Social Security Code 2020: Gig Workers and the New Law

The Social Security Code 2020, enacted as Act 36 of 2020, did the largest single consolidation in Indian welfare law: it repealed nine statutes, carried provident fund, state insurance, gratuity, maternity benefit, compensation and the construction workers’ cess into one text of fourteen chapters and one hundred and sixty-four sections, and then did something none of the nine had ever attempted. In its definitions it named a workforce that Indian law had previously been unable to see. A gig worker, a platform worker and an aggregator became statutory terms, and attached to those terms came a money obligation: the platforms that organise the work must pay a contribution, fixed as a percentage of turnover and capped against what they pay their workers, into a Social Security Fund from which benefits are to flow. The Code therefore holds two achievements of very different kinds. The consolidation is real and mostly mechanical, a tidying of a body of law that had grown by accretion across ninety-seven years, from the compensation statute of 1923 to the unorganised workers’ statute of 2008. The recognition of platform work is real and entirely new. But the benefit that the recognised worker can actually claim is a third thing, and it is not yet either of the first two: it waits on schemes that the statute authorises but does not itself write. That gap, between recognition in the text and entitlement in the hand, is what this article examines, because it is the gap on which the Code will be judged.

Nine statutes, and the workforce that stood outside all of them

The nine Acts repealed by Section 164 of the Code were the Employees’ Compensation Act of 1923, the Employees’ State Insurance Act of 1948, the Employees’ Provident Funds and Miscellaneous Provisions Act of 1952, the Employment Exchanges (Compulsory Notification of Vacancies) Act of 1959, the Maternity Benefit Act of 1961, the Payment of Gratuity Act of 1972, the Cine-Workers Welfare Fund Act of 1981, the Building and Other Construction Workers’ Welfare Cess Act of 1996 and the Unorganised Workers’ Social Security Act of 2008. Read as a list, they describe the architecture of Indian social security as it actually existed. Each statute protected a defined relationship. Provident fund and state insurance attached to employment in establishments above a numerical threshold. Gratuity rewarded long continuous service with a single employer. Maternity benefit and compensation presumed an employer who could be made to pay. The cess statute was the exception that proved the design: construction workers had no stable employer, so the law taxed the construction itself and pooled the proceeds in a welfare fund. The unorganised workers’ statute of 2008 was the other exception, a framework law that named the unprotected majority and then left their protection to schemes the government might frame.

The design had a population problem that grew worse every decade. The statutes were built for a workforce of employees in identifiable establishments, and most Indians have never worked in that form. Agriculture, casual construction, domestic work, street vending, home-based piece work and self-employment together account for the overwhelming majority of Indian livelihoods, and the employment relationship on which provident fund and state insurance depend is precisely what those livelihoods lack. Thresholds compounded the exclusion. A provident fund chapter that begins at twenty employees and an insurance chapter that begins at ten will not reach a workshop of six, however hazardous its work. The result was a two-storey system: a well-built floor of contributory protection for the organised minority, resting on a ground floor where the majority had a framework Act, a welfare board on paper in many states, and very little else that a worker could point to and claim by right.

Platform work arrived inside this gap and made it impossible to ignore. The delivery rider and the ride-hailing driver work for no establishment in the statutory sense. Their engagement is mediated by an application, their instructions arrive through an algorithm, their earnings are calculated per task, and the company that organises the whole arrangement describes itself, in its contracts, as a mere intermediary between the worker and the customer. Under the nine statutes, such a worker was not excluded by anyone’s decision; the worker was simply illegible. No definition in any of the nine Acts described the arrangement, so no obligation in any of them attached to it. The National Institution for Transforming India, in its 2022 report on the gig and platform economy, estimated the platform workforce at seventy-seven lakh workers in 2020 to 2021 and projected it to reach two crore thirty-five lakh by 2029 to 2030. A workforce of that projected size, growing inside the blind spot of the welfare statutes, was the practical pressure behind the Code. The legal pressure was older. The Second National Commission on Labour, which reported in 2002, had recommended that India’s labour law be regrouped into a small number of codes and that social security be placed on a footing that reached beyond the organised sector. The Code on Social Security is where that recommendation met the rider at the traffic signal.

It does not, and the distinction matters for everything that follows. Provident fund under Chapter III is a benefit of employment in a covered establishment, financed by employer and employee contributions on wages. Gig workers are placed in a separate channel altogether: government-framed schemes, financed in part by the aggregator contribution, delivered outside the employment relationship.

A rider may one day receive benefits worth as much as provident fund coverage, but the legal route is different, the financier is different, and the claim, until a scheme exists, does not exist at all in enforceable form.

The 2008 rehearsal: a framework that delivered very little

Any honest reading of the Code’s ninth chapter has to begin with the statute it most resembles, because Chapter IX repeats an architecture that India has already run once and watched underperform. The Unorganised Workers’ Social Security Act of 2008 was enacted to considerable goodwill. It defined the unorganised worker, constituted a National Social Security Board to advise on schemes, provided for state boards, required registration of workers, and empowered the Central Government to frame schemes on life and disability cover, health and maternity benefits and old age protection. What it did not do was entitle any worker to anything. The benefits lived in the schemes, the schemes lived at the discretion of the executive, and the money lived in budgetary allocations that were never tied to a dedicated revenue stream. In the years that followed, the Act’s machinery produced registrations and advisory boards, and a shelf of schemes of varying generosity administered under other programmes, but the judgment recorded by commentators and by parliamentary scrutiny over the following decade was consistent: as a delivery instrument, the 2008 Act changed the lives of very few of the workers it named.

The comparison is not decorative, and the drafters of the Code knew the record. Chapter IX of the Social Security Code keeps the 2008 design almost piece for piece: definition of the worker, registration as the gateway, a national board and state boards, and benefits that exist only inside schemes the government frames. What the Code adds to the design is the one element the 2008 Act conspicuously lacked, a revenue line. The aggregator contribution is meant to answer the question that defeated the earlier statute: who pays. Whether a dedicated revenue line is enough to convert a framework into an entitlement machine is the substantive question about the Code, and this article returns to it after the machinery is described, because the answer depends on details of rate, cap and commencement that are easy to miss and impossible to overstate. The reason for dwelling on the 2008 Act here is disciplinary. A reader who knows that the framework approach has already been tried, under a statute whose text promised much the same things, will read Chapter IX’s promises at their correct weight: as authority to act, generously drawn, whose value will be settled entirely by what is notified under it.

The 2008 Act also supplies the Code’s most revealing continuity. Among the schemes run under the old dispensation were the welfare arrangements for cine workers financed by the Cine-Workers Welfare Fund Act of 1981 and the vast, unevenly spent construction welfare funds built by the cess of 1996. The Code repeals both statutes but preserves their logic: sectoral welfare financed by a levy on the activity rather than by an employer’s contribution on wages. The aggregator contribution is, in design terms, the cess idea carried into the platform economy. Where the construction cess taxed the building because the builder changed weekly, the platform contribution taxes turnover because the platform denies being an employer at all. In both cases the legislature, unable to find the employment relationship its older statutes required, chose to attach the obligation to the transaction instead. That is a genuinely inventive line in Indian welfare law, and the Code’s ninth chapter is its second and third use.

What the 2008 Act actually delivered

The predecessor’s record deserves a paragraph of specifics, because the Code’s scheme architecture inherits both its method and its warning. The schemes notified under the Unorganised Workers’ Social Security Act and its surrounding programmes included health insurance for below-poverty-line families through the Rashtriya Swasthya Bima Yojana, life and disability cover through the Aam Aadmi Bima Yojana, and old age pensions administered through the National Social Assistance Programme, whose origins predate the Act. These were real benefits received by real households, and an account of the 2008 Act that called it a dead letter would overstate. What the record shows instead is a pattern of delivery by programme rather than by entitlement: cover existed where a scheme had been designed, budgeted and pushed through the states, lapsed where administrative energy lapsed, and changed shape when governments changed, health insurance for the unorganised poor being reorganised into the Ayushman Bharat insurance architecture in 2018 without any worker acquiring, in the transition, a right that survived the reorganisation. Registration under the 2008 framework never became the master key the design promised, because the schemes it unlocked were administered on their own registers and their own eligibility rules.

The lesson for the Code is exact. A register earns its keep only when schemes are built to run on it, and a scheme architecture earns the name of social security only when the benefit does not depend for its existence on the programme cycle. The e-Shram register is a stronger instrument than anything the 2008 Act produced, and the Code’s fund and contribution are stronger financing ideas. Strength of instruments, however, is what the 2008 comparison is meant to test. The predecessor did not fail for want of definitions or boards, both of which it had. It under-delivered at the joint between framework and scheme, and it is at exactly that joint that the Code, in Chapter IX, remains unbuilt.

The road to the Code: a commission, two bills and one week in September 2020

The consolidation itself was two decades in the making. The Second National Commission on Labour, chaired by Ravindra Varma, recommended in its 2002 report that the dozens of central labour statutes be regrouped into a handful of codes organised by subject: wages, industrial relations, social security, and safety and working conditions. Successive governments accepted the architecture and postponed the drafting. When the drafting came, social security proved the hardest of the four, because it had to do two unlike things at once: restate the mature contributory systems of provident fund and state insurance without disturbing the savings of crores of members, and invent a provision for workers whom those systems had never covered.

The first attempt, the Code on Social Security Bill of 2019, was introduced in the Lok Sabha in December 2019 and referred to the Standing Committee on Labour, which examined the draft through the following year and reported with recommendations that included sharper definitions and clearer funding for the new categories of workers. The government withdrew the approach into a revised text. The Code on Social Security Bill of 2020 was introduced in the Lok Sabha on 19 September 2020, passed by that House on 22 September, passed by the Rajya Sabha on 23 September, and received the President’s assent on 28 September 2020, becoming Act 36 of 2020. The passage formed part of an extraordinary legislative week in which three of the four labour codes cleared Parliament in the monsoon session, in sittings marked by an opposition boycott that the government’s defenders answered by pointing to the two decades of consultation behind the drafts and that critics answered by pointing to the speed of the final passage. Both descriptions are accurate about different things: the ideas were old and heavily consulted; the enacted text was debated at the end with unusual brevity.

Two features of the enacted text show the marks of that history. The first is its uneven depth. The chapters that restate provident fund, state insurance, gratuity, maternity benefit and compensation are long and detailed, because they consolidate statutes with decades of amendment and case law behind them. Chapter IX, on unorganised, gig and platform workers, occupies sections 109 to 114, six sections in a Code of one hundred and sixty-four. The brevity is partly a choice of technique, framework drafting for a fast-changing field, and partly a measure of how much remained genuinely undecided when the text was finalised: rates, dates, scheme contents and machinery are all left to notification. The second feature is the commencement design. Section 1 of the Code empowers the government to bring its provisions into force on different dates, which converted enactment from an event into a process, and deferred the Code’s real beginning by more than five years. That deferral, and the rules process that eventually closed it, is treated in the section on commencement below; the point for the drafting history is that Parliament in 2020 enacted an architecture whose occupation was always going to be a second, executive act.

