The Child Labour Amendment Act 2016 and the statute it inherited

An amendment trail is read by measuring distance: how far the finished text stands from the text it replaced, and in which direction the movement ran. The Child Labour Amendment Act 2016 moved the law of child work in India in two directions at once, and any honest account of it has to hold both movements in view at the same time. In one direction, the amendment produced the widest prohibition India has ever enacted on the work of young children. A statute that for thirty years had forbidden the employment of a child only in occupations and processes named on a list came to forbid the employment of a child in any occupation or process whatsoever, so that the question in a prosecution is no longer whether the work appears on a schedule but whether an exception covers it. In the other direction, the same amending Act wrote into the general prohibition an exception for a child who helps the family or a family enterprise after school hours or during vacations, and it reduced the schedule of hazardous work, which had grown over three decades to eighty-three entries, to three. The first movement tightened the law beyond anything its critics had demanded in form. The second movement loosened it at the point where, on the evidence of where Indian children actually work, the pressure is greatest.

The scale of the underlying problem is best stated with the caution the subject deserves. The Census of 2011 counted about 10.1 million children between the ages of five and fourteen as workers, main and marginal workers combined, out of a child population of about 259.6 million in that age band, and the Census of 2001 had counted about 12.6 million. Those figures are products of census definitions, which count work in the census’s own categories and do not map neatly onto the categories of the labour statute, so they measure the terrain rather than the offence. What they establish beyond argument is the direction and the order of magnitude: the number fell across the decade, and it remained in the tens of millions. They also establish the distribution that drives this article’s argument. Census and survey accounts of working children place the large majority of them in agriculture, in household industry and in family-mediated production of other kinds, rolling beedis, weaving carpets, making glass bangles, finishing garments, picking and sorting produce, working at home or on the family plot under arrangements in which no stranger employs them in any recognisable sense. The visible child worker, the boy in the roadside eatery or the girl carrying loads at a construction site, is the smaller and more reachable part of the phenomenon. The statute the amendment rewrote was designed around the visible part, because a list of named occupations and processes is a list of workplaces an inspector can enter.

This article therefore proceeds as an amendment trail in the strict sense. It states the law as the Child Labour (Prohibition and Regulation) Act, 1986 first built it, takes each change the 2016 amendment made in turn, and states on both sides of every change what the position was and what it became: the definitions of a child and of the new category of an adolescent, the architecture of the prohibition, the two exceptions, the schedule of hazardous work, the hours regime for the work that remains lawful, the penalty scheme and its treatment of parents, the new machinery of cognizance, compounding, district responsibility and rehabilitation, and the survival of the amended statute when the labour codes consolidated most of the rest of Indian labour law around it. Only after that record is laid out does the article take up the controversy the amendment carries, the family enterprise exception, and give each side its strongest form before reaching a verdict. The verdict, stated here so the reader can test it against the record as it accumulates, is that the amendment tightened the law at every point where tightening was easy to verify and loosened it at the one point where verification is close to impossible, and that this pattern, rather than either half of it alone, is what the trail actually shows.

The 1986 settlement: a prohibition built as a list

The statute the amendment inherited was the Child Labour (Prohibition and Regulation) Act, 1986, Act 61 of 1986, and its design cannot be understood apart from the constitutional text it implemented. Article 24 of the Constitution provides that no child below the age of fourteen years shall be employed to work in any factory or mine or engaged in any other hazardous employment. The founding text thus drew the line at hazard, not at work itself: it forbade the dangerous employments outright and left the remainder to legislation. Article 39, among the directive principles, required the State to direct its policy towards securing that the health and strength of workers and the tender age of children are not abused and that children are given opportunities to develop in a healthy manner and in conditions of freedom and dignity, protected against exploitation and against moral and material abandonment. For decades the legislative response to that mandate was sectoral. Separate statutes fixed ages and conditions industry by industry, in factories, mines, plantations, merchant shipping, motor transport and beedi and cigar establishments, with the Employment of Children Act, 1938 as an early general attempt covering a short list of occupations. A child working outside the named sectors stood outside the law’s attention altogether, and a child working inside them was protected only in the terms the sectoral statute chose.

The patchwork is worth a paragraph of its own, because the 1986 Act’s design can only be judged against the incoherence it replaced. The Factories Act, 1948 barred the employment of a child who had not completed fourteen years in any factory, and regulated the hours of the young person above that age. The Mines Act, after its amendment of 1983, barred persons below eighteen from employment in mines, which made mining law stricter for the older band than the general statute would be for decades. The Apprentices Act fixed fourteen as the minimum age for an apprenticeship, and the beedi and cigar legislation prohibited the employment of children in its industrial premises while the industry’s actual production sat in homes outside the premises concept. Each statute defined its own protected ages by its own vocabulary, child here, young person there, adolescent nowhere, and each policed a premises type rather than the child’s situation. A child rolling beedis at home for a contractor was outside the beedi statute’s premises, outside the factories statute, and inside nobody’s prohibition, although the work was identical to work the law forbade a mile away. The general statute of 1986 was Parliament’s answer to that structure, and its chosen technique, a prohibition keyed to occupations and processes wherever carried on, was a genuine advance on the premises technique. The technique’s cost, which took thirty years to become the dominant fact about the Act, is that a list of processes ages as production reorganises itself around it, and home-based putting-out work is precisely the form of organisation that reorganises fastest.

The general statute came on the recommendation of the committee chaired by M. S. Gurupadaswamy, whose 1979 report on child labour proposed a comprehensive central law, and Parliament enacted it in December 1986. Its long title described its own compromise exactly: an Act to prohibit the employment of children in certain occupations and processes and to regulate the conditions of work of children in others. Prohibition and regulation were the two halves of a single settlement, and the boundary between them was a document, the Schedule. Section 3 of the Act as enacted provided that no child, defined as a person who had not completed the fourteenth year of age, shall be employed or permitted to work in any of the occupations set forth in Part A of the Schedule or in any workshop wherein any of the processes set forth in Part B of the Schedule is carried on. As the Schedule stood on the eve of the 2016 amendment, Part A named eighteen occupations, among them work connected with the transport of passengers, goods or mails by railway, construction work, work in mines, the handloom and powerloom industry, employment as a domestic worker and work in dhabas, restaurants and similar establishments, and Part B named sixty-five processes, among them beedi making, carpet weaving, the manufacture of matches, explosives and fireworks, glass manufacture including bangles, tanning, and processes involving toxic substances. Eighty-three entries in total marked the frontier of the criminal law. Inside the frontier, employment of a child was an offence. Outside it, employment of a child was lawful, and the Act’s second half regulated its conditions.

The Schedule was not static. Section 4 empowered the Central Government, after giving notice of not less than three months, to add any occupation or process to the Schedule, acting on the advice of the Child Labour Technical Advisory Committee constituted under Section 5, and the list grew by notification across the decades. The addition that mattered most in social reach came in 2006, when with effect from 10 October 2006 the employment of children as domestic workers and in dhabas, restaurants, hotels, motels, tea shops and similar establishments was brought within Part A, a change aimed at the urban middle-class household and the roadside eatery, the two most visible sites of child work in Indian cities. The mechanism of growth by notification is the first thread to hold for later comparison, because the 2016 amendment kept the mechanism, enlarged it to include omissions, and then used a parliamentary substitution to shrink the list the mechanism had built.

The regulatory half of the settlement sat in Part III of the Act, brought into force in May 1993, some years after the rest. Part III applied to establishments in which none of the scheduled occupations or processes was carried on, and it regulated the work of the children whom the Act permitted to be there. Its conditions were concrete. A child was not to work for more than three hours at a stretch without an interval of rest of at least one hour, and the period of work including the rest interval was not to spread over more than six hours. No child was to work between 7 in the evening and 8 in the morning, or on overtime, or in more than one establishment on the same day. Every child was entitled to a weekly holiday of one whole day. The occupier owed duties of notice to the Inspector, of maintaining a register of children employed, and of displaying an abstract of the prohibition and penalty provisions, and the appropriate government held power to make rules on health and safety. Enforcement ran through inspectors, and the penalty chapter, Section 14 in its original form, punished employment in contravention of Section 3 with imprisonment of not less than three months and up to one year, or a fine of not less than ten thousand rupees and up to twenty thousand rupees, or both, with a repeat offence drawing six months to two years, and the remaining contraventions of the Act drawing simple imprisonment of up to one month or a fine of up to ten thousand rupees. Prosecution depended on an official complaint, which kept initiation in the hands of the enforcing department.

Two features of the old Section 3 deserve emphasis because the amendment’s originality is measured against them. The first is the proviso. Even within the listed occupations and processes, the enacted Section 3 did not apply to any workshop wherein a process was carried on by the occupier with the aid of his family, or to a school established by or receiving assistance or recognition from the government. A family carve-out therefore existed from the beginning, narrow in form: it lived in the proviso to a list-based prohibition, it spoke of a workshop and an occupier, and it operated only where the work was already prohibited. The second feature is what the design required a prosecution to prove. The offence was constituted by the character of the work, occupation by occupation and process by process, so every expansion of protection required an administrative act of listing, and every unlisted trade, however harsh, stood outside the prohibition no matter how young the child. By the time the education statute rewrote the expectations placed on children between six and fourteen, that architecture was the anomaly. A law that permitted a ten-year-old to work lawfully in an unlisted occupation, subject to hours conditions that presupposed the child’s availability for work, coexisted with a fundamental right to full-time elementary education for the same child. The amendment trail begins at that contradiction.

The long road from a bill of 2012 to a commencement in 2016

The Right of Children to Free and Compulsory Education Act, 2009, Act 35 of 2009, converted Article 21A into an enforceable scheme under which every child between the ages of six and fourteen has a right to free and compulsory education in a neighbourhood school. The alignment argument that later carried the amendment is visible in the commissioning documents in its simplest form: a corollary of the education right is that a child in a workplace during school hours is a child out of school, so a labour statute that licenses the child’s presence at work below fourteen contradicts the education statute’s promise, and aligning the two means prohibiting employment up to the same age at which schooling is compulsory. The full account of that education settlement, its own passage and its age architecture, belongs to the guide to the Right to Education Act 2009, and this article takes from it only the premise the amendment itself used, which is that fourteen became the load-bearing age in Indian law because the Constitution and the education statute had already put the weight there.

The legislative vehicle was the Child Labour (Prohibition and Regulation) Amendment Bill, 2012, introduced in the Rajya Sabha in December 2012 by Mallikarjun Kharge, then the Minister of Labour and Employment. The bill proposed the core of what later became law: a general prohibition below the education age, a prohibition of hazardous work for the band between fourteen and eighteen, and penalties recast upwards. It was referred to the Department-related Parliamentary Standing Committee on Labour, chaired by Dara Singh Chauhan, and the committee’s examination is the point at which the objections that define the amendment’s reception were first placed on the parliamentary record in their strongest official form. The committee reported in December 2013. On the family exception it was blunt in terms that have been quoted in every subsequent round of the argument: the Ministry, in the committee’s view, was itself providing loopholes by providing the provision, and the committee was not able to understand how the Ministry proposed to keep a check on children working in their homes. On the alignment justification it answered that while aligning with the education Act, the Ministry had missed the spirit of that Act, since the education statute’s protection of a child’s schooling does not end neatly at the fourteenth birthday in the way the labour bill’s structure assumed. A parliamentary committee chaired by a member of the then opposition, examining a bill of the then government, had thus fixed the two terms of the debate, verification inside the home and the meaning of alignment, before the bill had a single clause agreed.

