An amendment trail is judged as much by what it leaves standing as by what it moves, and the Maternity Benefit Act, 1961 supplies the cleanest illustration in Indian labour law. In 2017 Parliament more than doubled the paid maternity leave that the Act guarantees, extended the benefit to two classes of mother the original text had never contemplated, imposed a crèche duty on larger establishments, and added a work from home provision of a kind no Indian labour statute had carried before. By any measure of entitlement, the amendment was a large reform. By the measure that decides how a statute behaves in the world, it changed almost nothing, because the provision that governs the entire economics of the Act, the rule that the employer of the woman pays for her leave out of the establishment’s own funds, was carried through the amendment word for word, unexamined and unaltered. The leave doubled and the payer did not change. Everything that has been reported about the amendment since, the gratitude of the women it covered, the anxiety of the small employers it charged, the staffing studies predicting that firms would hire fewer women, and the persistence of non-compliance in small establishments, follows from that single untouched feature of the design.

This article follows the trail in order. It sets out the Act as Parliament enacted it in 1961, the entitlement the original text promised and the machinery that delivered it, the small amendments of the intervening decades, and then the amendment of 2017 provision by provision, including its staged commencement and the conditions buried in its provisos that most summaries omit. It then turns to the two questions the trail actually raises. The first is why a reform willing to double the leave was unwilling to move its cost off the individual employer, and what evidence exists about the consequences of that choice. The second is where the entitlement sits after the labour codes commenced, since the Act of 1961 has been repealed and its content carried into the Code on Social Security, 2020, a transition a reader who relies on the old text will miss entirely. The reader who finishes should know exactly what maternity entitlement Indian law provides, where that entitlement is located in the statute book, who is obliged to fund it, and why the identity of the payer, rather than the length of the leave, is the feature of the scheme that explains its record.

The Maternity Benefit Act of 1961 as enacted

The Act did not begin the subject. Maternity protection in India started in the provinces, with the Bombay Maternity Benefit Act of 1929 as the earliest of the provincial statutes, and the central legislature added the Mines Maternity Benefit Act, 1941 for women working underground and on the surface of mines. Several provinces enacted maternity benefit statutes on the Bombay pattern in the decades that followed, and the central statute for mines added a second, industry specific layer, so that by the time the Constitution came into force a working woman’s maternity rights were a function of geography and industry in combination. The result by the middle of the century was a patchwork: a woman’s entitlement to paid absence around childbirth depended on the province she worked in, the industry she worked in, and whether a statute covered that industry at all. Article 42 of the Constitution, among the directive principles of state policy, directed the state to make provision for just and humane conditions of work and for maternity relief, and the Act of 1961 was Parliament’s general answer to that direction, a single central statute that consolidated the field. On its application to mines, the Mines Maternity Benefit Act, 1941 stood repealed, and the Bombay Act of 1929 stood repealed for the factories in Delhi to which the central Act was applied, so the new statute was designed from the outset to displace the patchwork rather than sit on top of it.

The Maternity Benefit Act, 1961 is Act 53 of 1961. It received the assent of the President on 12 December 1961 and was published in the Gazette of India the following day. Its commencement was staged rather than general: the Act was to come into force on dates notified for different classes of establishment, by the Central Government for mines and performance establishments and by state governments for the rest, and in relation to mines it commenced on 1 November 1963. That structure, an Act whose operative life began establishment class by establishment class and state by state, is worth remembering, because the statute’s coverage has always been an administered fact rather than a self-executing one, and the same habit of staged commencement reappears at the end of the trail when the labour codes replace it.

Section 2 fixes the field. The Act applies in the first instance to every establishment that is a factory, a mine or a plantation, including establishments belonging to government, and to every establishment in which persons are employed for the exhibition of equestrian, acrobatic and other performances, the circus establishments that the Act’s rules have always named. It applies further to every shop or establishment within the meaning of the law in force in a state in relation to shops and establishments, in which ten or more persons are employed or were employed on any day of the twelve preceding months. Beyond that core, the appropriate government may, with the approval of the Central Government and after giving not less than two months’ notice of its intention, extend the Act’s provisions to any other establishment or class of establishments, industrial, commercial, agricultural or otherwise. The ten person threshold for shops and establishments is the boundary that matters most in practice: below it, the woman in a small office, a small shop or a small service firm holds no statutory maternity entitlement against her employer under this Act at all, however long her service, and no amendment in the trail, including the amendment of 2017, ever lowered that gate.

