Contract Labour Act 1970: Regulation Without Absorption

The Contract Labour Act 1970 carries its ambition in its title and its disappointment in its text. The full name is the Contract Labour (Regulation and Abolition) Act, 1970, and a reader who takes the second half of that name at face value will expect a statute that ends the practice of engaging workers through contractors, at least in the work that an establishment performs all year round. What the statute actually built was a licensing and welfare regime for the practice it named itself against, together with a power, handed to governments and hedged with conditions, to prohibit contract labour in a particular process. Then, three decades after enactment, a Constitution Bench of the Supreme Court settled the question the text had left open, and settled it against the workers: when a government does prohibit contract labour in a process, the workers engaged in that process do not become employees of the establishment. They simply cease to be engaged. The abolition limb of the statute therefore removes jobs rather than converting them, and everything else in this article follows from understanding why that is so and what, if anything, a contract worker can do instead.

The Contract Labour Act 1970, showing how a statute named for abolition built a regime of regulation instead - Insight Crunch

That single point is the one test this article sets itself. A reader who finishes it should be able to explain the whole working model of the Act: which establishments and contractors it binds, what registration and licensing require, what the prohibition power in Section 10 actually examines, who must provide the canteen and the drinking water and the wages when each actor in the chain defaults, what penalties stand behind the scheme, and how the courts first promised absorption and then withdrew the promise. The reader should also be able to explain the structural point that matters more than any single provision: contract labour in India grew enormously under a statute enacted to abolish it, and it grew because the surrounding body of labour law made direct employment expensive and rigid at defined thresholds while contracting avoided those thresholds altogether. The statute did not fail to notice this. It was designed as a compromise, and compromises of this kind regulate the thing they cannot bring themselves to forbid.

What existed before the Contract Labour Act 1970

Contract labour was not a marginal arrangement waiting for Parliament to notice it. In the mills, mines, docks, plantations and construction sites of the early twentieth century, the contractor, known in different regions as the sardar, the mukadam, the maistry or the jamadar, was the ordinary channel through which unskilled and semi-skilled labour reached the workplace. The establishment dealt with the intermediary; the intermediary recruited, supervised and paid the workers, and took his margin out of the difference between what the establishment paid per head and what the worker received. The worker’s legal relationship, such as it was, ran to the intermediary, who owned no factory, kept no reserves, and could disappear at the end of a season leaving wages unpaid and no one answerable. Official inquiries returned to this arrangement repeatedly, and their language hardened over time.

The Royal Commission on Labour, reporting in 1931, examined the contract system across Indian industry and recommended that the practice of engaging labour through contractors be abolished, because the intermediary’s interest lay in extracting the largest possible margin from the labour he supplied and the establishment disclaimed responsibility for people it treated as another man’s employees. The Bombay Textile Labour Enquiry Committee, reporting in the late 1930s, applied the same conclusion to the industry it knew best and recommended that the contract system of engaging labour be abolished as soon as possible and that workers for every department of a mill be recruited and paid directly by the management. The Bihar Labour Enquiry Committee of 1940 condemned the practice in terms that later committees kept quoting: contractors ordinarily lacked the sense of moral obligation towards labour that employers or managers were expected to have, and did not hesitate to exploit the helpless position of the workers in their charge.

The Labour Investigation Committee of 1946, chaired by D. V. Rege and known as the Rege Committee, made the recommendation that shaped the eventual statute. It found the contract system very much in vogue, endorsed the abolition that the Royal Commission and the provincial committees had urged, and then added the qualification that became the law’s architecture: where abolition was not possible, the conditions of contract labour should be regulated, so that wages were actually paid and essential amenities actually provided. The Second Five Year Plan took up the same formula, recommending studies to identify where contract labour could be progressively eliminated, responsibility for wages and conditions to be placed on the principal employer in addition to the contractor where elimination was not feasible, and decasualisation schemes where those fit. Tripartite discussions between the Centre, the states and the two sides of industry produced a consensus in the same words: the system should be abolished wherever possible and practicable, and where it could not be abolished, regulated.

Independence did not wait for a single statute to act on these recommendations. Instead, protection arrived in fragments, statute by statute. The Factories Act 1948, the Mines Act 1952 and the Plantations Labour Act 1951 each defined the worker they protected widely enough to include a person employed through a contractor, so that hours, safety and welfare standards reached contract workers inside those workplaces even though the man who engaged them was the contractor. The Employees’ State Insurance Act 1948 reached contract labour through its definition of an immediate employer. The Minimum Wages Act 1948 applied to contract labour in scheduled employments. Each fragment protected the contract worker for one purpose and ignored him for every other: a man might be a worker for the Factories Act, insured under the insurance statute, and entitled to the scheduled minimum wage, and yet have no security of tenure, no claim to continuity, and no identified employer when the season ended and the contractor moved on. The patchwork varied by industry, by state and by category of work, and its gaps were exactly the gaps the intermediary system was built to exploit.

Two features of that pre-statute landscape explain the statute that eventually emerged. The first is that every inquiry, without exception, recommended abolition and regulation in the same breath, so the political question was never whether contract labour was defensible but how much of it a legislature was prepared to prohibit outright, and the answer, each time the question was put, was that prohibition would be discretionary and gradual rather than general and immediate. The second is that the rest of the labour code was simultaneously building the incentives that would defeat a general prohibition. The Industrial Disputes Act 1947 subjected retrenchment and closure in larger establishments to conditions and, in its later chapters, to prior government permission. The Factories Act attached its obligations to headcount thresholds. Provident fund and insurance statutes attached contribution obligations to establishments of defined sizes. Every obligation that bit at a threshold taught the rational establishment the same lesson: work performed by another man’s employees sits outside several of those thresholds at once. A statute aimed at contract labour was therefore legislating against a current that the rest of the statute book was generating, and the legislators of 1970 knew it, which is why the Act they passed regulates the current and reserves a power to dam it in particular places rather than pretending to reverse its flow.

The sectoral spread of the system matters to the statute’s design, because contract labour was never one practice but several wearing a common legal disguise. In the mines, the contractor raised labour for workings the mine owner preferred not to staff directly, and the raisings system drew repeated condemnation from mining inquiries for its wage deductions and its insecurity. In the docks, cargo was handled by labour engaged through intermediaries on a daily basis, and the state’s eventual answer in that sector was not the contract labour statute at all but the dock labour boards, which decasualised registered dock workers by pooling their employment, a model that demonstrated an alternative the general law never adopted. In construction, the contractor is not an intermediary inserted into an otherwise stable workforce but the industry’s normal organising form, the project itself being temporary even where the builder is permanent, and construction labour’s exclusion from establishment-based welfare machinery eventually forced a separate statute, as the discussion of the neighbouring statutes below records. In the plantations, the sardari system of recruitment carried workers across provinces under arrangements whose abuses stretched back to the indenture era. And in the factories, the sector the general statute would most affect, contract labour began at the periphery, in loading, packing, cleaning and canteen work, and migrated over the decades towards the production line itself.

That migration is the point at which the pre-history joins the statute’s later failure. The committees of the 1930s and 1940s were describing a system concentrated in ancillary and seasonal work, and a prohibition philosophy aimed at that system could plausibly promise abolition: the work was bounded, the numbers were counted in gangs, and direct engagement was an administrative change rather than a structural one. By the time Parliament legislated in 1970, and far more by the 1990s, contract labour had become a general staffing strategy reaching into perennial production, and abolition by notification, process by process, had become a much heavier instrument than the one the Rege Committee had imagined. The Act’s text froze the committee-era design at the moment the practice was outgrowing it. Readers who ask why the legislature did not simply prohibit contract labour in perennial work across the board should hold both halves of that history together: the consensus formula was framed when abolition looked administratively feasible, and it survived into the statute because no later consensus replaced it, not because anyone in 1970 demonstrated that discretionary prohibition would catch a practice already learning to live at the centre of production.

One state had, by then, already demonstrated a different instrument for labour of this character. Maharashtra’s Mathadi, Hamal and Other Manual Workers (Regulation of Employment and Welfare) Act, 1969 created boards for unorganised manual workers in employments such as loading and unloading, under which workers register with the board, employers requisition labour from the board’s pool, wages are routed through the board, and a levy on employers funds welfare for workers whose engagement is casual and shifting. The board model attacked the problem from the opposite end to the contract labour design: instead of licensing the intermediary and backstopping his duties against the establishment, it displaced the intermediary with a statutory pool and attached the worker’s entitlements to his registration rather than to any single engagement. The model’s reach was confined to its scheduled employments and its state, and it carried administrative costs of its own, but its existence before 1970 matters for the assessment of the central choice. Parliament legislating in 1970 was not choosing between regulation and nothing; a working Indian precedent for pool-based decasualisation existed, the dock labour boards supplied another, and the Act’s licensing design was selected over both. The selection is defensible, licensing scales nationally in a way boards did not obviously do, but it was a selection, and the article’s verdict returns to it.

The compromise Parliament enacted in 1970

The Contract Labour (Regulation and Abolition) Act, 1970 is Act 37 of 1970. It was passed by both Houses of Parliament, received the assent of the President on 5 September 1970, and was brought into force on 10 February 1971. Its long title states the design without adornment: an Act to regulate the employment of contract labour in certain establishments and to provide for its abolition in certain circumstances and for matters connected therewith. The Statement of Objects and Reasons tracked the Rege formula almost word for word. The system of contract labour, it recorded, lent itself to exploitation; the consensus of the inquiries and the tripartite bodies was that it should be abolished wherever possible and practicable and that, where abolition was not possible, the working conditions of contract labour should be regulated so as to ensure payment of wages and provision of essential amenities. The Act is that sentence converted into machinery.

The machinery has two halves, and the title names them in the right order. The regulation half occupies most of the text. It applies the Act to establishments and contractors above a numerical threshold, requires the principal employer to register the establishment, requires the contractor to hold a licence, fixes the contractor’s obligations on wages, canteens, rest rooms, drinking water, latrines, washing facilities and first aid, and makes the principal employer the backstop for every one of those obligations, with a right to recover what he spends from the contractor. The abolition half is a single section. Section 10 empowers the appropriate government, after consulting the Central Advisory Board or the State Advisory Board, to prohibit by notification the employment of contract labour in any process, operation or other work in any establishment, having regard to four stated considerations. Nothing else in the Act abolishes anything. There is no schedule of prohibited processes, no date by which perennial work must be directly staffed, and, decisively, no provision stating what happens to the workers when a prohibition issues. That silence was not an oversight discovered later. It is the space in which the entire subsequent litigation was fought, and the courts filled it first one way and then the other.

The Act is arranged in eight chapters. The opening chapter states the short title, the extent, the commencement and the definitions on which everything else turns. A short chapter creates the advisory boards. Then come registration of establishments, licensing of contractors, welfare and health of contract labour, penalties and procedure, and a closing chapter of miscellaneous provisions covering inspectors, registers, rule-making and the Act’s relationship with other laws. The arrangement matters for a practical reason: the regulation half is self-executing machinery that an inspectorate can operate, while the abolition half is a political decision dressed as a section, exercisable establishment by establishment and process by process. The first half therefore operated everywhere the Act applied, from the day it commenced. The second half operated only where a government chose to move, and governments, answerable for the employment consequences of a prohibition, moved rarely and narrowly.

The Act’s constitutional validity was challenged early and upheld. In Gammon India Limited v Union of India (1974), the Supreme Court sustained the Act and the Central Rules made under it against the contention that the scheme was unreasonable, accepting that the legislature could rationally decide to regulate the contract system and abolish it selectively rather than prohibit it generally. The decision matters to the article’s argument because it confirms that the compromise was a choice the Constitution permitted, not a drafting failure a court was waiting to correct. From 1974 onwards, the meaning of the Act would be determined by the two questions the text genuinely left open: the reach of the prohibition power, and the fate of the prohibited workers. The first of those questions the statute answered itself, in Section 10. The second it refused to answer at all.

One further feature of the original settlement deserves emphasis before the provisions are examined, because readers habitually miss it and it explains the Act’s later history. The Act regulates the employment of contract labour; it does not prohibit the engagement of a contractor as such, does not cap the proportion of an establishment’s workforce that may be engaged through contractors, and does not require that work of a perennial character be performed by direct employees. An establishment could therefore comply with every provision of the Act, register, use only licensed contractors, provide every amenity and pay every wage on time, and still perform the whole of its perennial production work through contract labour for fifty years. Compliance and abolition were never the same project. The Act’s critics and its defenders both sometimes write as though the statute promised the second and delivered the first. The text promised the first and reserved the second to executive discretion, and the distance between those two things is the statute’s real subject.