The bill then crossed a change of government without dying, which is itself part of the trail, because amendment bills that lapse in the memory of one administration are usually reintroduced in fresh form by the next. This one was instead carried forward and reshaped by official amendments. In May 2015 the Union Cabinet approved official amendments to the pending 2012 bill, and it is at that point that the enacted text acquired its distinctive features in their final form: the exception for a child who helps the family or a family enterprise, the exception for a child artist, the treatment of parents, which softened the bill’s original penal posture towards them, and the reduction of the hazardous schedule to three heads. The government’s stated reasoning, recorded in the Cabinet decision, was that in considering a total prohibition on the employment of children it would be prudent to keep in mind the social fabric and the socio-economic conditions of the country, and officials explained the family exception as protecting the transmission of skills, children learning by participating in work with family elders, with learning at home presented as a seed of entrepreneurship. The bill’s passage followed in July 2016 under Bandaru Dattatreya, the Minister of Labour and Employment in the government that carried it through. The Rajya Sabha passed it on 19 July 2016 and the Lok Sabha on 26 July 2016, after debates in which the exception and the schedule were the contested ground, and it received the President’s assent on 29 July 2016. As enacted it is the Child Labour (Prohibition and Regulation) Amendment Act, 2016, Act 35 of 2016, published in the Gazette of India on 30 July 2016. Because its commencement was left to notification, it entered into force on the date the Central Government appointed, 1 September 2016, by notification S.O. 2823(E). Between the bill’s introduction and the amendment’s commencement, three years and nine months elapsed, and the text that commenced was materially different from the text introduced, which is why the trail has to be followed clause by clause rather than inferred from the bill’s original objects.

One episode on that road deserves recording because it rehearsed, a decade early, the amendment’s central difficulty. The 2006 notification that brought domestic work and work in dhabas and eateries into the old Schedule was the list mechanism’s boldest reach: it carried the prohibition out of the factory and the workshop and into the employer’s home, the one workplace no inspectorate visits in the ordinary course. The announcement produced strong public reaction in both directions, prosecutions remained rare relative to the practice it targeted, and the episode taught two lessons the 2016 debate drew on without always naming them. The first is that a prohibition’s social reach and its enforceable reach are different quantities, and legislation that ignores the gap accumulates unenforced commands. The second is that the household is not simply another workplace in the eyes of the households concerned, many of whom experienced the domestic-work prohibition as the State entering the home on the side of a child they regarded as helped rather than employed. The family enterprise exception of 2016 is, in part, the legislative deposit of that episode: a general prohibition was paired, from the moment of its design in the official amendments, with a recognition that the home would have to be treated by a different technique than the factory. Whether the technique chosen was regulation or abdication is the debate taken up later in this article, but the design problem it answered was real, and it had been visible since 2006.

A prohibition rebuilt: Section 3 before and after

The amending Act worked on the principal Act’s identity before it worked on its offences. It substituted a new long title, so the statute describes itself as an Act to prohibit the engagement of children in all occupations and to prohibit the engagement of adolescents in hazardous occupations and processes and the matters connected therewith or incidental thereto. It renamed the statute itself, substituting for the Child Labour (Prohibition and Regulation) Act, 1986 the name by which it is cited, the Child and Adolescent Labour (Prohibition and Regulation) Act, 1986. The year in the short title stayed at 1986, because the statute remains the statute of 1986 as amended, and the renaming records the structural fact the whole amendment turns on, which is that the Act governs two populations with two different rules rather than one population with a list.

Section 2 carries the two definitions. Before the amendment, the Act knew one protected category, the child, defined as a person who has not completed his fourteenth year of age, together with the machinery definitions of occupier, establishment, workshop and the rest. The amendment inserted a new definition of an adolescent as a person who has completed his fourteenth year of age but has not completed his eighteenth year, and it substituted a new definition of a child. A child is a person who has not completed his fourteenth year of age or such age as may be specified in the Right of Children to Free and Compulsory Education Act, 2009, whichever is more. The final clause is doing quiet but consequential work. By defining the protected age by reference to the education statute’s age, whichever is higher, the amendment built the alignment into the definition itself, so that if the education age were raised, the labour prohibition would follow it upwards without a further amendment to this Act. It is the definitional form of the government’s alignment argument, and it is more durable than a bare number because it hard-wires the two statutes together. The same definitional widening appears across the cluster of child statutes in divergent forms, since the juvenile justice statute and the sexual offences statute fix eighteen without gradation while this statute grades its protection at fourteen and eighteen, and the map of those divergences, which definition governs which question, is drawn in the child protection laws guide. What matters here is the labour statute’s specific choice: prohibition follows schooling to fourteen, and hazard follows minority to eighteen.

Section 3 was substituted wholesale, and the substituted text is the centre of the amendment. The old Section 3 prohibited employment in the occupations of Part A and the workshops of Part B of the Schedule. The new Section 3(1) provides that no child shall be employed or permitted to work in any occupation or process. The list has vanished from the operative prohibition. Every element of the old offence that depended on classifying the work has gone with it: the prosecution no longer identifies an entry in Part A or Part B, no longer proves that a process is carried on in a workshop, and no longer confronts the answer that the trade in question, glass polishing in a home unit, sorting in a godown, stitching in a tenement room, was never listed. The offence is constituted by the employment of a child, full stop, in any occupation or process, which includes the employments the old law reached, the employments it reached only by notification, and the employments it never reached at all. Agricultural work on another’s land, domestic work, work in shops, eateries, workshops, factories, construction, waste picking, vending: the prohibition covers the field because it no longer enumerates the field. That is the tightening half of the amendment, and in formal terms it is complete. Section 3(2) then opens the two exceptions, and the balance of this article is an account of how much of the field the exceptions hand back.

What the substitution did to proof is a change practitioners felt before commentators named it. The offence’s operative words, employs or permits to work, were carried across from the old Act, and the courts’ rescue jurisprudence had long treated the presence of a child at work in an establishment as the occupier’s responsibility to explain. Under the list system that responsibility was only the second half of the prosecution’s task; the first half was taxonomy, establishing that the work fell within a Part A occupation or a Part B process, carried on in a workshop where the process limb was invoked, and defended against the argument that the particular variant of the trade in the particular premises was not the listed thing. Taxonomy defences are the natural product of list legislation, and three decades of them are the practical record behind the government’s claim that the old scheme was unenforceable in the cases that mattered. The substituted Section 3 abolishes the taxonomy stage. Once age and the fact of work in an occupation or process are established, the only remaining questions are the exceptions, and the exceptions are framed by the statute as carve-outs from the prohibition, so the argument in a contested case after 2016 is about the character of the household or the production, not about the character of the trade. The burden of the trial moved from the Schedule to the family, which is another way of stating this article’s central finding: the amendment relocated the contest from ground the prosecution could survey to ground it largely cannot.

A second structural feature of the substituted prohibition deserves notice, because it explains some of the enforcement pattern the machinery section records. Section 3 speaks universally, of any occupation or process, but the statute around it remains built for establishments. The definitions of occupier and establishment, the notice and register duties, the inspector’s powers of entry and examination, and Part III’s hours code all presuppose a workplace with an occupier who can be identified, served and prosecuted. The prohibition of 2016 is wider than the machinery of 1986 in exactly the way the prohibition of 1986 was wider than the sectoral machinery it replaced, and the same consequences follow: the text reaches the home, the field and the street, while the instruments reach the registered premises first and the rest by exception and complaint. This is not a criticism unique to this statute; it is the standard geometry of Indian social legislation, in which prohibition is drafted against conduct and enforcement is organised around premises. But it fixes the questions an honest assessment must ask of the amendment. Not whether the ban below fourteen is complete, since textually it is, subject to the exceptions. Rather, how much of the field the premises-based machinery can carry the ban into, and what carries it the rest of the way: the complaint of a neighbour, the vigilance of a school over an absent child, the rescue operations of organisations of the kind Bachpan Bachao Andolan built, and the Child Welfare Committee’s intake. The amendment strengthened the first of those carriers a little, through cognizable offences, and leaned on the others without building them, which is why the detection figures the Ministry reported sit where they do.

The two exceptions, and the rules that narrowed them

The first exception, in Section 3(2)(a), provides that nothing in the prohibition applies where the child helps his family or family enterprise, which is other than any hazardous occupations or processes set forth in the Schedule, after his school hours or during vacations. Every phrase in the clause carries weight, and the clause rewards being read slowly. The activity protected is help to the family or to a family enterprise, not employment by a stranger; the work must not be a hazardous occupation or process on the Schedule; and the time condition confines the help to after school hours or during vacations. An Explanation appended to the section then defines the terms. Family, in relation to a child, means his mother, father, brother, sister and father’s sister and brother and mother’s sister and brother. Artist is defined for the second exception. Family enterprise bears the definition on which the controversy turned, and it is set out in the direct-answer block below.

What counts as a family enterprise under the amended Act?

A family enterprise is any work, profession, manufacture or business which is performed by the members of the family with the engagement of other persons. The definition names the kin group, parents, siblings and paternal and maternal uncles and aunts, and then describes the enterprise by who performs its work, the family members together with other persons engaged alongside them.

The final limb of that definition is the most consequential drafting choice in the amending Act. A family enterprise is not defined as a business performed only by family members. It is defined as work performed by the members of the family with the engagement of other persons, which means the presence of hired or engaged outsiders does not take a unit outside the definition. A unit can therefore employ adults from outside the family, sell into a supply chain, and take orders from traders, while remaining a family enterprise for the purposes of the exception, provided the child found working there is helping his own family within the defined kin circle. The critics’ reading, stated in its strongest form later in this article, is that the definition describes the standard organisation of India’s home-based production rather than a narrow carve-out from it: a beedi contractor’s household rolling work, a carpet loom in a weaver’s home, a bangle furnace run by an extended family with engaged helpers are the characteristic forms of the trades in which child work concentrates, and each of them fits the definition comfortably. The government’s reading is that the definition reflects how family production is actually organised and that a definition confined to unaided family labour would have described almost no real enterprise in the country, making the exception unusable for the harmless cases it exists to protect. Both readings are readings of the same enacted words, and the words were left in that form through the bill’s passage despite the objection having been taken on the record in committee three years earlier.

Measured against the old law, the exception is both a continuation and a large expansion. The old proviso sheltered the workshop carried on by the occupier with the aid of his family, but only against a prohibition that itself applied only to listed processes, so its practical footprint was small: it protected family workshops inside the frontier of the Schedule. The new clause shelters family help against a prohibition that applies everywhere, so its footprint is the whole field of work minus the Schedule. The exception did not change its character between 1986 and 2016; what changed was the size of the prohibition it is an exception to, and an exception’s real width is a product of both. The time condition, after school hours or during vacations, is the clause’s principal limiting device, and it has the peculiar property of being fully observable at the margins and unobservable at the centre. An inspector who finds a child working at noon on a school day has found a breach of the condition. An inspector who finds a child working in the evening has found nothing provable either way, because the child’s school hours, attendance and the duration of the help all sit inside the household’s own account of the day, and no register, notice or record duty in the Act reaches into the home to test it.

