Most accounts of the Industrial Relations Code, 2020 describe it as a deregulatory statute, and most union accounts describe it as an anti-worker one, and both descriptions fail the same test: neither can say, item by item, what the Code handed to employers and what it handed to workers. That itemised account is the only honest way to read this statute, because the Code moved in both directions inside a single text. The same Act that raised the threshold for prior government permission before lay-off, retrenchment and closure from one hundred workers to three hundred also imposed a strike notice requirement on every industrial establishment in India, a restriction that the repealed law had confined to public utility services. The same Act that wrote fixed term employment into the statute book also attached to it a parity condition and an accelerated gratuity entitlement that no earlier central law had contained. Flexibility and restriction were not traded across different statutes or different years. They arrived together, in one enactment, as the two halves of a single settlement.

The claim this article advances can be stated in one line: flexibility bought with process. The Code gave employers the threshold change they had campaigned for since 1976, and it gave workers a statutory bargaining agent and direct access to adjudication that the old regime had withheld, and it paid for both concessions by making lawful strike action conditional on advance notice everywhere in the country. Read that way, the statute is not a one-directional transfer of power. It is a settlement between organised interests in which each side received something it had sought for decades and surrendered something it had defended for just as long, and in which the price of the whole arrangement was spontaneity: the unannounced stoppage, the instant walkout, the strike called on the morning it begins. Whether that price was fairly charged is the argument the rest of this article works through, provision by provision, before reaching a verdict.

The Code is also one quarter of a larger consolidation. India carried twenty-nine central labour statutes into the reform process that the Second National Commission on Labour set in motion with its 2002 report, and emerged with four codes: the Code on Wages, 2019, the Industrial Relations Code, 2020, the Code on Social Security, 2020 and the Occupational Safety, Health and Working Conditions Code, 2020. What that four-code architecture did to the older statute book as a whole is a comparison with its own dedicated treatment in the labour codes versus old laws comparison, and this article stays inside its own quarter of the settlement: the law of trade unions, standing orders and industrial disputes. Within that quarter, the Code subsumed three statutes of very different ages and characters, the Trade Unions Act, 1926, the Industrial Employment (Standing Orders) Act, 1946 and the Industrial Disputes Act, 1947, and replaced their combined machinery with a single text of one hundred and four sections. Everything that follows is an account of what that replacement actually did.

The three laws the Industrial Relations Code replaced

The oldest of the three predecessors was the Trade Unions Act, 1926, the colonial-era statute that first gave Indian unions a legal personality. It provided for registration of trade unions, conferred immunity from certain criminal conspiracy and civil liability for acts done in furtherance of a trade dispute, and regulated union funds and their political objects. For nearly a century it did one conspicuous thing and omitted another. It made unions lawful and registrable, but it never required any employer to recognise a union, and it never said which union, among several rivals in one establishment, spoke for the workforce. Recognition at the central level rested on a voluntary Code of Discipline adopted in 1958 and on a patchwork of state legislation, of which Maharashtra’s statute was the best known. That omission is the gap the negotiating union machinery of the new Code was designed to fill, and it explains why a registration statute from 1926 belongs inside a code whose real subject is bargaining power.

What the 1926 Act did supply, and what the Code’s Chapter III carries forward in substance, is the protective shell within which unionism operates. A registered union and its office-bearers and members hold immunity from criminal conspiracy liability for agreements made to further the union’s lawful objects, and a qualified immunity from civil suit for acts done in contemplation or furtherance of a trade dispute, the twin shields without which collective action would be actionable at every step. The chapter carries the old Act’s financial architecture as well: the general fund applicable to the union’s objects, the option of a separate political fund raised by voluntary contribution, the rules for amalgamation of unions, for change of name and for dissolution, and the Registrar’s supervisory jurisdiction over registration and its cancellation. None of this was controversial in 2020, and its near-verbatim carriage is the clearest instance in the Code of the repeal-and-replace method at its most conservative: where the predecessor law worked, the Code kept it, and reserved its changes for the places where the predecessor law had failed or fallen silent.

The second predecessor, the Industrial Employment (Standing Orders) Act, 1946, governed the written conditions of service in larger industrial establishments. Its premise was that the terms on which a worker is employed, the classification of workers, shift working, attendance, leave, misconduct and the means of redress against unfair treatment, should not rest on managerial say-so but should be reduced to certified standing orders, drafted by the employer, scrutinised by a certifying officer and open to objection by the union. The Act originally applied to establishments employing one hundred or more workmen, and several states had lowered or varied that figure by amendment. Its subject matter, the internal rulebook of the establishment, survives in the Code as Chapter IV, with the applicability threshold reset at three hundred workers and a new reliance on model standing orders issued by the Central Government.

The third and largest predecessor was the Industrial Disputes Act, 1947, the statute that structured Indian industrial conflict for over seven decades. It defined the industrial dispute, created the machinery of works committees, conciliation officers, boards of conciliation, courts of inquiry, labour courts and industrial tribunals, regulated strikes and lock-outs, and, through amendments layered across the decades, built the elaborate permission regime of Chapter VB under which larger establishments could not lay off, retrench or close without prior government approval. Its definition of industry, expounded by a seven-judge bench in Bangalore Water Supply and Sewerage Board v A Rajappa in 1978, became one of the most contested phrases in Indian statute law. The record of that Act, its thresholds, its case law and the nine-judge reference that finally addressed the Rajappa test, belongs to the Industrial Disputes Act 1947 explained article in this series, which owns that record; this article takes the old Act’s positions as given and examines what the successor did with them.

Repeal, when it came, was effected by Section 104 of the Code, and its mechanics matter because thousands of disputes, settlements and awards were alive under the old statutes on the day the Code commenced. Section 104 repeals the three Acts and accompanies the repeal with the standard savings architecture of Indian repeal drafting. Acts done under the repealed statutes are treated as done under the corresponding provisions of the Code. Settlements and awards in force continue to bind. Proceedings pending before labour courts, tribunals and national tribunals under the old dispensation are not extinguished by the change of statute; the savings scheme carries the pending adjudicatory load across, and the general savings principle of the General Clauses Act, 1897 reinforces the same result for rights and liabilities already accrued. The consequence is a statute book with a clean break and a courtroom without one: the law governing a dispute filed the day before commencement continues to decide it, while every dispute arising after commencement runs on the Code. Employers and unions therefore operate two legal generations at once for a transitional period measured in years, and any account that treats the old Acts as simply gone misstates the position of every pending case in the country.

One feature of the repeal deserves emphasis because it recurs throughout this article. The Code did not reproduce the old law with new numbering. In several places it carried provisions across nearly verbatim, and in those places the predecessor case law carries across with them, because courts read a re-enacted provision against the construction it had already received. In other places it deliberately changed the words: the definition of worker widened at the supervisory ceiling, the definition of strike widened to catch concerted casual leave, the permission threshold moved, the reference machinery vanished. Where the words changed, the old judgments lose their grip, and the litigation of the next decade will be fought over exactly which category each provision falls into. The article flags that distinction at each point where it matters.

How the Code was assembled, and the five years it waited

The intellectual origin of the Code is the report of the Second National Commission on Labour, submitted in 2002 under the chairmanship of Ravindra Varma. The Commission proposed grouping the central labour statutes into a small number of codes organised by subject, and on industrial relations it proposed a bargain whose outline the 2020 Code follows with remarkable fidelity: a higher threshold below which the permission regime would not bite, set in exchange for stronger and more formal machinery for representation and dispute resolution above it. The Commission’s threshold suggestion was debated for eighteen years before a legislature enacted a version of it. Successive governments of both national coalitions accepted the code architecture in principle and shrank from the threshold question in practice, because the threshold was the one item in the entire labour reform programme that organised labour treated as existential and organised industry treated as non-negotiable. The states, acting on the Concurrent List entry for labour, moved first: Rajasthan amended the Industrial Disputes Act in 2014, with presidential assent, to raise its permission threshold to three hundred, and Gujarat and Uttar Pradesh followed with variations of their own. By the time Parliament acted, the three hundred figure was no longer a theory. It had been the working law in several large industrial states, and the Code generalised a change that had already been tested in the federal laboratory.

The legislative vehicle arrived in two attempts. An Industrial Relations Code Bill was introduced in 2019 and referred to the Parliamentary Standing Committee on Labour, whose examination took evidence from unions, employers and state governments and recommended a series of adjustments. A revised Bill was introduced in the Lok Sabha in September 2020, passed by the Lok Sabha on 22 September 2020 and by the Rajya Sabha the following day, in a session conducted under pandemic conditions and marked by the absence of opposition members who were boycotting proceedings over unrelated farm legislation. The President assented on 28 September 2020, and the Code became Act 35 of 2020. The manner of its passage is part of its political history and is regularly cited by its critics; the fact of its passage, by both Houses with the required majorities, is not in legal doubt, and no court was asked to undo the enactment on procedural grounds.

The Code did not travel alone, and its place in the convoy matters for reading it. The Code on Wages had gone first, enacted as Act 29 of 2019 with assent on 8 August 2019. The remaining three codes were enacted together in the September 2020 session: this Code as Act 35 of 2020, the Code on Social Security as Act 36 of 2020 and the Occupational Safety, Health and Working Conditions Code as Act 37 of 2020, each receiving assent on 28 September 2020. The sequencing explains several features of this text. Because the wages definition had been settled in the 2019 Code and adopted across the family, this Code could build its compensation arithmetic on a defined term it did not need to renegotiate. Because the social security code was being written in the same session, the re-skilling fund could be designed as this Code’s own chapter without waiting for the larger welfare architecture. And because all four codes shared a commencement strategy, the five-year wait described below was a family decision about the whole settlement, not a judgment on this Code alone.

Enactment and commencement then separated for five years, because the Code, like its three sibling codes, was drafted to commence on dates notified by the Central Government, and the government withheld notification while the rule-making on which the codes depend was prepared. The delay had a statutory logic and a political one. The statutory logic was that codes administered largely by state governments, under a Concurrent List subject, cannot function until state rules exist, and drafting four codes’ worth of central and state rules is the work of years. The political logic was that commencing the codes during an election cycle invited the general strike that central trade unions had repeatedly called against them. The full chronology of that five-year interval, the draft rules published and withdrawn, and the state-by-state state of readiness belongs to the labour codes implementation delay article in this series. For this article the operative fact is the endpoint: the Code was brought into force on 21 November 2025, together with the other three codes, and the Industrial Disputes Act, the Trade Unions Act and the Standing Orders Act stood repealed from that date, subject to the savings described above. The Industrial Relations (Central) Rules, 2026 followed on 8 May 2026 for establishments within the central sphere, while rule-making in a number of states remained incomplete, which means the Code operates through a combination of its own text, the new central rules where they apply, and, in the states, transitional arrangements that vary by jurisdiction.

That staged reality is the correct frame for everything said below about the Code in operation. A code of this kind is not a switch. Its thresholds bite immediately because they are self-executing numbers in the text. Its machinery, negotiating union verification, tribunal constitution, re-skilling fund administration, depends on rules, notifications and appointments that arrive over time. The article therefore distinguishes, at each point, between what the enacted text itself settles and what awaits the subordinate legislation that gives it hands and feet.

