qualifier tied to the resources of the party performing it is a discretion. The 2016 Act kept the first kind and abolished the second in the accessibility field, and keeping the distinction in view prevents the lazy conclusion, heard whenever the new Act is criticised, that all qualification is evasion. The vice of the 1995 formula was never that it acknowledged limits. It was that it made the obligated party the judge of them, in advance, in every case, forever.

Residues of the old habits also survive inside the new drafting, and the comparison is incomplete without them. Section 16, the inclusive education provision, obliges the appropriate Government and local authorities to endeavour that all funded or recognised educational institutions provide inclusive education, and “endeavour” is a soft verb of the same family as the old devices, though the harder provisions around it, the right to free education in Section 31 and the reserved seats in Section 32, limit the damage. The exemption power survived in both Section 20(1), for non-discrimination in employment, and Section 34, for reservation, so the door that stood in the old Section 33 stands in the new Act too, narrowed by the requirement of consultation with the Commissioner, by the confinement of the reservation exemption to Government establishments, and by the condition that it rest on the type of work carried on, but it remains a door, whose use will bear watching exactly as its predecessor’s did. The accessibility deadlines, for their part, run from the notification of rules under Section 40, which places the starting gun in the hand of the executive whose performance the deadlines measure; a government slow to notify rules postpones the five years of Section 45 and the two years of Section 46 without breaching either. Inside the rights chapter itself, the reasonable accommodation duty is bounded by the definition in Section 2(y), which excludes modifications imposing a disproportionate or undue burden; that is a standard tied to the individual case rather than a plea of general capacity, and so belongs to a different family from the deleted formula, but Indian authority on how the burden is to be assessed remains thin, and the thinness leaves employers, examination bodies and persons with disabilities alike uncertain where the duty ends in any given case. None of this restores the old statute’s position. A soft verb in one section, a consultation-bound exemption, and a deadline keyed to rule-making are defects of degree in a text whose central duties are unqualified. They are recorded here because the article’s claim is about drafting, and drafting claims are only worth making at the level of the sentence.

Weighing the whole transition, the replacement did what its theory required and its drafters mostly delivered the theory in operative text. The definition moved from diagnosis to interaction. The rights chapter gave the person claims that do not depend on a certificate. Legal capacity and limited guardianship recast the person’s status before the law. The fiscal plea was expelled from the duties owed to identifiable persons and confined, with a comparative floor in the largest case, to the design of schemes. The quantified duties were raised and their judicial construction codified. Behind the duties stand a penal chapter, Special Courts and Funds that the old Act never had. Set against that record, the implementation complaint that the new statute is stronger on paper than in practice is both true and incomplete: it is true of the accessibility deadlines and the awareness of private duty-bearers, and it is incomplete because it treats paper and practice as independent, when the history of the old Act shows that paper of the wrong shape guarantees the practice that follows. The 1995 Act was not badly implemented. It was implemented as drafted, which is the most damning sentence that can be passed on a statute, and the one its replacement was written to escape.

What changed and what only appeared to

The comparison can be drawn together in the form in which it is most useful, one row per subject, with each change classified by kind. Four classifications are used. An expansion of coverage means the same kind of provision reaching more persons or more ground. A removal of a qualifier means the operative change was the deletion of words that had conditioned the duty. A creation of a new obligation means a duty with no counterpart in the earlier text. A new enforcement power means machinery of consequence that the earlier statute lacked. The rows classified as removal of a qualifier are the argument of this article in tabular form.