The second exception, in Section 3(2)(b), deals with the child as artist. Nothing in the prohibition applies where the child works as an artist in an audio-visual entertainment industry, including advertisement, films, television serials or any such other entertainment or sports activities except the circus, subject to such conditions and safety measures as may be prescribed, and a proviso adds that no such work shall effect the school education of the child, the statute’s own verb carrying the sense that the work must not be allowed to damage schooling. The circus exclusion is unexplained in the text and is best understood against the circus’s long record as a site of trafficked and bonded child performers, which made it the one entertainment setting Parliament was unwilling to license. For the rest of the entertainment field, the amendment chose regulation over prohibition: the child performer is lawful, and the safeguards are delegated to rules. Those rules, the Child Labour (Prohibition and Regulation) Amendment Rules, 2017, supplied the detail the section promised. For a child artist they require the producer to obtain the permission of the District Magistrate of the district where the activity is to be performed before the child participates, on an undertaking in the prescribed form naming the participants, the parents’ or guardian’s consent and the individual responsible for the child’s safety and security; they cap the child’s work at five hours in a day and not more than three hours without rest; they protect schooling by requiring arrangements for the child’s education and by barring continuous engagement beyond twenty-seven days; they require at least twenty percent of the income earned by the child from the production to be deposited in a fixed deposit account in a nationalised bank in the child’s name, to be credited to the child on attaining majority; and they require screenings of productions involving a child to carry a disclaimer that all measures were taken to ensure no abuse, neglect or exploitation of the child occurred during the shooting. The same rules reached back to the first exception and confined it in a way the Act itself had not: a child helping the family or a family enterprise under the rules may do so only where the work does not affect the child’s education, is not a hazardous occupation or process, forms no part of a manufacturing, production or supply process, and occupies no more than three hours in a day. That fourfold confinement, added by subordinate legislation a year after commencement, is a genuine narrowing of the enacted exception, and the article returns to it in assessing the debate, because it changed at least one prominent critic’s stated position.

The table below states the resulting position for each of the Act’s two populations across the questions a reader, an employer or an inspector actually asks, with the difficulty of verifying compliance stated in the same cell as the legal position, since in this statute the two cannot be separated.

Question	Child, below fourteen	Adolescent, fourteen to eighteen
The general rule	Employment in any occupation or process is prohibited under Section 3(1); verification turns on proving the child’s age and presence at work, and age proof is itself the recurring contest.	Employment is lawful except in the hazardous occupations and processes of the Schedule under Section 3A; verification turns on the character of the work and on age.
Helping the family or a family enterprise	Permitted by Section 3(2)(a) where the work is not hazardous and the help is given after school hours or during vacations; the rules add a three-hour daily cap and exclude manufacturing, production and supply processes; none of these conditions can be verified inside a home.	The exception is unnecessary, since non-hazardous work is open to adolescents in any event; family work by an adolescent is governed by the regulation of Part III, whose records duties apply to establishments and not to homes.
Entertainment and sports work	Permitted by Section 3(2)(b) as an artist in audio-visual entertainment and sports, the circus alone excluded, on prescribed conditions and without affecting schooling; the rules require the District Magistrate’s permission, hourly caps and a deposit of twenty percent of earnings; permission creates a paper trail, making this the most verifiable of the exceptions.	Work in entertainment and sports is lawful on the same footing as other non-hazardous work, subject to Part III conditions where an establishment is involved.
Hazardous occupations and processes	Prohibited without exception; the family exception itself stops at the Schedule, so hazardous family work is outside the exception; verification is obstructed by the contraction of the Schedule, since work hazardous in fact may fall outside its three heads.	Prohibited under Section 3A across the Schedule’s three heads: mines, inflammable substances or explosives, and hazardous processes as defined in the Factories Act, 1948; verification turns on a borrowed definition administered by a different inspectorate.
Hours and periods of work	No hours regime applies, because lawful presence at work exists only inside the exceptions; the rules cap family help at three hours a day and artist work at five, and enforcement of either cap rests on complaint rather than inspection.	Part III applies: work in spells of not more than three hours before a rest interval of at least one hour, a spread-over including rest of not more than six hours, no work between 7 in the evening and 8 in the morning, no overtime, no second establishment on the same day, and one whole day of weekly holiday.
The employer’s obligations	An employer of a child outside the exceptions commits the principal offence; inside a family unit there is no employer in the ordinary sense, which is the structural reason the exception is hard to police.	The occupier owes notice to the Inspector, a register of adolescents employed, display of the abstract of Sections 3A and 14, and compliance with health and safety rules; these duties bite in registered establishments and thin out in informal ones.

The adolescent: a category the old Act did not have

Before 2016, the fourteenth birthday was the boundary of the statute’s world. A worker of fifteen stood outside the prohibition chapter entirely, and outside Part III as well, since Part III regulated children, so the general labour law of factories, mines and shops, with their own age and hours provisions, was the only law speaking to a young worker of fifteen, sixteen or seventeen outside the sectoral statutes. The amendment changed that by inserting Section 3A, which provides that no adolescent shall be employed or permitted to work in any of the hazardous occupations or processes set forth in the Schedule, with a proviso empowering the Central Government, by notification, to specify the nature of the non-hazardous work to which an adolescent may be permitted to work under the Act. For the first time the statute speaks to the whole band from fourteen to eighteen, and it speaks in the register the Constitution reserves for the young: hazard. Article 24’s own prohibition of hazardous employment stops at fourteen, so the extension of a hazard bar to eighteen is a legislative choice that goes beyond the constitutional floor, taken to bring the domestic law into line with the Worst Forms of Child Labour Convention, which protects everyone below eighteen against work likely to harm health, safety or morals.

The second half of the adolescent settlement is the retargeting of Part III. The amending Act renamed Part III so that it governs adolescents rather than children, substituted the word adolescent for child across Sections 6 to 13, and repointed Section 6, the application provision, from Section 3 to Section 3A, so that the regulatory code applies to establishments in which none of the scheduled hazardous occupations or processes is carried on. The practical content of Part III has been stated in the table, and its significance deserves a sentence of its own: the amendment converted the old regulation of permitted child work into a regulation of permitted adolescent work without rewriting its substance, so the hours code drafted for children in 1986, the three-hour spell, the six-hour spread-over, the night-work bar between 7 in the evening and 8 in the morning, the weekly holiday, the notice, the register and the displayed abstract, was carried over to discipline the employment of the fourteen-to-eighteen band in establishments the system can see. The machinery of age proof travels with it: disputes about whether a worker is a child or an adolescent are referred for decision to the prescribed medical authority, whose determination settles the question the offence turns on, because in this statute almost every prosecution begins as an argument about a birth date for which documents are missing, contested or disbelieved.

The adolescent settlement has an asymmetry that its critics identified and its text confirms, and it belongs in the trail because it is the point where the amendment’s alignment logic stops. The education statute’s entitlement runs from six to fourteen. The labour statute’s prohibition runs to fourteen in step with it, and its hazard protection runs on to eighteen. But no corresponding education entitlement runs from fourteen to eighteen: secondary schooling is not a fundamental right, and the adolescent whom the labour statute permits to work in non-hazardous employment is an adolescent the education statute has released. The Standing Committee’s remark that alignment missed the spirit of the education Act has this concrete content. Alignment was performed at the lower boundary, where the two statutes could be made to meet at fourteen, and not attempted at the upper boundary, where meeting would have required either an education right to eighteen or a labour prohibition to match, neither of which the amendment proposed. The result is a lawful space, from fourteen to eighteen, in which full-time work is permitted and full-time schooling is not owed, and the population that space holds is the population least likely to be in school in any event. Defenders of the settlement answer that a prohibition to eighteen without an education entitlement behind it would be the purest form of the unenforceable law this amendment set out to avoid, and that the hazard bar to eighteen takes from the settlement the part of the older band’s protection that safety requires. The answer is coherent. It also concedes the structural point, which is that the amendment aligned the statutes where alignment was cheap and left them unaligned where alignment would have cost the State an obligation.

Two consequences follow for employers, and both are commonly missed. The first is that the adolescent category creates liability where none existed: an establishment that lawfully employed a sixteen-year-old in a listed process before September 2016 must test that employment against the new Schedule, and the proviso to Section 3A, under which the Centre may specify by notification the nature of non-hazardous work open to adolescents, gives the government a continuing instrument for shaping the permitted field without further amendment. The second is that the category is invisible in the statute’s title to most casual readers and in much compliance practice, because the amendment’s publicity concentrated on the ban below fourteen. The employer and compliance payload of this article can be put plainly: the compliance unit in any establishment employing young workers has two age tests to run, not one, a prohibition test below fourteen subject to the two exceptions, and a hazard test from fourteen to eighteen governed by a schedule of three heads. This account states the general position under the statute; an employer facing an inspection or a prosecution needs advice on its own facts, because the exceptions turn on the precise arrangement under which a young person was present at work, and the penalty chapter turns on distinctions between employer, occupier and parent that a general account cannot apply to a particular case.

A schedule cut from eighty-three entries to three

The Schedule is where the amendment loosened the law in the most measurable way, because a list can be counted. The old Schedule, built by the Technical Advisory Committee’s advice and three decades of notifications, stood at eighteen occupations in Part A and sixty-five processes in Part B. It named the trades in the vocabulary of the trades themselves: beedi making, carpet weaving, the manufacture of matches and fireworks, glass and bangle manufacture, brick kilns, cotton ginning, tanning, the handling of toxic and inflammable substances, work in slaughterhouses, rag picking and scavenging, diamond cutting, zari making, electroplating, food processing, warehousing, mechanised fishing. Whatever its untidiness, it was a description of where children in India were actually found working, accumulated case by case. The amending Act substituted an entirely new Schedule, and the new Schedule reads in full: mines; inflammable substances or explosives; hazardous process. Three heads replaced eighty-three entries, and the two-part structure of occupations and processes disappeared with them.

What does the Schedule mean by a hazardous process?

The third head borrows its meaning from another statute. An Explanation to the Schedule gives hazardous process the meaning assigned by clause (cb) of Section 2 of the Factories Act, 1948: a process in a scheduled industry where, unless special care is taken, materials, products, wastes or effluents would impair health or pollute the environment.

In the statute’s own words, the definition covers a process or activity in relation to an industry specified in the First Schedule to the Factories Act where, unless special care is taken, the raw materials used or the intermediate or finished products, by-products, wastes or effluents would cause material impairment of the health of the persons engaged in or connected with the process, or result in the pollution of the general environment. The full treatment of that definition and of the factories code it lives in belongs to the Factories Act 1948 explained, and the point for this trail is narrower: the adolescent’s protection in the amended Act is exactly as wide as a definition drafted in 1948 for registered factories, administered by a factory inspectorate, in a statute most home-based and informal production never encounters.

The borrowing has consequences that the parliamentary debate identified and that the text confirms. A process can be dangerous to a fifteen-year-old without being a hazardous process within the Factories Act definition, because that definition is anchored to scheduled industries and to a factory model of production. Tobacco processing and beedi making, which stood in the old Schedule by name, do not appear in the Factories Act’s hazardous process list in terms, and the parliamentary committee’s criticism of the regression in the hazardous definition rested on exactly this gap: the new Schedule protects adolescents against the hazards of the organised factory sector, described in that sector’s own statute, while the named hazards of the informal economy, in which adolescents overwhelmingly work, drop out of the prohibition unless they can be brought within mining or inflammable substances and explosives. The government’s defence of the substitution was a defence of focus: the long list had grown arbitrary and unwieldy, entries overlapped, and a short schedule anchored to genuinely hazardous work, capable of expansion by notification, was said to be more enforceable than a catalogue nobody applied consistently. Both descriptions of the same three lines are accurate, which is why the schedule question cannot be settled by counting alone; it turns on whether the notification power in fact restores what the substitution removed.