The objects clause against the operative text

The series thesis for articles of this kind is that a statute’s real character is found by reading its objects clause against its operative provisions, and the Industrial Relations Code rewards that reading more than most. The Bill’s stated objects were the vocabulary of consolidation: to amalgamate, simplify and rationalise the law relating to trade unions, conditions of employment and the investigation and settlement of industrial disputes, to promote ease of compliance, and to supply machinery, including recognition of negotiating unions, two-member tribunals, the removal of the reference system and a re-skilling fund, in the service of harmonious industrial relations. Presented in those terms, the Code is housekeeping: three overlapping statutes, three sets of definitions, three procedural worlds, folded into one coherent text for a labour market that had outgrown the statute book of the nineteen forties.

The operative provisions tell a more interesting story, and the difference between the two accounts is the difference between a consolidation and a settlement. A consolidation changes the arrangement of the law. This Code changed the allocation of power under it: it moved a licensing line that had been defended and attacked for forty-four years, it withdrew from the executive a gatekeeping power over adjudication that governments had held since 1947, it created a statutory bargaining agent where none had existed in central law since 1926, it attached a levy to retrenchment that no predecessor statute contained, and it subjected the strike weapon to a universal precondition. None of those five decisions is a simplification, and none of them follows from the logic of tidiness. Each was a choice between organised interests, made in the knowledge of which interest would gain and which would pay. That is why the objects clause, read alone, misleads: it describes the architecture and conceals the bargain, and the bargain is the statute. The sections that follow therefore take the operative text as the authority and treat the objects clause as evidence only of how the settlement preferred to describe itself.

One genuinely consolidating achievement should be recorded before the settlement analysis proceeds, because it is the part of the objects clause the operative text fully honours. The definitions are unified across the four codes where unification matters most. Wages bears one meaning for the Code on Wages, for this Code, for the social security computations and for the safety code, so that the structuring games played across differently defined statutes lose their purchase. The central actor of this Code is the worker, a term that replaces the workman of 1947 without the gendered and manual connotations the older word carried, and the establishment concepts interlock across the codes so that an employer’s obligations can be mapped from a single definitional foundation. Consolidation of language was delivered. It is the consolidation of power that the statute never attempted, and was never, despite its presentation, designed to attempt.

Who counts as a worker, what counts as wages, what counts as an industry

Every operative provision in the Code is gated by three definitions, and the drafting of each of the three carries a decision that a reader of the old law needs to notice. Section 2(zr) defines the worker, Section 2(zq) defines wages across all four codes, and Section 2(p) defines industry. Between them they decide who may raise a dispute, on what sum compensation is calculated, and which establishments are inside the industrial relations system at all.

The worker definition is the successor to the workman definition of the Industrial Disputes Act, and the change of noun is the least of its changes. A worker is any person, other than an apprentice engaged under the Apprentices Act, 1961, employed in any industry to do manual, unskilled, skilled, technical, operational, clerical or supervisory work for hire or reward, whether the terms of employment are express or implied. The definition expressly includes working journalists and sales promotion employees by reference to their own statutes, and it includes, for the purposes of a dispute, a person who has been dismissed, discharged or retrenched in connection with that dispute, so that the act of dismissal cannot itself strip the dismissed person of standing to challenge it. It excludes members of the armed forces, police and prison staff, persons employed mainly in a managerial or administrative capacity, and persons in supervisory capacity whose wages exceed eighteen thousand rupees per month or a higher figure the Central Government may notify. Two comparisons with the old law matter. The supervisory ceiling under the Industrial Disputes Act had stood at a figure fixed in a different economic era, and its low level pushed large numbers of senior supervisory staff outside the workman definition and therefore outside retrenchment protection; the Code’s ceiling of eighteen thousand rupees redraws that line substantially higher, while keeping the principle that the genuinely managerial employee is outside the protective chapters. And for the purposes of Chapter III, the trade union chapter, the definition is wider still, extending to all persons engaged in any trade or industry, because the right to organise was never meant to stop at the factory floor’s edge.

Why does the definition of wages matter at the moment of retrenchment?

Retrenchment compensation is fifteen days of average pay for every completed year of service, and average pay is computed on wages. The Code’s wages definition caps the allowances and exclusions that can be stripped out of the wage figure at one half of total remuneration. The definition therefore decides, in rupees, what the fifteen days are fifteen days of.

An employer who loads a salary structure with allowances cannot shrink the base on which compensation, gratuity and other wage-linked entitlements are calculated, because the definition corrects the structure arithmetically. The fuller working of that definition runs as follows. Wages means all remuneration payable to a person employed for work done, expressed in money or capable of being expressed in money, and it includes basic pay, dearness allowance and retaining allowance where they exist. It then excludes a list of items: statutory bonus, the value of house accommodation and utilities, the employer’s contribution to pension and provident funds, conveyance allowance, sums paid to defray special expenses of employment, house rent allowance, remission payable under an award or settlement, and overtime, commission and gratuity payable on termination. The guard-rail is the proviso: where the excluded items together exceed one half of total remuneration, the excess is added back and treated as wages. Remuneration in kind counts toward wages up to fifteen per cent of the total. Because the same definition serves all four codes, its effect at retrenchment is one instance of a general redesign: the old statutes each defined wages or average pay in their own way, and structuring pay to minimise the statutory base was correspondingly easier. Under the Code the fifty per cent ceiling is a hard arithmetical limit, and a salary structure drawn with seventy per cent allowances produces, by operation of the definition itself, a wage figure equal to half the total package for every computation the labour codes require.

The third definition, industry in Section 2(p), is where the Code meets the longest-running definitional litigation in Indian labour law. The statutory words enact the core of the triple test from Bangalore Water Supply: an industry is a systematic activity carried on by cooperation between an employer and workers for the production, supply or distribution of goods or services calculated to satisfy human wants and wishes that are not merely spiritual or religious in nature. Around that core the Code writes express exclusions: institutions owned or managed by organisations wholly or substantially engaged in charitable, social or philanthropic service; activities of the appropriate government relatable to its sovereign functions, with the defence, atomic energy and space departments named in the text; and domestic service. The exclusions do work the 1978 judgment was criticised for not doing, because the Rajappa test, applied without statutory carve-outs, had swept hospitals, universities, clubs and municipal functions into the definition on reasoning that many later benches found unworkable.

The interaction between Section 2(p) and the old case law was addressed by the nine-judge bench in State of Uttar Pradesh v Jai Bir Singh, which heard the reference on the correctness of Rajappa over three days in March 2026, reserved judgment on 19 March 2026 and delivered it on 20 August 2026. As the predecessor article in this series records, the opinions differed on whether reconsideration of the 1978 test was required and on how any reformulated test should run, with the Chief Justice proposing a reformulation expressly as a hypothesis that left the existing law undisturbed; but the bench was unanimous on the two operative points. Disputes governed by the Industrial Disputes Act continue to be decided by the Rajappa triple test for as long as that statute governs them, and the Code’s own definition in Section 2(p) stands on its own text, to be interpreted independently of the 1978 precedent. The practical consequence is considerable: the definitional war of half a century does not transfer into the new Code by inheritance. Arguments about Section 2(p) begin from its words, its named exclusions and its structure, and the enormous Rajappa jurisprudence is background rather than authority for the new text. The question on this definition in the FAQ section below turns on exactly that point.

Two further features of the definitional scheme govern everything downstream. The first is the appropriate government, the concept that assigns each establishment to the Centre or to a state for the Code’s purposes: the Central Government for establishments within its sphere, railways, major ports, mines and oilfields, banking and insurance companies and central public sector undertakings among them, and the state government for the great majority of factories and establishments. The assignment matters because the permission decisions of Chapter X, the appointment of conciliation officers and tribunals, the recognition of state trade unions and the power to raise thresholds by notification all belong to whichever government is appropriate for the establishment in question, and the same multinational employer may answer to the Centre for one plant and to a state for the next. The second feature is the indexation weakness built into money figures in a statute. The supervisory ceiling of eighteen thousand rupees is a 2020 judgment about where supervision ends; prices and salaries move, and a ceiling left unaltered hardens into a different policy from the one enacted, as the old Act’s ceiling demonstrated over the decades in which it stood unrevised while the supervisory workforce’s pay multiplied around it. The Code’s answer is the notification power, the Central Government may raise the figure without legislation, and the discipline of the settlement therefore depends, here as at the three hundred line, on whether governments use the adjustment powers the text hands them or allow the enacted figures to drift into meanings nobody chose.

The permission threshold: Chapter X and the move from 100 to 300

Chapter X of the Code, Sections 77 to 82, is the chapter around which the entire politics of the statute revolves. It carries forward the principle of Chapter VB of the Industrial Disputes Act: in larger establishments, lay-off, retrenchment and closure are not merely compensable events but licensed ones, requiring the prior permission of the appropriate government. What the Code changed is the population to which that licensing applies. Section 77 confines Chapter X to an industrial establishment, not being of a seasonal character or one in which work is performed only intermittently, in which not less than three hundred workers were employed on an average per working day during the preceding twelve months, or such higher number as the appropriate government may notify. The old threshold of one hundred, itself the product of a 1982 amendment that had lowered an original 1976 figure of three hundred, is replaced by the figure with which Chapter VB began. The threshold architecture of Indian labour law, in which protection is switched on and off by headcount lines drawn at different heights for different statutes, is a subject with its own map in the Indian labour law framework guide; the point inside this article is narrower, namely what crossing this particular line, in either direction, does to the parties.

For the establishment above the line, the Chapter X discipline has three limbs. Lay-off is prohibited unless the permission of the appropriate government has been obtained, with exceptions for lay-off caused by shortage of power or natural calamity and, in a mine, by fire, flood, excess of inflammable gas or explosion. Retrenchment requires that the worker be given three months’ notice in writing, or wages for the notice period in lieu, and that prior permission be obtained on an application made to the appropriate government. Closure requires an application for prior permission made at least ninety days before the intended date of closure, stating the reasons, with a copy served on the representatives of the workers. In each case the government must decide after giving an opportunity of hearing to the employer, the workers and the persons interested, having regard to the genuineness and adequacy of the reasons, the interests of the workers and the public interest. That hearing structure is inherited from the recast Chapter VB that the Constitution Bench upheld in Meenakshi Mills in 1992, after the original 1976 closure provision had been struck down in Excel Wear in 1978 for dispensing with reasons and review. The Code kept the constitutionally repaired version of the machine and changed only the headcount that switches it on.

What happens if the government does not decide a permission application in time?

The Code answers executive delay directly. Where the appropriate government does not communicate its order granting or refusing permission within sixty days of the application, the permission applied for is deemed to have been granted on the expiry of that period. Silence operates as consent.