Who the Contract Labour Act 1970 binds

The Act extends to the whole of India, and Section 1(4) states its application in two limbs. It applies to every establishment in which twenty or more workmen are employed, or were employed on any day of the preceding twelve months, as contract labour. It applies also to every contractor who employs, or who employed on any day of the preceding twelve months, twenty or more workmen. The two limbs matter separately: the establishment-facing duties of registration and the welfare backstop attach through the first, and the licensing duty attaches through the second, so a contractor cannot escape the Act because his client is small, and a principal employer cannot escape it because his contractor is large. The count is of contract labour, not of the establishment’s whole workforce: a factory with five hundred direct workmen and fifteen contract workers falls outside the first limb, while a warehouse with no direct workmen at all and twenty-five contract workers falls squarely inside it.

The twenty-workman threshold was the figure in the Act as enacted, and it remains the figure in the Central text. But the brief for this article speaks of the threshold as amended in application, and that phrase captures a real feature of the statute’s life. The states, within whose sphere much of the Act’s administration lies, amended the figure upwards. Maharashtra raised the threshold from twenty to fifty by a state amendment of 2017, and Andhra Pradesh raised it to fifty by an amendment of 2015. Other states moved in the same direction through the years of the labour reform debates. The pattern is worth pausing on, because the direction of amendment is itself evidence in the assessment of the Act: given the chance to extend the Act’s coverage, legislatures chose to contract it, reasoning that the compliance burden fell hardest on smaller establishments and smaller contractors. Whatever the merits of that reasoning, the Act a reader meets on the ground in a large industrial state may bind only establishments engaging fifty contract workers, even though the bare central text on the page still says twenty.

Two provisions complete the applicability picture. The proviso to Section 1(4) lets the appropriate government, after giving not less than two months’ notice of its intention, apply the Act to establishments or contractors employing fewer than the threshold number, so the floor is a default and not a constitutional feature of the scheme. And Section 1(5) keeps the Act away from establishments in which work of an intermittent or casual nature is performed. The exclusion is defined with some care: work is not treated as intermittent if it was performed for more than one hundred and twenty days in the preceding twelve months, and seasonal work is not treated as intermittent if it was performed for more than sixty days in a year. The exclusions were aimed at genuinely occasional engagement, the harvest gang and the breakdown crew, and their drafting shows the legislature attempting, within a threshold design, to catch duration rather than labels. A process staffed through contractors for eleven months of every year does not become intermittent because the contract documents describe it as seasonal.

Who is a contract labourer under the Act?

A contract labourer is a person hired in or in connection with the work of an establishment by or through a contractor, whether or not the principal employer knows of the engagement. The classification turns on the channel of hire alone, not on the work’s character, the worker’s seniority, or the years he has served the same establishment.

A person is contract labour when he is hired in or in connection with the work of an establishment by or through a contractor, with or without the knowledge of the principal employer. The definition in Section 2(1)(b) turns entirely on the channel of engagement and not on the character of the work, the length of service or the place where the work is done, which is why a worker may serve the same establishment through the same contractor for twenty years and remain, in the statute’s eyes, contract labour on the last day as on the first. The definition expressly excludes a person regularly employed by the contractor for the contractor’s own activity and only deputed for a particular job, a carve-out for the genuine service business whose employees happen to work at a client’s premises. It also excludes the out worker, the person to whom articles or materials are given to be made up, cleaned, washed, altered, finished or adapted for sale on premises not under the control and management of the principal employer, the home-based and workshop-based piece worker of the older putting-out trades. The exclusion mattered in the beedi, garment and light engineering trades, where work moved out of establishments into homes and small workshops; those workers’ protection, such as it was, came through the welfare-fund statutes enacted sector by sector rather than through this Act, one more boundary at which the general statute stops and the patchwork resumes.

The definitions distribute the duties, so they deserve the reader’s full attention. A contractor, under Section 2(1)(c), is a person who undertakes to produce a given result for the establishment by engaging contract labour, or who supplies contract labour for any work of the establishment, and the term includes a sub-contractor, so the chain cannot be lengthened to break the definition. A workman, under Section 2(1)(i), is any person employed in or in connection with the work of an establishment to do skilled, semi-skilled, unskilled, manual, supervisory, technical or clerical work for hire or reward, excluding managerial and administrative staff and supervisory staff above a wage ceiling fixed in the original text at five hundred rupees a month, a figure whose stagnation across five decades quietly narrowed the definition’s reach. The principal employer, under Section 2(1)(g), is the person responsible for the control of the establishment: the head of a government department where the establishment is a government office, the owner, occupier or manager where it is a factory, the owner or agent where it is a mine, and in every other case any person having ultimate control over the affairs of the establishment. An establishment itself is defined broadly, covering offices and departments of government and local authorities, any place where industry, trade, business, manufacture or occupation is carried on, which sweeps in service establishments and warehouses that the older factory-centred statutes never reached.

The phrase “appropriate government” carries a load in this Act out of proportion to its length. The original text made the Central Government the appropriate government for establishments controlled by or under the authority of the Centre and the state governments for the rest. An amendment of 1986, in force from 28 January 1986, recast the definition to track the Industrial Disputes Act, so that the Centre is the appropriate government for establishments owned, controlled or managed by central public sector undertakings, autonomous bodies owned or controlled by the Centre, and the like, and the state for everything else. The recasting sounds technical. Its consequences were not, because the identity of the appropriate government decides who registers the establishment, whose inspectorate inspects it, whose advisory board must be consulted before a prohibition, and, as the litigation later showed, whether a prohibition notification issued for a central undertaking was valid at all. Readers of the Steel Authority case will meet this definition again; it was one of the two points on which that judgment turned.

The advisory boards and the administrative machinery

Chapter II of the Act creates the advisory boards through which the prohibition power must pass. Section 3 empowers the Central Government to constitute a Central Advisory Contract Labour Board, and Section 4 requires each state government to constitute a State Advisory Contract Labour Board, to advise on matters arising out of the administration of the Act and referred to them, and to carry out other functions assigned by the Act. The boards are tripartite in character, drawing representation from government, from the industries concerned and from the workmen engaged as contract labour, with the statutory scheme contemplating both official and non-official membership and a rule-making power over their composition. Section 5 empowers each board to constitute committees for specific purposes.

The boards are easy to skim past, and skipping them misreads the Act. Their consultation is not decorative: Section 10 makes consultation with the board a step in the prohibition process, and the courts treated a prohibition notification issued without the requisite consultation, or by the wrong government, as vulnerable. The boards also express the statute’s founding theory, which is tripartism, the belief carried over from the plan-era committees that the pace of abolition was a question for negotiation between the state and the two sides of industry rather than for a rule of law. That theory explains both the boards and the discretion in Section 10: abolition was to be agreed process by process, in establishments where the boards found the four statutory considerations satisfied, and not decreed across industry at a stroke. Whether that theory was realism or abdication is the question the assessment at the end of this article takes up; its presence in the design is not in doubt.

In operation, the boards’ composition explains their record. Their members are drawn from the government of the day, from the industries in which contract labour is engaged and from the workmen, and their recommendations carry no binding force; a board can find the four factors satisfied for a process and the government remains free not to notify a prohibition. Where the board system functioned as designed, it produced the evidentiary groundwork for notifications, surveys of conditions in particular processes, findings about comparators in similar establishments, which the government then weighed politically. Where it did not function, consultation degenerated into a procedural step attested in the file, and prohibition demands died in committee rather than in decision. The design placed the abolition question in the one forum in Indian public life structurally inclined to compromise, a tripartite table at which the employers of contract labour sat as members, and the reader who wonders why the power slept for long stretches need look no further than the forum through which a demand had to pass before it reached a government that was itself, in the public sector, an interested party.

Below the boards stands the operational machinery: registering officers appointed under Section 6, licensing officers appointed under Section 11, and inspectors appointed under Section 28, whom the appropriate government may invest with powers of entry, inspection, examination and seizure appropriate to an inspectorate statute. The division of labour is plain. The registering officer receives the principal employer’s application, issues the certificate of registration and may revoke it. The licensing officer receives the contractor’s application, grants the licence with conditions, and may revoke, suspend or amend it. The inspector visits, examines registers and records, and sets the penal machinery in motion when the regulation half is breached. The whole structure presumes a functioning state inspectorate, and the Act’s uneven record across states owes as much to the capacity of that machinery as to the text.

The counting question concealed inside Section 1(4) generated its own avoidance practice, and the reader should understand it because it shaped the Act’s real coverage. The threshold counts contract labour employed in the establishment, or by the contractor, on any day of the preceding twelve months, so an establishment cannot escape the Act by thinning the workforce on the day an inspector calls; a single day at or above the line in the year brings the establishment in. The rational avoidance was therefore structural rather than daily: split the engagement among several contractors, each below the contractor-side threshold, rotate engagements between firms, or keep each discrete contract’s headcount small while the aggregate contract workforce on the premises grew large. The establishment-side limb answers part of this, since it counts the contract labour in the establishment regardless of how many contractors supply them, and adjudicators examining coverage looked to that aggregate. But the fragmentation strategy raised the cost of enforcement in every case, because coverage had to be established from the registers and muster records of multiple intermediaries before any substantive duty could be enforced at all, and it taught establishments a lesson the successor code had to answer by fixing the establishment and the manpower supplier at a common figure of fifty and by treating the supply of manpower itself as the licensable activity.

A connected counting subtlety is the identity of the day count for seasonal industries. Sugar factories, brick kilns and salt works engage their largest workforces in a crushing or production season and run lean for the rest of the year. Section 1(4)’s reference to employment on any day of the preceding twelve months catches the season at its peak, so a sugar factory engaging forty contract workers through its crushing season is inside the Act for the year, while Section 1(5)’s intermittent exclusion, with its one hundred and twenty day and sixty day tests, decides whether the work itself is the kind the Act reaches. The two provisions working together produce a coverage pattern that follows the season rather than the calendar, which is why the seasonal industries feature so prominently in the Act’s enforcement history and why their cases so often turned on day counts reconstructed from wage records.

State rules and the federal administration of the Act

The Act is a central statute administered in two spheres, and a reader who treats it as a single national regime will misunderstand much of its record. For establishments under the Central Government as appropriate government, the Central Government’s rules of 1971 govern, its registering and licensing officers act, and its inspectorate inspects. For the far larger number of establishments in the state sphere, each state government makes the rules, appoints the officers, fixes the fees and security deposits within the statute’s frame, and runs the inspectorate. The text is one; the administered law is many. Licence conditions, fee scales, the forms of registers and the energy of inspection vary from state to state, and the same contract, lawful in its paperwork in one state, may be defective in the next for reasons located entirely in the rules.

This federal division explains features of the Act’s history that a purely textual account misses. The upward threshold amendments were state acts because the states held the administrative burden and felt the compliance politics. The quality of the wage and welfare backstops in a given establishment depended on whether its state’s inspectorate visited, and inspection capacity varied with state finances and administrative priorities rather than with the size of the state’s contract workforce. The advisory boards, whose consultation gates the prohibition power, were correspondingly uneven institutions: in some states they met, examined processes and recommended prohibitions; in others they were constituted on paper and consulted as a formality. The prohibition record of the Act is thus partly a record of state administrative capacity, and the Supreme Court’s insistence, in the Steel Authority litigation, on the identity of the appropriate government was an insistence on taking the federal structure seriously rather than treating India as one labour jurisdiction.

The division also shaped the drafting of the successor code. The OSH Code’s designers, writing with five decades of this experience behind them, standardised the registration and licensing core, moved applications into electronic form, and provided for a licence capable of operating across the contractor’s work rather than fragmented establishment by establishment, while leaving administration divided between Centre and states as the Constitution’s distribution of the labour subject requires. The reader comparing the two regimes should therefore look past the thresholds, which attract the headlines, to the administrative machinery: the 1970 Act’s deepest weakness was never the content of its duties but the variance of the apparatus meant to enforce them, and the Code’s wager is that uniformity of process will succeed where uniformity of text did not.

Registration of establishments employing contract labour

Chapter III places the first regulatory duty on the principal employer. The appropriate government appoints registering officers and defines the limits within which each exercises his powers. The principal employer of an establishment to which the Act applies must apply for registration in the prescribed form, within the period the appropriate government fixes, and the registering officer, if satisfied that the application is complete, registers the establishment and issues a certificate of registration. Section 8 permits revocation where the certificate was obtained by misrepresentation or suppression of material fact, or where its conditions have ceased to exist or have been contravened, after giving the principal employer a reasonable opportunity to show cause, and with the prior approval of the appropriate government.

Section 9 states the consequence of non-registration with unusual directness: a principal employer who has no certificate of registration in force cannot employ contract labour in the establishment. The provision converts registration from a census into a licence to use the contract system at all, and its placement matters. The Act could have punished non-registration with a fine and left the engagement lawful. Instead it made the unregistered engagement itself unlawful, so that a principal employer who skips the first duty of the regulation half is in breach from the first day the first contract worker walks through the gate. In practice, enforcement of Section 9 depends on inspection, and the registers of contract labour establishments in many states remained thin compared with the visible scale of contracting, a gap between the statute’s coverage and its registration record that official reviews of the Act noted repeatedly and that readers should keep in mind when registration figures are quoted as a measure of the system’s size.