That power matters, and the amendment enlarged it. Section 4 as amended empowers the Central Government not only to add to the Schedule any hazardous occupation or process but to omit from it, where the old power ran only to additions after three months’ notice. The ratchet that had operated in one direction for thirty years runs in both directions, and changes in either direction are made by notification, which is to say by executive act laid before Parliament rather than by legislation debated and voted as the 2016 substitution itself was. The advisory machinery was renamed in step: the Child Labour Technical Advisory Committee became the Technical Advisory Committee, and its function of advising the government on the Schedule’s contents survives.

Who advises the Centre on adding entries to the Schedule?

The Technical Advisory Committee, constituted under Section 5 and renamed by the 2016 amendment from the Child Labour Technical Advisory Committee, advises the Central Government on additions to and omissions from the Schedule, while Section 4 vests the power to change the Schedule in the Central Government by notification.

The committee’s advice and the notification together are therefore the route by which the three heads grow or shrink, and the route runs entirely within the executive: no bill, no debate on the perimeter as such, and no vote, although notifications are laid before Parliament in the ordinary way.

The structural point is one this series has had repeated occasion to make about Indian legislation: where the operative content of a prohibition sits in a schedule amendable by notification, the legislature has decided the principle and delegated the perimeter, and the political contest moves from the floor of Parliament to the process of the ministry. For the child labour statute the delegation has a particular edge, because the contested question between 2012 and 2016 was precisely the perimeter, what counts as hazardous for a worker of fifteen. Parliament settled that question in 2016 by substituting three heads and handing the perimeter to notification, so the debate the statute appeared to end was in fact relocated, and the record of notifications under the amended Section 4 is where a future reader should look to see which side of the 2016 argument prevailed in practice.

Penalties, cognizance and the price of a second conviction

The penalty chapter shows the amendment tightening in numbers that can be set side by side. Under the original Section 14, employing a child in contravention of the prohibition drew imprisonment of not less than three months and up to one year, or a fine of not less than ten thousand rupees and up to twenty thousand rupees, or both, and a repeat offence drew imprisonment of six months to two years. The substituted Section 14(1) punishes whoever employs any child or permits any child to work in contravention of Section 3 with imprisonment of not less than six months and up to two years, or a fine of not less than twenty thousand rupees and up to fifty thousand rupees, or both, and a new Section 14(1A) applies the same range to employing an adolescent in contravention of Section 3A. Every figure at least doubled, and the minimum terms became real minima rather than nominal ones. The residuary clauses of the old Section 14(3) were omitted in the substitution, and the scheme concentrates its force on the two principal offences. The detail is worth one sentence more, since compounding later turns on it. The original Section 14(3) punished a schedule of defaults, the failure to give the Inspector the required notice, the failure to maintain the register of children employed or a false entry in it, and the failure to display the abstract of the prohibition and penalty provisions, together with a residuary limb for contravention of any other provision of the Act or its rules, the whole carrying simple imprisonment of up to one month or a fine of up to ten thousand rupees or both. The amending Act omitted the three specific limbs and left the residuary limb standing, so the minor contraventions of the regulatory code, an unmaintained register, a missing notice, a breached hours rule where an adolescent is lawfully employed, remain punishable at that modest level rather than disappearing from the penal map. It is first offences under that residuary provision, and offences by parents and guardians, that the compounding power in Section 14D takes up, and the pairing is coherent: the defaults of paperwork and the defaults of poverty are priced for composition, while the offence of putting a child to work is priced for prosecution.

What happens to an employer convicted a second time?

A second conviction is punished under the substituted Section 14(2): whoever, having been convicted of an offence under Section 3 or Section 3A, commits a like offence afterwards is punishable with imprisonment for a term which shall not be less than one year but which may extend to three years.

The comparison with the old law is worth holding beside it. The original Section 14 punished a repeat offence with imprisonment of six months to two years, so the substituted repeat band of one year to three years doubles the floor and raises the ceiling by half, and the repeat provision carries no fine alternative in its terms, which means a court sentencing a second conviction under it is sentencing to imprisonment or nothing.

The treatment of parents is the penalty chapter’s distinctive construction, and it was one of the official amendments of 2015 rather than a feature of the bill as introduced. The provisos to Sections 14(1) and 14(1A) state that the parents or guardians of a child shall not be punished unless they permit the child for commercial purposes in contravention of Section 3, and correspondingly for an adolescent under Section 3A. A new Section 14(1B) then provides that, notwithstanding the rest, the parents or guardians of any child or adolescent shall not be liable to punishment in the case of the first offence. Where a parent or guardian, having been convicted, commits a like offence afterwards, Section 14(2A) caps the punishment at a fine which may extend to ten thousand rupees. The scheme thus separates three figures the old law had treated more nearly alike: the employer, who faces doubled imprisonment and fines; the parent who permits work for commercial purposes, who is reachable in principle; and the parent in the ordinary case, whom the first-offence immunity and the commercial-purposes proviso together place outside the penalty scheme altogether. The defence of the construction is the defence of the family exception in penal form: a statute that imprisons poor parents for their children’s work punishes poverty and drives the practice out of sight. The objection is the mirror image: where the employer and the parent are the same household, as the exception contemplates, the penal chapter’s force lands on the stranger employer and dissolves at the household door, which is where most child work is organised.

Two procedural changes complete the penal picture. Section 14A, newly inserted, provides that notwithstanding anything in the Code of Criminal Procedure, 1973, any offence committed by an employer and punishable under Section 3 or Section 3A shall be cognizable. Under the old law the offences were non-cognizable and prosecution moved on an official complaint through the enforcing department, which throttled initiation at the inspector’s desk; a cognizable offence empowers the police to register and investigate without waiting for that channel, and it converts every rescue conducted with police participation into a potential prosecution from the first hour. Section 14D, also new, qualifies the severity in the other direction: the District Magistrate may, on the application of the accused, compound an offence committed for the first time under the residuary penalty provision, and any offence committed by an accused who is a parent or guardian, in the prescribed manner and on payment of the prescribed amount to the appropriate government, with prosecution barred where composition precedes it and discharge following where it succeeds after prosecution has begun. Compounding is the scheme’s safety valve, and its allocation is telling: it is open to first offenders in the minor category and to parents and guardians generally, and closed to the employer facing the principal offences, which is a defensible distribution that nonetheless confirms how much of the penal architecture is arranged around keeping the household outside the criminal process.

The artist exception as regulatory design

The second exception rewards separate treatment because it is the amendment’s one essay in a different legislative technique. For the child performer, Parliament did not choose between prohibition and permission; it chose permission with architecture. The statutory text withholds the prohibition from the child who works as an artist in the audio-visual entertainment industry and in sports, the circus alone excluded, and subjects the permission to prescribed conditions and safety measures and to the proviso that the work shall not be allowed to affect the child’s schooling. The 2017 Rules then built the architecture the section contemplated: advance permission from the District Magistrate on a producer’s undertaking, named responsibility for the child’s safety, hourly and daily caps, a limit on continuous engagement, schooling arrangements, a screen disclaimer, and the fixed deposit of at least twenty percent of the child’s earnings in the child’s own name until majority. Each element answers a known failure of unregulated child performance: the permission creates a record before the work begins rather than an offence to be discovered after it; the deposit answers the dissipation of a child performer’s earnings by the adults around the child, the classic injury in this industry worldwide; the schooling proviso with its twenty-seven-day ceiling answers the serial engagement that quietly ends a child’s attendance. The National Commission for Protection of Child Rights later carried the framework further in guidelines for children in the entertainment industry, holding the permission, hours and deposit structure and extending its logic to children’s monetised appearances in online content, which the guidelines treat through the family enterprise lens where the content is produced by the family for gain. The contrast with the family enterprise exception is instructive and deliberate in this article’s construction. In entertainment, the work is organised by identifiable producers, in places that can be named, on schedules that exist in writing, for money that moves through accounts, and the amendment’s regulatory technique fits it exactly. In family production, none of those handles exists in the same form, and the amendment reached instead for description and exception. The artist exception therefore proves something about the family exception that neither side’s rhetoric quite states: where the State could see the work, it regulated the work in detail and the detail has largely held; the looseness of the family exception is not a drafting preference but an admission about visibility, and the argument between the sides is really an argument about whether an admission of that kind belongs in a prohibition statute at all.

The sports limb of the same exception deserves its sentence, because it is the part of Section 3(2)(b) least examined in the debate. The statutory definition of an artist expressly includes a child who performs or practises work as a sports person, so the young athlete in a residential academy, training on a schedule set by coaches and competing in events organised by federations and sponsors, stands inside the exception on the same footing as the child actor, subject to the prescribed conditions and to the proviso that the work not be allowed to affect schooling. The fit is imperfect in an instructive way: the rules built under the exception are drafted around productions, producers and District Magistrate permissions for audio-visual work, while the athlete’s engagement is continuous rather than production-based and the responsible adults are academies and federations rather than producers. The exception’s breadth here is a reminder that Section 3(2)(b) was drafted for an industry of shoots and schedules and then extended by its own words to a world of training camps, and that the prescribed conditions, not the section, are where the protection of the child in sport will be decided.

The provisions applied: four workplaces

Amendment trails are tested by application, and four situations, each commonplace, show where the substituted text bites and where it runs out. The first is the roadside eatery. A boy of twelve washes dishes and serves tables in a dhaba whose owner is no relation to him. Before September 2016 the prosecution had to bring the case within the Schedule, which the 2006 notification had made possible by listing such establishments, and then still faced the Act’s official-complaint procedure and the three-month minimum sentence. Under the substituted law the analysis shortens to three questions: is the worker a child within Section 2, was he employed or permitted to work, and does any exception apply. No exception does, since the help is not given to his own family and he is not an artist. The offence is complete under Section 3, cognizable under Section 14A, punishable under Section 14(1) at six months to two years or twenty thousand to fifty thousand rupees or both, not compoundable at the District Magistrate’s desk because the principal employer offences sit outside Section 14D, and on the employer’s conviction the realised fine feeds the district fund with the government’s fifteen thousand rupees behind it. The case is the amendment working exactly as designed, against the visible employer in premises an inspector can enter.

The second situation is the home-based carpet unit. A girl of eleven works at a loom in her parents’ dwelling; the unit sells through a trader, takes his orders and his yarn, and an unrelated adult weaver is engaged alongside the family. Under the law before 2016 the case turned on whether carpet weaving, a listed process in Part B, was being carried on in a workshop, and the old proviso then removed from the prohibition a workshop run by the occupier with the aid of his family, so the household’s own case was fragile for the prosecution from the start. Under the substituted law the threshold is easier and the exception is the whole contest. Section 3 prohibits the employment, and Section 3(2)(a) withdraws the prohibition where the child helps the family enterprise, the enterprise being performed by family members with the engagement of other persons, which describes the engaged neighbour precisely. The contest then moves to the exception’s conditions: the work must not be a hazardous occupation or process on the Schedule, and carpet weaving is not within the three heads as substituted, so that condition is satisfied in the household’s favour; the help must be after school hours or during vacations, a fact located in the household’s account; and the 2017 Rules add that the work must not affect her education, must occupy no more than three hours a day and must form no part of a manufacturing, production or supply process, a condition a loom selling into a trader’s chain appears to fail, although its breach is a breach of rules made under the Act rather than a clean route back into the Section 3 offence. A prosecution here is possible in law and forbidding in practice, and every forbidding element traces to a phrase in the exception rather than to any failure of the prohibition.