The deeming provision converts the permission regime from a potential pocket veto into a time-bound decision, and it applies across the lay-off, retrenchment and closure limbs of Chapter X. The device deserves a moment’s attention because it illustrates the Code’s general drafting temperament, which is to keep controls but to attach clocks to them. Under the old Act a corresponding deeming provision existed, and governments learned to live with it; its retention signals that the reform’s target was never the existence of a decision but the population required to seek one and the consequences of administrative inaction. Allied to it is the review and reference mechanism: an order granting or refusing permission may be reviewed by the government itself on application, or referred to a tribunal for adjudication, and the Code compresses the life of such a reference by requiring the tribunal to render its award within thirty days of the reference. The permission, once granted or deemed, is not a permanent licence; it governs the act applied for, and an employer who retrenches on a permission obtained for different numbers, or who lets the factual basis of the application dissolve, acts outside the protection of the order.

The headcount itself repays attention, because thresholds are games as well as lines. The three hundred is a count of workers within the meaning of Section 2(zr), employed on an average per working day over the preceding twelve months. Managerial and administrative staff outside the worker definition do not swell the count, nor do supervisory staff above the wages ceiling, which means the chapter’s reach is measured on the protected population rather than on total payroll. The twelve-month averaging smooths seasonal fluctuation and prevents a single month’s hiring from switching the regime on or off, and the exclusion of seasonal establishments and establishments where work is performed only intermittently keeps the licensing chapter aimed at continuous industrial operations. An establishment hovering near the line has incentives the drafters understood: to hold the average below three hundred by distributing work among units, by substituting contract labour for direct employment, or by automation, and the history of the hundred-worker line under the old Act is full of establishments that organised themselves to sit just beneath it, a distortion of firm size that economists studying Indian manufacturing have documented for decades and that formed part of the empirical case for raising the line. A threshold does not merely regulate the establishments above it; it shapes the establishments that choose to remain below it, and any honest account of Chapter X includes the shadow it casts downward.

Below the line, the legal position is the single most consequential deregulatory fact in the Code. An establishment employing fewer than three hundred workers may lay off, retrench or close without asking any government for permission, provided it complies with the general conditions in Chapter IX, which the next section sets out: notice, compensation, seniority ordering and the rest. The unions’ description of this as hire and fire is a political characterisation of a precise legal change. What was removed is the prior executive veto over the decision. What was not removed is everything else, and the everything else is substantial enough that an employer who treats a sub-three-hundred establishment as an unregulated space will accumulate liabilities quickly: retrenchment without the statutory notice and compensation is void in the same way it always was, and the tribunal’s power to set it aside and order reinstatement with back wages survives the repeal intact.

Why did the Code keep the permission principle at all?

Because abolition was never the settlement on the table. The Second National Commission on Labour proposed raising the threshold rather than scrapping the permission idea, the states that experimented raised rather than scrapped, and the 2020 text follows that line: permission survives for the largest establishments, where closure displaces hundreds of households at once.

The argument, across four decades, was over where the line sits, and the Code moved the line. In the largest establishments the state claims a legitimate interest in testing the reasons before the event rather than compensating after it, and that claim, whatever its abuses in operation, was not one the reform’s architects were prepared to abandon. That retention is also the honest answer to the strongest version of the employer complaint, which is that permission in practice meant prohibition, because governments, unwilling to own the political cost of a closure, refused or delayed decisions regardless of the commercial merits. The complaint had evidence behind it: the jurisprudence from Excel Wear through Meenakshi Mills is a record of courts disciplining a system in which the decisive question, whether a failing unit could close, was answered by an executive process with weak obligations of candour. The Code’s response was surgical rather than abolitionist. It shrank the licensed population to the largest establishments, kept the hearing and reason-giving structure the courts had imposed, kept the sixty-day deeming rule so that indecision cannot substitute for refusal, and left governments free to raise the threshold further by notification if experience justifies it. A government that believes even three hundred is too low a line may move it upward by notification without fresh legislation; no government may move it downward by that route, because the statute fixes three hundred as the floor of the chapter’s reach.

What the threshold change did not touch: notice, compensation and lay-off pay

Chapter IX of the Code contains the general law of lay-off, retrenchment and closure that applies to industrial establishments regardless of the Chapter X line, and its continuity with the old Act is the factual core of the answer to the hire-and-fire claim. Section 70 states the conditions precedent to retrenchment of a worker who has completed one year of continuous service, continuous service being measured, as under the old law, by the two hundred and forty days rule. The worker must be given one month’s notice in writing indicating the reasons for retrenchment, or wages in lieu of notice; retrenchment compensation equal to fifteen days’ average pay for every completed year of continuous service, or part of a year in excess of six months, must be paid at the time of retrenchment; and notice must be served on the appropriate government. Only when those conditions are met is the retrenchment valid, in an establishment of thirty workers exactly as in one of three thousand. The Chapter X establishment adds the permission and the longer three-month notice on top of this base; it does not substitute for it.

Lay-off follows the same pattern of continuity. The Code defines lay-off by the employer’s failure, refusal or inability to give employment for reasons of shortage of power, breakdown of machinery, accumulation of stocks, natural calamity or other connected causes, and Section 67 carries the compensation rule: a laid-off worker who is not a casual worker and whose name stands on the muster rolls is entitled to compensation equal to fifty per cent of the total of basic wages and dearness allowance for the days of lay-off, with the Code addressing the position where lay-off stretches beyond forty-five days in a twelve-month period by opening the retrenchment route and setting off the compensation already paid. Closure outside Chapter X requires sixty days’ notice to the appropriate government of the intention to close, and compensation to workers on the same fifteen-days-per-year basis as retrenchment, with a reduced basis where closure is on account of unavoidable circumstances beyond the employer’s control. The last-come-first-go ordering of retrenchment, the employer’s duty to maintain a seniority list, and the right of a retrenched worker to be offered re-employment when the employer re-hires, all travel into the Code from the old Act. The seniority rule deserves its gloss, because it is the part of Chapter IX that workers feel first in any downsizing: within each category of workers, the employer must ordinarily retrench the worker engaged last, and may depart from that order only for reasons recorded in writing, a discipline that converts favouritism in selection into a documented and therefore challengeable act. The rule protects the longest-serving at the price of the newest recruit, and it survives the threshold reform untouched, so that a two-hundred-worker establishment that needs no permission to retrench still retrenches in the Code’s order or not lawfully at all. So does the notice-of-change discipline in Chapter V: under Section 40, an employer proposing to change conditions of service on the matters in the Third Schedule must give notice of the change and wait twenty-one days before effecting it, a cooling interval the old Act’s Section 9A made familiar.

Chapter IX also carries the transfer rule, a provision whose absence would have opened a plain evasion route. Where the ownership or management of an establishment is transferred, whether by agreement or by operation of law, a worker who has completed one year of continuous service is entitled to notice and compensation as if retrenched, unless the transferee continues the worker’s service without interruption on terms no less favourable than before, with the transferee bound to the continuity. The rule prevents the corporate form of the transaction from defeating the employment substance of it: a sale of the undertaking cannot be used to launder seniority out of the workforce, and a buyer who wants the workforce takes the workforce’s accrued position with it. Allied provisions govern the closure of part of an establishment and the compensation payable where an undertaking closes for reasons beyond the employer’s control, and the chapter’s scheme is rounded by the employer’s obligation, when re-employing after retrenchment, to give retrenched workers the first opportunity in the prescribed manner, an obligation that converts retrenchment from an absolute severance into a severance with a recall right attached.

The drafting conclusion is inescapable, and the article states it as a finding rather than an argument: the Code’s threshold reform is a reform of authorisation, not of entitlement. It decides who must ask before acting. It does not decide what the worker receives when the act occurs, and on that second question it left the 1947 settlement essentially intact, down to the fifteen-day formula that has computed Indian retrenchment compensation for generations. Commentaries that describe the threshold change as the removal of retrenchment compensation commit an elementary and consequential error, and any reader who retains one distinction from this section should retain that one.

What each side got: the ledger

The One Test for this article is whether the reader can identify precisely what the Code gave each side, and the honest way to discharge that test is to set the exchange out as a ledger. Each entry below names the gain, the side that received it and the countervailing cost that arrived in the same statute, because no entry in this Code is free. The table is the article’s findable artifact; the prose before and after it explains why the entries are stated as they are.

Gain	Received by	Countervailing cost in the same Code
Permission threshold for lay-off, retrenchment and closure raised from 100 to 300 workers	Employers	Permission retained above the line, with hearing, reasons, and tribunal review; compensation duties untouched below it
Statutory fixed term employment with a written contract for a fixed period	Employers	Parity of hours, wages, allowances and benefits with permanent workers, and gratuity after one year of service
Standing orders threshold raised to 300 workers, with Central model standing orders deemed certified when adopted	Employers	Establishments above the line remain under certified standing orders on First Schedule matters
Strike and lock-out notice extended to every industrial establishment	Employers	Lock-outs are subjected to the identical notice regime, so the employer’s counter-weapon is equally fenced
Direct access to the Industrial Tribunal after conciliation fails, ending government reference gatekeeping	Workers	The reference route’s removal also ends the possibility of government pressing an unwilling employer’s dispute into adjudication
Worker Re-skilling Fund contribution of fifteen days’ wages on every retrenchment	Workers	The fund pays a fixed fifteen days’ amount; it supplements, and does not enlarge, the fifteen-days-per-year compensation formula
Fixed term parity and gratuity eligibility after one year	Workers	The expiry of a fixed term is placed outside retrenchment, so parity is the protection and continuity of employment is not
Negotiating union and negotiating council machinery with statutory recognition	Workers	Recognition is monopolistic: one union at fifty-one per cent takes the table, and unions below twenty per cent are outside the council

Read as a whole, the ledger shows why the two loudest descriptions of the Code both fail. A pure dilution does not contain a new fund, a new bargaining agent, a new route to the tribunal and a new parity right. A mere simplification does not contain a universal strike notice, a mass casual leave rule and a recognition threshold that can exclude small unions from the table altogether. The statute is a bargain, and like most bargains it is best judged by asking who paid for what, which is the question the sections that follow answer provision by provision.

Strikes and lock-outs: notice for every establishment

Section 62 of the Code is the provision that converts the settlement’s price into law. Under the Industrial Disputes Act, the elaborate notice requirements for strikes and lock-outs, in Sections 22 and 23 of that Act, applied in their strict form to public utility services; in the generality of industry, a strike could be called without advance notice, subject to the bars on striking during the pendency of proceedings. Section 62 abolishes that distinction. In every industrial establishment, a strike is prohibited unless a notice of strike has been given to the employer within sixty days before striking, and the strike does not begin within fourteen days of giving the notice, and does not begin before the date specified in the notice. A notice is therefore not a fourteen-day warning in the simple sense; it opens a window. The strike must fall after the fourteenth day and within the sixty-day currency of the notice, and a union that lets the window lapse must serve notice again. Section 63 completes the machinery by requiring the employer who receives a notice of strike, or who gives a notice of lock-out, to report it to the appropriate government and the conciliation officer within five days, which ensures that the conciliation apparatus is alerted by the notice itself rather than by the stoppage.