Registration also performs a quieter function in the Act’s logic. It identifies, on a public record, the establishment that stands behind the contractor as the statute’s backstop. The welfare and wage chapters do not ask who the contractor’s client is; they presume the registered principal employer of the establishment where the work is done, and the registration certificate is what ties that legal identity to a gate, a premises and an occupier an inspector can visit. The amendment of particulars, the display of the certificate and the maintenance of registers under the rules all serve that same end: fixing the establishment as a known point of responsibility in a system whose daily reality is that the immediate employer is a firm with a licence, an office that may move, and assets that may not survive a bad season.

Licensing of contractors

Chapter IV turns to the contractor himself. Section 12 provides that no contractor to whom the Act applies shall undertake or execute any work through contract labour except under and in accordance with a licence issued for the purpose. The licensing officer appointed under Section 11 receives the application, made in the prescribed form and containing particulars of the establishment, the nature of the process, operation or work for which contract labour is to be employed, and the duration for which the engagement is proposed. Section 13 governs the grant: the licence is issued on payment of the prescribed fee and on the deposit of security for the due performance of its conditions, it is valid for the period specified in it, and it may be renewed. The conditions the statute itself names go to the substance of the engagement: hours of work, fixation of wages and other essential amenities to be provided to the contract labour. The licence is thus not a mere trade permission. It is the instrument by which the terms on which contract labour may lawfully be supplied are written down in advance and attached to a named contractor, a named establishment and a defined number of workmen.

Section 14 empowers the licensing officer to revoke, suspend or amend a licence where it was obtained by misrepresentation or suppression of material fact, where its conditions have been contravened, or where the contractor has failed without reasonable cause to comply with its conditions or has contravened the Act or the rules. Section 15 gives a person aggrieved by an order under the licensing chapter an appeal to the appellate officer within thirty days. Revocation is the regulation half’s sharpest weapon against a contractor, because a contractor without a licence cannot lawfully execute work through contract labour at all, but the weapon has a recoil the reader should notice: cancelling the licence stops the lawful engagement, and the workers engaged under it are the people standing in the workplace when it falls. The Act addresses that problem in the wage and welfare chapters by fastening continuing duties on the principal employer, but the employment itself, the engagement through that contractor, ends. The regulation half can discipline the intermediary only by interrupting the livelihood that runs through him, which is one more reason inspectorates historically preferred fines, warnings and renewal conditions to cancellation.

The security deposit attached to the licence deserves a word, because it expresses the chapter’s theory of the contractor. The deposit, scaled under the rules to the number of workmen for whom the licence is sought, stands as a financial hostage for the licence conditions: a contractor who defaults on wages and amenities risks not only prosecution and revocation but the fund he lodged to obtain the licence. The theory is that the intermediary, whose weakness in the old system was precisely his lack of substance, must post substance as the price of admission, and the licensing officer holds it. In practice the deposit amounts fixed under the rules, like the penalty figures, aged badly, and a deposit set at the scale of a small engagement offered thin security against the wage bill of a large one; but the design idea, that access to the labour supply business should be conditioned on demonstrated financial standing, anticipates the bonding and security devices that later regulatory schemes in the labour field and beyond adopted as standard.

The paperwork of this chapter is heavier than it looks from the statute, because the rules prescribe the forms for applications, licences, certificates, notices and registers, prescribe the security deposit scaled to the number of workmen, and require the contractor to issue employment cards and wage slips and to maintain registers of the persons he employs. An establishment that runs this paperwork properly generates a documentary trail connecting each worker to a contractor, a licence, a wage rate and a disbursing representative, and that trail is frequently the only evidence the worker will ever have. Principal employers and contractors who keep these records in order find it worthwhile to open and check the underlying licence conditions, registration certificates and wage documentation with care, and reviewing the licence and wage documents directly in an office file viewer serves exactly that kind of checking, letting a reader inspect the files in which licences, registers and wage records are actually kept rather than relying on summaries of them. The point is not clerical. When a dispute reaches an adjudicator, the question is often what the paper says, and in this field the paper is usually generated by the side that later denies the relationship.

Registers, records and the evidentiary paper trail

The rules under the Act prescribe a documentary apparatus that deserves treatment as part of the statute’s substance rather than its stationery. The principal employer maintains a register of contractors engaged, recording the contractor’s particulars, the nature of work, and the number of workers, so that the establishment’s contract engagements exist as a single inspectable list. The contractor maintains registers of the persons he employs, issues an employment card to each contract worker recording his engagement, and issues wage slips recording the wages for each wage period, while muster and wage records document attendance and payment. The licensing and registration files hold the applications, certificates, security deposits and renewal records. Taken together, the apparatus was designed so that every element of a later dispute, who engaged the worker, at what rate, for which establishment, under which licence, and what he was paid, has a document that answers it.

The reason this matters is that contract labour litigation is evidence litigation. The sham inquiry asks who controlled the worker, and control is proved from supervision records, gate passes, muster rolls and the pattern of wage payment far more often than from testimony. The wage backstop asks what was paid, and the certified disbursement under Section 21 and the wage slips answer it. The threshold question asks how many contract workers an establishment engaged across the preceding twelve months, and the register answers it. Establishments that kept the paper trail in order litigated from strength; workers whose engagements lived entirely in the contractor’s cash book faced the burden of proving, years later, facts the statute had ordered to be recorded contemporaneously. The paper trail is therefore best understood as the regulation half’s enforcement multiplier: every duty in the Act becomes cheaper to enforce when its documentary shadow exists, and the inspectorate’s first and most productive demand on any visit was always for the registers.

The abolition half: Section 10 and the prohibition power

Section 10 is the provision on which the Act’s reputation rests, and it is short enough to paraphrase in full. The appropriate government may, after consultation with the Central Board or, as the case may be, the State Board, and by notification in the Official Gazette, prohibit the employment of contract labour in any process, operation or other work in any establishment. Before issuing the notification, the government shall have regard to the conditions of work and benefits provided for the contract labour in that establishment and to four other relevant factors. Those factors are: whether the process, operation or other work is incidental to, or necessary for, the industry, trade, business, manufacture or occupation carried on in the establishment; whether it is of a perennial nature, that is to say, of sufficient duration having regard to the nature of the establishment’s activity; whether it is done ordinarily through regular workmen in that establishment or an establishment similar thereto; and whether it is sufficient to employ a considerable number of whole-time workmen.

Read the factors as a test and their design becomes visible. Each limb pushes the inquiry towards the same underlying question: is this work, in truth, a permanent function of the establishment dressed up as a contract engagement? Work that is necessary to the establishment, performed continuously, done elsewhere by direct employees and substantial enough to occupy whole-time staff is work that the factors mark for prohibition. Work that is occasional, auxiliary, specialised or genuinely fluctuating fails one or more limbs and stays within the regulated zone. The factors are mandatory considerations and not mere recitals: the government must apply its mind to them, and a prohibition issued without regard to them, or issued by a government that is not the appropriate government for the establishment, invites judicial review. But the power remains a power. Satisfaction of all four factors does not itself prohibit anything; it empowers the government to prohibit, and a government that does nothing breaches no duty the courts were willing to enforce by directing prohibition across an industry.

Each factor repays individual attention, because prohibition litigation was fought limb by limb. The first factor, that the work is incidental to or necessary for the establishment’s activity, separates the establishment’s own functions from services that merely occur on its premises: the production process and its direct auxiliaries satisfy it readily, while a visiting service, present for its own clients across many premises, does not present the same case. The second factor, perennial nature, is the temporal test, and it is assessed against the establishment’s activity rather than the individual contract’s term, so a process staffed through a succession of eleven-month contracts for thirty years is perennial in substance whatever the documents say. The third factor imports a comparator: if comparable establishments staff the work directly, or the same establishment does so in other units, the claim that the work is inherently contract work weakens, and establishments resisting prohibition were put to evidence about practice across their industry. The fourth factor, sufficiency for a considerable number of whole-time workmen, excludes the genuinely fragmentary task, the specialist crew needed for a shutdown fortnight, from the prohibition logic, reserving the power for work substantial enough to constitute real employment. Applied together, the factors describe, with some precision, the profile of the engagement the committees had condemned: continuing, necessary, mainstream work, performed by people whose only distinction from the establishment’s own workmen is the piece of paper that names their employer. The critique writes itself: a test that identifies its object that exactly, and then leaves the consequence to discretion, had located the problem and declined to solve it.

The factors also explain the pattern of prohibitions actually issued. They operate process by process and establishment by establishment, which suits discrete, well-bounded operations: sweeping and cleaning in particular establishments, loading and unloading, watch and ward in defined premises. In December 1976 the Central Government issued a notification prohibiting contract labour in specified operations, including sweeping, cleaning, dusting and watching of buildings, in a range of central establishments, and prohibition notifications of that character, bounded and categorical within named premises, became the characteristic use of the power. What the factors do not readily produce is the transformation of a production process inside a large factory, because each limb then has to be fought over against the establishment, with its evidence about technology, fluctuation and comparators, and the consequence of losing is the displacement of the workforce engaged in that process. The power’s exercise record across five decades was accordingly modest against the growth of the practice, and the explanation is structural rather than conspiratorial: the section asks governments to make, case by case, the decision the legislature in 1970 declined to make generally.

The deeper point is the one the introduction flagged. Section 10 prohibits the employment of contract labour. It does not direct the principal employer to employ the affected workers, does not deem them his employees, does not preserve their wages or continuity, and does not even require the employer to prefer them when he staffs the work directly. A complete, lawful, unimpeachable exercise of the abolition power therefore leaves the workers it concerns with fewer rights the day after the notification than the day before, when at least the regulation half attached to their engagement. This was visible on the face of the Act from 1971 onwards, and it is why the labour movement’s litigation strategy moved so quickly from demanding prohibitions to demanding absorption as their automatic consequence, a demand the text does not contain and which the courts would eventually refuse to find in it.

Prohibition in practice: what the notification record shows

The prohibition power’s biography can be read from the notifications themselves. The characteristic prohibition is specific: a named operation, in named establishments, within the issuing government’s sphere. The Central Government’s notification of 9 December 1976, the best known of the early exercises of the power, prohibited contract labour in operations including sweeping, cleaning, dusting and watching of buildings in a range of central establishments, and state notifications followed the same pattern into canteens, loading and unloading, security and housekeeping functions in government offices, public sector premises and large industrial establishments. The pattern confirms the reading of the factors offered above: the power worked where the work was bounded, easily described and detachable from the production process, and where the establishment, usually a public one, could absorb the administrative consequence of staffing the function directly.

The production floor of a factory is where the record thins. A prohibition there required the government to find, against the establishment’s evidence, that the specific process was necessary to the industry, perennial, ordinarily done by regular workmen elsewhere and sufficient for considerable whole-time employment, and then to accept the political cost of the displacement or conversion that followed. Governments of every complexion rationed such decisions, and the advisory boards, the statutory forum for the question, processed demands slowly and establishment by establishment. The demands themselves were substantial: unions pressed prohibition claims for perennial processes in steel, coal, ports, railways and heavy engineering across the 1970s and 1980s, and the litigation that eventually reached the Constitution Bench in Steel Authority grew out of precisely this stratum of disputes, involving establishments of the Steel Authority, the Oil and Natural Gas Commission and the Food Corporation of India. The gap between the number of processes in Indian industry that satisfied the four factors and the number of prohibitions actually gazetted is the single clearest measure of the abolition half’s performance, and no reading of the Act that ignores that gap can explain why a Constitution Bench was eventually needed to say what a prohibition does and does not do.

A second feature of the record deserves note: prohibition, where it issued, frequently produced litigation rather than settlement. Establishments challenged notifications on competence, consultation and consideration grounds; workers, armed after 1997 with Air India, sued for absorption as the notification’s automatic consequence; and the validity of particular notifications oscillated through the courts for years, with workforces in a suspended position throughout. The Steel Authority judgment’s insistence that contravention sounds in penalties and not in status, and that absorption claims belong in industrial adjudication, was an attempt to drain this swamp by assigning each consequence to its proper forum. It succeeded doctrinally. Its human cost fell on the generation of workers whose claims had been framed, in reliance on Air India, as entitlements arising from notifications, and who learned in 2001 that the entitlements had never existed.

Welfare, health and the amenities chapters

Chapter V addresses the conditions under which contract labour works, and its structure is a chain of primary and backstop duties. Section 16 empowers rules requiring the contractor to provide canteens where a hundred or more workmen are employed as contract labour and the work is likely to continue for six months or more, the canteens to be run on a no-profit basis with the cost of maintenance borne as the rules direct. Section 17 requires the contractor to provide suitable rest rooms, or other suitable accommodation, where contract labour is required to halt at night in connection with the work, the rooms to be sufficiently lighted, ventilated, clean and comfortable. Section 18 obliges the contractor to provide and maintain a sufficient supply of wholesome drinking water at convenient places, a sufficient number of latrines and urinals of the prescribed types, conveniently situated and accessible, and washing facilities. Section 19 requires a first-aid box equipped with the prescribed contents, readily accessible during all working hours, at every place where contract labour is employed.