The third situation is the adolescent in the brick kiln. A boy of sixteen moulds and carries bricks at a kiln run by a contractor. Before 2016 he stood outside the Act altogether: not a child, so neither prohibited nor regulated by it. Under the substituted law he is an adolescent, and the question is whether the work falls within the Schedule’s three heads. Brick making as such is not mining, involves no inflammable substances or explosives on the facts stated, and is a hazardous process within the Factories Act definition only if the industry and process can be brought within that Act’s First Schedule and definition, an inquiry framed for factory premises that a seasonal open-air kiln fits awkwardly. If the work is within the Schedule, his employment is an offence under Section 3A carrying the Section 14(1A) range. If it is not, his employment is lawful and Part III governs its conditions, the three-hour spells, six-hour spread-over, night-work bar, weekly holiday, notice and register, duties framed for establishments and enforceable in proportion to the establishment’s visibility. The same sixteen-year-old, doing the same work, is thus either the victim of a serious offence or a regulated worker, and the switch is a borrowed definition from 1948. No provision in the amended Act lets the decision-maker ask the direct question, whether the work is in fact dangerous for a sixteen-year-old, because the statute has delegated that question to the Schedule.

The fourth situation is the child artist. A girl of nine is cast in a television serial, filming on weekdays across two months. Before 2016 the old Act’s list said nothing specific about performance, and her position turned on the establishment she worked in and the general law. Under the substituted law her work is within the second exception, subject to the conditions the 2017 Rules prescribe, and those conditions create the only paper trail the exceptions generate: the producer’s prior permission from the District Magistrate, the undertaking naming her and the responsible individual, the five-hour day and three-hour stretch, the bar on continuous engagement beyond twenty-seven days, the schooling arrangements, and the fixed deposit of at least twenty percent of her earnings in her name. An inspector or a parent testing this engagement tests documents, and documents can be checked. Set beside the carpet unit, the contrast is the article’s argument in miniature: where the amendment chose to regulate a lawful child activity through permission and records, as it did for the artist, verification exists; where it chose to except an activity by describing it, as it did for the family enterprise, verification dissolves into the household’s own account of itself.

Rescue, the District Magistrate and the Rehabilitation Fund

What happens to the child is a separate question from what happens to the employer, and the amendment gave it a statutory frame for the first time. A new Section 14B requires the appropriate government to constitute in every district, or for two or more districts, a Child and Adolescent Labour Rehabilitation Fund, to which the amount of fine realised from the employer shall be credited. The government itself must then credit an amount of fifteen thousand rupees to the Fund for each child or adolescent for whom a fine has been credited, and the amount so accumulated, with the interest on its deposit or investment, is to be paid to the child or adolescent in whose favour it was credited, in the prescribed manner. A companion Section 14C provides that the child or adolescent employed in contravention of the Act and rescued shall be rehabilitated in accordance with the laws for the time being in force, which connects the labour statute’s rescue to the care and protection system rather than building a parallel one. Under the juvenile justice statute, a child found working in contravention of labour laws is a child in need of care and protection, to be produced before the Child Welfare Committee and dealt with under that statute’s machinery of inquiry, placement and restoration, a route traced in the Juvenile Justice Act 2015 explained. The labour amendment deliberately borrows that destination instead of inventing its own, and Section 14C is the hinge between the two statutes: the labour inspectorate finds and removes, the juvenile justice system receives and rehabilitates, and the Fund attaches money to the child at the point of rescue.

The enforcement design also has a federal structure the trail should record. Labour welfare sits in the Concurrent List, and the Act’s own scheme of the appropriate government divides administration between the Centre for its sphere and the state governments for theirs, so the amended statute is executed overwhelmingly by state inspectorates, state-constituted funds and District Magistrates who are state officers, under a text written in New Delhi. The amendment changed none of that division; what it changed was the load the division carries. A cognizable offence engages the state police, a district fund engages the state treasury’s credit of fifteen thousand rupees per child, periodic inspection engages the state’s labour department staffing, and the age and rescue questions engage the state’s juvenile justice and education machinery, all at once and in the same district. The variance in the statute’s real force between states is therefore not an accident of implementation but a prediction of its design, and the Ministry’s own enforcement returns, compiled from the states, show the variance in the raw, with inspection and prosecution activity concentrated in the states with functioning child labour machinery. A central amendment that doubles penalties and multiplies duties without altering the enforcing tier has legislated the ceiling of the system rather than its floor, a pattern the implementation record of the parent Act had already established and which no provision of the 2016 text escapes.

Beneath the District Magistrate stands the cadre the Act has always run on. Section 17 provides for the appointment of Inspectors for the purposes of the Act, who are deemed to be public servants, and the inspection power is exercised through them under the rules: entry into establishments, examination of registers and premises, and the initiation of the processes the penalty chapter completes. The amendment left the cadre’s statutory basis untouched while multiplying what the cadre is asked to do, adding a cognizable-offence regime that engages the police alongside it, a permission regime for child artists that runs through the District Magistrate, a fund that has to be accounted child by child, and a monitoring duty in Section 17B that is framed as an obligation of the government rather than of any countable staff. No figure for the strength of this cadre is asserted here, because the returns are state subjects compiled unevenly and a national number would be an invention; the structural point does not need one. A statute whose detection depends on inspection will detect in proportion to its inspectors, and the amended Act’s own reported figures, hundreds of thousands of inspections against detections in the low thousands, describe a system inspecting at breadth rather than at depth, visiting many premises briefly because the premises it can visit are not where its subject principally is. The 2016 settlement’s wager, visible across Sections 14A, 14B, 17A and 17B read together, is that district accountability plus police cognizance plus the school system’s attendance records can substitute for inspector density. That wager is the amendment’s real implementation theory, and the returns of the coming years, rather than the text, are where its success or failure will be recorded.

The district is the amendment’s chosen unit of enforcement, and two further inserted sections build it out. Section 17A empowers the appropriate government to confer such powers and impose such duties on a District Magistrate as may be necessary to ensure that the provisions of the Act are properly carried out, with the District Magistrate able to specify a subordinate officer to exercise those powers within prescribed local limits. Section 17B obliges the appropriate government to make or cause periodic inspection of the places at which the employment of children is prohibited and hazardous occupations or processes are carried on, and to monitor the issues relating to the Act’s provisions. The District Magistrate is also the compounding authority under Section 14D and, through the 2017 Rules, the authority whose permission a child artist’s producer must obtain, so the amended statute concentrates rescue oversight, compounding, permission and inspection responsibility in a single district office that already carries the district’s general administration. The design judgement is that child labour enforcement failed as a departmental routine of a thin inspectorate and had a better chance as a named responsibility of the district’s chief officer; the countervailing consideration is that the District Magistrate is the most burdened office in Indian administration, and a duty added to it competes with every other duty it holds.

The paper trail this machinery generates is considerable, and it has to survive movement between offices that do not share systems: the inspection report, the age-determination record, the fund ledger entry for fifteen thousand rupees of government credit plus the realised fine, the committee’s orders in the child’s case, and the school enrolment record that closes the file. A district officer handed such a bundle in mixed formats can open the working files directly in the browser through ReportMedic’s office file viewer rather than lose days to incompatible software, which is a small practical point with a real consequence in a process where the child’s placement often waits on the file. The official enforcement figures the Ministry of Labour and Employment placed before Parliament in November 2019, drawn from state government returns for 2018, give the scale at which this machinery operated in a single year: 338,696 inspections conducted, 1,636 violations detected, 1,168 prosecutions launched and 690 convictions. The ratio in those attributed figures is the enforcement problem in one line: inspections in the hundreds of thousands produced detections in the low thousands, which is consistent either with widespread compliance, which the census figures contradict, or with an inspection system that visits the establishments it can see, the registered and the visible, while the work the statute most needs to reach sits in homes and family units where inspection does not go. The amendment’s own architecture predicts exactly that distribution of results, since its prohibitions are at their strongest in premises an inspector can enter and at their weakest in the household its exception describes.

The rehabilitation promise itself has an institutional history the statutory fund joins rather than begins. The National Policy on Child Labour of 1987 committed the government to a project-based approach in areas of high concentration, and the National Child Labour Project, running from 1988, built the machinery of special schools and rehabilitation centres in the districts where child work concentrated, withdrawing children from work, bridging them through non-formal education and stipendiary support, and mainstreaming them into the formal system, under district project societies chaired by the District Magistrate. The amended Act’s choice of the district as its unit, and of the District Magistrate as its named officer, follows the grain of that older scheme architecture, and Section 14C’s direction that the rescued child be rehabilitated in accordance with the laws for the time being in force hands the child to a system with three doors: the special-school and mainstreaming route inherited from the project approach, the care and protection route of the juvenile justice statute through the Child Welfare Committee, and the education statute’s entitlement to age-appropriate admission with special training. The fund’s contribution is to attach a sum of money to the individual child at the moment of rescue, where the older machinery attached services. In practice the rescued child’s file moves through all three doors at once, and the recurring failure recorded in the enforcement literature is not the absence of a door but the loss of the child between them, through migration, re-trafficking or the family’s economic need reasserting itself, which is the practical form of the premise the Supreme Court stated in 1996: the work was income, and any machinery that does not replace the income is negotiating with the reason the child was there.

Age proof deserves a final word in this section, because it is the hinge on which both prohibition and rehabilitation swing. The Act’s two categories turn entirely on the fourteenth and eighteenth birthdays, and the population the Act most concerns is the population least able to document a birthday: birth registration among poor and migrant households is incomplete, school records may not exist for a child who has never been enrolled, and the employer and the family may each have an interest in the age the record shows. The Act answers with its dispute mechanism, referral to the prescribed medical authority whose determination settles the question, and medical age estimation carries margins of error measured in years at exactly the boundary ages the statute cares about. Every protection in the amended Act is therefore administered through an evidentiary bottleneck, and the bottleneck is narrowest for the children with the weakest documents, which is to say for the children the Act most exists to reach. An amendment trail that ignored this would mistake the text for the system; the 2016 text is considerably stronger than the system that has to prove a birthday before it can use it.

What the courts had already built

The amendment was not written on a blank judicial record. The Supreme Court’s decision in People’s Union for Democratic Rights v. Union of India, arising out of the construction projects for the Asian Games of 1982, had established that Article 24 is enforceable against private employers and not only against the State, since the prohibition on hazardous child employment is addressed to whoever engages the child, and the Court treated the employment of children on construction work as within the Article’s reach. Between that establishment of principle and the Sivakasi judgment stands Bandhua Mukti Morcha v. Union of India, decided in 1984, in which the Supreme Court, dealing with bonded labourers in the stone quarries of Faridabad, read Articles 21 and 23 together to hold that labour exacted without payment of the minimum wage was forced labour within the constitutional prohibition, and treated the identification, release and rehabilitation of bonded workers, children among them, as a continuing judicial concern rather than a single order. Its importance for this trail is the template it set: in the child and bonded labour field the Court’s characteristic remedy has been the supervised scheme, a fund, a direction to employ or compensate, a committee to report back, rather than the bare declaration, and the statutory machinery of 2016, fund, District Magistrate, periodic inspection, is recognisably the legislative descendant of that remedial style.