What the Code kept: the contributory core, restated

Chapters III to VIII of the Code carry forward the substance of the six contributory and benefit statutes, and a reader mapping old law to new will find the continuity deliberate. Chapter III, sections 14 to 23, restates the provident fund: it applies to establishments employing twenty or more persons, provides for the provident fund, pension and deposit-linked insurance schemes to be framed under it, and preserves the institutional position of the Employees’ Provident Fund Organisation that administered the 1952 Act. Chapter IV, sections 24 to 52, the longest in the Code, restates employees’ state insurance: it applies to establishments employing ten or more persons, extends to hazardous establishments even where fewer are employed, and carries forward the Employees’ State Insurance Corporation and its medical and cash benefit structure. Contributions under both chapters are computed on wages, and here the Code leans on the project of the four codes taken together, since the definition of wages it uses, built on basic pay, dearness allowance and retaining allowance with the value of excluded components capped at one half of total remuneration, is the uniform definition shared with the Code on Wages 2019, which supplies the common wages definition for all four codes. The wage ceilings that governed mandatory coverage under the repealed Acts, fifteen thousand rupees a month for provident fund and twenty-one thousand rupees a month for state insurance in the closing years of the old dispensation, are not written into the Code’s chapters; they operate, as in substance they did before, through the schemes and notifications under the Code, which is why the coverage question for any given establishment is answered by the scheme in force and not by the bare text of the chapter.

Chapter V, sections 53 to 58, restates gratuity for establishments employing ten or more persons, and contains one of the Code’s few substantive liberalisations of the old benefit law. Under the Payment of Gratuity Act of 1972, gratuity was earned by five years of continuous service, a condition that the Code retains as the general rule, but Section 53 disapplies the five-year condition where employment ends on death, on disablement, or on the expiry of a fixed term engagement, and directs that the fixed term employee be paid gratuity on a pro rata basis for the service actually rendered. Guidance issued by the Ministry of Labour and Employment expresses the resulting position as eligibility after one year of service for the fixed term employee, which is the pro rata principle read with the Code’s definition of continuous service. The change answers a structural unfairness that the growth of fixed term hiring had made conspicuous: an employer could engage a worker on successive short contracts for work of a permanent character and never cross the five-year line at which gratuity fell due.

Chapter VI, sections 59 to 72, restates maternity benefit for establishments employing ten or more persons, carrying forward the entitlement architecture that the 2017 amendment had built into the 1961 Act, including the twenty-six weeks of paid leave at the centre of the Maternity Benefit Act as amended in 2017, the provisions for adopting and commissioning mothers, and the creche obligation graded by establishment size. Chapter VII, sections 73 to 99, replaces the Employees’ Compensation Act of 1923, the oldest of the nine, preserving the principle of no-fault employer liability for employment injury while restating the computation and the machinery for determining compensation. Chapter VIII, sections 100 to 108, preserves the building and other construction workers’ cess, a levy of between one and two per cent of the cost of construction at the rate notified, and with it the welfare fund model for a workforce without stable employers. Chapter II, sections 4 to 13, performs the corresponding institutional restatement, providing for the social security organisations, among them the National Social Security Board under Section 6, to which this article returns, and Chapter X and Chapter XI supply the finance, accounts, authorities, assessment and recovery machinery, including the Inspector-cum-Facilitator, the Code’s softer-named successor to the inspector of the old statutes.

What this restatement does not do is as important as what it does. The Code does not merge the provident fund and state insurance systems into each other; they remain distinct schemes with distinct funds, distinct contribution structures and distinct administering bodies, sheltering together under one cover. It does not lower the numerical thresholds that kept small establishments outside the contributory net; the twenty and the ten survive. And it does not disturb accrued rights, a matter the savings provision handles expressly. The consolidation is therefore best understood as a change of address for the organised sector’s law, executed carefully, rather than a redesign of it. The redesign, such as it is, sits in Chapter IX.

It preserves their separate welfare route rather than folding them into provident fund. Chapter VIII continues the cess of between one and two per cent of construction cost that the 1996 Act created, and the cess continues to feed welfare funds for registered construction workers in the states.

The design concedes what the sector has always shown: where the employer changes with every site, a levy on the work itself is the only contribution base that stays still long enough to be collected.

Inside the restatement: what actually changes for the covered worker

A consolidation that changed nothing would still be worth having, but the Code’s restatement is not inert, and the changes that touch the already covered worker are worth separating from the relabelling. The first is the extension of state insurance to hazardous work without a numerical floor. Under the 1948 Act, coverage followed the establishment’s size and character; the Code’s Chapter IV reaches every establishment in which a hazardous or life-threatening activity is carried on, whatever the headcount, which brings the single worker in a dangerous process inside the insurance net that his neighbour in a ten-person unit already had. The second is the treatment of plantation and other employments historically at the margin of the insurance scheme, which the chapter structure handles by applicability and notification rather than by exclusion. The third is the retention, in modernised form, of the two escape routes from the thresholds: application of the provident fund and insurance chapters to smaller establishments by notification, and voluntary coverage where employer and majority of employees agree, so that the boundary of the contributory net is adjustable by executive act and by consent rather than only by amendment.

The fourth change is the one employees will feel in their payslips, and it comes from the shared definition of wages rather than from any chapter of this Code. Because contributions are computed on wages as uniformly defined, with excluded allowances capped at one half of total remuneration, an employer can no longer shrink the provident fund and insurance base indefinitely by reclassifying salary into allowances. The pay structures built under the old dispensation, in which basic pay was a modest fraction of cost to company precisely because contribution and gratuity liabilities keyed off it, face arithmetic revision under the codes taken together: the contribution base widens, take-home pay adjusts downward against higher deferred benefits, and gratuity, computed on the same definition, rises. The size of the effect varies with each employer’s structure, and honest accounts of the Code should say so; but the direction is fixed by the definition, and it is the largest financial consequence of the consolidation for the organised workforce.

The fifth change is quieter and administrative. The Code moves the system toward a single registration identity for establishments and a common compliance vocabulary across chapters, replacing the separate registrations and returns that the nine statutes had multiplied. Chapter XI supplies assessment and recovery machinery with time limits the old Acts lacked or scattered, and the Code repeatedly prefers electronic process, web-based inspection and randomised allocation of the Inspector-cum-Facilitator to the territorial inspectorate of the old regime. None of this is the stuff of headlines, and all of it decides whether the restated law is easier to obey than the law it replaced. Consolidation’s promise to the covered worker was never new benefits; it was the same benefits with less friction, fewer forms and a wider base. On the text, that promise is kept. Its delivery depends on the rules and the portals, which is the recurring sentence of the Code’s whole story, true of the chapters nobody argues about, no less than of the chapter everybody does.

Provident fund under the Code: the coverage question in detail

Because provident fund is the benefit most Indian employees actually hold, the Code’s treatment of it is where consolidation touches the most lives, and the detail repays attention. The chapter applies, as its predecessor did, to establishments with twenty or more employees, and it carries the three-scheme structure of the old dispensation: a provident fund into which employee and employer contributions accumulate as the member’s own savings, a pension scheme carved from the employer’s contribution that converts part of the accumulation into a monthly pension, and a deposit-linked insurance scheme that pays a benefit on the member’s death in service. Accumulations, pension entitlements and insurance cover earned under the 1952 Act continue under the savings, administered by the same organisation through the same Universal Account Number that employees have carried between jobs since the old Act’s own modernisation. The coverage questions that recur in practice are three. Establishments below the threshold remain outside unless they volunteer or are notified in. The wage ceiling for mandatory membership continues to be a creature of the scheme rather than the chapter, so the boundary of compulsory coverage can be moved by notification without amendment, as it was under the old Act when the ceiling was revised. And the employee who crosses into platform or self-employed work leaves the chapter behind entirely, which is the sharpest statement of the Code’s two-world design: the same person is a member with an account on one side of the line and a registrable gig worker awaiting a scheme on the other, and the Code, which contains both worlds between one pair of covers, builds no bridge of portability between the account the worker has and the fund the worker might one day claim from.

What the Code discarded

Three discards deserve notice. The first is the proliferation of separate welfare statutes and funds for individual occupations, of which the Cine-Workers Welfare Fund Act was the surviving example: the Code’s wager is that a general framework with a standing fund and board machinery will serve occupational groups better than a bespoke Act for each, a wager whose wisdom depends on how the fund is used. The second discard is the employment exchange as the state’s labour-market instrument. The 1959 Act had required employers to notify vacancies to exchanges, a compulsion honoured unevenly for sixty years; Chapter XIII of the Code, sections 139 and 140, replaces the exchanges with career centres whose statutory functions are information, guidance and the collection of employment data, an honest demotion of an institution whose register had long ceased to be where hiring happened. The third discard is penal. The old statutes had accumulated imprisonment provisions of varying severity for defaults that ranged from fraud to late filing. Chapter XII, sections 133 to 138, grades the offences, leans on fines for first defaults in the generality of cases, and provides for compounding, continuing a decriminalisation of routine compliance failure that runs across the four codes. None of these discards attracted much argument in 2020; the argument gathered, then and since, around what the Code added.

The cine workers’ fund and the end of bespoke welfare

One of the nine repeals tells the Code’s whole design story in miniature. The Cine-Workers Welfare Fund Act of 1981 was Parliament’s answer to a workforce that, like the riders of the platform age, had no stable employer: film workers engaged production by production, paid by the schedule, invisible to provident fund. The answer was a cess on the industry, collected into a dedicated fund, from which welfare for the registered occupational group could be financed. The mechanism worked well enough to be imitated for construction workers in 1996 on a far larger scale, and both funds embody the insight that where the employment relationship cannot be found, the transaction can be taxed instead. The Code repeals the cine workers’ statute without ceremony, folding its logic into the general framework, and in doing so declares the end of an era of Indian welfare legislation: the era of the bespoke Act, one occupation at a time, each with its own cess, fund and board. The declaration is defensible, since a general framework with a standing fund can in principle serve any occupation faster than a Parliament that must legislate group by group. It also carries a quiet risk that the cine workers illustrate better than anyone. A bespoke Act, for all its narrowness, named its beneficiaries in primary legislation and fenced its cess for them alone. A general framework names everyone and fences nothing in particular; the occupational group that once had its own statute competes, inside one fund and one board’s agenda, with every other constituency the framework covers. Whether the generalisation is progress for the smallest groups will be decided, once again, in the schemes.

Consultation, criticism and the committee’s unfinished business

The Code’s text was not produced without scrutiny, and the scrutiny left a paper trail that helps a reader locate the live controversies. The Standing Committee on Labour, examining the 2019 Bill, heard the two arguments that have framed the subject ever since. The government’s witnesses presented consolidation as simplification and the gig provisions as a first in Indian law, which they were. Trade union witnesses answered that a first which pays nothing is a beginning whose value is unproven, and pressed the points that remain live: the absence of a guaranteed minimum benefit anywhere in Chapter IX, the dependence of every entitlement on schemes, the low effective ceiling on the aggregator contribution, and the exclusion of the classification question. The committee’s report, submitted in 2020, recommended among other things that the definitions be sharpened and the funding of the new welfare provisions be made firmer, and the revised Bill of 2020 adjusted drafting at the margins without altering the framework character of Chapter IX. The executive’s design choice, framework first and entitlements by notification, survived its parliamentary examination intact, which is why the argument about the Code has never really been an argument about its text since. It is an argument about the use the executive would make of the text, conducted first in the committee room and later, when notifications lagged, in the Supreme Court.