The bars that follow the notice conditions are as important as the notice. No strike or lock-out may be resorted to during the pendency of conciliation proceedings before a conciliation officer and for seven days after their conclusion; during the pendency of proceedings before a Tribunal or a National Industrial Tribunal and for sixty days after their conclusion; during the pendency of arbitration proceedings, where the reference to arbitration has been notified by the appropriate government, and for sixty days after; and during the period in which a settlement or award is in operation, in respect of the matters covered by that settlement or award. The combined effect deserves to be stated without varnish: a lawful spontaneous strike is, as a practical matter, impossible anywhere in India. Between the fourteen-day interval, the conciliation bar that the notice itself triggers, and the long tails after proceedings conclude, the Code constructs a corridor of delay around every stoppage, and a strike called outside the corridor is illegal, with consequences that run beyond the strike’s own legality.

Those consequences are priced in Section 86. A worker who commences, continues or acts in furtherance of an illegal strike is punishable with a fine of not less than one thousand rupees and up to ten thousand rupees, or imprisonment up to one month, or both. An employer who commences or furthers an illegal lock-out faces a fine of not less than fifty thousand rupees and up to one lakh rupees, or imprisonment up to one month, or both. Instigating others to join an illegal strike or lock-out, and knowingly financing one, each carry fines between ten thousand and fifty thousand rupees with the same one-month imprisonment ceiling. A lock-out declared in consequence of an illegal strike, and a strike declared in consequence of an illegal lock-out, are expressly not deemed illegal, a mutual provocation rule inherited from the old Act. Alongside the penal provisions sits the civil consequence that Indian labour law has always treated as the sharper instrument: wages for the strike period depend on the strike being both legal and justified, on the principle the Supreme Court worked out in the old Act’s cases, and an illegal strike forfeits that claim at the threshold.

The Code also widened what counts as a strike. The definition in Section 2(zk) covers the conventional cessation of work and refusal to continue, and then adds a clause aimed at a tactic that had grown inside establishments where formal strikes were risky: concerted casual leave on a given day by fifty per cent or more of the workers employed in the establishment is a strike. Mass casual leave had functioned as a stoppage that wore the costume of individual leave applications, and the old definition, drafted before the tactic matured, struggled to reach it. The Code reaches it by a headcount test. The same definitional section performs the symmetrical service for lock-outs, defined as the temporary closing of a place of employment or the suspension of work or the refusal to continue to employ, so that an employer who achieves a lock-out’s effects by simply not assigning work is within the definition.

The doctrinal frame within which Section 62 operates is set by two positions the Supreme Court fixed under the old law and which the Code does not disturb. The first is that the right to strike, while a legitimate and recognised weapon of collective bargaining, is a statutory right rather than a fundamental right: the Court held in T.K. Rangarajan v Government of Tamil Nadu, in 2003, in the setting of a mass action by government employees, that no fundamental right to strike exists, and the broader case law treats the strike as a right the statute confers and may therefore condition. Section 62 is an exercise of exactly that conditioning power, which is why constitutional challenge to the notice requirement faces the headwind that the underlying entitlement is itself a creature of legislation. The second position is the wages doctrine of Syndicate Bank v K. Umesh Nayak, decided in 1994, under which strikers earn wages for the strike period only if the strike is both legal and justified, legality being a matter of the statute’s conditions and justification a matter of the merits and the employer’s conduct. The doctrine doubles the stake of Section 62 compliance: a strike that breaches the notice conditions is not merely penal in the abstract, it is a strike for which the workforce’s wage claim fails at the first limb, and union legal advisers who once treated notice as a formality for public utilities alone must retrain their practice for every establishment in the country.

The restrictions have drawn named institutional criticism that a balanced account must record. The Committee of Experts on the Application of Conventions and Recommendations of the International Labour Organisation has deprecated the practice of imposing penal sanctions, including imprisonment, on workers for participation in strikes, treating such penalties as inconsistent with the principles of freedom of association; commentators on the Code, writing in Live Law among other forums, have measured Section 86’s imprisonment clauses against that position and found the Code exposed on the point. India’s position in this debate is complicated by its non-ratification of the ILO’s core freedom of association conventions, which it has historically resisted on the ground of its own constitutional and statutory settlement; but the criticism is on the record, it comes from the body charged with supervising the conventions system, and it identifies the genuinely vulnerable flank of the strike chapter, which is not the notice requirement itself, a precondition many systems impose, but the attachment of imprisonment to breaches of it.

The Code also carries forward the prohibition on financing illegality: no person may knowingly expend or apply money in direct furtherance or support of an illegal strike or lock-out, and Section 86 prices the offence between ten thousand and fifty thousand rupees with imprisonment up to one month. The provision matters most to federations, whose strike funds are the financial skeleton of any sustained stoppage, and it converts the legality question from a concern of the workers on the gate into a concern of every office-bearer who signs a disbursement.

Fairness requires the symmetry to be stated plainly, because it is the strongest point on the government’s side of this argument. The notice regime is not aimed at one party. Every bar in Section 62 has a mirror image for lock-outs, the employer’s weapon is placed under the same fourteen-day and sixty-day discipline, and the penalties for an illegal lock-out are set markedly higher than for an illegal strike at the level of the individual fine. The union objection is therefore not, at its strongest, an objection to asymmetry. It is an objection to the notice principle itself: that a strike’s power lies substantially in its timing and surprise, that a fourteen-day interval with conciliation triggered by the notice converts the strike into a managed event whose pressure dissipates in process, and that extending the discipline from public utilities to every establishment in the country universalises that conversion. That objection is taken up in the argument section below, where the article gives it its full force before answering it.

Who speaks for the workers: negotiating unions and councils

For ninety-four years, from the Trade Unions Act of 1926 to the commencement of this Code, central Indian law contained no answer to the question every employer and every union actually asks: with whom must the employer bargain? Registration conferred legality, not representativeness. In a multi-union establishment, an employer could sign a settlement with a pliable minority union, rival unions could outbid each other in demands no employer could accept, and the resulting inter-union warfare was itself a recognised cause of industrial unrest. Section 14 of the Code supplies the missing answer, and its design is deliberately majoritarian. Where there is a single registered trade union in an industrial establishment, that union is recognised as the sole negotiating union. Where there are several, the union having the support of fifty-one per cent or more of the workers on the muster roll of the establishment, verified in the prescribed manner, is recognised as the sole negotiating union, and it negotiates with the employer to the exclusion of the others. A settlement reached with the sole negotiating union then carries the authority of a workforce mandate behind it, which is precisely what the old regime could not produce.

Where no union crosses the fifty-one per cent line, Section 14 constitutes a negotiating council instead. The council is drawn from the registered unions that have the support of not less than twenty per cent of the workers on the muster roll, each such union receiving one representative for every twenty per cent of support and one for the remainder. An agreement reached in the council requires the support of a majority of its representatives, so the council cannot act by the vote of a single large member against the rest. Recognition, whether of a union or a council, runs for three years from the date of recognition, and the parties may by mutual decision extend it, subject to a ceiling of five years in total. The employer must extend the recognised union the facilities prescribed for the conduct of negotiations. The Industrial Relations (Central) Rules, 2026 add a floor of their own for the single-union case within the central sphere, requiring a minimum membership of thirty per cent of the total workers for recognition as sole negotiating union, a qualification that matters in the many establishments where a single registered union exists on paper with a thin enrolled base.

Verification is the point on which the whole machinery will stand or fall, and the Code knows it. Support is measured against the muster roll and verified as prescribed, which in practice means a membership verification exercise conducted under official supervision, of the kind state recognition statutes have run for decades. Such exercises generate the most sensitive documents in industrial relations: lists of named workers and the union each has joined, in establishments where joining the wrong union can be a career event. Unions guard membership lists the way news organisations guard source identities, and for the same reason, that the list is the people. The discipline involved, control over who may see a roster and on what terms, is the same discipline described in accounts of how journalists use VaultBook to protect sources, and an employer or a verifying authority that treats verification records casually will find unions refusing, rationally, to expose their members to the process at all.

The vacuum Section 14 fills is worth reconstructing, because its length explains the provision’s importance. The Trade Unions Act, 1926 registered unions and then fell silent on recognition. The First National Commission on Labour, reporting in 1969, recommended statutory recognition, and the recommendation was not enacted at the centre. What filled the gap was the Code of Discipline of 1958, a voluntary compact under which employers were expected to recognise the union with the largest verified membership, and a handful of state statutes, of which the Maharashtra Recognition of Trade Unions and Prevention of Unfair Labour Practices Act, 1971 was the most complete, building a recognition jurisdiction administered by industrial courts in that state. Everywhere else, recognition was a fact of industrial strength rather than of law: an employer recognised the union it could not avoid, refused the union it could, and signed settlements whose representative authority a rival could always contest. The consequences ran through the entire system. Settlements bound the parties who signed them and persuaded nobody else; recognition strikes, called not over wages but over the preliminary question of who may speak, consumed a measurable share of industrial conflict; and the conciliation machinery spent its effort on the threshold quarrel rather than the merits. Section 14 is the first central statutory answer to that preliminary question in ninety-four years of union law, and its lateness is itself a finding about what the old settlement preferred to leave to strength.

The design choices inside Section 14 repay scrutiny, because each one allocates power. Fifty-one per cent is a winner-takes-all threshold: it produces a single, unambiguous bargaining counterpart and an end to competitive outbidding, and it does so by disenfranchising, for bargaining purposes, every worker whose union falls short, including a workforce split forty-nine to forty-eight in which nearly half the workers’ chosen agent has no seat. The twenty per cent floor for the council performs the same exclusion at a lower height, keeping very small unions out of the room in the name of a workable table. The three-to-five-year recognition term gives the recognised union a period of stability in which to bargain without perpetual re-verification, and correspondingly freezes the representational position against shifts in worker allegiance between verifications. Defenders of the design, including the Standing Committee witnesses who supported statutory recognition, answer that some exclusion is the price of any determinate answer, and that the alternative the old law offered was not inclusion but manipulation, in which the employer chose its counterpart. Critics answer that a threshold set at a majority of the muster roll, rather than a majority of those voting in a secret ballot, systematically understates support for unions among workers who fear identification, and that winner-takes-all recognition imports into industrial relations a majoritarianism that Indian electoral law itself tempers with proportional elements. Both answers are live, and the Code’s text simply chooses: muster roll, fifty-one per cent, one union.

One consequence of recognition deserves separate statement, because it is where the machinery touches workers who belong to no union at all. Under the old law, a settlement signed outside conciliation bound only the parties to it, and a workforce could live under overlapping settlements with different unions claiming different coverage. The Code’s design answers that fragmentation at the recognition point: the negotiating union negotiates for the establishment, and a settlement reached with the sole negotiating union, or an agreement concluded through the negotiating council, is the establishment’s settlement, the instrument against which individual contracts and rival claims are measured. That is the consideration the employer receives for accepting a single counterpart it did not choose: finality. A recognised settlement closes the question it settles for its term, and the Section 62 bar on strikes over matters covered by an operating settlement is the enforcement of that closure. Minority unions experience the same finality as exclusion, and both descriptions are accurate, which is precisely why the threshold figures of Section 14, fifty-one and twenty, carry the political weight they do: they decide whose signature settles the establishment and whose dissent becomes, for the term, industrially irrelevant.