Inside factories, these duties overlap with a second regime. The definition of a worker in the Factories Act 1948 reaches a person employed directly or by or through any agency, including a contractor, so the occupier of a factory owes the Act’s safety, health and welfare obligations to contract workers on his premises whether or not the contractor performs his own duties, an overlap examined in this series’ account of the Factories Act 1948 and its worker definition. The overlap is instructive: where two statutes fix responsibility on two actors for the same canteen and the same drinking water, the worker’s chance of actually receiving them rises, and the drafters of 1970 built the same doubling into their own chapter deliberately.

The amenity list repays a slower reading, because each item answers a documented condition of contract work rather than a draftsman’s checklist. The canteen requirement bites where a hundred or more contract labour are employed on work of some duration, because the large contract workforce typically worked at distance from any bazaar and ate what the contractor’s arrangements provided or went without; the rules surround the canteen with management and pricing disciplines and forbid its use as a profit centre charged to the workers. The rest room requirement answers the practice, pervasive in process industries and transport-linked work, of contract gangs halting at the site overnight between shifts or awaiting the next day’s engagement, sleeping where they worked; the statute insists on a room that is lighted, ventilated, clean and comfortable, adjectives that read as modest and were, against the baseline, substantive. Drinking water, latrines and washing facilities are the sanitary minimum of any workplace, and their express inclusion, with accessibility and sufficiency specified, reflects the inquiries’ findings that contract workers were routinely excluded from facilities the establishment maintained for its own rolls, or provided with facilities of a standard the establishment would not have tolerated for its direct workmen. The first-aid box, readily accessible during all working hours, acknowledges that contract workers were concentrated in the manual operations where injuries occur and that the employer’s medical apparatus, where it existed, was organised around its own employees. Read as a group, the four sections are the Rege Committee’s regulation formula reduced to a schedule: if the system could not be abolished, its daily conditions would at least be specified, priced and assigned to a solvent obligor.

Section 20 is the doubling mechanism. If any amenity required by Sections 16 to 19 is not provided by the contractor within the prescribed time, it shall be provided by the principal employer, who may recover the expenses incurred from the contractor, whether by deduction from money payable to the contractor under any contract or as a debt payable by the contractor. The design deserves a moment’s attention, because it solves the oldest problem of the intermediary system, the vanishing obligor. Duties are placed on the contractor because he is the immediate employer and usually controls the workplace arrangements day to day. But the contractor is exactly the party most likely to lack the means or the will to perform, so the statute makes the principal employer liable to step in, and buys his cooperation by giving him a complete right of recovery against the contractor. The worker-facing duty cannot be defeated by the contractor’s default; the financial consequence travels back down the chain by set-off. Later labour statutes in India reused this pattern of primary duty on the intermediary, backstop duty on the principal, and recovery by deduction so consistently that its appearance here should be read as one of the Act’s genuine contributions to legislative technique.

Wages and the principal employer’s backstop

Section 21 performs the same design for wages, the obligation that matters most. The contractor is responsible for payment of wages to each worker he employs, before the expiry of the prescribed period. The principal employer must nominate a representative, duly authorised, to be present at the time of disbursement of wages by the contractor, and it is the duty of the contractor to ensure that wages are disbursed in the presence of that authorised representative. The representative certifies the amounts actually paid to the workmen, in the prescribed manner. Section 21(4) then states the backstop: if the contractor fails to make payment of wages within the prescribed period, or makes short payment, the principal employer shall be liable to make payment of the wages in full or the unpaid balance due, and to recover the amount so paid from the contractor, again by deduction from contract money or as a debt.

The certification machinery is the provision’s quiet strength. Wage default by contractors flourishes in undocumented settings, where the worker cannot prove the rate, the days or the shortfall. The Act responds by making disbursement a witnessed, certified event: the principal employer’s representative watches the money move, and his certificate fixes what was paid. A worker short-paid in the representative’s presence has an official record created by the party best placed to bear the liability. The model connects to the older wage statutes with which the contract worker must also deal, since the wage itself, its permissible deductions and its payment period are governed by the payment of wages regime and, in scheduled employments, by the Minimum Wages Act 1948 and its fixation machinery, while the contract labour statute’s contribution is the channel: it tells the worker whose pocket answers when the contractor’s pocket is empty, and it tells the principal employer that he cannot buy immunity from that answer by buying labour at arm’s length.

Can a contract worker claim the same wages as a regular worker doing identical work?

Yes, under the Central Rules. Rule 25(2)(v)(a) makes it a condition of the contractor’s licence that a contract workman doing the same or similar work as a directly employed workman receives the same wage rates, holidays, hours and conditions. In practice the claim is enforced through the inspectorate and industrial adjudication rather than by individual suit.

The entitlement is narrower in practice than it reads. Rule 25(2)(v)(a) of the Contract Labour (Regulation and Abolition) Central Rules 1971, which states conditions the licensing officer attaches to licences in the central sphere, provides that where a workman performs the same or similar kind of work as a workman directly employed by the principal employer of the establishment, the wage rates, holidays, hours of work and other conditions of service of the contract workman shall be the same as those applicable to the directly employed workman performing the same or similar work. The parallel state rules carry corresponding conditions. The entitlement exists as a condition of the contractor’s licence and as a rule the inspectorate can enforce, which is a different thing from a right the individual worker can sue on in the ordinary way, and adjudication of wage parity claims usually arrived through industrial disputes rather than through the Act’s own machinery. The constitutional courts separately developed an equal pay for equal work doctrine for daily-rated and temporary employees performing the same duties as regular staff, most comprehensively in State of Punjab v Jagjit Singh (2017), and contract workers’ parity claims often travel on that broader doctrine as well as on the rule. The honest summary is that parity is a stated rule of the scheme, an enforceable licence condition, and a litigated question of fact in each establishment, all at once.

The rules complete the wage chapter with display and notice duties that are easy to overlook and were litigated out of proportion to their size. The contractor must display notices showing the wage rates, the hours of work, the wage period, the dates of payment and the name and address of the inspector for the area, and must send prescribed returns to the licensing officer, so that the terms of the engagement exist in public, on the wall of the workplace, and not only in the contract file. A worker who cannot read his wage rate off any document he is shown is a worker whose underpayment cannot be proved even by him, and the display duty is the Act’s answer at the level of the individual engagement, just as the certified disbursement is its answer at the moment of payment. Inspectors enforcing the chapter began with the notices, because an establishment that had not posted its rates was telling the inspector, in advance, how the rest of the inspection would go.

One more feature of the wage chapter deserves record, because it explains litigation patterns. The principal employer’s liability under Section 21(4) is a liability to pay, not an admission of employment. Paying the contractor’s defaulted wages does not make the worker the principal employer’s employee, any more than providing his canteen does, and the courts consistently refused to convert backstop liability into employer status. That refusal is doctrinally tidy and humanly hard: the statute treats the principal employer as the safest pocket for every duty except the one duty, recognition as employer, that carries security of tenure, provident fund continuity on the establishment’s own rolls, and protection against arbitrary termination. The reader who grasps that the backstops are deliberately capped at amenity and money, and stop short of status, has grasped the regulation half’s ceiling.

The Contract Labour Act is a conditions statute, and the contract worker’s retirement and insurance entitlements arrive, where they arrive, through two older statutes whose definitions reach through the contractor. The Employees’ Provident Funds and Miscellaneous Provisions Act, 1952 applies to establishments of prescribed sizes and classes, and its scheme places the duty of depositing contributions for employees employed by or through a contractor on the principal employer, who recovers the amount from the contractor; the Supreme Court sustained that allocation early and the provident fund organisation enforced against principal employers for their contractors’ defaults as a matter of routine. The Employees’ State Insurance Act, 1948 works through a matching pair of definitions: the contractor is the immediate employer of the workers he supplies, and Section 40 of that Act makes the principal employer responsible in the first instance for paying contributions in respect of employees engaged through an immediate employer, with a right of recovery. Medical care and cash benefits under the insurance scheme therefore reach the contract worker through the principal employer’s payment duty, not through the contractor’s solvency.

The pattern across the three statutes is the architecture this article has already described in the wage chapter, extended to deferred wages and insurance: the law declines to make the worker’s entitlement depend on the intermediary’s performance, fixes the establishment as the paying point, and gives the establishment its remedy against the contractor by deduction. The practical weakness is equally consistent. Contribution compliance requires the worker to be on somebody’s rolls with a recorded wage, and the fragmentation and cash practices of the contract system produced systematic under-recording, so that the contract worker’s provident fund accumulation and insurance record were frequently thinner than the statutes promised even where the establishment was large and compliant in its own direct workforce. The lesson the codes drew, visible in the social security code’s extension of its frame to unorganised, gig and platform workers, is that entitlement systems built on establishment rolls underserve precisely the workers whose employment is intermediated, and that a portable, worker-attached record answers the problem the backstop architecture could only mitigate.

Penalties and the enforcement architecture

Chapter VI contains the penalties and procedure. Section 22 punishes whoever obstructs an inspector in the discharge of his duties, or refuses or wilfully neglects to afford him reasonable facility for inspection, entry, inquiry or examination, with imprisonment for a term which may extend to three months, or with fine which may extend to one thousand rupees, or with both. Section 23 punishes contravention of any provision of the Act, or of any rule, prohibiting, restricting or regulating the employment of contract labour, and contravention of licence conditions, with the same scale of punishment, and adds that where the contravention continues after conviction, an additional fine extending to one hundred rupees may be imposed for every day of the continued offence. Section 24 covers offences for which no separate penalty is provided, at the same scale. Section 25 deals with offences by companies, deeming the person in charge of and responsible to the company for the conduct of its business guilty unless he proves the offence was committed without his knowledge or that he exercised all due diligence to prevent it, and reaching directors, managers and officers where the offence occurred with their consent, connivance or neglect. Section 26 confines cognizance to complaints made by, or with the previous sanction of, the inspector, and Section 27 bars courts from taking cognizance of an offence after three months from the date on which the alleged commission came to the knowledge of the inspector.

Two features of this architecture stand out. The first is the scale. The fines of 1970, one thousand rupees at their ceiling and one hundred rupees a day for a continuing offence, were meaningful when enacted and trivial within a generation, because Parliament never revised them; the statute that governed a vast and growing practice carried monetary penalties smaller than a single day’s wage bill for the establishments it regulated. Imprisonment of up to three months existed on paper, and prosecutions were possible, but the system’s working sanctions remained administrative: refusal or revocation of licences, refusal of registration, and the commercial pressure of the backstop liabilities. The second feature is the gatekeeping. Only an inspector can complain, prosecutions die unless lodged within three months of the inspector’s knowledge, and tripartite advice, registration and licensing all sit with the appropriate government, so the entire enforcement chain runs through the same under-resourced inspectorate on whose capacity the Act’s record always depended. Reform proposals across the decades returned to both features, higher penalties and an enforcement design less dependent on individual prosecution, and the successor code eventually took up both, in the manner the comparison below records.

Enforcement on the ground: inspection, prosecution and the licence economy

Section 28 empowers the appropriate government to appoint inspectors and to define their local limits, and an inspector under the Act may enter premises where contract labour is employed, examine persons found there, require information about the names and addresses of workers and the nature of their work, inspect registers and records, and seize or take copies of registers, records of wages and notices as evidence. The design is the standard Indian inspectorate model, and the Act’s working life followed the model’s known trajectory: inspection coverage thinned as the regulated population grew, inspections concentrated on documentation offences, the missing register, the unrenewed licence, the absent wage slip, rather than on the substantive questions of wage suppression and absent amenities, and prosecutions, gated by the inspector’s complaint and the three-month cognizance window of Section 27, formed a small and slow tail to the inspection process. Official reviews of labour administration across the decades returned to the same arithmetic for the labour statutes generally: an inspectorate sized for the factory economy of the 1950s carrying a regulated population that had multiplied in size and changed in kind, with contract labour, the least visible and most mobile segment, receiving the least inspection of all.

The licence economy compounded the problem. A licensing regime with fees scaled to headcount, security deposits and periodic renewals creates, around the licence, a small industry of intermediaries, agents and facilitators, and the contractor’s compliance energy is predictably spent where the regime measures it: on the renewal file rather than on the rest room. None of this is an argument that the regulation half was worthless. Establishments of reputation registered, licensed their contractors, witnessed disbursements and provided amenities, and the contract workers of the organised public sector and the large private sector received measurably more of the statute’s protection than those in the dispersed small-establishment economy. The enforcement record sorts by establishment size almost exactly as the threshold design predicts, which is the deepest criticism of the design available: a statute whose protection strengthens with the size and visibility of the establishment protects most where the need, though real, is least desperate, and least where the intermediary system operates in its rawest form.