Industry’s response followed the same division. Aggregators and their associations welcomed a settlement that recognised their workforce without reclassifying it, and their public statements at passage dwelled on the Code’s modernity, India joining a small set of countries that had legislated for platform work at all. The concerns the industry recorded were computational rather than constitutional: the meaning of turnover for a marketplace, the interaction of a central contribution with state levies, and the compliance mechanics of paying into a fund for workers who serve several platforms in the same week. Those concerns were well founded, as the entitlement chain shows, and they remain unanswered in the same notifications that workers await for different reasons, a symmetry worth noticing. The statute’s silences are not distributed to the advantage of one side; they have left the payer as uncertain of the bill as the payee is of the benefit. What differs is the parties’ exposure to the delay. A platform can provision for a contingent levy and price it when it lands. A rider cannot provision against an unframed accident insurance scheme, and the interval between recognition and entitlement is carried, in the meantime, entirely on the worker’s side of the arrangement.

What the Code added first: three definitions and an honest difficulty

Indian law had tried before to reach workers who stand outside the employment relationship, most elaborately through the Contract Labour Act of 1970, which regulated workers supplied to an establishment by a contractor, but that statute still worked through an establishment and a principal employer. Platform work had neither. The Code’s response was to define the arrangement instead of the relationship. Section 2, clause 35 defines a gig worker as a person who performs work or participates in a work arrangement and earns from such activities outside of a traditional employer-employee relationship. Section 2, clause 60 defines platform work as a work arrangement outside a traditional employer-employee relationship in which organisations or individuals use an online platform to access other organisations or individuals to solve specific problems, to provide specific services, or to perform other activities that the Central Government may notify, in exchange for payment. Section 2, clause 61 defines a platform worker as a person engaged in or undertaking platform work. And Section 2, clause 2 defines the aggregator, the entity on which the money obligation will fall, as a digital intermediary or a marketplace for a buyer or user of a service to connect with the seller or the service provider. The Seventh Schedule then lists the aggregator’s habitats in nine categories: ride sharing services, food and grocery delivery services, logistics services, e-marketplaces for goods or services in both marketplace and inventory models, professional services providers, healthcare services, travel and hospitality services, content and media services, and any other goods and service provider platform.

The definitions do their recognitional work, and the reader should notice exactly how they do it. Each definition is constructed by negation: the gig worker is defined by the absence of the traditional employer-employee relationship, platform work by the same absence plus the presence of an online platform. The Code thereby performs a careful manoeuvre. It brings the platform workforce inside a labour statute without deciding the question that has consumed platform litigation across the world, whether that workforce is employed. The worker is covered as a worker of a new statutory kind, and the cover does not disturb the platform’s contractual position that no employment exists. Recognition and classification, which most legal systems have treated as one question, are split into two, and the Code answers only the first.

PRS Legislative Research, in its analysis of the Bill, identified the cost of the drafting: the three definitions overlap. A ride-hailing driver is a gig worker, because the work sits outside a traditional employment relationship; a platform worker, because the work is accessed through an online platform; and an unorganised worker within Section 2, clause 86, which takes in a wage worker in the unorganised sector and also a worker in the organised sector who is not covered by the Industrial Disputes Act of 1947 or by Chapters III to VII of the Code. Overlap in a definitions section is not a pedant’s objection. Which definition a worker falls under determines which board advises on the worker’s schemes, which fund pays, and which registration record the worker’s claim travels on; a worker who is three things at once may be administered as any of them, and the Code supplies no tie-breaker because, at the level of entitlement, it expects the schemes to do the sorting. The unorganised worker definition is also the Code’s widest net: it expressly includes the home-based worker and the self-employed worker, which means Chapter IX’s unorganised schemes and Chapter IX’s gig schemes stand side by side for populations that substantially coincide. The tidy statement that the Code covers three new categories of worker should therefore be read with the overlap in view. The statute names a territory more confidently than it maps it.

What is the difference between a gig worker and a platform worker?

A gig worker is defined by the work arrangement: work performed outside a traditional employer-employee relationship, with earnings from that activity. A platform worker is defined by the route to the work: the arrangement must run through an online platform connecting the worker with those who need a service provided.

Every platform worker is therefore also a gig worker, but a gig worker engaged directly, without a platform between the worker and the customer, is not a platform worker.

Yes, within the same scheme architecture. The Code defines the inter-state migrant worker to include a worker who obtains employment through a contractor and also a worker who has come on the worker’s own and then obtained employment, widening the older picture that saw the migrant as a contractor’s recruit.

For the migrant, as for the gig worker, the operative benefit still depends on registration and on the schemes framed, and the portability of a registration record across states is the practical point on which the definition will be tested.

The Seventh Schedule read as an economic map

The Seventh Schedule’s nine categories reward a slow reading, because together they describe the platform economy as the legislature saw it in 2020 and mark the boundary of the contribution obligation. Ride sharing and food and grocery delivery, the first two categories, are the visible platform economy of the street. Logistics follows, capturing the warehouses-to-doorstep chains in which the rider’s employer-shaped counterpart is a fleet. The e-marketplace category is drafted with unusual care: it covers marketplaces that merely connect buyers and sellers and also inventory models in which the platform itself holds the goods, a distinction that matters enormously for the turnover question, since an inventory platform books the full sale value as revenue while a pure marketplace books only its commission. Professional services, healthcare, travel and hospitality extend the net to the white-collar and care segments of platform work, the freelance consultant, the home nurse booked by application, the salon worker dispatched to a flat. Content and media services reach the creator economy’s intermediaries. The ninth category, any other goods and service provider platform, is the drafter’s admission that the schedule describes a moving target, and it converts the schedule from a closed list into a principle with examples: if an online platform intermediates the supply of a service, the presumption of the Code is that its workers are within Chapter IX’s concern.

Two consequences follow for the contribution. First, the breadth of the schedule means the contribution, when activated, will be an economy-wide levy on intermediated services, not a tax on two or three famous delivery companies, and its yield and its politics will be correspondingly larger than the public argument, which has mostly been about riders, suggests. Second, the schedule’s breadth throws the whole weight of the system onto the definitions of aggregator and turnover in Section 2 and Section 114, because the boundary disputes will be fought there: whether a platform that also employs staff directly is an aggregator for its gig segment, whether a franchised outlet’s turnover counts at the brand or the outlet, and whether the value of goods passing through a marketplace is the platform’s turnover at all. These are the questions the scheme and rules must answer, and they are the same questions the industry raised in 2020. The schedule, in short, is a map of an economy drawn boldly and priced vaguely, and the vagueness is not accidental: the price of bringing low-margin, high-volume intermediation inside a welfare statute was leaving the base negotiable, and the negotiation is still pending.

Chapter IX read closely: the provisions for the unorganised worker

Chapter IX is usually discussed as if it were only about platforms, but its first provisions concern the far larger population of unorganised workers, and they deserve reading on their own terms. Section 109 empowers the Central Government to frame and notify schemes for unorganised workers on matters that include life and disability cover, health and maternity benefits, old age protection and any other benefit the government determines, and it contemplates state governments framing schemes of their own within their sphere. Section 110 addresses the funding of such schemes in the same permissive grammar as the gig provisions: funding may come wholly from the Central Government, partly from the Centre and partly from the states, partly from contributions of beneficiaries or employers, or from corporate social responsibility funds and other sources. Registration under Section 113, the boards under Section 6, and the fund under Section 141 complete the chapter’s machinery. The reader will recognise the design. It is the Unorganised Workers’ Social Security Act rebuilt inside the Code, with its boards, its register and its discretionary schemes, and the rebuilding changes the legal neighbourhood without changing the legal nature of the promise, which remains that the government may provide, on terms it sets, rather than that the worker may claim, on terms the statute sets.

That continuity cuts in two directions, and fairness requires both to be stated. In the Code’s favour, the unorganised worker provisions no longer stand in a neglected corner statute; they sit in the same text as the provident fund, are served by the same organisations, and are attached to a register, e-Shram, that has in fact been built at national scale and linked to the delivery of general programmes, from insurance enrolments to ration portability drives. A framework’s value is not exhausted by the schemes framed under its own sections, and administrators have used the register as infrastructure for benefits designed elsewhere. Against the Code, the counting problem is unchanged: India’s unorganised workforce numbers in the tens of crores on any estimate, the register’s thirty-one crore enrolments, impressive as administration, still rest on self-declaration and uneven state effort, and the schemes that would convert registration into income security in old age, disability and death remain a patchwork of general programmes rather than an entitlement under this Code. The honest summary for the unorganised worker mirrors the one for the gig worker, with the proportions altered. For the platform worker, recognition was new in 2020 and entitlement is the missing piece. For the unorganised worker, recognition dates to 2008, the register is the Code’s genuine addition, and entitlement is the missing piece for the second time.

The aggregator contribution, computed

Section 114 is the Code’s most consequential single provision, and it repays slow reading. Its first sub-section empowers the Central Government to frame and notify schemes for gig and platform workers in relation to matters that include life and disability cover, accident insurance, health and maternity benefits, old age protection and creche facilities, together with any other benefit the government determines. Its third sub-section lists the possible financiers of such schemes: the Central Government wholly or in part, the state governments in part, the aggregators, the beneficiaries themselves, and corporate social responsibility funds or other sources. Its fourth sub-section then fixes the aggregator’s share with unusual precision for a framework provision: the scheme may provide for a contribution by the aggregator at a rate notified by the Central Government, being not less than one per cent and not more than two per cent of the aggregator’s annual turnover, subject to a ceiling of five per cent of the amount paid or payable by the aggregator to its gig and platform workers. Annual turnover for this purpose excludes taxes, levies and cess paid by the aggregator to the Central Government. Its fifth sub-section adds the sentence on which everything in this article turns: the date from which the contribution takes effect is itself to be notified by the Central Government.

Work the arithmetic, because the two limits pull in opposite directions and which one binds depends on the shape of the platform’s business. Take a hypothetical aggregator with an annual turnover of one thousand crore rupees and payments to gig and platform workers of three hundred crore rupees. The band produces a figure between ten crore rupees at one per cent and twenty crore rupees at two per cent. The ceiling produces fifteen crore rupees, five per cent of the payments to workers. A notified rate of one per cent would set the contribution at ten crore rupees, the band binding; a notified rate of two per cent would be cut back by the ceiling to fifteen crore rupees. Take next a food delivery platform whose turnover, understood as the commission and charges it books rather than the value of food ordered, is small relative to the payouts that flow to riders: the ceiling will bind at almost any notified rate, and the contribution will behave like a levy of up to five per cent on worker payouts. Take instead a professional services platform that books large revenues and deploys few workers: the band binds, and the contribution behaves like a turnover tax of one to two per cent. The statute’s single formula thus operates as two different taxes on two different business models, which is either sophisticated calibration or an unresolved ambiguity, depending on the reader’s charity. The word turnover is not further resolved for this purpose in the provision, and for platforms whose accounting separates the value of goods sold from the commission earned, the difference between the two readings is the difference between a contribution that is material and one that is merely noticeable. That question, too, awaits the notification of the rate and the scheme that interprets it.