The leading child labour judgment is M.C. Mehta v. State of Tamil Nadu, decided on 10 December 1996 and reported at (1996) 6 SCC 756, which arose from the match and fireworks industry of Sivakasi and surveyed the position across the country. Its directions built, by judicial order, the rehabilitation architecture the 2016 amendment later built by statute. An employer found employing a child in hazardous work was directed to pay twenty thousand rupees per child as compensation into a Child Labour Rehabilitation-cum-Welfare Fund, to be used for the child’s benefit; the State was directed to secure employment for an adult member of the child’s family in place of the child or, failing that, to contribute five thousand rupees per child to the fund; and the income from the fund was to support the child on condition that the child attended school. Read against Section 14B, the ancestry is direct: a fund fed by employer payments, supplemented by a government contribution per child, with the proceeds tied to the rescued child, was the Supreme Court’s construction in 1996 and became Parliament’s construction twenty years later, with the government’s per-child credit set at fifteen thousand rupees and the employer contribution taking the form of the realised fine rather than a fixed compensation sum. The judgment also fixed the analytical premise this article has followed, when it located the cause of child labour in poverty and held that prohibition without an alternative income for the family would not hold, which is the same premise the family exception rests on and the same premise its critics answer.

The rescue jurisprudence did not stop in 1996. The Sivakasi litigation itself had an earlier chapter, in which the Court in 1991 gave directions addressed to the quality of life of children employed in the factories there, before the comprehensive directions of 1996 settled the compensation and fund architecture. In Bachpan Bachao Andolan v. Union of India, decided in 2011 and reported at (2011) 5 SCC 1, the Supreme Court dealt with missing children and the trafficking pipeline that feeds child labour, directing among other things the registration of first information reports in cases of missing children and the strengthening of the juvenile justice machinery through which rescued and trafficked children pass. The organisation that brought that litigation is the same organisation whose founder’s statement on the 2016 bill anchors the critics’ case later in this article, and the continuity is worth noticing: the movement’s method for three decades was litigation that built machinery judgment by judgment, and its objection to the amendment was in part the objection of a litigant watching a statute hand back, through an exception, ground that litigation had treated as won. None of this decides the legislative question, which was Parliament’s to decide, but it explains the temperature of the objection, which purely textual readings of the debate tend to miss.

The interaction with the education statute in rescue work is practical rather than doctrinal. A rescued child of nine or twelve is a child entitled to admission to an age-appropriate class under the education Act, with special training to reach the level of the class, and the rehabilitation that Section 14C promises by reference to the laws in force is in large part the working of that entitlement plus the care and protection procedures of the juvenile justice statute. The doctrinal point the cases collectively make about the amended Act is about the liability of the employer where a child is found on premises: the prohibition speaks of employing or permitting a child to work, and the courts’ approach in the rescue cases treats the presence of a child at work in an establishment as the occupier’s responsibility to explain, which is why the Act’s penalty sections are drafted against whoever employs or permits. That drafting survived the amendment unchanged in its operative words, and it is the reason the family exception had to be drafted as an exception to the prohibition rather than as a defence to be proved, because outside the exception the presence of the working child is close to the whole of the offence.

Where the children are: the pattern the exception had to fit

The debate over the exception is unintelligible without the distribution of child work itself, and the distribution can be stated from the official and attributed record without inventing precision. The Census of 2011 counted about 10.1 million working children between five and fourteen, main and marginal workers combined, against about 12.6 million in the Census of 2001, a fall of about two and a half million across a decade in which the child population grew. The International Labour Organization’s estimates, built on national survey data and a wider age band, put the number of child workers between five and seventeen at 5.7 million in its 2015 global report, lower than the census count because the methods differ, and the honest conclusion from the two together is that the phenomenon is counted in millions by every method and that the methods disagree about the millions. Geographically the census count concentrates: five states, Uttar Pradesh, Bihar, Rajasthan, Maharashtra and Madhya Pradesh, account for about fifty-five percent of it. Socially it concentrates further: UNICEF India, drawing on the census, recorded child labour rates highest among Scheduled Tribes at 6.7 percent and Scheduled Castes at 3.9 percent of the respective child populations, with rural rates above urban ones in both groups, which is the factual substrate of Satyarthi’s argument that the exception entrenches hierarchy, since the families whose enterprises the exception protects are drawn disproportionately from the communities at the bottom of it.

Sectorally, the reporting of the International Labour Organization’s estimates placed more than half of child workers in agriculture, toiling in cotton, sugarcane and rice fields with exposure to pesticides and to sharp tools and heavy equipment, and at least a quarter in manufacturing, very often confined to poorly lit and barely ventilated rooms, embroidering garments, weaving carpets, making matchsticks or rolling beedis, with others helping their parents in brick kilns and mines and still others in services, in shops, restaurants and hotels and in middle-class homes. Two features of that distribution decide the legislative argument. The first is the share of work that is family-mediated. The child in agriculture works overwhelmingly on the family plot or alongside parents in others’ fields; the child in home-based manufacturing works at the family loom, the family rolling board, the family furnace; the child in the mine-head mica belt works in family collection units; even much service work runs through kin networks. The employer-mediated child worker, hired by a stranger into a registered establishment, is the minority case, and it is the minority case that the prohibition, the inspection system and the penalty chapter are shaped to reach. The second feature is the invisibility gradient. Work inside the home is invisible to the census in part, since census categories capture household industry imperfectly and miss unpaid domestic labour in the child’s own home almost entirely, a blind spot that falls heaviest on girls, whose work as carers, cleaners and cooks within the household is the least counted labour in the entire field. UNICEF India’s statement on the bill named the point directly when it warned that the family exception would leave children exposed to exploitation in invisible forms of work, and it listed the home-based trades: work in cotton fields, making bangles and beedis, rolling tobacco, carpet weaving and metal work. The exception’s defenders and its critics were thus not describing different economies. They were describing the same economy and disagreeing about what a prohibition can honestly claim to govern within it, which is why the empirical section of this article is placed here, immediately before the debate: the debate is about the legal treatment of a distribution neither side disputes.

The family enterprise debate, stated at full strength on both sides

The government’s case for the amendment, as Bandaru Dattatreya made it in Parliament and as the Cabinet’s reasoning recorded it, rests on enforceability and on a claim about Indian social life. The enforceability claim is that the purpose of the Act is that it should be capable of practical implementation, the minister’s own formulation being that the exemptions were given so that the law could be practically implemented, and that a prohibition which criminalises a child helping on the family farm after school, or learning the family trade in a potter’s or weaver’s household, would be a prohibition the State could not honestly enforce and would therefore not enforce, leaving the whole statute discredited. The social claim is the balance argument: the exemptions strike a balance between education and the economic reality in which parents rely on children’s help in farming and artisanal work to fight poverty and to pass on a family trade, and the social fabric of a country in which children learn by participating in work alongside family elders is not a defect the law should set out to redraw. Supporting considerations run with these. The prohibition below fourteen is total outside the exceptions, which is more than the old law ever attempted, and the alignment with the education age gives the ban a coherence inspectors and courts can use. The adolescent hazard bar to eighteen, the doubled penalties, cognizable offences and the rehabilitation fund are tightenings no previous amendment attempted. And the amendment made possible what three decades of the old law had not: India’s ratification of the International Labour Organization’s Minimum Age Convention, 1973 (No. 138) and Worst Forms of Child Labour Convention, 1999 (No. 182), which followed on 13 June 2017, with India specifying fourteen years as the minimum age for admission to employment under Convention No. 138, a ratification path the government had identified as an object of the reform from its introduction.

The political use of the census decline formed part of the government’s passage case and deserves a caution this article has applied to its own figures. In the Lok Sabha the minister presented the fall in child work as a fall from 12.6 million children in 2001 to about 4.5 million in 2011, a presentation reported at the time and repeated since. The two numbers do not count the same population: the 2001 figure counts main and marginal workers together, while the lower 2011 figure corresponds to the census’s main workers alone, the main and marginal total for 2011 being the 10.1 million figure used in this article. The trend the minister cited is real, since both series fell, but the comparison as presented overstates its steepness by changing the measure midway, and the episode is a small demonstration of why this field’s numbers must always be quoted with their definitions attached, a discipline the map article in this series states for the whole cluster and which the amendment’s own debate repeatedly breached on both sides.

The critics’ case was made earliest and most authoritatively by the Standing Committee on Labour, whose objections have been quoted earlier in this article, and it was made most publicly by Kailash Satyarthi, the founder of Bachpan Bachao Andolan and a Nobel Peace laureate, in his statement on the bill’s passage. Satyarthi called the amendment a missed opportunity and said that despite its progressive elements its lacunae were self-defeating: the definition of family and family enterprise was flawed, the bill used Indian family values to justify the economic exploitation of children, and it blurred the line between learning in a family and working in a family enterprise. He argued that the bill reinforced the status quo by hindering the socio-economic mobility of the marginalised and furthering rigid norms of social hierarchy, since the children who learn a family trade under the exception are, in the structure of Indian production, the children of beedi rollers, carpet weavers, bangle makers and agricultural labourers, for whom the family trade is the poverty the school exists to interrupt; that the exception opened a back door through which large numbers of children would enter the workforce; and that children of any age, under the garb of family enterprises, could in his words legally work in brick kilns, slaughterhouses, beedi making and glass furnaces. In a further interview he described the amendment as regressive. UNICEF India stated its position in the week of passage: it strongly recommended the removal of children helping in family enterprises from the bill, arguing that the exception would leave children exposed to exploitation in invisible forms of work, to trafficking and to dropping out of school through long hours of work, and it urged an exhaustive list of hazardous occupations with a system for periodic review in place of the contracted Schedule. Members in the parliamentary debates added the verification point in concrete form, pointing to mica mining in the producing states, where collection is organised through families and small units, and arguing that the State has no means of verifying whether every child found working forms part of the family owning the enterprise, so the exception functions as an escape route for anyone employing children behind a family facade.

The two cases are not symmetrical in kind, and the honest assessment has to notice it. The government’s strongest argument is about the behaviour of the legal system: unenforceable prohibitions decay, and a law that pretends to ban what it cannot reach teaches disregard for the parts it can reach. The critics’ strongest argument is about the location of the harm: the exception is not distributed randomly across the economy but is concentrated, by its own terms, in the household production where the census and every survey locate the bulk of child work, so its width is not a marginal leak in the prohibition but a channel cut through its centre. Both arguments can be true together, and the record since commencement supports elements of each. The narrowing of the family exception by the 2017 Rules, confining family help to three hours a day, excluding any part of a manufacturing, production or supply process and requiring that the child’s education be unaffected, is an administrative concession that the enacted clause was wider than its defenders’ examples suggested; and Satyarthi’s own stated position shifted after those rules, when he told The Hindu that he had not been happy earlier but that the rules had addressed the weaknesses, since on the rules the interpretation was that a child could help only in a few occupations and hazardous occupations stood clearly defined. The trajectory is instructive for an amendment trail: the enacted text of 2016 drew the widest exception, the rules of 2017 cut it back, and the political contest moved accordingly from the statute book to the rule book, where the schedule question also waits on notification under Section 4.

The claim that the amendment legalised child labour, which circulated widely in commentary on the bill, can be answered precisely, as the brief for this article requires. It overstates the change in one direction and understates it in another. It overstates it because the prohibition below fourteen became general for the first time, because the exceptions are conditions-laden rather than open, because hazardous work is outside the family exception altogether, and because the penalty for employing a child doubled while the offence became cognizable. It understates it because the exception as enacted, before the rules confined it, covered the organisational form in which most child labour in India occurs, defined that form to include enterprises engaging outside labour, and attached conditions that no inspectorate can test inside a home. Precision about both halves is the only version of the argument worth having, and it is the version the text supports.