Subject	1995 position	2016 position	Classification
Recognised disabilities	Seven conditions listed in Section 2(i), a closed list with no power to extend it	Twenty-one specified disabilities in the Schedule, feeding the benchmark definition in Section 2(r)	Expansion of coverage
The threshold	Not less than forty per cent of a listed disability, certified by a medical authority, defined who counted under the Act at all	Not less than forty per cent of a specified disability defines the benchmark class for benefits, while rights attach to the wider class in Section 2(s)	Removal of a qualifier
Reservation in employment	Three per cent of vacancies under Section 33, one per cent each for three groups, in identified posts, with an exemption proviso	Four per cent of vacancies in the cadre strength in each group of posts under Section 34, across five categories, with identification a duty under Section 33	Removal of a qualifier
Reservation in education	Three per cent of seats in government and aided institutions under Section 39	Five per cent of seats in higher education under Section 32, with five years of age relaxation, and free education as a right under Section 31	Expansion of coverage
Accessibility	Transport, roads and built environment duties in Sections 44 to 46, each governed by the economic capacity formula	Standards by rule under Section 40, a permission gate in Section 44, five years for existing public buildings under Section 45, two years for service providers under Section 46, and no fiscal plea	Removal of a qualifier
Legal capacity and guardianship	No provision on legal capacity or guardianship in the Act	Equal legal capacity in Section 13 and limited guardianship as joint decision-making in Section 14, with existing guardians deemed limited guardians	Creation of a new obligation
Non-discrimination	A chapter on facilities and one prohibition on reducing an employee in rank in Section 47	A general bar on discrimination in Section 3, denial of reasonable accommodation defined as discrimination, and employment protection with supernumerary continuation in Section 20	Creation of a new obligation
Enforcement powers and penalties	Commissioners who inquire and take up the matter, reports laid with reasons for non-acceptance, and one offence, fraud by a claimant, in Section 69	Commissioners with enlarged functions, a general contravention fine and an atrocities offence in Chapter XVI, Special Courts, and statutory Funds	New enforcement power

Read down the classification column and the transition explains itself. The famous change, the lengthening of the list, is one row of eight, and it is an expansion of coverage, the least consequential kind of change a statute can make, since it widens the gate without altering what happens to those who pass through it. The threshold row shows the subtler operation: the same forty per cent figure, retained, but stripped of its role as the definition of the statute’s subject, which is why it is classified as a removal of a qualifier rather than a continuity. The reservation and accessibility rows carry the argument’s weight. In employment, the percentage rose by a point, but the operative events were the writing of the cadre strength base into the text, against the executive’s narrower computation, and the conversion of identification from a precondition pleaded in defence into a duty with an expert committee and a review cycle. In accessibility, nothing in the 1995 chapter was quantified at all, and the 2016 chapter is nothing but quantities: standards, a gate, five years, two years. Legal capacity and non-discrimination are new obligations, created rather than amended, because the old text held nothing on those subjects to amend. Enforcement is a new power in form, though its recommendatory core persists inside the larger machinery, as the assessment above records. A table cannot argue, but it can prevent a miscount, and the miscount this one prevents is the counting of disabilities as the measure of the reform.

The table’s title promises a second category, the changes that only appeared to happen, and three deserve naming so that the classification is not mistaken for a claim that everything else was transformed. The Commissioners appear in the new Act as enlarged officers of a rights regime, and their functions are genuinely wider, extending to the review of other laws for consistency with the Act; at the core of the office, though, the 2016 Commissioner inquires, takes up the matter and recommends, as his predecessor did, and the coercive additions arrived around the office, in the penal chapter and the three-month reporting duty, rather than inside it. The threshold appears, in much commentary on the Act, to have been swept away by the barriers definition; it stands exactly where it stood in number, governing every distributive provision, and only its jurisdiction contracted. The coordination committees appear to have been abolished with the old Act; their function continues in the advisory boards, in a consultative register and with persons with disabilities seated as members, which is a change of composition and temper rather than of existence. Placing the apparent changes beside the real ones is not pedantry. A reform’s reputation is built from its announcements, and the announcements of 2016 emphasised the list and the rights chapter. The text supports a more precise reputation: a statute whose machinery mostly continued, whose percentages rose, whose definitions were rebuilt, and whose enforceability was transformed by the removal of the words that had licensed non-performance.