Three ways a contract worker can become an employee, and why two of them fail

Route to employee status	Legal basis	What the worker must show	Governing authority	Realistic prospect
Absorption following a Section 10 prohibition notification	Section 10 of the Contract Labour Act 1970	That the government has prohibited contract labour in his process	Steel Authority of India v National Union Waterfront Workers (2001)	Fails: the Constitution Bench held absorption does not follow automatically from a prohibition
Absorption through industrial adjudication of a dispute raised against the principal employer	Section 10 of the Industrial Disputes Act 1947 reference machinery, applied in Steel Authority of India (2001)	That the adjudicator should direct regularisation on the facts of the establishment	Steel Authority of India (2001); Gujarat Electricity Board v Hind Mazdoor Sabha (1995)	Available in principle and occasionally won, at the price of years of litigation and proof about one establishment at a time
Proof that the contractor arrangement is a sham or camouflage	The general law of employment applied through adjudication	That the principal employer in reality controls, supervises and can dismiss him, and the contractor is a name-lending intermediary	Steel Authority of India (2001); Workmen of Nilgiri Cooperative Marketing Society v State of Tamil Nadu (2004)	The only route that succeeds with any regularity, and only where the paper structure and the real control diverge

The table states, in compressed form, the article’s central finding. The first route is the one the statute’s title invites the reader to expect, and it is the one the Supreme Court closed. The second route exists and is real, but it converts a statutory promise into a litigation lottery, establishment by establishment. The third route succeeds precisely because it does not depend on the statute at all: it asks the adjudicator to disregard a false paper structure and find the true employer, which is a power courts and tribunals have always had. The Contract Labour Act’s abolition half, measured by what it delivered to the workers engaged under it, is therefore best understood as a power that changed the employer’s compliance position when exercised, without changing the worker’s legal position at all.

The courts and the absorption question

The judicial history divides into a first period of expansion, a brief interval in which absorption seemed to follow abolition, and a long correction. In the first period, the courts built the doctrine that policed the boundary between genuine and false contracting. In Gujarat Electricity Board v Hind Mazdoor Sabha (1995), the Supreme Court held that only the appropriate government can abolish contract labour, and that neither a court nor a tribunal can direct abolition of a genuine contract labour system; but where the contract is found to be a sham, a camouflage for direct employment, the tribunal can declare the workers to be the employees of the principal employer. That division, government abolishes, adjudicator unmasks, survived everything that followed and frames the table above.

The interval of expansion came with Air India Statutory Corporation v United Labour Union (1997), where a three-judge Bench held that on the issuance of a prohibition notification under Section 10, the contract labour engaged in the prohibited process became entitled to absorption as employees of the principal employer. The reasoning drew on the Act’s object: abolition, the Court reasoned, was meant to improve the workers’ condition, not to extinguish their livelihood, so the notification’s consequence must be a change of employer rather than a termination of work. For the four years it stood, Air India was the most consequential labour judgment in India. Unions demanded prohibitions precisely because prohibitions then carried absorption with them, employers resisted notifications with corresponding intensity, and the appropriate-government question, which government could prohibit in which undertaking, acquired decisive importance because the wrong government’s notification could not trigger the consequence.

A Constitution Bench of five judges reconsidered the question in Steel Authority of India Limited v National Union Waterfront Workers (2001) 7 SCC 1, decided on 30 August 2001, and overruled Air India on its central holdings. The propositions for which Steel Authority stands, stated at the level at which later courts apply them, are these. First, the Act neither expressly nor by necessary implication provides for automatic absorption of contract labour on the issuance of a prohibition notification, and the notification does not create an employer and employee relationship between the principal employer and the workers. Second, where the appropriate government prohibits contract labour in a process, the workers’ remedy is to raise an industrial dispute, before which the adjudicator examines whether the contract was genuine or a mere sham or camouflage; if it was a sham, the workers are in law the direct employees of the principal employer, and if it was genuine, the engagement ends with the prohibition, subject to the Bench’s observation that the principal employer should give preference to the displaced contract labour, as far as possible, when it recruits regular workmen for the same work. Third, a contravention of the Act, such as employing contract labour in a prohibited process, attracts the statute’s penal consequences; it does not, by itself, create an employment relationship. Fourth, the amended definition of appropriate government, tracking the Industrial Disputes Act, governs which government may act for central public sector establishments, and notifications issued by a state government for such establishments could not stand.

The litigation that produced the judgment illustrates the notification system’s practical workings. The appeals decided together as Steel Authority arose out of prohibition notifications and absorption claims across major public sector establishments, including the Steel Authority’s steel plants, the Oil and Natural Gas Commission and the Food Corporation of India, and one strand of the dispute concerned the Central Government notification of 9 December 1976 prohibiting contract labour in operations such as sweeping, cleaning, dusting and watching of buildings, and whether that notification could reach establishments after the 1986 amendment had settled the appropriate government question in the Industrial Disputes Act’s terms. Another strand concerned state notifications directed at central undertakings, which the amended definition placed beyond state competence. The Constitution Bench’s clubbing of these strands into one judgment was itself significant: absorption, appropriate government, the effect of contravention and the reach of the 1976 notification were decided together, giving the field a single authoritative restatement rather than a generation of fragments.

The reasoning is as important as the holdings. The Court read the Act as a statute that creates the relationship between the principal employer and the contract worker only for the specific purposes the text states, wages, welfare and regulatory responsibility, and declined to extract from a prohibition power a status the legislature had not conferred. Where Air India had reasoned from the statute’s beneficent purpose, Steel Authority reasoned from its text and structure: a legislature that knew how to deem workers employees, as other Indian statutes deem in defined situations, had not done so here. The judgment also took care to preserve the sham doctrine in full, observing that where the contractor is a mere veil the adjudicator’s power to lift it is untouched. The practical law after 2001 is therefore a bifurcated system. Paper structures are tested by control; genuine contracts are terminable by prohibition without consequence for status; and the demand for absorption migrated out of the Contract Labour Act altogether, into industrial adjudication about sham and into the political arena of regularisation policy, where it remains.

Which government is the appropriate government for a central public sector establishment?

The Central Government. The 1986 amendment recast the Act’s definition of appropriate government to follow the Industrial Disputes Act, under which the Centre acts for establishments owned, controlled or managed by central public sector undertakings and central statutory bodies, and the state government acts for all other establishments.

The point was worth a Constitution Bench’s time. In Steel Authority of India (2001) this definition disposed of prohibition notifications that state governments had issued for central undertakings: the wrong government had acted, after consultation with the wrong board, and its notifications could not operate. For a worker or a union contemplating a prohibition demand, the definition is the first thing to check, because a notification from the state government covering a central public sector plant is void at the threshold, whatever the merits of the four Section 10 factors.

A special strand of the sham litigation concerned canteen workers, and it deserves record because it shows the doctrine working at its boundary. Where a statute obliged the establishment itself to maintain a canteen, as the Factories Act does for factories of prescribed size, workers engaged in the canteen through a contractor argued that the canteen was the establishment’s own statutory function, that the contractor was interposed into a duty the law laid on the occupier, and that they were therefore the establishment’s employees. Some courts accepted arguments of that family; others treated the statutory duty as discharged by providing the canteen through an agency. The Supreme Court’s later treatment, in Balwant Rai Saluja v Air India Limited (2014), resolved the matter back onto the control test: the canteen’s statutory character did not by itself make its workers the establishment’s employees, and the decisive question remained who engaged, paid, supervised and could dismiss them. The canteen cases thus mark the limit of the purposive argument that Air India had embodied: after Steel Authority, even a statutory duty owed by the establishment to the workers did not manufacture an employment relationship the evidence of control did not support.

How does a court decide whether a contract is genuine or a sham?

By looking past the documents to the reality of control. The adjudicator asks who engages and can dismiss the worker, who pays him and from whose funds, who supervises his daily work, whose premises and tools he uses, and whether the contractor runs any independent business beyond supplying labour to this establishment.

The adjudicator looks past the contract document and the contractor’s licence to the working arrangement: who engages and can dismiss the worker, who pays his wages and from whose funds, who supervises his daily work and fixes his hours, whose tools and premises he works with, whether the contractor has any independent business or exists only to supply labour to this establishment, and whether the worker is integrated into the establishment’s organisation. The Supreme Court restated the inquiry in Workmen of Nilgiri Cooperative Marketing Society Limited v State of Tamil Nadu (2004), treating the determination as one of fact on the whole evidence, and applied the control test again in Balwant Rai Saluja v Air India Limited (2014), where canteen workers engaged through a contractor were held not to be employees of the airline because the contractor, and not the airline, exercised the decisive control, notwithstanding that the canteen itself was a statutory obligation of the establishment. A sham finding converts the workers into direct employees from the arrangement’s true character; a genuine finding leaves the contract standing in full, with the regulation half as the workers’ only statutory shelter.

The aftermath of Steel Authority is part of the statute’s honest record. Prohibition notifications continued to be issued, but their human meaning had changed: after 2001 they threatened the workers’ engagement itself, so the constituency for abolition shifted, and unions in several industries became ambivalent about demanding the very remedy the Act’s title promised. Industrial adjudication on sham contracts became the main battlefield, with its fact-heavy inquiries, its decades-long references in some cases, and its outcomes turning on supervision records and wage trails that the establishments themselves usually kept. The Second National Commission on Labour, reporting in 2002, recorded the framework’s inadequacies within its wider case for consolidating the labour statutes, and the long legislative answer, when it eventually came, was not to repair the Contract Labour Act but to replace it, inside a code whose contract labour chapter rewrites the threshold, the machinery and, in the core activity prohibition, the very logic of Section 10.

The public sector paradox

The litigation record of this Act carries a lesson that a private-sector framing would miss: the largest and most contested deployments of contract labour in the statute’s history were in the public sector. The establishments whose names recur in the reports, steel plants, the Oil and Natural Gas Commission, the Food Corporation of India, railways and ports, government offices and public utilities, were the state’s own. The reasons are structural. Public establishments faced sanctioned-post ceilings and recruitment freezes that froze their direct rolls while their workloads grew, and contracting was the channel through which the work got done without a sanctioned post being created; the same establishment was simultaneously the appropriate government’s own undertaking, the regulated party, and, in prohibition proceedings, the respondent resisting the notification its own government was pressed to issue. The Steel Authority litigation’s appropriate-government strand is unintelligible without this background: the question whether the Centre or the state could prohibit contract labour in a central undertaking mattered because the undertakings were the system’s heaviest users.

The paradox shaped the politics of abolition. A prohibition in a private factory displaced a private employer’s contract workforce; a prohibition in a public undertaking displaced workers who regarded the state as their real employer, who were frequently organised, and whose absorption claims were pressed as claims on public employment with constitutional equality arguments in support. Governments consequently faced, in their own establishments, the concentrated form of the dilemma Section 10 created everywhere: prohibit and displace, or tolerate and be charged with running the system the statute condemned. The record shows governments choosing toleration with regulation in their own houses as reliably as private employers chose it in theirs, and the contract workforces of the public sector grew across the same decades in which the abolition power slept. When the Constitution Bench in 2001 closed the absorption route, the class most immediately affected was the public sector’s contract workforce, and the regularisation politics of the following decades, in state after state, is the continuation of that closure by other means.

The doctrinal aftermath: regularisation and the Uma Devi climate

Steel Authority did not operate alone, and the contract worker’s legal climate after 2001 was set by its combination with the Supreme Court’s regularisation jurisprudence in public employment. In Secretary, State of Karnataka v Uma Devi (2006), a Constitution Bench held that regularisation is not a mode of recruitment, that appointments made without following the constitutional scheme of public employment confer no right to absorption or permanence, and that courts should not direct regularisation of irregularly appointed workers as a matter of course, save for a one-time measure for a defined class of long-serving irregular appointees against sanctioned posts. Uma Devi concerned direct engagement by the state rather than engagement through contractors, and its facts differ from the contract labour setting. Its gravitational effect did not. Public sector establishments, which employed contract labour on the largest scale, defended absorption claims in its language, adjudicators absorbed its caution about converting engagements into permanence by judicial direction, and the combined doctrine of the two Constitution Benches made the years after 2001 a period in which every road from contract engagement to employee status ran uphill.

The later case law worked within that climate rather than against it. The sham doctrine continued to produce findings of direct employment where the evidence of control justified them, and courts remained willing to examine the reality of arrangements in canteens, security and facility services where the contractor’s independence was doubtful. Wage parity claims advanced on the equal pay doctrine. What disappeared was the transformative claim, the argument that a class of workers becomes the establishment’s employees by operation of a notification, a scheme or the sheer passage of time. For the purposes of this article, the aftermath confirms the assessment the artifact table states: the law offers the contract worker routes that are individual, evidentiary and slow, and withholds the route that is collective, automatic and fast. A statute that had wanted the second kind of route available would have had to enact it, and the whole history from 1970 to the codes is the history of legislatures declining, at each opportunity, to enact it while continuing to regulate the first kind elaborately.