The conventions test: what ratification required and what it left open

Since the amendment’s defenders rested part of its case on the international instruments, the trail should record what those instruments actually required, because ratification is a test with a text. The Minimum Age Convention, 1973 (No. 138) requires a ratifying state to specify a minimum age for admission to employment, not less than the age of completion of compulsory schooling and in any event not less than fifteen, with a developing-economy facility to specify fourteen initially, the facility India used when it specified fourteen on ratification. The Convention also contemplates, in its article on light work, that national laws may permit the employment of children in the band just below the minimum age in work which is not likely to be harmful to their health or development and does not prejudice their attendance at school or their capacity to benefit from instruction. The family exception of Section 3(2)(a) resembles that light-work idea in its conditions, schooling protected, hazard excluded, but differs from it in two respects the ratification debate largely passed over: it is not confined to the upper part of the band, reaching a child of eight as readily as a child of thirteen, and its operative limits, the three-hour day and the exclusion of manufacturing and supply processes, live in the rules of 2017 rather than in the Act that was ratified against. The Worst Forms of Child Labour Convention, 1999 (No. 182) requires the elimination of the worst forms for all persons below eighteen, including work which by its nature or circumstances is likely to harm the health, safety or morals of children, with the types of such work to be determined by national laws or regulations after consultation with employers’ and workers’ organisations. The three-head Schedule is India’s determination for the labour statute’s purposes, and its critics’ deepest point can be restated in the Convention’s own terms: a determination that reaches the adolescent in the factory and misses the adolescent in the tobacco field is a determination whose adequacy the Convention leaves to national judgement but whose factual basis the national record contradicts. Ratification followed in June 2017 because the conventions’ architecture is one of specified ages, national determinations and progressive elimination rather than uniform prohibition, and the amended Act fits that architecture. Fitting it is a real achievement and a limited one, and the amendment’s trail is incomplete if either half of that sentence is dropped.

Verdict: tightened at the edges, opened at the centre, and still standing

An amendment trail ends by judging the direction of travel, and the record laid out above supports a verdict with two limbs. The amendment genuinely strengthened the statute in its prohibitory form, its penalties, its procedure and its machinery: a general prohibition in place of a list, a protected category extended to eighteen for hazardous work, minimum imprisonment doubled, cognizable offences, a District Magistrate charged with implementation, periodic inspection made a statutory duty, compounding confined to the minor and the familial, and a rehabilitation fund with a government contribution written into the Act rather than left to scheme design. It also completed the statute’s external alignment, first with the education Act through the definitional reference, and then with the two International Labour Organization conventions ratified in June 2017. Measured as a text against the text of 1986, this is one of the more substantial single amendments in Indian labour legislation. Against that stands the trail’s other limb. The exception the amendment carries is addressed to the household, and its enacted definition of a family enterprise, work performed by the members of the family with the engagement of other persons, describes the characteristic employer of Indian child labour rather than an edge case beside it. The schedule contraction moved the adolescent’s hazard protection from a working description of Indian informal industry to a borrowed definition from the factories code, and the power to restore the perimeter was delegated to notification. The amendment therefore chose, as its own defenders said, enforceability over coverage, and the series thesis this article set out to test is borne out by the text: an exception drafted narrowly on its face can be wide in operation because of where the conduct actually occurs, and this is the clearest instance in the series because the conduct’s location is not a matter of inference but of the census’s own distribution.

The mica question shows the schedule and the exception interacting in the way the critics predicted and the text, read strictly, answers. Mica collection in the producing belts of Jharkhand and Bihar is organised substantially through families and small informal units, and members in the 2016 debates argued that the family exception would operate as an escape route there, since no inspector can verify whether a child collecting mica is helping a family enterprise or working for a trader’s chain. Read against the substituted text, however, mining is the first head of the new Schedule, and the family exception by its own terms does not extend to hazardous occupations or processes set forth in the Schedule, so a child helping in a mine, including a mica mine, stands outside the exception altogether, and the Ministry of Labour and Employment has maintained exactly that position in answers to Parliament, that children are prohibited from helping in mines including mica mines. The text therefore answers the mica case in the prohibition’s favour. What the text cannot answer is the verification problem the members actually raised, which is not whether mica work is prohibited but whether its prohibition can be proved against a household’s account, and that problem is the exception’s general problem appearing at a site where the exception does not even apply, which suggests the deeper difficulty is evidentiary rather than definitional.

Two closing matters belong to the trail rather than to the controversy. The first is survival. When Parliament consolidated the central labour statutes into four codes, on wages, industrial relations, social security and occupational safety and health, twenty-nine enactments were subsumed and repealed, and the child and adolescent labour statute was not among them, so it continues as a free-standing law outside the code architecture, unrepealed and unconsolidated. The comparison between the code statutes and the laws they replaced, and the treatment of the survivors, is carried in the labour codes comparison, whose record this article confirms from the child labour side: the statute’s survival means its amendment trail remains a live one, and the questions the 2016 settlement deferred, the schedule’s perimeter and the exception’s width, will be answered if at all by notification, by rules and by further amendment of this Act rather than by any code. The second matter is what the trail predicts. The pressure points are identifiable because both sides named them: demands to expand the Schedule by notification towards the named hazards of the informal economy, of which tobacco and beedi work are the standing examples; demands that the family exception be reined in by statute rather than by rules, so that its limits do not rest on subordinate legislation a ministry can relax; and the standing administrative question of whether district-level responsibility changes the detection ratios the Ministry reported for 2018. A statute that was amended once in thirty years was then amended comprehensively, and the pattern of its first amendment suggests the second will be fought on the same ground, the width of the exception and the length of the list, because those are the two choices the 2016 text made that its critics never accepted and its defenders never quite denied.

Frequently Asked Questions

Q: What did the Child Labour Amendment 2016 change?

It replaced a list-based prohibition with a general one, and then it qualified the general prohibition in two ways. Under Section 3 as substituted, no child below fourteen may be employed or permitted to work in any occupation or process, where before the ban applied only to the occupations and processes named in the Schedule’s eighteen occupations and sixty-five processes. It created the category of an adolescent, fourteen to below eighteen, whose employment is prohibited in hazardous occupations and processes under the new Section 3A, and retargeted Part III’s regulation of working hours, weekly holiday, notice, registers and health and safety to that age band. It cut the hazardous schedule to three heads, doubled the penalties for employing a child or adolescent, made employer offences cognizable, gave parents a first-offence immunity, added compounding by the District Magistrate, charged the District Magistrate with implementation, and created the district-level rehabilitation fund. The statute was renamed the Child and Adolescent Labour (Prohibition and Regulation) Act, 1986.

Q: Can a child work in a family business?

In narrowly defined circumstances, and the narrowness is contested. Section 3(2)(a) permits a child below fourteen to help the family or a family enterprise, which is other than any hazardous occupations or processes in the Schedule, after school hours or during vacations. The statute defines family as parents, siblings and paternal and maternal uncles and aunts, and family enterprise as work, profession, manufacture or business performed by family members with the engagement of other persons. The 2017 Rules tightened the exception substantially: the work must not affect the child’s education, must not be hazardous, must form no part of a manufacturing, production or supply process, and may occupy no more than three hours a day. So helping in a family shop or on a family farm outside school time can be lawful, but placing a child in family-based carpet weaving, beedi rolling or furnace work is not, and any inspector can test the hours and the nature of the work where it is visible.

Q: Who is an adolescent under the child labour law?

An adolescent is a person who has completed his fourteenth year of age but has not completed his eighteenth year. The category did not exist before the 2016 amendment, which inserted the definition into Section 2 and built the statute’s second chapter of protection around it. An adolescent may work, unlike a child, but the work must be non-hazardous: Section 3A prohibits the employment of adolescents in the hazardous occupations and processes set forth in the Schedule, which since 2016 means mines, inflammable substances or explosives, and hazardous processes as defined in the Factories Act, 1948. Where the work is lawful, Part III of the Act regulates its conditions, capping spells of work at three hours before a rest interval, spreading the working day to no more than six hours including rest, barring night work between 7 in the evening and 8 in the morning, and requiring a weekly holiday, notice, a register and health and safety compliance.

Q: Why was the hazardous occupations list reduced?

The official position is that the old Schedule had grown unwieldy and arbitrary through decades of notifications, with entries that overlapped and a structure, eighteen occupations in Part A and sixty-five processes in Part B, that nobody applied consistently, and that a short schedule anchored to genuinely hazardous work, mines, inflammable substances or explosives and hazardous processes borrowed from the Factories Act, 1948, is more enforceable, especially since the Central Government can add to or omit from it by notification on the Technical Advisory Committee’s advice. The critics’ position, pressed by the Standing Committee on Labour and by child rights organisations, is that the old list was a working description of where children were actually found, beedi making, carpet weaving, glass furnaces, tanning, rag picking, and that the substitution protects the organised factory sector’s hazards while dropping the named hazards of the informal economy, where adolescents overwhelmingly work. Both positions concern the same three lines; the notification power decides which prevails in practice.

Q: What punishment does employing a child attract?

Employing a child in contravention of Section 3, or permitting the child to work in such contravention, draws imprisonment of not less than six months and up to two years, or a fine of not less than twenty thousand rupees and up to fifty thousand rupees, or both, under Section 14(1) as substituted. The same range applies to employing an adolescent in contravention of Section 3A under Section 14(1A). A repeat offence draws imprisonment of not less than one year and up to three years under Section 14(2), with no fine alternative in the provision. Before the amendment the first offence drew three months to one year and a fine of ten thousand to twenty thousand rupees, so every figure in the principal penalty roughly doubled. The offence by an employer is cognizable under Section 14A, which lets the police register and investigate it without waiting for an official complaint. Parents are treated separately, with a first-offence immunity and a commercial-purposes proviso in place of the employer’s full range.

Q: Did the child labour law survive the four labour codes?

Yes. When the four labour codes on wages, industrial relations, social security and occupational safety and health were enacted, they subsumed and repealed twenty-nine central enactments between them, and the Child and Adolescent Labour (Prohibition and Regulation) Act, 1986 was not among them. It continues as a free-standing statute, unrepealed and unconsolidated, so its prohibitions, its penalty chapter and its machinery of the District Magistrate, inspection and rehabilitation funds operate outside the code architecture. That survival matters for anyone reading the statute’s future: the questions the 2016 settlement deferred, the hazardous schedule’s perimeter and the family enterprise exception’s width, will be answered, if at all, by notification, by rules and by further amendment of this Act rather than by any code provision. The statute’s long title after 2016 names the whole of its present ambition, the prohibition of engagement of children in all occupations and of adolescents in hazardous occupations and processes.

Q: Can a child work as an actor or performer?

Yes, under regulation rather than prohibition. Section 3(2)(b) excepts a child who works as an artist in an audio-visual entertainment industry, including advertisement, films, television serials or any such other entertainment or sports activities except the circus, subject to prescribed conditions and safety measures, with a proviso that the work must not affect the school education of the child. The circus alone is excluded, a choice that tracks its record as a site of trafficked and bonded child performers. The 2017 Rules supplied the prescribed conditions: the producer must obtain the District Magistrate’s permission before the child participates, the child may not work more than five hours in a day or more than three hours without rest, continuous engagement beyond twenty-seven days is barred with schooling protected throughout, at least twenty percent of the child’s earnings must be deposited in a fixed deposit in the child’s name payable on majority, and productions must carry a disclaimer that no abuse, neglect or exploitation of the child occurred during the shooting.