The verdict on the replacement

The series of which this article forms part has repeatedly found a statute’s real effect in a phrase rather than a provision, in a definition, a threshold or a qualifying clause whose placement decided what the imposing remainder of the text was worth. The passage from the 1995 Act to the 2016 Act is the largest instance of the pattern in Indian social legislation, because here the phrase was not incidental to the statute. The qualifiers were the statute. A law that recognised seven disabilities and then conditioned prevention, transport, roads, buildings, rehabilitation and allowances on the economic capacity of the very governments being obliged had not made a drafting error distributed across its chapters. It had chosen, chapter by chapter, a theory of obligation under which the state’s duties were real exactly as far as the state found them convenient, and twenty years of reports, recommendations and unbuilt ramps were that theory working itself out. The courts salvaged what the drafters had left unqualified, and their salvage operation, three per cent enforced to the last vacancy in the cadre, is the proof that the failure was textual. Nothing about the governments, the budgets or the bench changed in 2016. The words changed.

What the words became is a statute that distributes its trust differently. It trusts percentages, dates and definitions to do the work that exhortation could not, confines its fiscal honesty to the scheme provisions where honesty about money belongs, and gives the person with a disability the two things the old Act never gave: a definition of her situation that names the barrier rather than her body as the problem, and claims that survive the duty-bearer’s unwillingness because a court, a Commissioner backed by a penal chapter, or a Special Court stands behind them. The reform will be mismeasured for as long as it is summarised as seven becoming twenty-one. The list grew by fourteen entries. The obligations grew by the deletion of a sentence fragment, and it is the deletion, repeated across the chapters where the fragment had lived, that a reader of the two statutes side by side should carry away.

Frequently Asked Questions

Q: What did the Persons with Disabilities Act 1995 provide?

The Act provided a national framework of welfare and coordination for persons with disabilities. It recognised seven disabilities, defined a person with disability by a forty per cent threshold certified by a medical authority, created coordination committees at the Centre and in the states, directed governments on prevention, education, employment and rehabilitation, reserved three per cent of government posts and of seats in government and aided educational institutions, maintained Special Employment Exchanges, protected an employee who acquired a disability during service from being discharged or reduced in rank, and created Chief and State Commissioners to look into complaints. Most of its substantive duties were conditioned on the economic capacity of the government concerned, and its only offence was fraud by a person claiming a benefit.

Q: Why was the 1995 disability law replaced?

It was replaced because India ratified the United Nations Convention on the Rights of Persons with Disabilities in 2007 and the 1995 Act could not be reconciled with it. The Convention required disability to be understood as the interaction between an impairment and barriers, required equality, non-discrimination, accessibility and equal legal capacity, and treated persons with disabilities as holders of rights. The 1995 Act rested on a medical list, a welfare model and duties qualified by the formula of economic capacity and development, which two decades of experience had shown to be unenforceable in the main. The mismatch lay in the older statute’s definitions and drafting technique, so Parliament repealed and replaced it rather than amending it clause by clause.

Q: What did the UNCRPD require India to change?

The Convention required India to move its disability law from welfare to rights. In substance that meant defining disability by the interaction of impairment with barriers rather than by diagnosis alone, guaranteeing equality and non-discrimination with reasonable accommodation, ensuring accessibility of buildings, transport and information, recognising the legal capacity of persons with disabilities on an equal basis with others, protecting them from abuse and exploitation, and securing inclusive education and participation in the community. The Rights of Persons with Disabilities Act 2016 was enacted to give effect to those requirements, as its long title states. The Convention permits progressive realisation for economic, social and cultural rights, and the 2016 Act reflects that distinction in the provisions where it retained a fiscal qualification.

Q: How many disabilities did each disability statute recognise?

The 1995 Act recognised seven disabilities: blindness, low vision, leprosy-cured, hearing impairment, locomotor disability, mental retardation, in the language of the time, and mental illness, listed in Section 2(i). The list was closed, with no power to add to it. The 2016 Act recognises twenty-one specified disabilities in its Schedule, including conditions the earlier list omitted such as autism spectrum disorder, cerebral palsy, dwarfism, muscular dystrophy, specific learning disabilities, multiple sclerosis, thalassemia, haemophilia, sickle cell disease, acid attack victims and Parkinson’s disease. The larger number is the most visible change between the statutes, though the operative change was the removal of the qualifying words that had emptied the older duties, and the separation of the list from the definition of disability itself.