Notice what the precision of sub-section (4) does not touch. It fixes the band and the ceiling; it does not fix the rate, the start date, the collection machinery, the treatment of a platform that operates through subsidiaries and franchisees, or the fund’s entitlement to interest on delay, each of which sits in the scheme and the rules. And it fixes nothing at all until the Central Government notifies both a scheme that carries a contribution clause and the date from which the contribution runs. Until that double notification, the obligation that headlines every summary of the Code is an obligation in design. The entitlement chain set out below exists to make that plain in one view: it follows a rupee and a claim from the platform’s turnover to the worker’s hand, and marks at each step whether the Code itself does the work or hands it on.

The contribution in scale: what employment costs, and what the levy asks

Scale clarifies what the aggregator contribution is and is not. For an employee inside the contributory chapters, the employer’s statutory load under the schemes in force in the closing years of the old dispensation stood at twelve per cent of wages into provident fund and pension and three and a quarter per cent of wages into state insurance, against the employee’s own twelve per cent and three quarters of one per cent. Those percentages built individual accounts and insurance entitlements that vest in the worker. The aggregator contribution, at its maximum lawful setting, is two per cent of the platform’s turnover, and where the five per cent ceiling binds, five per cent of what the worker was paid, building no individual account and vesting nothing; it is a welfare levy pooled in a fund, from which schemes may pay benefits the worker has no personal balance against. Even on the most generous lawful assumptions, the platform’s obligation per worker is a fraction of the employer’s obligation per employee, and unlike the employer’s, it is not yet switched on. The comparison is not an argument that the two should be equal; the employment package purchases a different and deeper protection, and the levy model deliberately prices platform work lower in exchange for flexibility the platforms prize. It is an argument for reading the Code’s settlement as a price. Parliament priced the recognition of platform work at between one and two per cent of turnover, capped, deferred and pooled. Each side of the debate the article has described is, at bottom, a claim about that price: the industry’s, that it is fair for work that is genuinely intermittent and multi-homed; the unions’, that it is the discount at which the employment relationship was sold.

The pricing also explains the design’s attraction for the exchequer. A contribution from aggregators, blended in the Section 114 funding list with central and state money and corporate social responsibility funds, lets a scheme be announced as substantially industry-financed while the fund’s actual composition remains a matter of annual budgeting. If the first framed schemes lean on budgetary support, the levy will have served as political cover for expenditure the state was making anyway; if they lean on the contribution, the notified rate and its enforcement become the whole ballgame. Either way, the metric that will reveal the settlement’s truth is boring and public: the fund’s accounts, showing what came in, from whom, and what went out, to how many. Social security systems everywhere are finally judged on such tables, and the Code has arranged, by constituting a named fund in Section 141, for its own judgment to be legible when the numbers begin.

The gig worker entitlement chain

Read the chain from the top. The early steps are complete in the statute; the later steps are where the Code delegates, and the delegation compounds, because each step depends on the one before it. A contribution cannot be credited before it is collected, a scheme cannot pay before it is framed, and a worker cannot claim under a scheme that does not yet exist, however completely the worker’s own steps, registration and declaration, have been performed.

Step	What must happen	Who must act	What the Code itself requires	What is left to rules and schemes
1	Platform work is performed outside an employment relationship	The worker and the platform	The definitions in Section 2 recognise the worker, the work and the aggregator	Nothing at this step; recognition is complete in the text
2	The worker registers on the designated portal	The worker, with a self-declaration	Section 113 makes registration the gateway to benefits, with Section 142 applying Aadhaar based identification	The portal, the form of declaration and the record the worker receives
3	The aggregator’s contribution falls due	The aggregator	Section 114 fixes the band of one to two per cent of annual turnover and the ceiling of five per cent of payments to workers	The notified rate within the band, and the very date from which contribution begins
4	The contribution is credited to the Social Security Fund	The collecting machinery under the Code	Section 141 constitutes the Social Security Fund as the destination of the contribution	Collection procedure, interest on delay and reconciliation across platforms
5	A scheme for gig and platform workers is framed and notified	The Central Government	Section 114 lists the matters a scheme may cover, from life and disability cover to old age protection	The benefits, the eligibility conditions, the funding mix and the delivery channel
6	The registered worker actually receives the benefit	The fund and the scheme’s administrators	The Code authorises the payment and the machinery	Practically everything that determines what arrives, when, and on what proof

The chain is the article’s argument in miniature. Steps one and two are law. Steps three to six are, in ascending order of importance, instructions to the future. This is not said in condemnation of framework drafting, which has its place where a field moves faster than a statute can be amended. It is said so that the reader measures the Code by the right unit: not by the breadth of its definitions, which is genuine, but by the notifications issued under sections 113 and 114, which are the acts that convert those definitions into a benefit a named worker can receive.

What a scheme under Section 114 still has to decide

Because the entitlement lives in the scheme, the scheme’s unmade decisions are the Code’s real unfinished business, and listing them shows how much law remains to be written by notification. A scheme must first name its benefits in claimable form: not health benefits in the abstract, but a defined insurance cover of a stated sum for stated events, an old age pension of a stated amount from a stated age, a maternity benefit of a stated quantum for a stated period. It must fix eligibility with precision a front-line office can apply: the registered worker’s age bands, any minimum period of platform work, the treatment of a worker who serves several aggregators, and the documents or data by which platform work is proved, where the platforms’ own records are the natural source and the worker’s self-declaration the statutory starting point. It must settle the funding mix that Section 114 leaves open among the Centre, the states, the aggregators, the beneficiaries and corporate social responsibility funds, because a benefit level is only as real as the share of its cost that is actually assigned. It must choose a delivery channel, insurance company, existing social security organisation or direct transfer, and a grievance route when a claim is refused. And it must say how the central scheme composes with the state schemes and state Acts that have grown up in the interval, so that a worker in Rajasthan is neither doubly registered to no purpose nor paid twice for the same risk while a worker in a state without an Act waits for the centre alone.

None of these decisions is exotic; every functioning scheme in Indian welfare history has had to make them, and the construction workers’ welfare boards have been making versions of them, well and badly, for a quarter of a century. The point of the list is diagnostic. When the notifications under Section 114 appear, they should be read against it, decision by decision, because a scheme that names a benefit without fixing eligibility, or fixes eligibility without assigning the funding shares, has moved the entitlement chain one link and left the rest. The interval the Code has spent in framework form will then be judged by a simple standard: whether the first schemes were drafted as if the chain mattered, end to end, from the turnover figure in a platform’s accounts to the credit in a named worker’s account.

Registration: the gate between recognition and benefit

Section 113 of the Code requires the registration of unorganised workers, gig workers and platform workers, and makes the register the gateway through which scheme benefits are reached. The design conditions are light by intention: the worker must have completed sixteen years of age, must furnish a self-declaration in the prescribed form, and is identified through Aadhaar, Section 142 being the provision that applies Aadhaar based authentication to registrations and benefits under the Code. It is a detail worth pausing on that Section 142 was the Code’s first provision to be brought into force, on 3 May 2021, more than four years before the Code’s general commencement: the identification plumbing was switched on long before the house it serves was occupied, because the register was always the foundation on which the scheme architecture would stand.

The register acquired a working form in the e-Shram portal, launched on 26 August 2021, which registers unorganised workers on self-declaration and issues each a Universal Account Number meant to make the record portable across states and employments. Platform based gig work was later added among the occupations the portal records. The scale of the general register is considerable, and the scale of the platform register within it is, so far, the most instructive number in the Code’s story. A statement of the Ministry of Labour and Employment in December 2025 put registrations on the portal at over thirty-one crore unorganised workers, of whom just over five lakh were gig and platform workers. Set against the NITI Aayog estimate of seventy-seven lakh platform workers in 2020 to 2021, the register in late 2025 held well under a tenth of the platform workforce it was built to serve. Registration is voluntary in operation, driven by the worker’s own application, and a worker registers when there is something to register for. A portal that precedes the schemes it will serve asks workers to enrol for benefits that have not yet been specified, and the five lakh figure is what that request has so far produced. It is the single best piece of evidence that the entitlement chain’s weak links are not at its recognitional start.

For the individual worker, the practical position described in this article is a statement of the general law and does not substitute for advice on a particular case; registration, a retained record of the Universal Account Number, and attention to state level welfare schemes, several of which are paying benefits while the central schemes wait, are the three steps available in the meantime. The portability promise deserves a word of its own. A Universal Account Number that follows the worker between platforms and states answers the defining feature of platform work, which is churn: riders and drivers move between aggregators, work several in the same week, and leave and return with the demand cycle. A scheme register that reset with each engagement would exclude precisely the workforce it names. The Code and the portal are designed against that failure; whether the delivery systems honour the design when money begins to move is a question the first framed schemes will answer.

Can a gig worker claim benefits while working for several platforms?

Yes, in design. The register is the worker’s own, built on self-declaration and a Universal Account Number under the e-Shram system, and it is not issued by any platform or tied to a single engagement. Contributions reach the fund at aggregator level, not as deductions matched to an individual account.

Working for several platforms therefore does not split or forfeit the worker’s place on the register. What the worker can claim still depends on the scheme in force, which may set its own conditions.

The register as infrastructure: identity, data and the state’s new memory

It is easy to read the Code’s registration provisions as clerical, and that would be a mistake, because the register is the one part of the Code’s new architecture that has actually been built, and what has been built is a piece of state infrastructure with consequences beyond welfare. A self-declared, Aadhaar-authenticated register of the unorganised workforce, carrying occupation, skill, income band and migration pattern for tens of crores of workers, gives the Indian state a memory of its informal economy that it has never possessed. The welfare uses are the declared ones: targeting schemes, paying benefits, counting the constituency. The register’s other uses follow from its design. It is the natural base for the identity cards announced for gig workers, for the portability of ration and insurance entitlements across states, for skill programmes seeking trainees, and, in an emergency of the pandemic kind, for direct transfers to a workforce the state previously could not enumerate, a failure of 2020 that formed part of the political background to the Code’s own passage that year. A register of this kind also concentrates questions that the Code’s text barely touches: the accuracy of self-declared data on which entitlements will turn, the correction rights of a worker wrongly recorded, the uses to which platform-supplied data may be put when schemes begin requiring aggregators to share their rosters, and the safeguards around a database that maps the livelihoods of the poorest workforce in the country. The gig schemes, when framed, will have to answer some of these questions operationally, since a scheme that pays on the register’s record makes the record a legal fact about the worker. The Code’s defenders are entitled to count the register as the consolidation’s most concrete achievement, the more so since it was built under the authority of the one section commenced early for the purpose. Its critics are entitled to answer that a memory is not a meal. Both are right, and the distance between the two sentences is, once more, the distance between the Code as built and the Code as promised.