Why contracting grew: the structural explanation

The fact to be explained is stark. A statute enacted in 1970 to abolish contract labour where possible presided over five decades in which contract labour became a standard, expanding feature of Indian organised industry, used not in the interstices of production but at its centre. Official inquiries, parliamentary answers and the reports of the labour administration record the pattern consistently: across organised manufacturing, the share of workers engaged through contractors rose over the decades while direct employment stagnated, and large establishments that once staffed their perennial processes on their own rolls came to staff substantial parts of them through intermediaries. The explanation is not that employers conspired against the Act. It is that the Act regulated the contract while the rest of the legal system priced the alternative.

Begin with the price of direct employment. A directly employed workman in an Indian establishment accumulates statutory protections that compound with size and service: security against retrenchment, with compensation and, in establishments above prescribed sizes under the Industrial Disputes Act, a requirement of prior government permission for retrenchment, layoff and closure; bonus under the payment of bonus legislation; provident fund and insurance contributions; gratuity; standing orders governing classification, discipline and termination; and the collective strength of a union able to raise disputes about the whole body of conditions. Each element was enacted for defensible reasons, and their accumulation produced an employment relationship that an employer could enter easily and exit, adjust or restructure only slowly, publicly and at considerable cost. The contract worker, by contrast, could be engaged and disengaged through the contractor as volumes moved, stood outside the establishment’s standing orders and retrenchment chapters, and counted, for many threshold purposes, as somebody else’s employee. Read the Industrial Disputes Act pillar in this series alongside this article and the mechanism becomes plain: once the disputes statute had made direct employment so consequential, contracting was the rational response for an establishment solving a cost and flexibility problem, and the Contract Labour Act, which taxed that response with registration, licensing and amenities but never prohibited it, could not reverse arithmetic of that kind.

The two-tier workforce that resulted inside a single establishment had consequences beyond the wage bill, and the labour movement’s internal history records them. Direct workmen and contract workers doing adjacent work belonged to different legal worlds, were organised, where they were organised at all, in different unions or in the contract workers’ own unions, and bargained, where they bargained, against different employers: the establishment’s union negotiated settlements that bound the establishment’s rolls, while the contract workers’ claims were directed at a contractor who could plead, truthfully, that his margins were fixed by the establishment’s contract price. Establishments dealt with the intermediary on price; unions learned that the real counterparty for every question that mattered, rates, numbers, continuity, was the establishment that had placed itself outside the relationship; and the industrial relations system built by the Industrial Disputes Act, organised around the employer of the workmen in dispute, strained at a structure in which the workers with the grievance and the party with the power to answer it were connected only by a commercial contract. The sham doctrine, in this light, was the system’s pressure valve: it let adjudicators reunite the grievance and the power in the cases where the separation was a pretence, while leaving the structure intact where the separation was real.

The employer’s case for the system, stated at its strongest, deserves to be put fairly. Indian industry faces fluctuating demand, seasonal peaks, specialised tasks requiring skills an establishment needs for a month rather than a decade, and construction and maintenance work that is project-shaped by nature. A firm that must staff every fluctuation on its permanent rolls either carries idle labour through troughs, which its competitors do not, or declines the work. The contractor, on this account, is a labour market institution, matching workers to shifting demand, absorbing the administrative burden of recruitment, payroll and compliance across many small clients, and giving small establishments access to manpower channels they could not build themselves. The case continues that the abuses attributed to contracting, suppressed wages, missing amenities, vanished contractors, are failures of enforcement of obligations that the Act itself imposed and the backstops it created, and that a prohibition-led policy destroys the jobs along with the intermediary, as the aftermath of Steel Authority demonstrated whenever a notification displaced workers who received no absorption in exchange.

The worker’s case, also at its strongest, answers that the flexibility described is purchased by a specific transfer of risk. The same work, at the same machines, for the same establishment, is performed at materially lower total cost when it is contracted, and the saving is not found in the contractor’s administration; it is found in the wage differential, the missing social security accumulation, the absence of job security and the weakened bargaining position of workers whose employer of record is an intermediary with no assets and no future at the site. The dual structure divides a single workforce into two classes doing comparable work under one roof, a division the wage parity rule in the Central Rules acknowledged and the system’s operation overrode. On this account, the Act’s failure is written into its compromise: it accepted the intermediary as legitimate, regulated his conduct, and reserved abolition to a discretion that governments, dependent on industrial investment and anxious about employment effects, predictably exercised sparingly. The four Section 10 factors themselves build the worker’s grievance into the law: they concede that perennial, necessary, whole-time work exists inside the contract system, and then make its liberation a matter of executive grace.

The structural reform point follows from whichever case the reader finds stronger, and it is the point this series’ thematic treatment of the labour codes develops at length. A regime that protects direct employment heavily and contracted employment lightly manufactures the demand for the second category; a regime that responds by prohibiting the second category, without touching the price of the first, manufactures displacement instead of conversion. The only exits from that triangle are to narrow the protection gap between the two categories, by extending real social security and wage floors to contract workers, or to lower the exit costs that make direct employment feel irreversible, or to prohibit by category where the work is the establishment’s own core, accepting the displacement cost openly rather than hiding it inside a licensing regime. Parliament, in the codes, eventually attempted elements of all three, as the successor section records.

Why did contract labour grow under a statute named for abolition?

Because regulation was cheap and direct employment was dear. The Act made the contract route lawful and licensable at modest compliance cost, while the surrounding statutes made each direct workman a long-term commitment hedged with retrenchment controls and threshold obligations. Employers complied with the Act and expanded the practice it named itself against.

Because the statute regulated the practice while the rest of the labour code priced direct employment out of reach at the margin. The Industrial Disputes Act’s retrenchment and permission chapters, the threshold-triggered obligations of the welfare statutes, and the standing orders regime made each direct workman a long-term, hard-to-adjust commitment, while the Contract Labour Act left the contract route lawful, licensable and, its compliance costs counted, cheaper. Abolition under Section 10 required a discretionary government notification for each process in each establishment, a decision governments had employment reasons to avoid, so the regulatory half operated and the abolition half slept. Employers did not need to defeat the Act; they needed only to comply with it.

The neighbouring statutes: migrant workmen and construction workers

The Contract Labour Act was the general statute, but Parliament twice legislated for contract-shaped workforces in specific settings, and the two enactments illuminate what the general statute could not do. The Inter-State Migrant Workmen (Regulation of Employment and Conditions of Service) Act, 1979, Act 30 of 1979, addressed the worker recruited by a contractor in one state for employment in an establishment in another, the sardari pattern carried across state lines. It built on the contract labour template, licensing the recruiting contractor and registering the establishment, and added protections aimed at the migrant’s specific exposure: a displacement allowance, journey allowances, and wage parity with local workers, enforceable against the principal employer through a backstop on the familiar pattern. Its subject matter overlapped the 1970 Act’s so completely that the same contractor might need both licences for the same gang, and the OSH Code eventually folded it, with the Contract Labour Act, into a single regime.

The Building and Other Construction Workers (Regulation of Employment and Conditions of Service) Act, 1996, Act 27 of 1996, answered a different gap. Construction workers were contract labour in form but worked for no continuing establishment: the building, the bridge and the road were the workplace, and the establishment-facing machinery of registration and backstop fitted awkwardly where the project ended and the site dissolved. Parliament’s response funded welfare outside the employment relationship, through a cess on construction cost paid into state welfare boards that register building workers as beneficiaries and deliver benefits to the worker as a person rather than as someone’s employee. The construction statute is the clearest evidence that, by the 1990s, the legislature understood the contract system’s central problem to be the fragility of employer-linked entitlements, and was experimenting, sector by sector, with detaching protection from the identity of the employer. That experiment, carried much further, becomes the social security code’s architecture for unorganised and gig workers; its seed is visible here, in the recognition that a workforce engaged through contractors on shifting sites needs a welfare floor that travels with the worker. The Contract Labour Act itself never made that move. Its protections remained exactly where its text put them, attached to establishments and contractors, which is why construction, the largest contract workforce of all, needed a statute of its own.

The newest chapter of the same structural story is the platform economy, and a brief word on it belongs in this article because it tests the Act’s categories. The delivery rider and the cab driver engaged through an aggregator’s application reproduce, in digital form, the elements the committees of the 1930s described: an intermediary between the worker and the organisation whose business he performs, remuneration set by the intermediary’s terms, no security of engagement, and the organisation’s insistence that the intermediary’s counterparty is an independent partner rather than its worker. The Contract Labour Act does not reach the arrangement in its classic form, because the platform denies being an establishment engaging labour through a contractor in the statute’s sense and the rider is classified outside the workman definition altogether. Parliament’s response, when it came in the labour codes, was not to force the platform worker into the contract labour chapter but to name him separately, in the Code on Social Security 2020, as a gig worker and a platform worker entitled to welfare schemes funded in part by aggregator contributions, the welfare-board logic of the construction statute carried into the digital economy. The sequence confirms this article’s reading of the general statute: where intermediated work escapes the establishment frame, Indian legislation protects the worker by building a fund and a floor beside the employment relationship, and reserves the employment relationship itself for the cases the control test can capture.

Fixed-term employment and the road not taken by the Act

One later development belongs in this article because it reframed the choice the Act had presented as binary. The Act’s world offered the establishment two doors: employ directly, with the full industrial relations apparatus, or engage through a contractor, with the Act’s regulation. Fixed-term employment opened a third door: direct employment for a stated term, with the worker on the establishment’s own rolls, entitled to the same wages, allowances, hours and statutory benefits as a permanent workman doing the same work, with gratuity eligibility on a pro rata basis and with the employment ending on the term’s expiry without the retrenchment machinery being engaged. The device was written into the Industrial Employment (Standing Orders) Central Rules by an amendment of 2018 for the central sphere, and was then carried into statute by the successor code, whose treatment of fixed-term employment in the Industrial Relations Code 2020 states the model in full. For the contract labour question, the significance is plain: fixed-term employment was the legislature’s eventual answer to the employer’s flexibility case, an answer that concedes the need for time-bound engagement while refusing the intermediary, and its arrival half a century after 1970 measures how long the Act’s two-door world governed Indian industry.

The three doors can be compared on the points that decide real engagements. On wages, the fixed-term employee must receive the same wages and benefits as the permanent workman on like work, while the contract worker’s parity rests on a licence condition in the rules and a contractor whose margin depends on the differential. On social security, the fixed-term employee accumulates provident fund and insurance on the establishment’s own rolls as a matter of course, while the contract worker’s accumulation depends on the principal employer policing a chain. On exit, the fixed-term engagement ends by its own terms without retrenchment proceedings, which is the flexibility the employer sought, while the contract engagement ends by ending a commercial contract, which is the same flexibility with the worker’s statutory identity attached to the wrong party throughout. On voice, the fixed-term employee belongs to the establishment’s bargaining unit in a way the contract worker structurally does not. The comparison shows why the reformers of the code era reached for fixed-term employment as the flexible form to legitimate: it concedes the employer’s one defensible demand, time-boundedness, and refuses the indefensible one, the interposition of a party whose business is the dilution of the worker’s entitlements. That the device arrived by a rules amendment in 2018, quietly, after decades in which the contract system carried the whole weight of Indian industry’s demand for flexibility, is a fair measure of how completely the 1970 settlement had foreclosed the imagination of alternatives.

The successor: contract labour under the OSH Code

The Contract Labour Act does not end by amendment; it ends by subsumption. The Occupational Safety, Health and Working Conditions Code, 2020, Act 37 of 2020, is one of the four labour codes enacted to consolidate the twenty-nine central labour statutes, and its repeal schedule carries the Contract Labour Act into the list of repealed enactments. This series sets the four codes and their predecessor statutes side by side in the labour codes against the old laws comparison, and records, consistently across its treatment of the Industrial Disputes Act, the Factories Act and the Minimum Wages Act, that the Government of India brought all four codes into force on 21 November 2025, so that the Act this article examines operates only through the Code’s savings, which preserve earlier registrations, licences and proceedings through the transition. The contract labour chapter of the OSH Code is therefore not a proposal but the working successor, and its architecture shows both how much of 1970 survived and what its drafters thought had failed.

The transition mechanics matter to any reader holding an engagement that straddles the change. The Code’s repeal and savings provisions preserve the validity of what was done under the repealed Act, so registrations granted, licences issued, notifications published and proceedings begun under the 1970 Act continue to be governed, through the transition, by the framework that created them until the Code’s corresponding machinery is substituted, and Section 10 notifications in force do not evaporate on commencement. The Code’s common registration for establishments replaces the separate registrations that the Factories Act and the Contract Labour Act each demanded of the same premises, and its licensing chapter contemplates the contractor’s application in electronic form with the licence framed for the contractor’s work rather than as a purely local permission. For establishments, the practical consequence is consolidation of the compliance file; for workers, continuity of the wage and welfare backstops is preserved by the savings, and the substantive upgrade, where it comes, arrives through the core activity prohibition and the payment trail rather than through any change of status.