Q: How did the qualifying words weaken the 1995 disability Act?

The qualifying words were the formula “within the limits of their economic capacity and development”, which opened the sections on prevention, employer incentives, transport, roads, the built environment, rehabilitation and the unemployment allowance, and which the Statement of Objects and Reasons announced as the design of the Bill. The formula weakened the Act by giving the obligated government a statutory answer to any complaint of non-performance: it could plead its own assessment of its capacity, and no court could substitute a standard the text did not contain. Duties that carried the formula produced almost no enforceable case law in twenty years, while the unqualified duties in the same Act, above all the reservation percentage, were enforced strictly. The weakness was therefore not in the Act’s intentions but in a sentence fragment repeated through its operative chapters.

Q: Did the 1995 disability Act create enforceable obligations?

It created a small number of enforceable obligations inside a much larger body of unenforceable direction. The quantified duties were enforceable: the three per cent reservation in employment under Section 33, the three per cent of seats in education under Section 39, the three per cent in poverty alleviation schemes under Section 40, and the prohibition in Section 47 on discharging an employee who acquired a disability during service. Courts enforced those provisions, culminating in the Supreme Court’s direction that reservation be computed on the cadre strength. The programmatic duties governed by the economic capacity formula were not enforceable in any practical sense, and the Commissioners who received complaints could take up the matter with the authorities and recommend, but could not order compliance or punish default.

Q: What happened to cases pending under the 1995 disability Act?

They continued under the replacement statute. Section 102 of the Rights of Persons with Disabilities Act 2016 repealed the 1995 Act and provided that anything done or any action taken under the repealed Act shall be deemed to have been done or taken under the corresponding provisions of the 2016 Act. A complaint pending before a Chief Commissioner or a State Commissioner therefore crossed the repeal without change of forum, the offices having been reconstituted under the new Act with the same names and wider functions. Rights accrued and liabilities incurred under the old Act are also preserved by the general savings doctrine of the General Clauses Act 1897. What did not continue was the old Act’s standard for future conduct, which the new Act replaced.

Q: Which constitutional power was used to enact each disability statute?

Each statute was enacted under Article 253 of the Constitution, Parliament’s power to make laws for implementing international agreements and decisions of international bodies. The power mattered because relief of the disabled is Entry 9 of the State List, a subject ordinarily belonging to the states. The 1995 Act used Article 253 to give effect to the Proclamation on the Full Participation and Equality of People with Disabilities adopted at Beijing in 1992, to which India was a signatory, as its long title records. The 2016 Act used the same power to give effect to the United Nations Convention on the Rights of Persons with Disabilities, which India signed in March 2007 and ratified on 1 October 2007, as its long title and preamble record.

Q: Is a disability certificate issued under the 1995 Act still valid after the repeal?

Yes. A certificate issued by a medical authority under the 1995 Act was an action taken under that Act, and Section 102(2) of the 2016 Act deems anything done or any action taken under the repealed Act to have been done or taken under the corresponding provisions of the new Act. The certificate therefore survived the commencement of the repeal on 19 April 2017, and its holder was not required to undergo reassessment merely because the issuing statute had been replaced. The 2016 Act separately provides, in Section 58(3), that a certificate issued under its own certification chapter is valid across the country. A holder whose older certificate is refused by an authority can answer the refusal with the text of Section 102(2).

Q: What happens to appointments, admissions and schemes completed under the 1995 Act?

They stand. Appointments made against the three per cent reservation, admissions taken against reserved seats in educational institutions, registrations of institutions, and allotments made under the affirmative action provisions were all acts done under the 1995 Act, and Section 102(2) of the 2016 Act deems them to have been done under the corresponding provisions of the new Act. Repeal does not unwind completed administrative action, and the savings doctrine of the General Clauses Act protects rights that had accrued. What changed on commencement was the standard governing future action: appointments after 19 April 2017 fall under the four per cent reservation of Section 34, admissions under the five per cent of Section 32, and schemes under the provisions of the new statute.