Boards, a fund, and schemes that must still be framed

The Code’s machinery for the new workforce has three moving parts. The first is the boards. Section 6 constitutes a National Social Security Board for unorganised workers, charged with recommending and monitoring schemes, and Section 6 also provides for State Unorganised Workers’ Boards. Section 114, sub-section (6) then adapts the national board to the platform economy: the same board is to function as the National Social Security Board for the welfare of gig and platform workers, with its composition altered for that business to include five representatives of aggregators and five representatives of gig and platform workers, alongside the Director General of the Employees’ State Insurance Corporation, the Central Provident Fund Commissioner, experts, five representatives of state governments taken by rotation, and the Joint Secretary in the Ministry of Labour and Employment as Member Secretary. The composition is a corporatist design of a familiar Indian kind, and it carries the same wager as the rest of the chapter: that a standing table at which platforms, workers and governments sit will settle by negotiation what the statute left open. Boards of this design are advisory. They recommend schemes; they do not frame them, fund them or pay benefits, and the history of the 2008 Act’s boards is a warning about mistaking the existence of the table for the delivery of the meal.

The state boards carry the heavier practical load, and their inheritance is mixed. The construction welfare boards that the cess regime built accumulated, in state after state, the pattern that audit scrutiny of welfare funds has repeatedly recorded: collections running ahead of disbursement, registers thinner than the workforce, and benefits announced faster than they were paid. The Code preserves those boards inside its architecture rather than rebuilding them, on the same continuity principle that governs the provident fund and insurance bodies. The principle is sound, since a worker registered in 2019 should not have to re-register because a code commenced in 2025, but it means the Code’s delivery machinery for its oldest welfare constituency begins life carrying the record of the regime it replaced, unspent balances and all. When the gig schemes are framed, they will be administered in the shadow of that record, and the boards’ defenders and critics will both reach for it: the defenders to show the machinery exists and can pay, the critics to show how slowly it has paid before. The fund accounts and the disbursement figures, published year by year, are the only arbiter either side will finally accept.

The second part is the fund. Section 141 of the Code constitutes a Social Security Fund, and the aggregator contribution under Section 114 is expressed to be credited to it. A named fund matters more than its modest place in Chapter XIV suggests. Money collected for gig worker welfare under a state law or a central scheme that is not fenced in a dedicated fund can be absorbed into general revenues and quietly spent elsewhere, a fate that befell portions of the construction cess in several states, where audit reports found collections unspent or diverted while registered workers went unpaid. The fence will be tested by the same pressures, but its presence in the text gives auditors, boards and courts a fixed point to test against.

The third part is the schemes themselves, and here the reader should hold two lists side by side. The list of matters is generous: life and disability cover, accident insurance, health and maternity benefits, old age protection, creche facilities, and anything else the Central Government determines. The list of framed central schemes that a gig or platform worker can claim under Section 114 is, at the time this article’s account closes, empty; the benefits that have actually reached platform workers have come from state legislation and from general programmes, not from a scheme notified under the section. Everything about the labour codes’ implementation delay converges on this point, because a code whose benefits live in schemes is delayed not once, at commencement, but again at every scheme the commencement was meant to enable. Draft central rules under the Code were circulated for comment in November 2020, and the rules process, like the commencement process, ran for years behind the text. None of this makes the scheme route illegitimate; social security for a heterogeneous workforce almost has to be assembled from schemes rather than a single uniform benefit. It does mean that the Code’s centre of gravity sits outside its own text, in notifications that carry no parliamentary moment and attract little of the scrutiny the parent statute received.

The institutions: who administers the Code

A consolidated text needs consolidated administration, and Chapter II of the Code, sections 4 to 13, performs the institutional settlement. The provident fund chapter is administered through the Central Board of Trustees and the Employees’ Provident Fund Organisation beneath it, continuing the tripartite board of government, employers and employees that has governed the fund since the 1950s. The insurance chapter runs through the Employees’ State Insurance Corporation, a body corporate with its own tripartite standing committee and its hospital network, whose medical infrastructure is the largest single asset in Indian social security and whose performance, praised for its reach and criticised for its quality, the Code inherits without redesigning. The National Social Security Board under Section 6 and the state boards under its sub-section (9) carry the unorganised, gig and platform mandates, advisory in character, tripartite-plus in composition, with the Section 114 adaptation adding aggregator and worker representatives when the board sits on platform welfare. Building workers’ welfare remains with the state boards built under the cess regime, whose accumulated funds and uneven disbursement record are the Code’s inheritance in that sector.

The institutional map explains a feature of the Code that puzzles first-time readers: a single statute administered by a federation of bodies that predate it, each with its own board politics, audit and administrative culture. The alternative, a unified social security authority, was canvassed in the long reform debate and rejected in the drafting, on the defensible ground that merging the provident fund’s investment machinery, the insurance corporation’s hospitals and the welfare boards’ disbursement habits into one new body would have put every existing beneficiary’s payment at transitional risk. The cost of the safer choice is coordination: the worker whose life crosses the categories, the construction worker who drives a cab in the off-season, the factory hand who retires into domestic work, will be administered by different bodies under one Code, and the promise that the register, at least, will be common is the thread meant to hold the federation together. On the register’s success, therefore, rests more than Chapter IX. It is the Code’s only entirely new institution, and the one on which the consolidation’s administrative claim will ultimately be graded.

Compliance redesigned: assessment, facilitation and penalties

The Code’s treatment of the employer-facing machinery deserves a section of its own, because it is where consolidation meets the daily experience of the establishments that fund the system. Chapter XI, sections 122 to 132, gathers the authorities, assessment, compliance and recovery provisions that the old Acts had scattered across nine texts with nine vocabularies. The Inspector-cum-Facilitator is its emblematic figure: the inspector renamed to signal an advisory first function, empowered to examine, to advise employers and employees on compliance, and to conduct inspections allocated through electronic systems rather than territorial discretion. Whether the rename changes conduct on the ground is an administrative question the text cannot answer, but the design intent is consistent with the rest of the codes: inspection as an exception triggered by risk and data, not as a routine visitation. Assessment of dues, across provident fund, insurance and cess, is given time-bound shape, and recovery follows the familiar coercive ladder of attachment and sale that the old Acts built, expressed in common terms.

Chapter XII, sections 133 to 138, then rationalises the consequences of failure. The scheme is graded: offences are sorted by gravity, fines carry the weight for first defaults in the generality of cases, imprisonment recedes to the serious and the repeated, and compounding is provided for offences not punishable with imprisonment alone, so that a default can be closed by payment rather than prosecuted as a matter of course. For the aggregator, the compliance chapter will matter in a specific way once the contribution is activated: the assessment machinery is the route by which a disputed turnover figure or a contested payout base becomes a determined liability, and the offences chapter is where failure to pay after determination is priced. The entitlement chain’s third step, the contribution falling due, is therefore not self-executing merely when the rate is notified; it depends on this machinery being staffed and used, a dependence the reader can add to the list of things the text arranges and the administration must animate. The redesign is genuinely modernising as drafting, and it shares the Code’s general condition: its quality will be known only in operation, and operation has a commencement date for the chapters but not yet a track record for the whole.

Savings, repeal and a five-year commencement

Section 164 performs the repeal of the nine Acts with the standard, necessary savings: things done under the repealed statutes, including registrations, contributions collected, funds constituted and schemes in operation, are preserved and treated as done under the corresponding provisions of the Code, so that a provident fund member’s account, an insured person’s benefit record and a construction worker’s registration do not fall with the statutes that created them. The Employees’ Provident Fund Organisation and the Employees’ State Insurance Corporation continue as the administrators of their chapters, the schemes framed under the repealed Acts continue until replaced, and the transition for the organised sector is therefore designed to be, and in the event has been, administratively quiet. The reader who wants the wider picture of what the four codes together kept and repealed will find it in the comparison of the labour codes with the old laws they replaced; the social security chapter of that story is the one where continuity for the already covered and novelty for the never covered had to be drafted into a single text.

Commencement followed the four codes’ common pattern, and it must be stated exactly. The four labour codes, this Code among them, were brought into force on 21 November 2025. The Central Rules under the codes were published on 8 May 2026. State rules remained incomplete at the close of this article’s account, with the rules under the repealed statutes continuing to operate where consistent with the codes. Within that common pattern the Social Security Code carries its own internal stagger: Section 142 on Aadhaar authentication had been in force since 3 May 2021, so that the registration infrastructure of e-Shram could lawfully be built on it, and the general commencement of November 2025 switched on the chapters whose schemes and rules were still being completed. Commencement of a framework code is a peculiar legal event. The provident fund and insurance chapters stepped into a running system and changed little overnight. Chapter IX stepped into an empty room: its commencement made the definitions law, the board provisions operative and the contribution provision available to be activated, and none of that, without the notifications the sections await, put a rupee into the fund or a benefit into a rider’s hands. The five years between assent and commencement were not idle; they contained the portal, the draft rules and two general elections’ worth of debate about the codes. But the interval also normalised a legislative habit worth naming, in which Parliament enacts, the executive schedules, and the citizen is asked to treat the announcement of a right and the availability of the right as separated by a period of administrative preparation that has no statutory limit.

The savings deserve one further word, because repeal in social security law is never only about the future. Proceedings pending under the repealed Acts, assessments half-completed, prosecutions launched and appeals filed, are carried through the repeal by the savings in Section 164 and by the general law of repeal, so that a default committed under the old dispensation is still pursued under it, and a benefit that had accrued is still payable out of it. Employers sometimes read a consolidation as a reset, a suggestion that pre-Code liabilities dissolve into the new text. The Code gives that reading no support: contributions due under the 1952 Act remain due, compensation awarded under the 1923 Act remains recoverable, and the inspector’s file crosses the repeal intact. Continuity, in the savings, runs in both directions, protecting the member’s account and the state’s claim with the same sentence.

The five years between assent and force

The interval between the Code’s assent in September 2020 and its general commencement in November 2025 is part of its meaning, and an account of the Code should describe what filled it. Legally, the nine old Acts remained the law; not a single provident fund contribution or insurance benefit in those years ran under the Code, save that Section 142, the Aadhaar provision, was brought into force on 3 May 2021 to undergird the register. Administratively, the interval was used to build e-Shram, to circulate draft central rules in November 2020 and revise them in consultation, and to press states on their draft rules, a process whose incompleteness is part of the commencement position recorded in this article. Politically, the interval contained the announcement in the Union Budget for 2021 to 2022 that social security benefits would be extended to gig and platform workers and that a portal would gather information on them, the portal duly arriving as e-Shram that August, so that the executive could point, in each year of the interval, to preparatory acts that were genuinely occurring.

The interval also had a jurisprudential cost that deserves recording. A statute book that carries both a repealing code not yet in force and the nine statutes it will repeal obliges every adviser, employer and worker to ask, for half a decade, which text governs, and the answer changed by section and by notification. Courts in the interval continued to decide provident fund and gratuity questions under the old Acts, as they had to, while litigants on both sides positioned for the new text. The Supreme Court’s direction of February 2025, requiring the government to file a definitive timeline for framing the rules, was the judiciary’s entry into the interval’s management: not a decision on the Code’s merits, but a refusal to let the executive’s timetable remain entirely its own. When commencement came in November 2025, it therefore closed an interval that had itself become a subject of litigation and commentary, and the Central Rules of May 2026 continued the pattern of the operative text trailing the proclaimed one. The reader who keeps one date from this article’s chronology should keep the right one for the right purpose: 2020 for the law’s making, 2025 for its force, and a date still unwritten for the gig worker entitlement that the making and the force were supposed, between them, to produce.