A second boundary in Section 2 is quieter and equally consequential. The Act stands aside where the Employees’ State Insurance Act, 1948 governs the establishment, subject to two bridging provisions, Sections 5A and 5B, which preserve a woman’s position when the insurance scheme’s coverage begins or ceases to reach her. A woman whose wages fall within the state insurance ceiling and whose establishment is covered by that scheme draws her maternity benefit from the insurance fund, financed by contributions, and not from her employer under this Act. The bridging provisions deserve their detail, because they govern the cases where the boundary moves under a woman’s feet. Section 5A addresses the woman whose establishment is inside the insurance scheme: her benefit is the insurance benefit, and the provision preserves the continuity of her entitlement within that scheme. Section 5B addresses the reverse movement: where the insurance scheme ceases to apply to a woman, because her wages cross the scheme’s ceiling or her establishment’s coverage changes, she does not lose maternity protection with the loss of insured status, and the provision entitles her to claim under the Maternity Benefit Act on the conditions it states, so the Act operates as a residual guarantee behind the insurance scheme rather than a stranger to it. The architecture is more thoughtful than the boundary’s critics usually allow; its defect is not incoherence but reach, since both tracks presuppose an establishment of a size and formality that the majority of working women never encounter. The 1961 settlement therefore contained two financing models from the beginning: social insurance where the insurance statute reached, and direct employer liability everywhere else. The amendment trail that follows is, in large part, the story of Parliament repeatedly enlarging the second model while leaving the boundary between the two exactly where 1961 had drawn it.

The entitlement as enacted: what the original text promised

The core of the original Act is Section 5. Every woman employed in a covered establishment was entitled to maternity benefit at the rate of her average daily wage for the period of her actual absence, the period immediately preceding the day of delivery, the day of delivery itself, and the period immediately following it. The maximum period for which benefit could be drawn was fixed by Section 5(3) at twelve weeks, of which not more than six weeks could precede the expected date of delivery. The entitlement was conditional. Under Section 5(2), no woman was entitled unless she had actually worked in an establishment of the employer from whom she claimed benefit for a period of not less than 80 days in the twelve months immediately preceding the date of her expected delivery, and the explanation to the provision counts days of layoff and holidays with wages toward the 80, so the test is attachment to the establishment rather than unbroken physical attendance. The benefit rate, the average daily wage, is computed from the wages payable to the woman for the days she worked in the three calendar months preceding her absence, or the statutory minimum wage, or a small floor figure, whichever is highest, a computation the rules administer and this article need not reproduce in full.

Around that core the Act built a set of protections, and their durability is part of the trail, because the amendment of 2017 left every one of them in place. Section 4 prohibits the employment of a woman during the six weeks immediately following the day of her delivery, miscarriage or medical termination of pregnancy, and prohibits an employer from knowingly employing her in that period; it also entitles a pregnant woman, in the period running from one month before the six weeks preceding her expected delivery, not to be required to do work that is arduous, involves long hours of standing, or is likely to interfere with her pregnancy or the normal development of the foetus or cause miscarriage. Section 6 governs notice and payment: the woman gives notice that she will absent herself and nominates the person to whom benefit should be paid if she cannot receive it; the amount for the period before delivery is paid in advance on production of proof of pregnancy, and the amount for the period after delivery within 48 hours of production of proof that she has been delivered of a child; and the failure to give notice does not disentitle a woman who is otherwise entitled, since an Inspector may direct payment. Section 7 carries the benefit through death: if a woman dies, benefit is payable for the days up to and including the day of death, and if she dies after delivery leaving the child, for the whole period, with payment to her nominee or legal representative.

The ancillary entitlements complete the original scheme. Section 8 provides a medical bonus where the employer does not provide pre-natal confinement and post-natal care free of charge. Section 9 gives six weeks of leave with benefit immediately after a miscarriage or medical termination of pregnancy, on production of proof. Section 9A, added later in the trail, gives two weeks of leave with benefit immediately after a tubectomy operation. Section 10 gives a woman suffering illness arising out of pregnancy, delivery, premature birth of a child, miscarriage, medical termination or tubectomy an additional period of leave with wages at the maternity benefit rate, up to a maximum of one month, over and above the other periods of absence the Act allows. Section 11 requires two nursing breaks of the prescribed duration in the course of her daily work, in addition to her interval for rest, until the child attains the age of fifteen months. Section 19 requires the abstract of the Act and the rules to be exhibited in a conspicuous place in every part of the establishment where women are employed. The general position described in this article is legal information about a statute whose trail has since moved into a successor code; an individual claim turns on the coverage of the establishment, the claimant’s service record and the instruments in force for her workplace, and is best pursued with advice on its own facts.

Section 12 is the protection on which the whole entitlement depends, because paid leave that an employer can answer with dismissal is not an entitlement at all. When a woman absents herself in accordance with the Act, it is unlawful for her employer to discharge or dismiss her during or on account of that absence, or to vary the conditions of her service to her disadvantage, and a notice of discharge or dismissal given during the absence that would expire while she is away does not take effect against the benefit. The single qualification is gross misconduct: an employer may, by order in writing for gross misconduct as prescribed, deprive a woman of maternity benefit or medical bonus or both, and the woman aggrieved by such an order may appeal to the prescribed authority within sixty days. Section 17 arms the Inspectorate: an Inspector may direct an employer to pay benefit wrongfully withheld, and either side may appeal. Section 21 attaches penalties of imprisonment and fine to contraventions, and Section 23 confines prosecutions to complaints made by or with the sanction of an Inspector, instituted within a year of the offence. Section 27 gives the Act overriding effect against anything inconsistent in an award, agreement or contract of service, while expressly preserving any benefit more favourable to the woman. The design is complete and coherent: a wage replacement entitlement, a dismissal shield around it, an inspectorate to enforce it, and the entire cost placed on the employer of the individual woman. That last element sat inside the design from the first day, largely unremarked, because at twelve weeks the cost was containable. The trail’s later chapters are unintelligible without seeing that it was there from the start.