The threshold moves first: the Code’s contract labour provisions apply to establishments, and to contractors supplying manpower, at fifty or more contract labour, completing at the national level the upward revision the states had begun by amendment. The licensing chapter rebuilds the 1970 design with a single licence concept workable across states, an electronic application process, and, for the wider code, a common registration for establishments replacing the separate registrations of the predecessor statutes. The principal employer’s backstop duties for welfare facilities, health and working conditions continue, as does the liability for wages the contractor fails to pay, with the Code directing payment of wages by bank transfer or electronic mode, a direct answer to the undocumented disbursement problem Section 21’s witnessed certification was designed to police. The core activity prohibition is the Code’s genuine novelty: Section 57 prohibits the employment of contract labour in the core activities of an establishment, subject to three exceptions, where the activity is ordinarily performed through contractors in the field, where the activity does not require full-time workers for the major portion of working hours or such longer periods, and where there is a sudden increase in the volume of work to be completed within a specified time. A statutory list of support activities, including sanitation, security and watch and ward, canteen and catering, loading and unloading, housekeeping and transport services, is deemed not to be core activity, and the appropriate government, advised by a designated authority, decides doubtful classifications.

The Code’s wage provisions deserve separate mention because they answer the oldest abuse in the system. Under the 1970 Act, the contractor’s wage duty existed alongside the payment of wages and minimum wages statutes, and the contract worker’s rate often rested on the contract documents rather than on any schedule he could inspect. The Code framework, read with the Code on Wages 2019, attaches a statutory floor wage and minimum wage structure across employments, obliges the principal employer to ensure the contractor’s wage payment, and routes payment through bank transfer or electronic mode, creating the payment record as a by-product of the payment itself. The witnessed disbursement ceremony of Section 21, the representative standing beside the cash table with his certificate, belongs to the cash economy the 1970 Act governed; the successor regime assumes an electronic trail and builds its enforcement on it. Whether the trail reaches the dispersed, multi-layer subcontracting chains of construction and services is the operational question the transition will answer, and the savings provisions, which keep the repealed Act’s registrations and licences alive until the Code’s machinery is fully substituted, acknowledge that the answer will arrive in stages rather than on a single commencement date.

The reader who knows Section 10 will see what Section 57 is: the four-factor test generalised into a standing rule. Instead of asking a government to prohibit perennial, necessary, whole-time work process by process, the Code prohibits contract labour in the establishment’s core outright, and spends its drafting energy on the exceptions and the classification machinery instead. Whether that inversion succeeds where the discretion failed is a question the Code’s enforcement record will answer; the design concedes, fifty years on, that the discretionary model was the failure point. What the Code does not do is equally notable for this article’s subject: it does not enact automatic absorption either. The contract worker of the Code gains a higher threshold of coverage, a wider welfare backstop and a statutory bar on core deployment, but his status still runs to the contractor, and the sham doctrine remains his route across the line. The codes changed the map of regulation; the central settlement of 1970, regulation without absorption, survived its own repeal.

The amendment arc in brief

For a statute that governed a contested practice for over fifty years, the Contract Labour Act was amended remarkably little at the Centre, and the amendments that matter most are quickly told. The significant central change was the amendment of 1986, in force from 28 January 1986, which recast the definition of appropriate government to follow the Industrial Disputes Act and thereby settled, legislatively, which government acts for central public sector establishments; its importance is disproportionate to its length because, as the Steel Authority litigation showed, the definition controlled the validity of prohibition notifications across the public sector. Beyond that, the central text stood essentially as enacted: the threshold stayed at twenty in the bare Act, the penalty figures stayed at their 1970 values, and the structure of registration, licensing, welfare and prohibition was never centrally revisited.

The amending energy ran through the states instead, in the threshold revisions already described, Maharashtra and Andhra Pradesh raising the figure to fifty in 2017 and 2015 respectively, and through administrative reform of rules rather than of the Act. At the Centre, the reform debate of the 2010s repeatedly produced proposals to raise the national threshold and to simplify licensing, and those proposals matured not as amendments to the 1970 Act but as the contract labour chapter of the OSH Code, where the fifty figure, the single licence concept and the core activity prohibition were finally enacted in one settlement. The amendment arc is therefore best described as stasis at the Centre, drift in the states, and replacement at the end: a pattern that itself testifies to the political difficulty of reopening the 1970 compromise. Any amendment bill had to take a position on absorption, thresholds and prohibition simultaneously, and governments found it easier to let the Act age than to re-fight the settlement its title advertises. Even the fossilised penalty figures survived unrevised into the Code era, a small fact that carries the whole story in miniature.

What survives the repeal

Repeal by subsumption is not erasure, and the reader closing this article should be able to say what of the 1970 settlement the successor regime carries forward. The registration idea survives, generalised into the Code’s common registration, because the premise that the state must know which establishments engage labour, and through whom, was never the contested part of the design. The licensing idea survives, standardised and electronic, because conditioning the labour supply business on demonstrated standing was likewise common ground. The backstop survives almost intact: the principal employer answers for the amenities and the wages when the contractor defaults, and recovers downwards, because fifty years of the system’s operation proved that no other allocation reaches the worker. The tripartite advisory idea survives in attenuated form, in the boards and designated authorities through which classification questions under the core activity rule are to be resolved. And the prohibition idea survives transformed, the four factors of Section 10 promoted from a checklist for discretionary notifications into the standing core activity rule of Section 57, with the discretion relocated from whether to prohibit to how to classify.

What does not survive is equally instructive. The twenty-workman threshold, the figure the states spent decades litigating and amending, is gone in favour of fifty. The witnessed cash disbursement, Section 21’s ceremony, is gone in favour of the electronic trail. The 1970 penalty scale is gone, unrevised to the end. And the title’s promise is gone: the successor statute does not call itself an abolition code, and its contract labour chapter is a regulation chapter that prohibits a defined zone, stated in those terms. The candour of the successor’s drafting is the final evidence for this article’s verdict. When Parliament next legislated for contract labour, it kept the regulation, kept the backstop, generalised the prohibition test, and dropped the word abolition from the enterprise altogether, an implicit concession that the 1970 title had promised what the 1970 text, and every text since, declined to deliver.

The honest assessment and the verdict

The Act’s strengths are real and should be recorded first. It named a system that official India had preferred to deplore in committee reports and ignore in legislation, and gave it a complete regulatory skeleton: coverage thresholds with anti-avoidance drafting in the intermittent work exclusion, registration that identified the responsible establishment, licensing that attached conditions to the contractor’s permission to operate, welfare duties with a solvent backstop, a witnessed wage disbursement mechanism, and a recovery-by-deduction device that later statutes borrowed. It reached establishments the factory-centred code never touched. Its prohibition factors were intelligently drawn and remain, through the OSH Code’s core activity rule, the intellectual frame in which Indian law still asks which work may be contracted. And its backstop architecture, the refusal to let the worker’s entitlements depend on the weakest firm in the chain, is a permanent contribution to Indian legislative design.

Its failures are equally legible, and they are failures of the compromise rather than of drafting incompetence. The abolition half was a discretion without a duty, exercised rarely, and after 2001 stripped of the consequence that had made it worth demanding. The penalties fossilised at 1970 values. The inspectorate through which everything ran was never built to the statute’s scale. The thresholds invited fragmentation of engagements below the line, and the states eventually competed to raise them. Above all, the Act regulated a practice whose growth was driven by the surrounding code, and no licensing regime can outrun an arbitrage created by neighbouring statutes. The worker at the centre of it received amenities, a wage backstop and a litigation route through the sham doctrine, and did not receive the thing the title led his unions to expect: a path from the contractor’s rolls to the establishment’s.

It is worth stating plainly what would have had to be different, because the alternatives debated across the Act’s life were concrete and each was rejected at an identifiable point. The first alternative was a general statutory prohibition of contract labour in perennial work, converting the four Section 10 factors from considerations for the executive into a definition binding on every establishment, with the prohibition operating by force of the statute itself. It was the committees’ original instinct, and it was rejected in the tripartite and plan-era consensus that produced the discretionary design, on the stated ground that Indian industry’s variety made a single rule unworkable, a ground whose strength faded as the factors proved capable of identifying perennial work in case after case. The second alternative was absorption as the statutory consequence of prohibition, the rule Air India briefly supplied by interpretation. It was rejected by the legislature’s silence in 1970 and by the Constitution Bench’s reading of that silence in 2001, and no government afterwards moved to enact it, which locates the responsibility for its absence in the legislature rather than in the Court that declined to invent it. The third alternative was an equal treatment floor strong enough to remove the arbitrage: wage parity, social security and bargaining access for contract workers set at the level of the establishment’s own workmen, so that contracting survived only where flexibility, and not cheapness, was its purpose. Rule 25’s parity condition gestures at it inside the licensing system, and the codes’ floor wage and electronic payment provisions move towards it, but no enactment across the Act’s life adopted it in full. The assessment of the statute is incomplete without this map, because the Act’s defenders are entitled to ask what else, in the politics available, could have been enacted, and its critics are entitled to answer that each alternative was available, was debated, and was declined.

For the reader who comes to this statute as a worker, an employer or a practitioner, the practical position can be stated without ornament. The Act’s duties are legal duties, their breach carries penal and licensing consequences, and the questions this article has discussed, prohibition, absorption, sham and liability, turn on documents, dates and the identity of the appropriate government in ways that general reading cannot settle; this article states the law’s design and record as legal information, not as advice on any particular engagement, and a live dispute under the Act or its successor code should be placed before a qualified legal practitioner with the contract, the licence and the wage records in hand. That caution is not a formality in this field. More cases under this statute were lost on the paper, the wrong government, the missing certificate, the unwitnessed disbursement, than on any contestable proposition of law.

The verdict on the Contract Labour Act is the one its own text earns. As a regulation statute, it was a serious, well-built instrument that arrived early and was then left unrevised while the economy it governed transformed around it. As an abolition statute, it was a promise the legislature made in the title and withdrew in the text, leaving the courts to fight over the gap for thirty years until a Constitution Bench closed the argument in favour of the text. The statute’s name promised the end of a system; its sections built that system’s rulebook. India ran its factories, ports, mines and offices under the rulebook for over half a century, and the successor code, for all its redesign, kept the rulebook’s central settlement intact. Regulation without absorption was not the Act’s accident. It was the Act.

Frequently Asked Questions

Q: What does the Contract Labour Act regulate?

The Act regulates the employment of contract labour in establishments and by contractors at or above its numerical threshold. Its regulation half requires the principal employer to register the establishment, requires the contractor to hold a licence with conditions on hours, wages and amenities, fixes the contractor’s duties on canteens, rest rooms, drinking water, latrines, washing facilities and first aid, makes the principal employer the backstop for those amenities and for unpaid wages, and backs the scheme with inspectors and penalties. Its abolition half, in Section 10, empowers the appropriate government to prohibit contract labour in a particular process after consulting the advisory board and weighing four statutory factors. The Act regulates the conditions of the engagement comprehensively; it does not regulate the proportion of an establishment’s work that may be contracted, and it confers no employee status on the workers it protects.

Q: When can contract labour be prohibited in a process?

Contract labour in a process can be prohibited when the appropriate government, after consulting the Central Advisory Board or the State Advisory Board as the case requires, issues a notification in the Official Gazette under Section 10, having regard to the four statutory factors: whether the work is incidental to or necessary for the establishment’s activity, whether it is of a perennial nature, whether it is done ordinarily through regular workmen in that establishment or a similar one, and whether it is sufficient to employ a considerable number of whole-time workmen. The government must also have regard to the conditions of work and benefits provided for the contract labour in the establishment. Prohibition is discretionary even where the factors are satisfied, it operates process by process and establishment by establishment rather than across an industry at a stroke, and it must come from the government that is the appropriate government for that establishment.

Q: Do contract workers become employees of the principal employer on abolition?

No. A Constitution Bench of the Supreme Court held in Steel Authority of India Limited v National Union Waterfront Workers (2001) that a prohibition notification under Section 10 does not make contract workers employees of the principal employer, and does not create an employer and employee relationship between them. The earlier decision in Air India Statutory Corporation v United Labour Union (1997), which had held that workers were entitled to absorption when contract labour was abolished in their process, was overruled on that point. On prohibition, the engagement of contract labour in the process must end; the workers may then raise an industrial dispute, in which the adjudicator can examine whether the contract was a sham, and the principal employer is expected to give preference to the displaced workers, as far as possible, when recruiting regular workmen for the same work.

Q: Who pays a contract worker if the contractor defaults?

The principal employer pays. Section 21 of the Act makes the contractor responsible for paying wages before the expiry of the prescribed period, and requires disbursement in the presence of a representative authorised by the principal employer, who certifies the amounts paid. Section 21(4) then provides that if the contractor fails to pay wages within the prescribed period, or makes short payment, the principal employer shall be liable to pay the wages in full or the unpaid balance due. The principal employer may recover the amount so paid from the contractor, either by deduction from money payable to the contractor under any contract or as a debt payable by the contractor. The worker’s claim for the money therefore does not depend on the contractor’s solvency or willingness to pay, while the financial burden travels back to the contractor through the recovery mechanism.