Around Section 14 the Code builds the supporting structure of the bargaining relationship. Chapter III carries forward registration of trade unions on the familiar conditions, a union of workers requiring the support of ten per cent of the workers or one hundred workers, whichever is less, subject to a minimum of seven members, and it adds an appeal to the Industrial Tribunal against refusal or cancellation of registration, replacing the old appeal to the civil court hierarchy. It empowers the Central and State Governments to recognise trade unions and federations as central and state trade unions for consultative purposes. It continues the immunities that make union activity possible and the prohibition of unfair labour practices, stated in Section 84 and particularised in the Second Schedule, which lists on the employer side the classic offences of interfering with union organisation, dominating a union, dismissing a worker for union activity and refusing to bargain in good faith with the recognised union, and on the union side coercion, picketing misconduct and refusal to bargain with the employer. Breach of the prohibition carries a fine under Section 86 of not less than ten thousand rupees and up to two lakh rupees. The list will be recognisable to any reader of the old Act’s Fifth Schedule, which it substantially reproduces; what is new is not the catalogue but the fact that the recognised counterpart against whom good faith is owed is, for the first time in central law, identified by the statute itself.

Fixed term employment, written into the statute

Fixed term employment is the Code provision with the longest future. Section 2(o) defines it as the engagement of a worker on the basis of a written contract of employment for a fixed period, and then writes three conditions into the definition itself, so that they are not policy aspirations but part of what the term means. The fixed term worker’s hours of work, wages, allowances and other benefits shall not be less than those of a permanent worker doing similar work. The fixed term worker is eligible for all statutory benefits available to a permanent worker, proportionately according to the period of service rendered, even where the period of employment does not extend to the qualifying period the benefit statute would otherwise require. And the fixed term worker is eligible for gratuity on rendering service under the contract for a period of one year, a striking acceleration against the five-year qualifying period that the gratuity law applies to the generality of employees. The placement of these conditions inside a definition section is a drafting decision with teeth: an engagement that denies parity is not merely a breach of some free-standing duty, it is arguably not fixed term employment within the Code’s meaning at all, with consequences for how the engagement is characterised if litigated.

The other half of the bargain sits in the definition of retrenchment. Section 2(zh) defines retrenchment as termination for any reason other than punishment, and then excludes, among other things, voluntary retirement, retirement on superannuation, termination resulting from non-renewal or expiry of the contract, and termination on the ground of continued ill-health. The expiry of a fixed term therefore falls outside retrenchment, and with it outside the retrenchment apparatus of notice, compensation and, in Chapter X establishments, permission. The legal route the Code constructs is a clean exchange: the employer gains a hiring form in which the end date is fixed at the start and the ending is not a retrenchment event, and the worker gains, for the duration of the term, terms equal to the permanent worker beside them and a gratuity entitlement after a single year. The contrast with the neighbouring hiring form is instructive, and the Contract Labour Act 1970 explained article in this series treats that older route, in which a worker engaged through a contractor historically stood outside the principal employer’s obligations; fixed term employment is the Code’s answer to the demand that produced contract labour’s spread, an attempt to give the employer flexibility inside a direct employment relationship with parity attached, rather than flexibility purchased by interposing a contractor between employer and worker.

The abuse the parity condition is drafted against has a name in the older law. The unfair labour practices catalogue that the Code reproduces in its Second Schedule includes, on the employer side, the practice of employing workers as casuals or temporaries and continuing them as such for years with the object of depriving them of the status and privileges of permanent workers. That entry existed under the old Act alongside a system that offered the employer every incentive to commit it, because the temporary worker cost less in wages, benefits and exit obligations than the permanent worker doing the same job. Fixed term employment re-prices the temptation: where parity is real, the flexible worker costs the same per day as the permanent worker, and the employer’s remaining gains from the form are the defined end date and the absence of a retrenchment event at expiry, gains that are legitimate incidents of a genuinely time-bound engagement. The litigation the Code invites is therefore predictable in outline. Employers will characterise engagements as fixed term; workers and unions will answer that the engagement was permanent in substance, renewed in chains to simulate temporariness, or stripped of the parity that the definition makes a condition of the form itself. Tribunals will decide those cases on documents and payrolls, and their accumulated decisions will determine, establishment by establishment, which of the two statutes latent in Section 2(o) India is actually operating.

Before the Code, fixed term engagement in Indian industry rested on a thinner basis. The Industrial Disputes Act did not recognise the category in terms; its footing came from the standing orders framework, where the Industrial Employment (Standing Orders) Central Rules were amended in 2018 to introduce fixed term employment as a classification across sectors, after an earlier introduction for the apparel sector, and from case law holding that a genuine fixed term engagement ending by efflux of time was not retrenchment. The result was a category that existed by subordinate legislation and judicial tolerance, with parity expressed as policy rather than enacted as definition. The Code converts that into primary legislation. Whether the conversion dignifies flexible hiring or licenses the conversion of permanent posts into rolling terms is the question that must be answered with discipline rather than enthusiasm, and the disciplined answer is conditional. If the parity condition is enforced, if inspectors and tribunals test whether the fixed term worker in fact receives the permanent worker’s wages, allowances and benefits, the form is a legitimate flexibility with real protections, superior in every respect to the informal and contract arrangements it competes with. If parity is not enforced, the same text becomes a mechanism for staffing permanent needs with terminable labour at lower effective cost, and the one-year gratuity reads as a consolation rather than a protection. The statute supplies the condition; enforcement decides which statute India actually has. The article states that dependence rather than predicting its resolution, because the evidentiary record of enforcement under the Code is still being written by the machinery described next.

Disputes without a gatekeeper: conciliation, tribunals and the re-skilling fund

The deepest structural change in the Code is the one that attracted the least public argument, and it runs in the workers’ favour. Under the Industrial Disputes Act, an industrial dispute reached a labour court or tribunal, in the general run of cases, only if the appropriate government chose to refer it under Section 10 of that Act. The reference power was a gate, and the government held the key. Governments declined references they considered inexpedient, delayed references for years, and referred questions framed so narrowly that the real dispute could not be adjudicated; a substantial body of Supreme Court litigation exists on nothing but the reviewability of reference refusals. The individual dismissal dispute had a partial escape route, since a dismissed workman could himself be treated as raising a dispute, but even that dispute ordinarily travelled to adjudication through the same governmental door. The Code removes the door. Conciliation remains the first station, before a conciliation officer appointed by the appropriate government, but when conciliation fails and the failure is reported, a party to the dispute may apply directly to the Industrial Tribunal for adjudication. For the individual worker whose dismissal, discharge, retrenchment or termination is the dispute, the Code sets the route in terms: the worker may apply to the tribunal after forty-five days from the date of applying for conciliation of the dispute, and the application must be made within two years of the dispute arising. No ministerial discretion stands between the grievance and the adjudicator. This account is legal information about the statute’s design and machinery, not advice on any individual dispute, and a worker or employer with a live case needs representation, not a summary.

Conciliation, the station at which the Code still expects most disputes to end, is conducted by conciliation officers appointed by the appropriate government, whose function is mediation rather than decision: to intervene, to investigate, to induce the parties toward a fair and amicable settlement, and, failing settlement, to send the government a report of the failure on which the parties’ direct route to the tribunal then opens. A settlement arrived at in conciliation binds not only the signatories but, in the design inherited from the old Act, the wider constituency the Code specifies, all parties to the dispute and, in the case of a settlement in the course of conciliation proceedings, the workforce at large, which is one reason the identity of the bargaining agent certified under Section 14 matters so much when a conciliation settlement is signed. The binding force of settlements and awards is not left to honour: Section 86 makes a breach of a term of a binding settlement or award an offence punishable with a fine of not less than twenty thousand rupees and up to two lakh rupees, or imprisonment up to three months, or both, with an additional daily fine for a continuing breach, and empowers the court to direct that the fine realised be paid over as compensation to the person affected by the breach. An award of a tribunal, once it becomes enforceable, stands on the same footing. The architecture is best read as a ladder with pressure at every rung: grievance committee inside the establishment, conciliation officer outside it, tribunal beyond, and at each transition the Code has removed a discretion that previously allowed a dispute to be parked indefinitely by an authority that did not wish to decide it.

The removal of the reference gate is the gain the union critique of the Code most consistently understates. Executive gatekeeping over adjudication was not a worker protection; it was a power that governments of every complexion used to manage industrial peace politically, and its routine casualties were unorganised workers and small unions without the access to move a secretariat. Direct application converts access to adjudication from a favour into a right, subject to the discipline of conciliation first and a two-year limitation. The employer side of the ledger is real but narrower: employers lose the possibility that a friendly government will quietly bury a dispute, and gain in exchange a faster, more predictable route to a binding award in disputes they themselves wish to see resolved, together with the abolition of the multiplicity of forums the old Act maintained. Boards of conciliation and courts of inquiry disappear. Labour courts and industrial tribunals merge into a single institution, the Industrial Tribunal, with a National Industrial Tribunal above it for the largest questions.

What does the National Industrial Tribunal decide?

The National Industrial Tribunal takes the disputes the Central Government refers to it: questions of national importance, and disputes in establishments situated in more than one state, where a single state’s tribunal would be the wrong forum. It is constituted, like every Industrial Tribunal under the Code, with two members, one judicial and one administrative.

It stands at the apex of the statutory adjudication structure the Code builds, and below it stand only the constitutional courts, whose writ jurisdiction over tribunal awards the Code does not displace. The two-member composition of the tribunals is a considered change from the single presiding officer of the old system. The judicial member brings the adjudicatory discipline; the administrative member, drawn from senior civil service ranks, brings administrative experience of the industrial world the tribunal regulates. The Code and the rules made under it divide the tribunal’s work between the two-member bench and a single member according to the nature of the case, with the categories of weightier disputes reserved for decision by both members together. There is no appeal from a tribunal’s award within the Code; an award, once published, binds the parties, and challenge lies through the writ jurisdiction of the High Courts and, from them, the Supreme Court, as it did under the old law. The tribunal system’s known weakness is not design but staffing, and the Code’s text cannot conceal it. Tribunals exist only when the appropriate government constitutes them and appoints their members, and the labour judiciary under the old Act carried chronic vacancies in which presiding officers held charge of multiple districts and references aged in registries. The Code answers the design half of that problem: one tribunal type instead of three, two members instead of one, direct application instead of reference, and a composition that pairs judicial members, appointed from the judicial service and the Bar on the qualifications the Code prescribes, with administrative members drawn from senior ranks of the civil service with experience of labour administration. It cannot answer the appointment half, which belongs to governments and budgets. The honest assessment is that the Code builds a better courthouse and leaves its furnishing to the same hands that under-furnished the old one, and the direct access right it creates will be worth exactly what tribunal capacity makes it worth in each state.

The tribunal inherits, too, the adjudicatory powers that made its predecessors consequential, including the power, in a dismissal dispute, to test the employer’s punishment against the misconduct proved and to substitute a lesser punishment or order reinstatement with appropriate relief where the dismissal cannot stand, the jurisdiction the old Act expressed in its Section 11A and whose case law carries into the Code with the re-enacted substance.