Q: What is the child labour rehabilitation fund?

It is a district-level fund that attaches money to the rescued child, created by the new Section 14B. The appropriate government must constitute the Child and Adolescent Labour Rehabilitation Fund in every district or for two or more districts, into which the fines realised from employers are credited. For each child or adolescent for whom a fine has been credited, the government itself must credit fifteen thousand rupees more to the fund, and the accumulated amount with interest on its deposit or investment is paid to the child or adolescent in whose favour it was credited, in the prescribed manner. The idea is not new to the statute: the Supreme Court’s 1996 directions in M.C. Mehta v. State of Tamil Nadu ordered employers to pay twenty thousand rupees per child into a Child Labour Rehabilitation-cum-Welfare Fund and directed a state contribution where an adult family member could not be given employment instead. Section 14B converts that judicial architecture into a statutory one, with the employer’s payment taking the form of the realised fine.

Q: When did the Child Labour Amendment Act 2016 come into force?

The bill’s passage ran through July 2016: the Rajya Sabha passed it on 19 July and the Lok Sabha on 26 July 2016, after debates dominated by the family exception and the schedule. The President gave assent on 29 July 2016, and as enacted the statute is Act 35 of 2016, published in the Gazette of India on 30 July 2016. Its commencement was left to notification, and the Central Government appointed 1 September 2016 as the date of coming into force by notification S.O. 2823(E) dated that day. The passage timeline is longer than the July dates suggest: the vehicle was the Amendment Bill of 2012, introduced in the Rajya Sabha in December 2012, examined by the Standing Committee on Labour, carried across a change of government, reshaped by the official amendments the Cabinet approved in May 2015, and only then passed. Three years and nine months passed between introduction and commencement, and the text that commenced differed materially from the text introduced.

Q: Why was the principal Act renamed by the 2016 amendment?

Because the statute after 2016 governs two populations with two different rules, and the old name described only one of them. The amending Act substituted the short title so the statute of 1986 is cited as the Child and Adolescent Labour (Prohibition and Regulation) Act, 1986, and it substituted a new long title describing the statute as an Act to prohibit the engagement of children in all occupations and to prohibit the engagement of adolescents in hazardous occupations and processes and the matters connected therewith or incidental thereto. The adolescent category, the substituted Section 3A, the retargeting of Part III from children to adolescents and the new penalty provisions for adolescent employment all belong to the second population, and a short title that said only child labour no longer described the Act’s field. The year 1986 stayed in the short title because the statute remains the 1986 enactment as amended, and its amending vehicle, Act 35 of 2016, is the year’s count in the numbering of central Acts.

Q: What work may an adolescent lawfully do?

An adolescent may work in any occupation or process that is not hazardous, which is the inverse of the child’s position. Section 3A prohibits the adolescent’s employment only in the hazardous occupations and processes set forth in the Schedule, and a proviso lets the Central Government specify by notification the nature of the non-hazardous work to which an adolescent may be permitted, giving the government a continuing instrument for shaping the field without fresh amendment. All lawful adolescent work sits under Part III’s regulation: spells of work not above three hours before an hour’s rest, the working day including rest not spread over more than six hours, no work between 7 in the evening and 8 in the morning, no overtime, no second establishment on the same day, one whole day of weekly holiday, notice to the Inspector, a register of adolescents employed and a displayed abstract of the prohibition and penalty sections. Mines, work with inflammable substances or explosives and the hazardous processes of the Factories Act definition are barred without qualification, and the occupier’s duties bite in registered establishments, thinning out in informal ones.

Q: What happens to a child who is found working?

The inspectorate and the police remove the child, and the law then divides the case into two tracks. The employer track runs through the penalty chapter: prosecution for the cognizable offence under Section 3, fines realised from the employer credited to the district rehabilitation fund, and, where the accused is the employer and the offence is a first offence of the residuary kind, possible compounding by the District Magistrate. The child track runs through rescue and care: Section 14C provides that the child or adolescent employed in contravention of the Act and rescued shall be rehabilitated in accordance with the laws for the time being in force, which in practice means production before the Child Welfare Committee under the juvenile justice statute, since a child found working in contravention of labour laws is a child in need of care and protection, with inquiry, placement and restoration to follow. The fund’s money follows the child, with the government’s fifteen thousand rupees per credited fine and the interest on the deposit, payable to the child in the prescribed manner, and the school entitlement under the education Act supplies the admission and training route back to class.

Q: Can a child be employed as a domestic worker after the 2016 amendment?

No, not as a hired worker. The general prohibition in Section 3(1) covers any occupation or process, and the only doors out of it are the two exceptions, neither of which admits employment as a domestic servant in someone else’s household. The family exception protects a child helping the family or a family enterprise, not a child placed to work in a stranger’s home, and the artist exception does not extend to domestic service. The position before the amendment had been a list-based one: the employment of children as domestic workers and in dhabas, restaurants and similar establishments was brought within the Schedule by notification with effect from 10 October 2006, which aimed the prohibition at the urban household but still required the prosecution to prove the child was in a listed occupation. The amendment removed that requirement entirely. Where the household claims the child is a relative helping the family, the dispute turns on the statutory definition of family, parents, siblings and paternal and maternal uncles and aunts, and on whether the time and hazard conditions of the exception are met.

Q: How many hours may an adolescent work in a day?

The answer comes from Part III, which the 2016 amendment retargeted to adolescents. No adolescent may work for more than three hours before an interval of rest of at least one hour, and the period of work including the rest interval may not spread over more than six hours. No adolescent may work between 7 in the evening and 8 in the morning, or on overtime, or in two establishments on the same day. Every adolescent employed is entitled to one whole day as a weekly holiday. These limits describe the registered establishment that maintains the register, gives notice to the Inspector and displays the abstract of the law, which the occupier’s duties require. For a child below fourteen there is no hours regime at all, because lawful presence at work exists only inside the family-help and artist exceptions: the 2017 Rules cap family help at three hours a day and artist work at five, but those caps are enforced by complaint rather than by the inspection routine that watches registered establishments. The hours architecture is thus two-tiered by design and two-tiered by visibility.

Q: What did the Standing Committee on Labour object to in the amendment bill?

The committee, examining the 2012 bill under the chairmanship of Dara Singh Chauhan and reporting in December 2013, placed the two defining objections to the reform on the parliamentary record. On the family exception it said the Ministry was itself providing loopholes by providing the provision, and that the committee could not understand how the Ministry proposed to keep a check on children working in their homes. On the government’s alignment justification it answered that while aligning with the Right to Education Act, the Ministry had missed the spirit of that Act, since the education statute’s protection of schooling does not map onto the labour bill’s clean division of ages as the bill assumed. The committee also criticised the regression in the definition of hazardous work, noting that the bill reached back to the 1948 Factories Act’s list rather than carrying forward the 1986 Act’s own accumulated schedule. These objections, made by the government’s own parliamentary committee three years before passage, became the template of the entire subsequent debate.

Q: Can parents be punished if their child is found working?

In the ordinary case, no. The amended penalty chapter builds a separate track for parents and guardians. Section 14(1B) provides that, notwithstanding the employer’s liability, the parents or guardians of any child or adolescent referred to in Sections 3 or 3A shall not be liable to punishment in the case of the first offence. The provisos to Sections 14(1) and 14(1A) add that parents or guardians shall not be punished at all unless they permit the child for commercial purposes in contravention of Section 3, and correspondingly under Section 3A for an adolescent. Where a parent or guardian has been convicted of an offence under Section 3 or Section 3A and commits a like offence afterwards, Section 14(2A) limits the punishment to a fine which may extend to ten thousand rupees, against the one to three years of imprisonment the repeat-offence provision imposes on others. The construction was part of the 2015 official amendments rather than the bill as introduced: the State punishes the employer who profits from child work and declines to imprison poor parents, while keeping a penal answer for parents who commercialise the child’s labour.

Q: Can an offence under the child labour Act be compounded?

Some of them can. The new Section 14D provides that the District Magistrate may, on the application of the accused, compound any offence committed for the first time under the residuary penalty provision, and any offence committed by an accused person who is a parent or a guardian, in the prescribed manner and on payment of the prescribed amount to the appropriate government. Where the composition precedes prosecution, no prosecution may be instituted for that offence against that offender, and where it follows the institution of prosecution, the composition is brought in writing to the notice of the court, which discharges the person on approval. The principal offences of an employer under Sections 3 and 3A, and any repeat offence, are outside compounding. The allocation is deliberate: the escape hatch is open to first offenders in the minor category and to parents and guardians, whose offences the scheme already treats leniently, and closed to the employer facing the main prohibition, so that the doubling of the penalties in the same amending Act is not undone at the District Magistrate’s desk.

Q: Why does the definition of a child refer to the Right to Education Act?

To bind the two statutes’ ages together so that the labour law follows the education law automatically. Section 2 defines a child as a person who has not completed his fourteenth year of age or such age as may be specified in the Right of Children to Free and Compulsory Education Act, 2009, whichever is more. The immediate effect is the alignment the amendment was designed to produce: since the education Act guarantees free and compulsory education to children of six to fourteen, the labour statute prohibits employment to the same age, so a child entitled to be in school cannot lawfully be at work. The deeper effect is the whichever-is-more clause. If Parliament were to raise the age of free and compulsory education, the labour prohibition’s boundary would rise with it without any amendment to this Act, because the definition borrows the education statute’s age rather than copying it. The committee’s objection that alignment misses the education Act’s spirit remains on the record, but the definitional device itself is the amendment’s most durable piece of drafting, and it is the mechanism through which the ban below fourteen stays consistent with the schooling promise.

Q: What does the District Magistrate do under the amended Act?

The amendment made the District Magistrate the district’s named owner of the statute’s enforcement. Section 17A lets the appropriate government confer powers and impose duties on the District Magistrate to ensure the Act’s provisions are properly carried out, with power to specify a subordinate officer for their exercise within prescribed local limits. Section 14D gives the District Magistrate the exclusive power to compound the compoundable offences on the accused’s application. Under the 2017 Rules, the producer of any audio-visual production or commercial event involving a child artist must obtain the District Magistrate’s permission beforehand, with an undertaking naming the participants, parental consent and the individual responsible for the child’s safety. Section 17B’s duty of periodic inspection and monitoring rests with the appropriate government but is discharged in the district through the same office. The design substitutes one accountable officer for a diffuse inspectorate, at the cost of loading the new duties onto an office that already carries the district’s general administration, which is why the statute’s detection record remains the honest measure of the design.

Q: What did the Supreme Court order in the Sivakasi child labour case?

In M.C. Mehta v. State of Tamil Nadu, decided on 10 December 1996 and reported at (1996) 6 SCC 756, the Supreme Court addressed child labour in the match and fireworks industry of Sivakasi and laid down directions for the whole country. It held that the employment of children in hazardous industries violated their fundamental rights, located the cause of child labour in poverty, and built a rehabilitation architecture by judicial order: an employer found employing a child in hazardous work had to pay twenty thousand rupees per child into a Child Labour Rehabilitation-cum-Welfare Fund for the child’s benefit; the State had to secure employment for an adult member of the child’s family in place of the child, or failing that contribute five thousand rupees per child to the fund; and the child had to attend school, with the fund’s income supporting that schooling. The judgment’s construction, employer money plus a government contribution per child, with proceeds tied to the rescued child, is the direct ancestor of Section 14B’s statutory fund of 2016, and the case remains the authority for treating a child found at work in an establishment as the occupier’s responsibility to explain.