The question the Code declines: employee, or not

Beneath the Code’s architecture sits the dispute it was drafted to avoid. If platform workers are employees, the entire apparatus of Chapter IX is a detour: employees get provident fund and state insurance by right under Chapters III and IV, financed by employer contributions an order of magnitude larger than a one to two per cent turnover levy. If they are independent contractors, the classification that platforms assert in their terms of engagement, then Chapter IX is the only protection they will ever have, and its scheme dependence is the ceiling of their welfare. The Code chooses neither proposition. It covers the worker without classifying the worker, which is politically deft and legally incomplete, because the classification question does not disappear when a statute declines to answer it; it migrates to the courts.

It has duly migrated. The Indian Federation of App-based Transport Workers has petitioned the Supreme Court contending that the control platforms exercise over the work, through allocation algorithms, pricing, ratings and deactivation, creates in substance an employer-employee relationship, and that the denial of employee status offends the equality, livelihood and anti-exploitation guarantees of Articles 14, 21 and 23 of the Constitution. The petition asks the court to do what the legislature would not. Separately, and more procedurally, the pace of the executive has itself come before the court: on 18 February 2025 the Supreme Court directed the government to file a definitive timeline for framing the rules under this Code, a judicial acknowledgement that a statute’s promise decays when its subordinate legislation does not arrive. This article predicts neither litigation; it records that both exist, and that each is, in its own way, a claim that the distance between the Code’s text and the worker’s entitlement has lasted too long.

The comparative picture shows that the classification question has been answered elsewhere, in opposite directions, and that India’s legislative sidestep is a third route rather than an evasion unique to it. In the United Kingdom, the Supreme Court held in 2021 that Uber drivers were workers within the British statutory category, reasoning from the reality of the platform’s control over price, terms and performance rather than from the contract’s labels, and worker status there carries minimum wage and holiday entitlements. In California, legislation in 2019 applied a strict test that treated drivers as employees, and a ballot measure in 2020, Proposition 22, then carved app-based drivers back out into independent status with a floor of benefits attached, a legislative compromise purchased at the ballot box. The European Union’s Platform Work Directive of 2024 builds a rebuttable presumption of employment into platform work and couples it with duties of algorithmic transparency. Each system has distributed the costs of platform work differently among platform, worker, consumer and state. India’s Code distributes them provisionally, through a fund and schemes not yet framed, and reserves the distribution question for later. A reader assessing the Code should therefore resist the claim, heard from both camps, that its silence on classification settles anything. Supporters of the Code present the silence as respect for a new form of work that employment law fits badly. The trade union critique, of which the federation’s petition is the legal expression, presents the same silence as the point of the exercise: recognition cheap enough to leave the cost structure of platform business untouched. Both readings are available in the text, which is precisely why the litigation matters.

One feature of the Code’s own drafting sharpens the litigation’s interest. The Code defines an employee broadly enough to include a person employed through a contractor, and its contributory chapters therefore already know how to protect a worker whose immediate paymaster is an intermediary rather than the enterprise that benefits from the work. The platform worker’s exclusion from those chapters is thus not a conceptual necessity flowing from intermediation itself; Indian social security had solved the intermediary problem once, for contract labour, by fastening obligations on the principal employer. Chapter IX declines to repeat that solution for platforms, preferring the levy on turnover to the fiction of employment through the intermediary. Defenders call the preference realism about a genuinely different work form. Critics call it the moment the statute chose the cheaper answer. The courts, when they rule, will be choosing between those descriptions with the contract labour precedent open on the desk.

The states legislate first

While the central scheme under Section 114 waited, the states moved, and their statutes are instructive because they show what legislating for platform workers looks like when a legislature intends money to move. Labour and social security sit in the Concurrent List of the Seventh Schedule to the Constitution, so the field was open to them. Rajasthan came first, with the Rajasthan Platform Based Gig Workers (Registration and Welfare) Act of 2023, which constituted a state welfare board for platform workers, required registration of aggregators and workers, and financed welfare through a welfare fee levied as a notified percentage of the value of each transaction on the platform, a per-transaction cess rather than a percentage of annual turnover. Karnataka followed, through an ordinance in 2025 and then an Act passed in August 2025, fixing its welfare fee as a band of one to five per cent of the payout on each transaction, and adding obligations the central Code does not contain at all: notice before a worker’s engagement is terminated or the worker deactivated, and duties of transparency about the automated systems that allocate work and rate performance. Other states, among them Bihar, Jharkhand and Telangana, have brought forward their own measures in Act or draft form. The direction of travel is consistent even where the instruments differ: fee on the transaction, register at the state board, benefits from a state fund.

The coexistence of state statutes with the central Code raises a question that neither level of government has yet had to answer in operation. A national platform whose riders work in Jaipur and Bengaluru will, on the texts as they stand, owe a welfare fee under each state Act and, once Section 114 is activated, a contribution under the central Code, computed on different bases, a percentage of transaction value or payout in the states and a percentage of annual turnover at the centre, payable into different funds for overlapping benefits. Nothing in the Code as enacted reconciles the two levies or credits one against the other. The likeliest reconciliations are administrative, a central scheme that treats state fees as satisfying the contribution in part, or legislative, an amendment once the double charge is litigated by an aggregator with standing and incentive. Until one arrives, the states’ head start has a further consequence for the central design: the state boards’ registers, built earlier and attached to paying schemes, may become the operative records of who India’s platform workers are, leaving the central register and the Universal Account Number to negotiate a federation the Code imagined as a hierarchy. Federalism was not an afterthought in this field; it has become the field’s leading edge, and any account of the Social Security Code that stops at the central text describes only the smaller half of the law in action.

Health cover for the platform worker: the benefit most likely to arrive by another road

Of all the matters listed in Section 114, health and accident cover are the ones a rider’s working life makes urgent, and they are also the ones where the Indian state’s existing machinery is deepest. The Employees’ State Insurance Corporation runs one of the largest occupational health systems in the world for the insured population, hospitals, dispensaries and cash benefits combined, but its door is the employment door, and Chapter IX does not open it to the gig worker. The general health insurance architecture, Ayushman Bharat’s hospitalisation cover for eligible households, is the other road, and it is the road the executive has in fact taken: the Union Budget for 2025 to 2026 announced that gig workers would be issued identity cards through the e-Shram register and brought within the health cover of the national insurance programme, a benefit for of the order of one crore gig workers on the government’s own estimate. The announcement is characteristic of the Code’s era and worth reading precisely. The constituency is the Code’s, the register is the Code’s infrastructure, the entitlement vocabulary is the Code’s, and the legal vehicle is a general programme entirely outside Section 114. Nothing in that is objectionable as delivery; a hospital bed does not ask which section authorised it. But it confirms the article’s central distinction from the worker’s side of the counter: what reaches the gig worker in the Code’s first years will arrive because a programme adopted the worker, not because a scheme under the Code conferred a claim, and programmes can be redesigned, retargeted and renamed by the same executive discretion that created them, without the worker’s statutory position moving at all.

The accident gap is narrower and harder. Platform work on two wheels carries a road risk that is the closest analogue in the new economy to the factory risks the compensation chapter was built for, and Section 114 lists accident insurance expressly among the matters for schemes. In the interval, the cover a rider actually has is the cover the platform voluntarily buys, the cover a state Act mandates, or none. An entitlement architecture that began with the risk most likely to hospitalise its constituency would look like accident insurance framed first, funded by the contribution, and payable on the register’s record. That no such scheme was the first fruit of Chapter IX, while health cover arrived by programme adoption instead, tells the reader something real about the framework method: it delivers in the order of administrative convenience, not in the order of the worker’s exposure, because no one’s statutory duty orders the sequence. The Code gave the sequence a possible author in the National Social Security Board, whose recommendations on schemes for this constituency will be the document to watch when it speaks.

Maternity, creche and the woman platform worker

Chapter VI’s maternity benefit, twenty-six weeks of paid leave and the protections around it, belongs to the employment world: it is the employer who pays, and the establishment that must provide the creche. The woman who delivers food or provides home services through a platform stands outside that chapter for the same reason she stands outside provident fund, and Section 114’s list is where her maternity protection is supposed to be built, health and maternity benefits and creche facilities being named expressly among the matters for gig and platform schemes. The naming matters, because the drafters could have confined the list to insurance against death and accident, the risks of the male rider who dominates the platform workforce’s image; instead the list tracks the full span of a working life, including the life events that push women out of platform work altogether. The NITI Aayog’s study of the platform economy recorded women as a growing share of the workforce and identified the absence of maternity and care support among the constraints on their participation. A framed scheme that paid a maternity benefit on the register’s record, without an employer’s leave to negotiate, would be the single most distinctive welfare instrument the Code makes possible, since no chapter of the employment law can deliver it to a worker with no employer. Its absence from the first years of the Code’s operation is therefore not a detail. It is the cleanest test of whether Chapter IX will be used for the things only it can do, or merely as a slower duplicate of the general programmes that already exist.

Old age: the pension strand the Code inherits and the programmes beside it

Old age protection exposes the framework habit most clearly, because India already runs, side by side, every technique the Code might have chosen. For the organised employee there is the contributory pension carved out of provident fund, an entitlement with an account behind it. For the unorganised worker there is a family of programmes: the assisted pensions of the National Social Assistance Programme for the elderly poor, and the Pradhan Mantri Shram Yogi Maan-dhan of 2019, a voluntary contributory pension promising three thousand rupees a month from age sixty to unorganised workers who enrol young and pay a small monthly contribution matched by the government. These programmes are real and their enrolments are counted in crores, yet none of them is an entitlement conferred by this Code, and the Code does not absorb them. Section 109 lists old age protection among the matters on which schemes for unorganised workers may be framed, and Section 114 repeats the item for gig and platform workers. What the repetition means in practice is that the pension future of a platform worker, on the law as it stands, runs through programmes that predate the Code, enrolment in a matching-contribution scheme designed for the unorganised generally, rather than through the fund to which the worker’s own aggregator will one day contribute. A framed scheme under Section 114 could change that by designing an old age benefit on the fund; until it does, the Code’s old age provision for its newest constituency is a pointer to other people’s programmes, and the reader should register the fact without the embellishment of calling it either betrayal or prudence. It is the framework technique applied to the risk that matters most and arrives last, and its adequacy will be visible, in enrolment and payout figures, long before it is visible in the statute book.