Before either conciliation or the tribunal, the Code interposes two workplace-level institutions. The works committee, under Section 3, is required in an establishment employing one hundred or more workers: a consultative body of employer and worker representatives, workers’ representatives being not fewer in number than the employer’s, charged with promoting measures to secure amity and good relations and to comment on matters of common interest, and expressly not charged with adjudicating the hard questions of wages and conditions that belong to bargaining. The Industrial Relations (Central) Rules, 2026 give the committee a maximum of twenty members, equal representation from the two sides, and a meeting at least once in three months. The grievance redressal committee, under Section 4, is the newer institution: every establishment employing twenty or more workers must maintain one, of not more than ten members with equal representation of employer and workers and representation of women workers in proportion to their share of the workforce, the chairpersonship rotating between the two sides. An individual worker’s grievance over discharge, dismissal, retrenchment, discontinuance or termination must first travel to this committee, which is required to complete its proceedings within thirty days, and the worker carries the right to move onward into conciliation if dissatisfied. The committee’s threshold of twenty workers, against fifty in the short-lived grievance arrangement the old Act had introduced in 2010, extends an internal remedy to establishments the old law never reached.

What is a works committee, and which establishments must have one?

A works committee is a consultative body required where one hundred or more workers are employed. Employer and worker representatives sit on it in numbers that give the workers’ side parity or better, and its business is the everyday friction of the workplace. It recommends and discusses; it does not bargain over wages and does not adjudicate disputes.

Amenities, safety, welfare, canteens, rest facilities and every other matter of common interest the two sides bring to it are its proper agenda. Those harder functions belong to the negotiating union and the tribunal respectively, and the committee’s value lies precisely in its exclusion from them: a forum that cannot award anything is a forum in which both sides can speak without conceding. Above the workplace institutions, the Code retains voluntary arbitration as an alternative track. Parties to a dispute may agree, before a tribunal application, to refer the dispute to arbitration, and the Code gives the arbitration statutory standing, including the power of the appropriate government, where a majority of each side is party to the agreement, to notify the reference so that its discipline extends across the establishment; it is that notification which engages the Section 62 bar on strikes during the arbitration and for sixty days after. Arbitration under the Code never acquired, under the old Act, the volume its designers hoped for, because neither side wished to surrender the dispute to a private arbitrator whose award binds like a tribunal’s. Its retention, with the notification device strengthened, reads as a standing invitation rather than a prediction.

Chapter XI then adds the settlement’s most original single provision: the Worker Re-skilling Fund, constituted under Section 83. The fund consists, in the part that matters to the parties, of an employer contribution equal to fifteen days’ wages last drawn by each worker who is retrenched, or such other number of days as the Central Government may notify. The contribution is in addition to, not in substitution for, retrenchment compensation, and the amount is to be credited to the account of the retrenched worker within forty-five days of the retrenchment, to be used for the worker’s re-skilling. The design is modest in scale and significant in principle. Fifteen days’ wages will not finance a new career; what the fund does is convert retrenchment, for the first time in central law, from a pure cash settlement of the past into a transaction that carries a funded obligation toward the worker’s future employability, and it does so by a levy on the retrenching employer rather than from general revenue. Its critics, including union federations that dismissed it as tokenism, make the arithmetic point fairly. Its defenders answer that a statutory fund with a defined contribution and a defined credit period is a foundation that notification can enlarge, since the Code itself empowers the Central Government to raise the number of days. Both are right about the provision as enacted; the argument is over what it becomes.

Standing orders at three hundred

Chapter IV of the Code, Sections 28 to 39, carries the standing orders regime into the new dispensation, and its threshold decision mirrors Chapter X exactly: the chapter applies to every industrial establishment in which three hundred or more workers are employed, replacing the older Act’s general line of one hundred, which several states had already varied. Below three hundred, the conditions of service are governed by the contract of employment, the general law and the other chapters of the Code; above it, the establishment must have standing orders covering the matters set out in the First Schedule: the classification of workers, including whether they are permanent, temporary, apprentice, probationer or fixed term; the manner of informing workers of working hours, holidays, pay days and wage rates; shift working; attendance and late coming; leave and holidays and how to apply for them; the closing and reopening of sections of the establishment; termination of employment and the notice to be given; suspension or dismissal for misconduct and the acts that constitute misconduct; and the means of redress available to a worker against unfair treatment or wrongful exactions by the employer.

The procedural heart of the old system was certification: the employer drafted standing orders, the certifying officer heard the union and adjudicated the draft’s fairness and reasonableness, and the certified text then bound the establishment like a statutory contract. The Code retains certification but builds a fast lane through it. The Central Government prepares model standing orders on the First Schedule matters. An employer who prepares standing orders based on the model, covering the matters relevant to the establishment, has standing orders deemed to be certified, without a certification proceeding. Only where the employer’s draft departs from the model must certification be sought, and then only the departing clauses need to travel through the certifying officer. A certified or deemed certified text must be communicated to the workers and posted as prescribed, amendments follow the same route, and an appeal lies from the certifying officer’s decision to the appellate authority appointed by the appropriate government. An employer who modifies standing orders otherwise than in accordance with the chapter, or who acts in contravention of certified standing orders, faces penalty under Section 86, with the fine for contravention of certified standing orders set at not less than one lakh rupees and up to two lakh rupees, and higher for repetition.

The content of certified standing orders has a legal character that employers occasionally underestimate. The Supreme Court’s jurisprudence under the old Act treated certified standing orders not as a mere employer’s handbook but as statutory conditions of service, binding on both sides, which could not be overridden by a contract of employment inconsistent with them, and against which dismissals and disciplinary actions were tested. That character travels into the Code with the re-enacted machinery: a dismissal effected in defiance of the establishment’s standing orders is defective at its root, whatever the merits of the misconduct alleged, and a worker’s obligations are equally those the certified text states, not those a manager announces. The model standing order route extends that statutory character to the Central Government’s template itself in adopting establishments, which is a quiet constitutional curiosity worth noticing: a rulebook drafted by the executive acquires, by deeming, the force that the 1946 Act reserved for texts that had survived adversarial certification, and the fairness scrutiny the certifying officer once applied at the front end is replaced by reliance on the model’s own drafting and on challenge after the event. Whether that substitution holds the level of worker protection the certification jurisprudence built is one of the Code’s open empirical questions, and it will be answered in the appellate litigation over the first generation of deemed certified texts.

The change is best understood as a shift in where uniformity comes from. The 1946 Act produced uniformity slowly, establishment by establishment, through thousands of individual certifications in which certifying officers and appellate authorities, over decades, developed a genuine jurisprudence of fairness in service conditions. The Code substitutes uniformity by template: the model standing order becomes, in practice, the national default rulebook for large establishments, deviations requiring justification through certification. For employers with operations across many states, the gain is obvious, one rulebook drafted once against a central model instead of dozens of local certification histories. For workers, the position is more double-edged than the threshold headline suggests. The matters of substance, classification, misconduct, dismissal procedure, redress, remain regulated above the line, and the model text the Centre issues will itself be a negotiated public document; but the three hundred threshold removes, at a stroke, the certified rulebook from establishments between one hundred and three hundred workers that the old Act covered, and in those establishments the discipline of a certified text gives way to contract and to whatever the employer’s own unilaterally issued policies provide, tested only when a dispute reaches a tribunal. The standing orders entry in the ledger records both sides of that exchange because both sides are in the statute.

The argument over the Code: dilution, simplification, or a settlement

The public argument about this Code has been conducted in two slogans, and the stance this article takes is that both slogans fail against the text, though they fail in different degrees. The strongest version of the union case is not the slogan but a sequence of specific propositions. The permission threshold’s move from one hundred to three hundred removes the prior restraint from the establishments where most organised workers actually are, leaving compensation, which employers can price, in place of permission, which they could not buy their way past. The strike notice regime in Section 62, extended to every establishment, hollows out the strike weapon by subjecting it to a fourteen-day interval, a conciliation process the notice itself ignites, and prohibitions that run for sixty days beyond the conclusion of proceedings, so that the lawful strike becomes a scheduled event that a prepared employer can ride out with stocked inventory and contracted substitutes. The negotiating union threshold of fifty-one per cent of the muster roll entrenches incumbent major unions and freezes out emerging ones, and the exclusion of unions below twenty per cent from the council silences minorities altogether. Fixed term employment, whatever its parity text says, hands employers a statutory route to staff permanent functions with terminable workers whose incentive to organise is destroyed by the expiry date in their contracts. Central trade union federations, including the Indian National Trade Union Congress, the All India Trade Union Congress, the Centre of Indian Trade Unions and Hind Mazdoor Sabha, opposed the codes on substantially these grounds, called repeated general strikes against them, and characterised the package as a charter for hire and fire.

The strongest version of the government and employer case is equally specific. The permission regime, in the form the old Act gave it, had decayed from a safeguard into an arbitrariness: permission refusals unreasoned or endlessly delayed, closures of unviable units blocked until the unit’s value was destroyed and with it the very jobs the refusal purported to protect, and an investment climate in manufacturing measurably damaged by the knowledge that entry was free and exit was licensed. The states that raised the threshold first, Rajasthan in 2014 and others after it, produced no evidence of the predicted wave of arbitrary retrenchment, because compensation and tribunal review remained. The strike notice requirement does no more than extend to all industry the discipline that public utility services, railways, posts, power, water, had lived under for seventy years without the strike weapon disappearing from those sectors; and it binds lock-outs identically. Statutory recognition of a negotiating union answers a demand unions themselves had pressed since the National Commission on Labour’s first report era, and direct tribunal access abolishes an executive veto over workers’ cases that unions had litigated against for decades. Employer organisations, including the Confederation of Indian Industry and FICCI, supported the codes on substantially these grounds, while pressing, in the case of several industry bodies, for thresholds higher still.

The hire and fire slogan, which did more than any other phrase to define the public reception of the Code, deserves the precision treatment on its own, because it is wrong in a specific and instructive way. It is accurate about authorisation: below three hundred workers, the state no longer pre-approves dismissals of a workforce, lay-offs or closures, and a decision that once required a file in a secretariat requires only compliance with the statute. It is inaccurate about consequence: a retrenchment carried out in breach of the notice, compensation or seniority conditions is not a cheap retrenchment, it is a void one, liable to be set aside with reinstatement and back wages years later, and the tribunal route to that outcome is, under the Code, faster for the worker to reach than the reference route ever was. An employer who prices the slogan rather than the statute will discover that the Code charges for exits in arrears, with interest in the currency of litigation. The political career of the slogan is also part of the record: the joint platform of central trade union federations placed the labour codes at the head of the charter of demands for the nationwide general strike of 26 November 2020, among the largest coordinated stoppages in the country’s history, and repeated countrywide strikes against the codes followed in the years of the commencement wait, so that the Code entered into force carrying an organised opposition that its defenders must count as a fact of its operation, not a footnote to its passage.