Q: Did the forty per cent threshold change when the 2016 Act replaced the 1995 Act?

The figure did not change; its function did. Under Section 2(t) of the 1995 Act, a person with disability was a person suffering from not less than forty per cent of any disability as certified by a medical authority, so the threshold defined the statute’s subject for nearly all purposes. Under the 2016 Act, Section 2(r) uses the same forty per cent figure to define a person with benchmark disability, the class entitled to free education for children, reserved seats, reserved posts and special schemes. The rights chapters of the 2016 Act, however, protect the wider class defined in Section 2(s), persons whose impairment interacts with barriers, whether or not they hold a benchmark certificate. The threshold was retained and demoted, from a definition of disability to a gateway for benefits.

Q: How did the Beijing Proclamation and the UNCRPD differ as foundations for the two Acts?

The Proclamation on the Full Participation and Equality of People with Disabilities in the Asian and Pacific Region, adopted at Beijing in 1992, was a regional declaration of goals associated with a decade of action running from 1993 to 2002. It was not a treaty and imposed no obligations of result, and the 1995 Act built on it inherited its register of targets and endeavour. The United Nations Convention on the Rights of Persons with Disabilities, adopted on 13 December 2006, is a treaty creating binding obligations, which India signed on 30 March 2007 and ratified on 1 October 2007. The difference in the instruments explains the difference in the statutes: a proclamation produced a welfare framework of qualified duties, and a convention produced a rights statute with unqualified duties, deadlines and offences.

Q: How did the legal meaning of disability change between the two Acts?

Under the 1995 Act, disability was a list. Section 2(i) named seven conditions, and a person was disabled in law if a medical authority certified not less than forty per cent of one of them. The condition was located in the person’s body and measured as a percentage. Under the 2016 Act, Section 2(s) defines a person with disability as a person with long term physical, mental, intellectual or sensory impairment which, in interaction with barriers, hinders full and effective participation in society equally with others, and Section 2(c) defines barriers to include attitudinal and structural factors. The list survived as the Schedule of specified disabilities for certification and benefits, but the legal meaning moved from diagnosis to interaction, which is why the newer statute regulates buildings, transport, information and examinations, and not only benefits.

Q: How did the 1995 Act’s exemption power let governments escape the reservation duty?

Section 33 of the 1995 Act required not less than three per cent of vacancies in every establishment to be appointed from persons with disabilities, but its proviso allowed the appropriate Government, having regard to the type of work carried on in an establishment, to exempt that establishment from the provision by notification. The duty therefore carried its own release: the government that owed the duty could discharge a particular establishment from it by an executive act, without amending the statute or answering to a court for the merits. Combined with the practice of treating identification of posts under Section 32 as a precondition, which the Supreme Court rejected in Ravi Prakash Gupta and National Federation of the Blind, the exemption power narrowed the reservation considerably in practice. The 2016 Act retains an exemption proviso in Section 34, but requires consultation with the Commissioner.

Q: How do the punishments for breaking the two Acts compare?

The 1995 Act contained one offence. Section 69 punished a person who fraudulently availed, or tried to avail, a benefit meant for persons with disabilities, with imprisonment up to two years or a fine up to twenty thousand rupees or both. No provision punished a government or an establishment for failing to perform a duty. The 2016 Act contains a penal chapter. Section 89 punishes contravention of any provision of the Act or the rules with a fine up to ten thousand rupees for a first contravention and between fifty thousand and five lakh rupees for later contraventions. Section 92 punishes atrocities, including public humiliation, assault and sexual exploitation, with imprisonment from six months to five years and a fine. Section 91 reprices the fraud offence at a fine up to one lakh rupees, and Section 84 provides Special Courts for trial.

Q: Why does the 2016 Act still carry the economic capacity formula in its scheme provisions?