What the Code does not attempt

A map of the Code should mark its outer boundary as carefully as its contents, because some of the largest risks of Indian working life remain outside it by design. The Code creates no general unemployment insurance. The insurance chapter inherits the limited unemployment allowances that the old state insurance dispensation had attached to insurable employment, relief for the insured worker who loses a covered job, but a worker who loses platform work, or a self-employed worker whose trade collapses, has no income-replacement claim anywhere in the text; the schemes for unorganised and gig workers list protection against death, disability, sickness and old age, not against the disappearance of the work itself. The Code guarantees no minimum income and sets no benefit floor for any Chapter IX scheme, so the adequacy of whatever is framed is a pure question of executive choice. It does not touch the contributory pension architecture for government employees, which lives in its own rules, nor the tax-financed social pensions beyond listing old age among scheme matters. And it leaves the largest single gap in Indian social protection, the cost of outpatient and chronic care for the uninsured, exactly where it found it, addressed by programmes outside the Code rather than by entitlement within it. These absences are not oversights to be laid at the drafters’ door individually; several reflect choices every Indian government has made about fiscal reach. But they discipline the claims that can honestly be made for the consolidation. The Code unifies the law of the protected, recognises the unprotected majority, prices their recognition modestly, and stops there. A reader who knows where it stops will not mistake its silence on unemployment or adequacy for an answer, and will listen to future announcements in this field with the right question ready: which door of the building does this open, the door of entitlement or the door of recognition, and what, precisely, is behind it.

The Code is the third of the four labour codes in the usual ordering, and its meaning is partly relational. The Code on Wages supplies the definition of wages on which its contributions are computed and the floor wage beneath which state minimum wages cannot fall, which indirectly disciplines the wage base of the contributory chapters. The Industrial Relations Code governs the employment relationship whose presence or absence sorts workers between the contributory chapters and Chapter IX, since every classification decision taken under that code’s definitions, or by a court reading them, moves workers across this Code’s central line. The Occupational Safety, Health and Working Conditions Code carries the establishment-facing regulation of the workplaces in which the insured and compensated injuries of Chapter VII occur. The four were conceived as one regrouping, and their staggered commencement into force on a single date in November 2025 was the executive’s acknowledgement that the regrouping only works as a set: a wages definition in force in one code and dormant in another would have produced contribution chaos. For the reader of this article, the relational point that matters is the sorting line. Indian social security after the codes is a two-door building. Behind one door, employment, stand funded entitlements of long standing, restated and widened at the edges. Behind the other, everything else, stand recognition, a register, a fund and a formula awaiting activation. The Social Security Code is the only one of the four that contains both doors, which is why it, and not its three companions, is the statute on which the Indian welfare state’s next decade will be argued.

The single commencement date also conceals a difference in what commencement meant for each code. For the wages and industrial relations codes, force changed the rules of ongoing relationships from the day they commenced. For this Code, force changed the address of the organised sector’s benefits and activated a framework whose most publicised constituency still waits for schemes, so the same gazette notification was, in one statute, an operative event and, in this one, largely a promissory one. That asymmetry explains why the social security code attracts a scrutiny its companions escape. Nobody asks whether the Code on Wages has begun to exist; its existence is visible in every payslip computed under it. The Social Security Code’s existence, for the worker it was named for in every headline, remains a matter of notifications rather than payslips, and until that changes it will be read, fairly, as two statutes bound in one cover: a finished consolidation and an unfinished promise.

Recognition without entitlement: the honest assessment

The case for the Code, put at its strongest by its defenders in government, begins with the proposition that first recognition is not a small thing. Before September 2020, no Indian statute defined the gig worker; a workforce counted in tens of lakhs had no legal existence as a workforce, and every conversation about its welfare began from analogy to statuses designed for factories. The Code gave the workforce names, a register, a board at which its representatives sit by statute, a fund fenced for its benefit and a contribution formula with a floor, a ceiling and a payer identified by name. The floor matters to the argument: unlike the 2008 Act, which left financing entirely to budgetary grace, Section 114 writes a minimum rate of one per cent of turnover into primary legislation, so that a future scheme cannot lawfully set the aggregator’s share at a token figure, and a future government cannot quietly zero the obligation without returning to Parliament. The platform industry, for its part, substantially welcomed the settlement at passage; a spokesperson for Uber described the company as pleased that the Code had made India one of the first countries to recognise the need for social security for independent workers, a welcome that reflects the Code’s other service to the industry, the classification question deferred. Consolidation, on this reading, was the overdue half of the reform and recognition the historic half, and the schemes are a matter of sequencing in a field where the government also had to build a national register from nothing.

The case against, put at its strongest by the Code’s critics, accepts the recognition and disputes its cash value. The entitlement chain, the critics observe, has six links and the worker controls only one of them; the rest wait on notifications whose absence carries no consequence for the notifier. The architecture is the 2008 Act’s architecture, and the 2008 Act is the standing proof that frameworks of exactly this kind, definition, board, register, discretionary schemes, can coexist indefinitely with negligible delivery. The one element said to distinguish the Code, the contribution, is itself doubly conditional: the rate within the band is unnotified, and the date from which any rate runs is unnotified, so the provision that was supposed to answer the question of who pays has so far answered only the question of who would pay, eventually, and within limits. The ceiling compounds the point for the critics: by capping the contribution at five per cent of what workers are paid, the Code ensures that the fund’s income falls as platforms squeeze payouts, and the turnover base invites the accounting contest over what a marketplace’s turnover is. Add the registration figures, just over five lakh platform workers enrolled against an official estimate of seventy-seven lakh in 2020 to 2021, and the critics’ summary writes itself: the Code has recognised a workforce it has not reached, through a register it has not filled, for benefits it has not specified, financed by a contribution it has not commenced. Trade unions that organise platform workers make the sharper version of the argument, that recognition without classification is a subsidy to the platform model, converting what should have been employer liability into a modest industry levy pooled with public money.

Between these cases, the evidence available supports a verdict narrower than either. The consolidation half of the Code is done and done competently; the organised sector’s law is genuinely simpler for the move into one text, and the savings are holding. The recognition half is done as a matter of text and not yet as a matter of administration, and the distinction is not a quibble, because in social security the administration is the right. The verdict this article leaves with the reader is therefore the claim it set out to test: recognition without entitlement, until the schemes exist. The sentence is neither an epitaph nor an acquittal. The Code contains every legal instrument needed to falsify it, a fund, a formula, a board and a register, and the states have shown that instruments of this family can be made to pay out within a single legislative term when a government chooses the transaction fee and the register over further consultation. What would change the verdict is specific and checkable: notification of the rate and the commencement date under Section 114, a framed scheme with named benefits and eligibility a worker can test against a payslip, contribution figures in the fund’s accounts, and registration numbers that move toward the workforce estimates rather than an order of magnitude below them. Until some of those events occur, the honest description of the Social Security Code for the worker it most famously names is of a statute that can see the rider clearly, and has not yet paid the rider anything.

For the three readers this article most expects, the residue is different. The platform worker should take the register seriously before the schemes arrive, because every design in this field, central and state alike, pays the registered and overlooks the unregistered, and enrolment costs an afternoon at most. The platform operator should take the contribution seriously before the rate is notified, because the difference between the band and the ceiling across a large payout base is a material number, and the state fees already accumulating will not wait for the central reconciliation. The student of Indian legislation should take the Code as the clearest available specimen of a legislative technique that dominates social law: the framework statute that enacts a constituency, a fund and a formula, and reserves every operative decision to the executive. The technique is neither fraud nor reform in itself. It is a transfer of the legislative moment, and the measure of each statute built this way is the interval between its enactment and its first notification. For the Unorganised Workers’ Social Security Act, the interval consumed the statute. The Social Security Code has better instruments and a fuller register than its predecessor had at the same age. It also has the same clock running, and the closing observation belongs to the arithmetic of that clock: assent in September 2020, a first provision in force in May 2021, general commencement in November 2025, central rules in May 2026, and, on the gig worker provisions that made the Code famous, the operative notifications still to come. Readers working through the Code’s text alongside the rules as they issue will find it easier to keep the chapter map, the repeal list and their own tables in view if they open the statute and the scheme documents together in the ReportMedic office file viewer, since the entitlement chain this article has traced is assembled from documents that no single one of them contains.

What would settle the argument: the revision paths

The debate about the Code will not be settled by commentary, and it is worth closing the analysis by identifying the events and instruments that would actually settle it, because each corresponds to a weakness the preceding sections have isolated. The first instrument is already in the Code: notification of the rate and the commencement date under Section 114, with a framed scheme behind them, would convert the contribution from design to revenue within a single gazette issue, and the first year’s fund accounts would then discipline every claim made in this article, including the critics’. The second is a reconciliation, by rule or amendment, between the central contribution and the state welfare fees, crediting one against the other or assigning each a distinct benefit, so that the federal experiment compounds instead of colliding. The third is the classification litigation working its way through the courts, whose outcome, whichever way it falls, will fix the legal meaning of the Code’s central silence: a finding of employment for any class of platform workers would move that class into the contributory chapters and expose Chapter IX as a waiting room, while a rejection would confirm Chapter IX as the destination and raise the stakes on its schemes accordingly. The fourth is legislative and the least discussed: Parliament could, if the notifications continue to lag, write a deadline into Section 114, a device Indian statutes have used elsewhere to discipline commencement and scheme-making, converting the executive’s open timetable into a statutory one. None of these events requires a new code, a new commission or a new consensus. Each is a single act by an institution that already holds the pen. That is the final measure of the Social Security Code as a piece of legislation: it is complete enough that everything missing is a decision, and the decisions are all, still, available to be taken.

Frequently Asked Questions

Section 164 of the Code repealed nine statutes: the Employees’ Compensation Act of 1923, the Employees’ State Insurance Act of 1948, the Employees’ Provident Funds and Miscellaneous Provisions Act of 1952, the Employment Exchanges (Compulsory Notification of Vacancies) Act of 1959, the Maternity Benefit Act of 1961, the Payment of Gratuity Act of 1972, the Cine-Workers Welfare Fund Act of 1981, the Building and Other Construction Workers’ Welfare Cess Act of 1996 and the Unorganised Workers’ Social Security Act of 2008. Their substance was restated in the Code’s chapters, and the repeal carried savings so that accounts, registrations and schemes built under the old Acts continue under the new text.

Section 2, clause 35 of the Code defines a gig worker as a person who performs work or participates in a work arrangement and earns from such activities outside of a traditional employer-employee relationship. The definition is built on the absence of employment rather than on any feature of the work itself, which is why it reaches far beyond application based work: a freelancer, a task worker or a home-based worker engaged outside an employment relationship can fall within it. Where the work is also accessed through an online platform, the worker is also a platform worker under clause 61, and where the worker labours in the unorganised sector, the worker is also an unorganised worker under clause 86.

Yes, in the sense that the Code recognises them, defines them and builds machinery for their welfare, and no, in the sense that recognition is not yet a claimable benefit. Chapter IX, sections 109 to 114, applies to gig and platform workers: they may register under Section 113, the National Social Security Board is adapted to their welfare under Section 114, aggregators owe a contribution toward their schemes, and the Central Government may frame schemes for them on life and disability cover, accident insurance, health and maternity benefits, old age protection and creche facilities. Unlike an employee’s provident fund or state insurance, which operate by force of the chapters themselves, the gig worker’s benefits operate only through a framed scheme, so coverage in the text runs ahead of entitlement in the hand.