Q: Which establishments must register under the Contract Labour Act?

Every establishment to which the Act applies must be registered by its principal employer. Under Section 1(4), the Act applies to every establishment in which twenty or more workmen are employed, or were employed on any day of the preceding twelve months, as contract labour, and state amendments have raised that figure to fifty in states such as Maharashtra and Andhra Pradesh, so the operative threshold depends on where the establishment is situated. The principal employer applies to the registering officer in the prescribed form, and on registration receives a certificate of registration. Section 9 gives the duty its edge: a principal employer who has no certificate of registration in force under the Act cannot employ contract labour in the establishment at all, so registration is not a formality that follows the engagement but a condition of lawfully engaging contract labour in the first place.

Q: What is the difference between contract labour and fixed term employment?

The difference is the identity of the employer and the path by which the worker is engaged. A contract worker is engaged by or through a contractor: his employer of record is the intermediary, the principal employer owes him only the duties the Contract Labour Act imposes, and his engagement can end when the contract between the establishment and the contractor ends. A fixed-term employee is engaged directly by the establishment itself for a stated period: he is on the establishment’s own rolls, entitled to the same wages, allowances, hours of work and statutory benefits as a permanent workman performing the same work, with gratuity eligibility calculated pro rata, and his employment simply expires when the term ends, without the retrenchment machinery being engaged. Fixed-term employment, introduced through the standing orders rules in 2018 and carried into the Industrial Relations Code 2020, was designed to give employers time-bound flexibility without the intermediary that the contract system interposes.

Q: Can core activities be outsourced to contract labour?

Under the Contract Labour Act 1970, the question had no direct statutory answer: the Act contained no category of core activity, and contract labour could lawfully be engaged even in an establishment’s central processes unless the appropriate government prohibited it in that process under Section 10. The successor code changed the position. Section 57 of the Occupational Safety, Health and Working Conditions Code 2020 prohibits the employment of contract labour in the core activities of an establishment, subject to three exceptions covering activities ordinarily done through contractors, activities that do not require full-time workers for the major portion of working hours, and a sudden increase in the volume of work to be completed in a specified time. A statutory list of support services, including sanitation, security, canteen and catering, loading and unloading, housekeeping and transport, is treated as non-core, and doubtful cases are decided by the appropriate government on the advice of a designated authority.

Q: What licence does a contractor need under the Contract Labour Act?

A contractor to whom the Act applies needs a licence under Section 12 before undertaking or executing any work through contract labour, and he must execute the work in accordance with the licence. He applies to the licensing officer in the prescribed form, stating the particulars of the establishment, the nature of the process, operation or other work for which contract labour is to be employed, and the duration of the engagement. The licence is granted under Section 13 on payment of the prescribed fee and on deposit of security for due performance of its conditions, it is valid for the period specified in it and may be renewed, and its statutory conditions cover hours of work, fixation of wages and other essential amenities for the contract labour. The licensing officer may revoke, suspend or amend the licence under Section 14 for misrepresentation, contravention of conditions or failure to comply with the Act, and an aggrieved person may appeal within thirty days under Section 15.

Q: What happens to the workers when a contractor’s licence is revoked or suspended?

Revocation under Section 14 ends the contractor’s lawful authority to execute work through contract labour, so the engagement run through that contractor cannot lawfully continue in its existing form. The Act does not convert the affected workers into employees of the principal employer when this happens, and their position follows the same rule as on a prohibition: status does not change by operation of the licensing action. What does continue is the protection of the regulation half for the period worked. Wages earned remain payable, with the contractor responsible in the first instance and the principal employer liable under Section 21(4) for any default or short payment, recoverable by him from the contractor, and the principal employer must step in under Section 20 for welfare amenities the contractor failed to provide. In practice the establishment must either bring in a licensed contractor, engage the work directly, or stop the work, and workers displaced by a revocation who believe the arrangement was a sham retain the industrial dispute route.

Q: What did the Air India judgment hold, and why did the Constitution Bench overrule it?

In Air India Statutory Corporation v United Labour Union (1997), a three-judge Bench of the Supreme Court held that when the appropriate government prohibited contract labour in a process under Section 10, the workers engaged in that process became entitled to absorption as employees of the principal employer. The reasoning was purposive: abolition was intended to better the workers’ lot, so it should change their employer rather than end their work. A Constitution Bench of five judges overruled that holding in Steel Authority of India Limited v National Union Waterfront Workers (2001), reasoning from the Act’s text and structure instead. The Act creates specific, limited relationships between the principal employer and the contract worker, for wages, welfare and regulation, and contains no provision deeming the workers employees on prohibition, although the legislature knew how to write such deeming provisions when it wished. A power to prohibit an engagement, the Bench held, cannot be read as a command to convert it.

Q: Does long service as contract labour create a right to regularisation?

No, not by itself. Neither the Contract Labour Act nor the general law converts length of service through a contractor into a right to be regularised as an employee of the principal employer. A worker who has served the same establishment through a contractor for many years remains, in the statute’s classification, contract labour on the final day as on the first, because the Act’s definition turns on the channel of engagement and not on duration. Long service matters in two narrower ways. It is powerful evidence in a sham inquiry, since work performed continuously for years under the establishment’s supervision fits badly with the claim that the engagement was genuinely the contractor’s independent undertaking, and adjudicators weigh it accordingly. It also informs the equities when an adjudicator or a government frames a regularisation scheme. But the entitlement itself must come from a sham finding, a settlement, a scheme or an adjudicated direction in the individual case; the calendar alone confers nothing.

Q: Can contract workers raise an industrial dispute against the principal employer?

Yes. The Steel Authority of India judgment expressly preserved this route: contract workers who seek absorption or regularisation may raise an industrial dispute, and the industrial adjudicator seised of the reference will examine whether the contract under which they are engaged is genuine or a sham or camouflage. The dispute runs against the principal employer because the relief sought, recognition as his employees, can only be granted against him, even though the workers’ contracts of engagement are with the contractor. If the adjudicator finds the contract genuine, he cannot direct abolition of the system, which is the appropriate government’s function under Section 10, and the workers remain the contractor’s employees with the Act’s regulatory protections. If he finds the contract a sham, he declares the workers to be the principal employer’s employees in law, with the consequences that follow for wages, continuity and conditions. This adjudication route, establishment by establishment, became the principal legal path to absorption after 2001.

Q: What is the role of the Central and State advisory boards under the Act?

The Central Advisory Contract Labour Board, constituted under Section 3, and the State Advisory Contract Labour Boards, constituted under Section 4, advise the respective governments on matters arising out of the administration of the Act that are referred to them, and perform the other functions the Act assigns. They are tripartite bodies, bringing together government, the industries in which contract labour is engaged, and representatives of the contract workmen themselves, and each board may constitute committees under Section 5 for specific purposes. Their most consequential function is procedural: Section 10 requires the appropriate government to consult the Central Board or the State Board before prohibiting contract labour in a process, so consultation is a step in the validity of a prohibition notification, and a notification issued by the wrong government, or without the consultation the section contemplates, is open to challenge. Beyond that gatekeeping role, the boards embody the Act’s tripartite theory that the pace of abolition should be negotiated between the state and the two sides of industry.

Q: What penalties follow a conviction under the Contract Labour Act?

The Act’s penalties are contained in Chapter VI and are modest by later standards. Obstructing an inspector, or refusing him reasonable facilities for inspection, entry, inquiry or examination, is punishable under Section 22 with imprisonment for a term extending to three months, or a fine extending to one thousand rupees, or both. Contravention of a provision or rule prohibiting, restricting or regulating the employment of contract labour, or of a licence condition, carries the same scale under Section 23, with an additional fine extending to one hundred rupees for every day a contravention continues after conviction. Section 24 applies the same scale to offences for which no penalty is separately provided, and Section 25 makes the person in charge of a company’s business answerable for its offences unless he proves absence of knowledge or due diligence. Only an inspector can set the prosecution in motion, by complaint or sanction, and Section 27 requires cognizance within three months of the offence coming to the inspector’s knowledge.

Q: Does the Act apply where the work is intermittent or seasonal?

Generally no, but the exclusion is narrower than its wording suggests. Section 1(5) provides that the Act does not apply to establishments in which work only of an intermittent or casual nature is performed. The same provision then defines the boundary: work is not deemed intermittent if it was performed for more than one hundred and twenty days in the preceding twelve months, and seasonal work is not deemed intermittent if it was performed for more than sixty days in a year. An establishment whose contractor-supplied workforce is present for most of the working year therefore cannot escape the Act by describing the engagement as seasonal in the contract documents; the statutory test counts days of actual performance rather than labels. The exclusion protects genuinely occasional engagement, the repair crew called for a breakdown or the gang engaged for a short harvest window, and its day-count drafting is one of the Act’s deliberate anti-avoidance devices, catching duration where a purely verbal test would have been gamed.

Q: Who is a principal employer, and how is the term defined?

Section 2(1)(g) defines the principal employer by reference to control of the establishment. Where the establishment is an office or department of the government or a local authority, the principal employer is the head of that office or department, or such other officer as the government or authority specifies. Where the establishment is a factory, it is the owner or occupier of the factory, or, where a person has been named as the manager of the factory under the Factories Act 1948, that person. Where the establishment is a mine, it is the owner or agent of the mine. In every other case, it is any person responsible for the supervision and control of the establishment. The definition matters because the Act’s heaviest duties, registration, the welfare backstop under Section 20 and the wage liability under Section 21(4), are fastened on this person, and identifying him correctly is the first step in enforcing any of them.

Q: What welfare amenities must a contractor provide, and who steps in if he does not?

Chapter V of the Act obliges the contractor to provide canteens where a hundred or more workmen are employed as contract labour on work likely to continue for six months or more, rest rooms or other suitable accommodation where contract labour is required to halt at night in connection with the work, a sufficient supply of wholesome drinking water at convenient places, a sufficient number of latrines and urinals of the prescribed types, washing facilities, and a first-aid box equipped with the prescribed contents and readily accessible during all working hours. If the contractor fails to provide any of these amenities within the prescribed time, Section 20 obliges the principal employer to provide it, and entitles him to recover the expense from the contractor, by deduction from money payable under any contract or as a debt. The worker-facing duty therefore survives the contractor’s default; the cost of the default is settled between the two employers afterwards, through the set-off mechanism, and not at the worker’s expense.

Q: Why does the Act set a threshold of twenty workmen, and have states changed it?

The threshold reflects the standard design of Indian labour statutes, which reserve their full machinery for establishments of a size at which record-keeping, inspection and compliance costs can reasonably be borne, and leave very small engagements to the general law. Section 1(4) as enacted set the figure at twenty contract workmen for both establishments and contractors, with a proviso allowing the appropriate government to extend the Act below the figure after two months’ notice. States changed the figure upwards during the labour reform period, judging that the compliance burden fell disproportionately on smaller establishments and contractors: Maharashtra raised the threshold to fifty by an amendment of 2017, and Andhra Pradesh raised it to fifty by an amendment of 2015, with other states moving in the same direction. The successor code completed the movement at the national level, since the contract labour provisions of the Occupational Safety, Health and Working Conditions Code 2020 apply at fifty or more contract labour.

Q: Can the government extend the Act to establishments with fewer than twenty contract workers?

Yes. The proviso to Section 1(4) empowers the appropriate government, after giving not less than two months’ notice of its intention to do so by notification in the Official Gazette, to apply all or any of the provisions of the Act to any establishment or contractor employing such number of workmen, less than the threshold figure, as may be specified in the notification. The power lets a government pull a sector of small but problematic engagements inside the regulatory scheme without amending the statute, and the two months’ notice requirement gives affected establishments and contractors time to arrange registration, licensing and records before the duties bite. The existence of the proviso also confirms the character of the twenty-workman figure itself: it is a default line drawn for administrative convenience, adjustable downwards by the executive as well as upwards by state amendment, and not a judgment that establishments below the line have no obligations to the workers they engage through contractors under the general law.

Q: How can a prohibition notification under Section 10 be challenged?

A prohibition notification is executive action under a statute and is open to judicial review on the ordinary grounds. The challenges that have succeeded in the case law cluster around three defects. The first is competence: the notification must issue from the appropriate government for the establishment, and notifications by state governments covering central public sector establishments were struck down on exactly that ground in the litigation culminating in Steel Authority of India (2001). The second is procedure: Section 10 requires consultation with the Central Board or the State Board before the notification issues, and a government that skips the consultation acts outside the section. The third is consideration: the government must have regard to the four statutory factors and to the conditions of work and benefits in the establishment, and a notification issued without application of mind to them, or on irrelevant considerations, is vulnerable. A challenger, whether an establishment resisting the prohibition or workers resisting its absence of consequence, proceeds by writ petition to the High Court, and the court examines the notification against the section rather than substituting its own view of the factors.