Against the text, the verdict this article reaches is the namable claim with which it opened: the Code is a settlement, flexibility bought with process, and neither slogan survives contact with the ledger. The union case proves the restrictions are real: the strike corridor is a genuine new fetter on a constitutionally significant activity, the freedom of association that Article 19(1)(c) protects and within which the strike has been treated by the courts as a legitimate, if not fundamental-right-guaranteed, weapon of collective bargaining; and the threshold change genuinely removes a prior restraint that compensation does not replicate, since a right to be paid is not a right to be consulted before the decision hardens. But the dilution thesis cannot explain the Code’s own contents: a statute bent on one-directional transfer does not abolish the executive’s gate over adjudication, does not create a re-skilling levy, does not write parity into the definition of the flexible hiring form, and does not give a majority union, for the first time in central law, a statutory right to be recognised and bargained with. The employer case, for its part, proves that the old regime’s pathologies were real, and then overclaims: a simplification does not need to fence the lock-out with the same notice it imposes on the strike, does not leave the permission principle standing for the largest establishments, and does not preserve, untouched, the compensation formulae, the seniority rule, the re-employment right and the notice-of-change discipline of 1947. What the Code actually did is visible only in the exchange taken whole. Employers bought the freedom to resize below three hundred without permission, and paid with a universal strike-notice symmetry that also fetters their own counter-weapon, with parity for the flexible form, and with a levy on every retrenchment. Workers bought adjudication without a gatekeeper, a statutory bargaining agent and a fund, and paid with the spontaneity of the strike and the narrowing of the permission shield. Settlements are judged by whether the consideration on each side is real. On the evidence of the text, on both sides, it is.

Two refinements belong in the verdict. The first concerns the strike: the characterisation of Section 62 as the effective abolition of the strike is arguable in a way the article should mark rather than adopt. Legal strikes remain possible, on notice, outside the barred periods, and Indian public utility workers struck legally under the same architecture for decades. What is abolished is the unannounced strike, and the honest union argument is that the unannounced strike is not an abuse of the weapon but a large part of its substance, because notice is precisely the interval in which an employer neutralises a stoppage. That argument deserves its force; it is an argument about what a strike is for, and evidence cannot settle it, though the comparative fact that most mature industrial relations systems impose some notice or ballot precondition on lawful strikes, and retain functioning strike activity, limits how far the abolition language can travel. The second refinement concerns fixed term employment, where the article has already given its conditional answer: parity enforced makes the provision the most worker-protective flexibility in the Code, parity unenforced makes it the most dangerous, and the provision’s own drafting, parity inside the definition itself, hands tribunals the tool to police it if they choose to pick the tool up.

Penalties and compounding: the price list of the settlement

A settlement is only as real as its sanctions, and Section 86 of the Code is where the reader can price each party’s obligations. The section is a graded tariff rather than a single penalty, and its gradations encode the settlement’s priorities. At the top of the tariff stand the offences against the employment security chapters: an employer who contravenes the provisions governing lay-off, retrenchment or closure is punishable with a fine that may extend to ten lakh rupees, and a repeat offence after conviction attracts a substantially higher fine and the possibility of imprisonment for up to six months. The figure is a considered signal. The old Act’s penalties for illegal retrenchment were trivial sums that bore no relation to the liabilities in issue, and much of the old system’s discipline came from the civil consequence of voidness rather than from prosecution; the Code keeps the civil consequence and adds a criminal price set at a level a board of directors must notice. Standing orders contraventions sit in the next band, a fine of not less than one lakh rupees extending to two lakh rupees for acting in contravention of finally certified standing orders, rising on repetition. Breach of a binding settlement or award, as noted above, draws a fine between twenty thousand and two lakh rupees with imprisonment up to three months. Unfair labour practices draw a fine between ten thousand and two lakh rupees under Section 86(5). The strike and lock-out offences, priced earlier in this article, sit deliberately lower for the individual worker and higher for the employer, and the instigation and financing offences sit between.

Against that tariff stands Section 89, the compounding provision, which carries into this Code the decriminalising philosophy that runs across the four codes and the wider regulatory reform of the period. Offences under the Code that are not punishable with imprisonment alone, or with imprisonment and fine together in the gravest categories the section reserves, may be compounded, on application, by the prescribed officer on payment of the composition sum the section fixes by reference to the maximum fine, before or after prosecution is instituted. An offence compounded is an offence closed: no proceeding continues in respect of it. Compounding is not available as a revolving door; an offence of the same nature committed again within the period the section specifies after a previous composition or conviction cannot be compounded a second time. The pairing of Sections 86 and 89 is the settlement’s enforcement philosophy in miniature. The Code raises the stated price of defiance well above the old Act’s levels, because a permission-free regime for most establishments requires credible deterrence against the abuses the permission system once caught in advance; and it simultaneously offers the employer who has defaulted on a technical or first-time obligation a priced exit from prosecution, because a regime whose every default became a criminal trial would reproduce, in the courts, the friction the reform removed from the secretariat. Whether the balance is struck correctly depends on enforcement choices the text cannot make: compounding offered too freely converts the tariff into a fee schedule, and compounding refused converts it into the old arbitrariness wearing new numbers.

Prosecution discipline completes the picture. Cognizance of offences is channelled through complaints by or with the sanction of the appropriate government and its authorised officers, including the inspector-cum-facilitator cadre the codes create in place of the old inspectorate, and the Code contains the standard provisions attributing offences by companies to the persons in charge of the business, subject to the due diligence defence. These are unglamorous provisions and they decide more real cases than the headline thresholds, because a penalty that cannot be prosecuted is a sentiment. The article records them for the same reason it records the rest of the machinery: the settlement lives in its enforcement design, and a reader who knows only the thresholds knows the smaller half of the Code.

The settlement in operation: rules, savings and pending disputes

A repeal and replacement statute is finally judged in its transition, and the Code’s transition has three working parts. The first is commencement discipline within the Code itself: its provisions operate from 21 November 2025, and establishments that had organised their affairs around the old thresholds, the old union arrangements and the old standing orders had to re-map themselves onto the new lines, the certification status of existing standing orders, the continuance of settlements and awards, and the treatment of notices given under the old law being governed by the savings in Section 104 and the corresponding-provisions principle. Settlements and awards in operation on the commencement date run their course and bind as before; a strike notice served under the old Act’s regime is judged by the law under which it was served; a retrenchment effected before commencement is tested by the old Act’s conditions, and one effected after it by the Code’s, which is why the savings question in the FAQ section below is the one practitioners ask first.

The transition also had a geography. Labour sits in the Concurrent List, and the Code’s central sphere covers railways, mines, major ports, banking, insurance and the central public sector, while the ordinary factory in an industrial estate answers to its state government for permission decisions, conciliation appointments and tribunal constitution. Two establishments on the same road, one a central public sector unit and one a private company, therefore operate under the same Code text administered by different governments with different rulebooks at different stages of readiness, and the state amendment route that produced the threshold experiments of 2014 onward remains constitutionally available beside the Code, subject to the repugnancy discipline of the Concurrent List and presidential assent where a state law conflicts with the central text. The threshold architecture of the settlement is accordingly not a single line but a floor with notified and legislative space above it, and the framework guide linked earlier in this article maps that architecture across the labour codes as a whole.

The second part is subordinate legislation. The Industrial Relations (Central) Rules, 2026, notified on 8 May 2026, supply the central sphere’s machinery: the composition and meeting discipline of works committees, the constitution of grievance redressal committees and the nomination of workers’ representatives to them by the negotiating union or council, or by the workers themselves where neither exists, the membership floor for a single union’s recognition, and the application timelines under Chapter X, including the periods within which applications for permission to lay off and to retrench must be made ahead of the intended action. State rules for the state sphere arrived unevenly, and in states where rules remained unframed, the Code’s self-executing provisions, its definitions, thresholds, prohibitions and compensation duties, operated while machinery provisions waited on the subordinate detail, with the old rules continuing so far as they were not inconsistent with the Code. That layered position, statute live, central rules live, state rules partial, is the accurate description of the Code’s first phase, and any reader applying the Code to a particular establishment must locate that establishment’s sphere before reaching for a rule number.

The third part is the docket. The adjudicatory inheritance the Code took over was large: industrial disputes pending before labour courts and tribunals across the country, references at every stage, and a body of workmen’s individual dismissal cases in which the parties had already spent years in the old machinery. The savings scheme’s promise is that none of that investment is destroyed; the same dispute continues, the same record counts, and the forum changes its name and composition rather than its function. Overlaid on the inherited docket is the constitutional litigation the codes attracted as a package, challenges to particular provisions and to the codes’ architecture pressed by union federations and others, on which the article takes the position the evidence supports: the Code is an enacted, commenced statute whose provisions govern, and challenges to it will be decided provision by provision by the courts in the ordinary way, as challenges to its predecessors were, from Excel Wear onward. The statute’s durability will be determined less by those challenges than by the unglamorous facts of verification exercises held, tribunals constituted, funds credited within forty-five days and parity inspected. That is where the settlement either keeps its word or does not, and it is the ground on which every claim made in this article can be checked.

A reader who wants a single sentence to carry away can take the claim in its full form. The Industrial Relations Code, 2020 is the statute in which India finally traded the permission raj over industrial exit for a regime of priced consequences and formal representation: employers received the three hundred line, statutory fixed term employment and a model rulebook, workers received a recognised bargaining agent with a statute behind it, a tribunal door no government can close, parity inside the flexible form and a levy toward their next skill, and both sides surrendered the unannounced stoppage and the unaccountable refusal that had been the old system’s twin currencies. Judged as consolidation, the Code is unremarkable. Judged as what it is, a negotiated peace between interests that had fought the same battle since 1976, it is the most consequential industrial relations statute since the Act it replaced, and its consequences will be decided, as this article has argued throughout, in verification rooms, tribunal registries and payroll offices rather than in the slogans under which it was enacted.

Frequently Asked Questions

Q: Which laws did the Industrial Relations Code replace?

The Code repealed and replaced three central statutes: the Trade Unions Act, 1926, which governed registration and immunities of trade unions; the Industrial Employment (Standing Orders) Act, 1946, which governed certified conditions of service in larger establishments; and the Industrial Disputes Act, 1947, which governed disputes, strikes, lay-off, retrenchment and closure. The repeal is effected by Section 104 of the Code and took effect when the Code commenced on 21 November 2025. Section 104 carries savings: acts done under the old Acts count as done under the corresponding provisions of the Code, settlements and awards in force continue to bind, and proceedings pending under the old law are not destroyed by the change of statute.

Q: Why was the retrenchment threshold raised to three hundred workers?

The figure restores the level at which the permission regime began. When Chapter VB entered the Industrial Disputes Act in 1976, prior government permission for lay-off, retrenchment and closure applied at three hundred workers; an amendment in 1982 lowered the line to one hundred. Employers campaigned for four decades against the lower line, arguing that permission in practice meant refusal or indefinite delay and deterred manufacturing investment. The Second National Commission on Labour in 2002 proposed raising the threshold as part of a bargain that strengthened representation and adjudication machinery. Several states, beginning with Rajasthan in 2014, raised the line to three hundred under the Concurrent List before Parliament generalised the change in the Code.

Q: What notice must workers give before a strike?