Section 2, clause 2 defines an aggregator as a digital intermediary or a marketplace for a buyer or user of a service to connect with the seller or the service provider. The definition is functional: it looks to what the platform does, standing between the customer and the person who performs the service, rather than to what the platform calls itself in its contracts. The Seventh Schedule illustrates the definition with nine categories of business, from ride sharing and food delivery to e-marketplaces, healthcare and professional services platforms. The aggregator is the Code’s chosen point of obligation for platform work: because the aggregator is not treated as the workers’ employer, its duty under the Code is not the employer’s contribution on wages but the welfare contribution fixed by Section 114.

Q: Which businesses are listed as aggregators in the Seventh Schedule?

The Seventh Schedule lists nine categories: ride sharing services; food and grocery delivery services; logistics services; e-marketplace platforms for goods and services, in both marketplace and inventory models; professional services providers; healthcare services; travel and hospitality services; content and media services; and any other goods and service provider platform. The closing category is the important one for the schedule’s reach, since it converts a list into a principle and allows the framework to extend to platform models that did not exist when the Code was drafted. A business may therefore be an aggregator for the purposes of the Code without resembling a delivery or ride platform, if in substance it digitally intermediates between those who need a service and those who provide it.

Q: What must an aggregator contribute?

Under Section 114, a scheme framed for gig and platform workers may require the aggregator to contribute at a rate notified by the Central Government, being not less than one per cent and not more than two per cent of the aggregator’s annual turnover, subject to a ceiling of five per cent of the amount paid or payable by the aggregator to its gig and platform workers. Annual turnover for this purpose excludes taxes, levies and cess paid to the Central Government. The contribution is credited to the Social Security Fund constituted under Section 141. Two qualifications define the obligation in practice: the rate within the band must be notified before anything is payable, and the date from which the contribution runs must itself be notified under Section 114, so the duty, though fully designed in the statute, awaits executive activation.

Section 141 of the Code constitutes a Social Security Fund, the dedicated destination for the money the Code raises for its welfare purposes, including the aggregator contribution under Section 114. The fund’s purpose in the design is fencing: contributions collected for the welfare of unorganised, gig and platform workers are to be held and applied for those workers, rather than absorbed into general revenues. Benefits under schemes framed for those workers are paid from the fund, alongside whatever budgetary support and other receipts the schemes assign to it. The fund is therefore the point at which the Code’s promises become accounting: once contributions begin, its receipts and disbursements will show, year by year, what the recognition of the platform workforce was worth in money.

Q: What happens if an aggregator does not pay the contribution required by the Code?

While no rate and commencement date have been notified, no aggregator can be in default, because the obligation has not been switched on. Once a scheme imposes the contribution and the date has run, non-payment is handled through the Code’s ordinary machinery: the authorities under Chapter XI assess the amount due and recover it by the coercive process the chapter provides, and the offences and penalties chapter, sections 133 to 138, prices continued default, with fines carrying the weight for first offences and compounding available for offences not punishable with imprisonment alone. The aggregator’s practical exposure is therefore not a single penalty but a sequence, assessment, recovery with the machinery of arrears, and prosecution reserved for the serious and the repeated. The design follows the Code’s general compliance philosophy of payment first and punishment last.

No. The Code deliberately covers platform workers without classifying them as employees. Its definitions proceed on the express footing that gig work and platform work are performed outside of a traditional employer-employee relationship, and Chapter IX builds a welfare route, registration, schemes, a fund and an aggregator contribution, that does not depend on employment status. A platform worker therefore does not, by reason of the Code, acquire an employee’s provident fund, state insurance or gratuity. Whether platform workers are in law employees, whatever the Code assumes, is a separate question that has been placed before the Supreme Court by the Indian Federation of App-based Transport Workers, which contends that platform control creates a de facto employment relationship. The Court’s answer, when it comes, will operate on the Code from outside.

Under the Code, the unorganised worker’s position is a framework position. Section 109 empowers the Central Government to frame schemes for unorganised workers on life and disability cover, health and maternity benefits, old age protection and any other benefit the government determines, with state governments able to frame schemes in their sphere, and Section 110 provides that such schemes may be funded by the Centre, shared with the states, supported by beneficiary or employer contributions, or financed from corporate social responsibility and other sources. Registration under Section 113 is the gateway to whatever is framed. In practice, the unorganised worker’s received benefits still come mainly from general programmes that stand beside the Code, such as assisted pensions and contributory pension enrolment for the unorganised, rather than from an entitlement the Code itself confers.

Q: What is the difference between an unorganised worker and a gig worker under the Code?

The two definitions answer different questions. An unorganised worker, under Section 2, clause 86, is a home-based worker, a self-employed worker or a wage worker in the unorganised sector, including a worker in the organised sector who is not covered by the Industrial Disputes Act of 1947 or by Chapters III to VII of this Code: the category is about the sector and the absence of other statutory cover. A gig worker, under clause 35, is defined by the work arrangement, work performed outside a traditional employer-employee relationship: the category is about the form of the engagement. The populations overlap heavily, and a platform rider is ordinarily both at once, but the distinction is not empty, because a self-employed artisan with no platform and no gigs is an unorganised worker and not a gig worker, while the schemes and funding lines in Chapter IX are drawn partly by category.

Q: What schemes can the Central Government frame for gig and platform workers?

Section 114, sub-section (1) authorises the Central Government to frame and notify schemes for gig and platform workers in relation to life and disability cover, accident insurance, health and maternity benefits, old age protection, creche facilities and any other benefit the government may determine. Sub-section (3) leaves the financing of such schemes flexible: they may be funded wholly or partly by the Central Government, partly by state governments, wholly or partly by aggregator contributions, partly by beneficiary contributions, or from corporate social responsibility funds and other sources. The width of the authorisation is the point to hold on to. The statute fixes the menu of subjects and the possible payers, but the benefit levels, eligibility conditions and delivery channels exist only once a scheme is framed, and a worker’s claim is against the scheme, not against the section.

Section 6 of the Code constitutes a National Social Security Board for unorganised workers, and Section 114, sub-section (6) provides that the same board is to function as the National Social Security Board for the welfare of gig and platform workers, with its composition altered for that business. Sitting on platform welfare, the board includes five representatives of aggregators and five representatives of gig and platform workers, the Director General of the Employees’ State Insurance Corporation, the Central Provident Fund Commissioner, experts, five representatives of state governments by rotation, and the Joint Secretary in the Ministry of Labour and Employment as Member Secretary. The board’s role is recommendatory and supervisory: it advises the Central Government on schemes, monitors their administration and reviews the registration and welfare machinery. It does not itself frame schemes, hold the fund or pay benefits.

Yes, registration is the gateway the Code builds between the worker and any benefit under Chapter IX. Section 113 provides for the registration of unorganised workers, gig workers and platform workers, on conditions that include completion of sixteen years of age and a self-declaration in the prescribed form, with Aadhaar based identification applied by Section 142. In operation, registration runs through the e-Shram portal, which issues the worker a Universal Account Number intended to make the record portable across states and engagements. The requirement is not a formality: schemes framed under sections 109 and 114 will pay the registered, and a worker who has not registered has no record against which a claim can be made. Registration is free and rests on self-declaration, so the barrier is not cost or proof but awareness, which the registration figures suggest remains the binding constraint.

Q: Does a fixed term employee get gratuity?

Yes. Gratuity under Chapter V of the Code is generally earned on five years of continuous service, but Section 53 removes the five-year condition where the employment ends on the expiry of a fixed term engagement, and directs that the fixed term employee be paid gratuity on a pro rata basis for the period of service actually rendered. Guidance from the Ministry of Labour and Employment expresses the position as eligibility after one year of service for the fixed term employee, reading the pro rata rule with the Code’s definition of continuous service. The change closes a gap that had widened as fixed term hiring spread: under the Payment of Gratuity Act of 1972, an engagement kept below five years could end without gratuity ever falling due, however permanent the work had been in substance.

The Code restates provident fund in Chapter III, sections 14 to 23, with the essential architecture of the 1952 Act preserved: applicability to establishments with twenty or more employees, the provident fund, pension and deposit-linked insurance schemes, and administration through the Employees’ Provident Fund Organisation, with existing accounts and entitlements carried across by the savings in Section 164. The changes are at the edges and in the base. Coverage extends by notification and by voluntary agreement beyond the threshold, the definition of wages on which contributions are computed is the uniform code definition with excluded allowances capped at half of total remuneration, which widens the contribution base in allowance-heavy pay structures, and the wage ceiling for mandatory membership continues to be set through the scheme rather than written into the chapter. Gig and platform workers, standing outside employment, are not brought into provident fund by the Code.

Q: What happened to provident fund and ESI accounts when the old Acts were repealed?

They continued. The repeal of the Employees’ Provident Funds and Miscellaneous Provisions Act of 1952 and the Employees’ State Insurance Act of 1948 by Section 164 was accompanied by savings under which things done under the repealed Acts, registrations, contributions, funds and schemes in operation, are treated as done under the corresponding provisions of the Code. A member’s provident fund accumulation, pension entitlement and deposit-linked insurance cover therefore survived the repeal undisturbed, as did an insured person’s state insurance record and benefit history, and the Employees’ Provident Fund Organisation and the Employees’ State Insurance Corporation continued as administrators under their chapters of the Code. The transition was designed to be invisible to the account holder, and no fresh enrolment or transfer was required of any existing member by reason of the repeal alone.

Q: What happened to employment exchanges?

The Code retired them in their old form. The Employment Exchanges (Compulsory Notification of Vacancies) Act of 1959, which had required employers to notify vacancies to employment exchanges, is among the nine statutes repealed, and Chapter XIII of the Code, sections 139 and 140 on employment information and monitoring, replaces the exchanges with career centres. The centres’ statutory functions are the collection and provision of information on employment and vacancies, vocational guidance and related services, reflecting the reality that the exchange register had long ceased to be the place where hiring was transacted. The change is one of honest re-description as much as machinery: the state keeps an employment information service, and drops the pretence, maintained since 1959, that notified vacancies were the labour market’s main channel.

In stages. The Code was enacted as Act 36 of 2020 and received assent on 28 September 2020, but its provisions were commenced separately under Section 1. Section 142, applying Aadhaar based identification to registration and benefits, was brought into force on 3 May 2021, and the registration portal built on it followed that year. The four labour codes, this Code among them, were brought into force generally on 21 November 2025, and the Central Rules under the codes were published on 8 May 2026, with state rules remaining incomplete and the rules under the repealed statutes continuing to operate where consistent. The staging matters for reading the Code: enactment in 2020, first provision in force in 2021, general force in 2025, and the scheme notifications under Chapter IX still to complete the structure the commencement began.

Three differences stand out. First, the revenue base: the state Acts finance welfare through a welfare fee on each transaction, in Karnataka fixed as a band of one to five per cent of the payout on the transaction, while the central Code fixes the aggregator contribution as one to two per cent of annual turnover capped at five per cent of payments to workers. Second, the obligations beyond money: Karnataka’s Act adds duties the Code does not contain, including notice before termination or deactivation of a worker and transparency about automated decision systems. Third, immediacy: the state Acts, Rajasthan’s of 2023 being the first, constitute their own boards and registers and are built to pay benefits, while the central entitlement still awaits a framed scheme. A national platform may in time owe both the state fee and the central contribution, a coexistence neither text yet reconciles.