Because the drafters of 2016 distinguished between duties owed to identifiable persons and programmes the state must design and fund. The formula “within the limit of its economic capacity and development” survives in Section 24 on social security schemes, Section 27 on rehabilitation programmes, Section 35 on incentives to private employers and Section 55 on assistance to registered institutions. In each, the government is framing a scheme whose size is a budgetary choice, and the Convention itself accepts progressive realisation for economic and social rights. What the 2016 Act removed was the formula’s presence in the rights, non-discrimination and accessibility provisions, where it had let the duty-bearer answer a claim by pleading its own capacity. Section 24 also adds a discipline the old Act lacked, requiring disability assistance under such schemes to be at least twenty-five per cent higher than comparable general schemes.

Q: What did the 2016 Act deliberately keep from the statute it repealed?

It kept the load-bearing parts of the old scheme. The forty per cent threshold survived as the benchmark definition in Section 2(r). The reservation structure survived, with identification of posts, a fixed percentage, allocation among categories, carry-forward of unfilled vacancies and an exemption proviso, at higher percentages and across more categories. The Chief Commissioner and State Commissioners survived in Chapter XII with enlarged functions, and the coordination committees were recast as advisory boards. Special Employment Exchanges survived in Section 36, the protection of an employee acquiring a disability during service survived and was strengthened in Section 20(4), and the administrative department responsible for the subject continued unchanged. Replacement was aimed at the old Act’s drafting technique and its theory of disability, not at machinery that twenty years of practice had built and the courts had already construed.

Q: Why did Parliament repeal and replace rather than amend the 1995 Act?

Because the defects were architectural rather than local. Amendment can raise a percentage, add a disability to a list or insert a section, but the 1995 Act’s problems sat in its definition of disability as a medical list, in a qualifying formula distributed through its operative chapters, in an enforcement design that ended in recommendations, and in the absence of any provision on legal capacity, discrimination or accessibility standards. Repairing each of those by amendment would have produced a patched statute whose chapters contradicted one another. A repeal and replacement let the drafters restate the subject on the Convention’s theory, reorganise the chapters around rights, certification, offences and funds, and carry the old Act’s completed acts across by the savings clause in Section 102. The drafting history reflects the scale of the change: a committee appointed in 2009, a draft in 2011, a Bill in 2014, and passage in December 2016.

Q: How did the Supreme Court’s reading of the 1995 reservation clause shape the text of the 2016 Act?

Directly, in the words of Section 34. Under the 1995 Act the Union had computed the three per cent reservation on vacancies in identified posts, under an office memorandum of 2005, and the Supreme Court in Union of India v. National Federation of the Blind, (2013) 10 SCC 772, held that the percentage fell on the total number of vacancies in the cadre strength and struck down the conflicting clauses. Section 34(1) of the 2016 Act enacts that construction, reserving not less than four per cent of the total number of vacancies in the cadre strength in each group of posts. The judgment in Government of India v. Ravi Prakash Gupta, (2010) 7 SCC 626, which denied that reservation waited on identification of posts, is likewise reflected in the structure of Sections 33 and 34, where identification is the government’s duty and not a condition of the percentage.

Q: What replaced the coordination committees and commissioners of the 1995 Act?

The committees were replaced; the commissioners were retained and strengthened. The Central and State Coordination Committees of the 1995 Act, executive bodies that reviewed and coordinated departmental work, gave way to the Central and State Advisory Boards on Disability under Chapter XI of the 2016 Act, broader consultative bodies that include persons with disabilities among their members, with District-Level Committees beneath them. The Chief Commissioner and State Commissioners continued under Chapter XII of the new Act, with functions enlarged to include identifying laws and policies inconsistent with the Act and recommending corrective steps, inquiring on their own motion into deprivation of rights, and monitoring implementation and funds. An amendment effective in January 2018 added that an authority receiving a Commissioner’s recommendation on a complaint must act and report within three months, or give reasons for non-acceptance to the Commissioner and the aggrieved person.