Under Section 62 of the Code, no strike is lawful in any industrial establishment unless notice of the strike has been given to the employer within sixty days before striking, and the strike must not begin within fourteen days of the notice or before the date the notice specifies. The notice requirement, previously confined in this form to public utility services, applies to every establishment. A strike is also barred during conciliation proceedings and for seven days after they conclude, during proceedings before a Tribunal or National Industrial Tribunal and for sixty days after, during notified arbitration proceedings and for sixty days after, and on matters covered by a settlement or award while it operates. Lock-outs carry the identical discipline.

Q: What is a negotiating union and what membership does it need?

A negotiating union is the trade union that Section 14 of the Code recognises as the sole bargaining counterpart of the employer in an industrial establishment. Where only one registered union exists, that union is the sole negotiating union. Where several exist, the union verified as having the support of fifty-one per cent or more of the workers on the establishment’s muster roll takes that position, and the employer negotiates with it to the exclusion of other unions. Where no union reaches fifty-one per cent, a negotiating council is formed from unions with not less than twenty per cent support each, one representative for every twenty per cent, and an agreement needs the support of a majority of the council’s representatives.

Q: What rights does a fixed term employment worker get?

Section 2(o) of the Code writes the rights into the definition of fixed term employment itself. A fixed term worker, engaged on a written contract for a fixed period, must receive hours of work, wages, allowances and other benefits not less than those of a permanent worker doing similar work. The worker is eligible for all statutory benefits available to a permanent worker, proportionately according to the period of service rendered, even if the engagement is shorter than the qualifying period a benefit statute would otherwise demand. The worker becomes eligible for gratuity on completing one year of service under the contract, against the five-year period that applies generally. The trade-off is that expiry of the term is not retrenchment.

Q: Who pays into the re-skilling fund and when?

The employer pays. Section 83 of the Code constitutes the Worker Re-skilling Fund and provides that, on the retrenchment of a worker, the employer must contribute an amount equal to fifteen days’ wages last drawn by that worker immediately before the retrenchment, or such other number of days as the Central Government may notify. The contribution is owed for every retrenched worker and is in addition to the retrenchment compensation payable under the Code’s general provisions; it does not reduce or replace that compensation. The amount is to be credited to the account of the retrenched worker within forty-five days of the retrenchment, and its statutory purpose is the worker’s re-skilling for further employment.

Q: How do standing orders work under the Industrial Relations Code?

Chapter IV of the Code, Sections 28 to 39, applies the standing orders regime to industrial establishments employing three hundred or more workers. Standing orders are the certified rulebook of the establishment on the matters in the First Schedule: classification of workers, working hours, holidays and pay days, shift working, attendance, leave, termination, misconduct and dismissal, and redress against unfair treatment. The Central Government issues model standing orders; an employer who adopts the model for the matters relevant to the establishment has standing orders deemed to be certified, and only clauses departing from the model need certification by the certifying officer, whose decision is appealable. The threshold replaces the older line of one hundred workers.

Q: Has the definition of industry changed in the new code?

Yes, in form and partly in substance. Section 2(p) of the Code enacts the core of the triple test from Bangalore Water Supply: a systematic activity carried on by cooperation between employer and workers for the production, supply or distribution of goods or services to satisfy human wants that are not merely spiritual or religious. The Code then adds express exclusions the old judge-made law lacked, covering institutions of charitable, social and philanthropic organisations, sovereign functions of government including the defence, atomic energy and space departments, and domestic service. The nine-judge bench in State of Uttar Pradesh v Jai Bir Singh, in its judgment of 20 August 2026, held unanimously that the Code’s definition is interpreted on its own text, independently of the 1978 precedent.

Q: How can a worker reach the Industrial Tribunal without a government reference?

The Code abolished the reference gate. Under the Industrial Disputes Act, most disputes reached adjudication only when the appropriate government referred them, and the government could delay or refuse. Under the Code, conciliation before a conciliation officer remains the first step, but once conciliation has failed a party to the dispute may apply directly to the Industrial Tribunal. For an individual dispute over dismissal, discharge, retrenchment or termination, the worker may apply to the tribunal after forty-five days from the date of applying for conciliation. The application must be made within two years of the dispute arising. No government order stands between the worker’s grievance and adjudication, which is among the largest worker gains in the Code.

Q: What is a grievance redressal committee under the Industrial Relations Code?

Section 4 of the Code requires every industrial establishment employing twenty or more workers to maintain a grievance redressal committee for disputes arising out of individual grievances. The committee has a maximum of ten members, with equal representation of the employer and the workers, representation of women workers in proportion to their share of the workforce, and a chairperson chosen alternately from the employer’s side and the workers’ side. A worker’s grievance concerning discharge, dismissal, retrenchment, discontinuance or termination goes first to this committee, which must complete its proceedings within thirty days. A worker dissatisfied with the outcome may proceed to conciliation and onward under the Code’s dispute machinery.

Q: What counts as a strike under the Industrial Relations Code?

Section 2(zk) defines a strike as the cessation of work by a body of workers acting in combination, a concerted refusal to continue work or accept employment, and, by an express addition, concerted casual leave on a given day by fifty per cent or more of the workers employed in the establishment. The casual leave clause closed a gap in the old law: coordinated mass leave had operated as a stoppage in practice while escaping the strike definition drafted generations earlier. A stoppage within the definition attracts the Code’s strike discipline, including the Section 62 notice conditions and the bars on striking during conciliation and tribunal proceedings, whether it takes the form of a conventional walkout or a synchronised day of leave.

Q: Who is a worker under the Industrial Relations Code?

Section 2(zr) defines a worker as a person, other than an apprentice under the Apprentices Act, 1961, employed in any industry for manual, unskilled, skilled, technical, operational, clerical or supervisory work for hire or reward. Working journalists and sales promotion employees are expressly included, and a person dismissed or retrenched in connection with a dispute counts as a worker for that dispute. The definition excludes armed forces personnel, police and prison staff, persons employed mainly in managerial or administrative capacities, and supervisory staff drawing wages above eighteen thousand rupees per month or a notified higher figure. For the trade union chapter, the term extends more widely, to all persons engaged in any trade or industry.

Q: Can the three hundred worker threshold be increased further, and by whom?

Yes. Section 77 of the Code applies Chapter X to establishments with not less than three hundred workers, or such higher number as the appropriate government may notify. The appropriate government, the Central Government for establishments in the central sphere and the state government for the rest, may therefore raise the permission threshold above three hundred by notification, without amending the Code. The power runs in one direction only: the statute itself fixes three hundred as the floor, so a notification can shrink the licensed population further but cannot enlarge it below the statutory line. The same notification technique appears elsewhere in the Code, including the power to vary the supervisory wage ceiling in the worker definition.

Q: What is the penalty for an illegal strike or lock-out under the Industrial Relations Code?

Section 86 prices the two offences differently. A worker who commences, continues or acts in furtherance of a strike that is illegal under the Code is punishable with a fine of not less than one thousand rupees, extending to ten thousand rupees, or imprisonment for up to one month, or both. An employer who commences or furthers an illegal lock-out faces a fine of not less than fifty thousand rupees, extending to one lakh rupees, or imprisonment for up to one month, or both. Instigating another person to take part in an illegal strike or lock-out, and knowingly spending money in direct support of one, each carry fines between ten thousand and fifty thousand rupees, also with imprisonment up to one month.

Q: What happens to disputes and settlements pending under the repealed laws?

The savings in Section 104 of the Code carry them across. The repeal of the Trade Unions Act, the Standing Orders Act and the Industrial Disputes Act does not extinguish proceedings pending under them: disputes already before labour courts, tribunals and conciliation officers continue, acts done under the repealed statutes are treated as done under the corresponding provisions of the Code, and rights and liabilities already accrued survive, reinforced by the general savings principle in the General Clauses Act, 1897. Settlements and awards in operation when the Code commenced continue to bind the parties for their term. New disputes arising after commencement on 21 November 2025 are governed wholly by the Code’s machinery and definitions.

Q: How is a trade union registered under the Industrial Relations Code?

Chapter III of the Code carries the registration system forward from the Trade Unions Act, 1926. A trade union may apply for registration to the Registrar of Trade Unions where it is supported by at least ten per cent of the workers of the establishment or industry, or one hundred workers, whichever is less, subject to a minimum of seven members, and where its rules satisfy the Code’s requirements on objects, funds, membership and office-bearers. Registration gives the union legal personality, the capacity to hold property and sue, and access to recognition as a negotiating union or council membership where its support qualifies. The Code adds an appeal to the Industrial Tribunal against refusal of registration or cancellation of a registration, replacing the older appellate route.

Q: How long does recognition of a negotiating union last?

Recognition under Section 14 of the Code runs for three years from the date on which the negotiating union or negotiating council is recognised. The employer and the recognised union or council may, by mutual decision, extend the recognition, but the total period cannot exceed five years. When the term ends, representativeness is tested again in the prescribed manner, and a different union that has crossed the fifty-one per cent threshold, or a reconstituted council, takes the seat. The fixed term is a deliberate design choice: it gives the recognised bargaining agent a stable period in which settlements can be negotiated and worked, while preventing any single verification from permanently freezing the representation of a changing workforce.

Q: Can an industrial dispute be settled by voluntary arbitration under the Code?

Yes. The Code preserves and strengthens the arbitration track: the employer and the workers, or their representatives, may agree in writing to refer an industrial dispute to an arbitrator of their choice before it is taken to a tribunal, and the arbitrator investigates the dispute and submits a signed award to the appropriate government. Where the parties to the arbitration agreement represent the majority of each side, the appropriate government may notify the agreement, which extends its discipline across the establishment and engages the bar on strikes and lock-outs during the arbitration and for sixty days after its conclusion. An arbitration award binds the parties to the agreement, and the route offers a faster, private alternative to tribunal adjudication where trust in a chosen arbitrator exists.

Q: What must an employer do before changing workers’ conditions of service?

Section 40 of the Code carries forward the notice-of-change discipline of the old law. An employer who proposes to change the conditions of service applicable to workers on any matter specified in the Third Schedule, which covers such matters as wages and allowances, hours of work, leave, classification and rationalisation of work, must give notice of the proposed change in the prescribed manner to the workers likely to be affected. The change cannot be effected within twenty-one days of the notice, giving the negotiating union or the workers a defined interval to object, negotiate or raise the proposal as a dispute. The appropriate government may exempt specified establishments from the requirement, but without an exemption, a change effected without the notice and interval is vulnerable in adjudication.

Q: What happens if a fixed term worker is dismissed before the contract ends?

Only the natural expiry of the fixed term stands outside retrenchment. A dismissal or termination by the employer before the term runs out is a termination of service like any other, and where it amounts to retrenchment the worker is entitled to the Code’s full retrenchment protection: one month’s notice in writing or wages in lieu, compensation of fifteen days’ average pay for every completed year of continuous service, and notice to the appropriate government, with Chapter X permission applying in establishments at or above the three hundred threshold. The worker also carries the parity rights earned during the engagement and, after one year of service under the contract, eligibility for gratuity, none of which the premature termination cancels.