The wage base and the machinery: what the benefit is worth and how it is delivered

An entitlement to paid absence is worth exactly what its wage definition lets it be worth, and the Act’s definition of wages in Section 3 is drawn to protect the payment against erosion by salary structure. Wages means all remuneration paid or payable in cash to a woman by her employer for her work, and the definition expressly includes cash allowances, including dearness allowance and house rent allowance, and incentive bonus; it excludes bonus other than incentive bonus, overtime earnings, the employer’s contribution to pension or provident funds and sums of that character, so the benefit base is the woman’s recurring cash wage rather than her cost to company and rather than her bare basic pay. The benefit rate is then fixed by Section 5 as the average daily wage for the period of actual absence, computed, in the ordinary case, by dividing the wages payable for the days on which she worked in the three calendar months immediately preceding the date from which she absents herself by the number of days she worked in that period, with statutory floors beneath the computation so that a woman on very low or irregular wages is not left with a nominal benefit. The design matters because it ties the benefit to earnings actually received in the quarter before absence, which for most women in regular employment is a fair measure, and because it forecloses the simplest method of shrinking the entitlement, the reclassification of pay into heads outside the definition.

The calendar of the benefit was settled by litigation of a kind that reveals how establishments first read the Act. Employers argued in the early years that a week of maternity benefit meant a week of working days, so that Sundays and holidays falling inside the period carried no payment. The Supreme Court rejected the argument in B. Shah v. Presiding Officer, Labour Court, decided in 1977, holding that the period of maternity benefit is a continuous period of calendar weeks and that the benefit is payable for all the days of the period, Sundays and holidays included, because the statute protects the woman’s absence and not her attendance. The proposition sounds elementary, and it is, but it fixed the unit of account for everything the amendment of 2017 later did: when Parliament doubled the leave from twelve weeks to 26 weeks, it doubled a benefit measured in unbroken calendar time, so the real value of the enhancement was the full fourteen additional weeks of wages and not a lesser number of paid working days.

The machinery that delivers the benefit is deliberately light, and its lightness is part of the enforcement record. The Act creates no fund, no board and no claims tribunal. It appoints Inspectors, charges them with ensuring that abstract summaries of the Act are exhibited where women work, requires employers to maintain registers and muster rolls of the women employed and to submit returns under the rules, and gives the Inspector the direction power in Section 17 when a benefit is withheld. Penalties under Section 21 punish contravention, and prosecutions under Section 23 need the Inspector’s sanction and a one year window. For a large establishment with a human resources department, this design is adequate: the claim is a payroll event, the Inspector’s visit is a formality, and the dismissal shield in Section 12 is reinforced by the woman’s own knowledge of it, which the intimation duty of 2017 later made systematic. For a small establishment, the same design depends on a chain of contingencies: the woman must know of the entitlement, must claim it in writing against an employer who may be her only source of livelihood in the locality, and must set the Inspectorate in motion if refused, in a system where inspectorates are thinly staffed across every labour statute at once. The trail’s recurring findings, women paid for fewer weeks than the text promises, women asked to resign and rehired after delivery, engagements structured as consultancy or rolling contracts to argue the establishment or the employment relation away, are failures of this machinery rather than of the entitlement text, and they explain why the amendment’s intimation duty, which costs an employer nothing but a letter, may have done more for actual receipt of the benefit in small establishments than the doubling of the period did. The entitlement and its delivery are different subjects, and the amendment of 2017 addressed the first comprehensively and the second only at this single point.

The long interval: small amendments and a static core

Between commencement and the amendment of 2017 the Act was amended several times, and the pattern of those amendments is itself evidence: Parliament was willing to adjust the statute’s edges for five decades while its core, twelve weeks of employer paid leave behind an 80 day qualifying condition, went untouched. The territorial settlement was completed early, when the words excepting the State of Jammu and Kashmir from the Act’s extent were omitted by an amendment of 1970, making the statute’s geography uniform. An amendment of 1973 substituted the commencement machinery of Section 1 and recast the application provision, tidying the staged structure through which the Act had been brought into force. The tubectomy entitlement in Section 9A responded to the family planning programme’s place in public policy, recognising that a sterilisation operation undertaken in the national interest should not cost a woman her wages, and it widened the Act’s idea of maternity related absence beyond childbirth and its failures for the first time.

What changed in the Act between 1963 and 2016?

Very little at the centre. The amendments of the interval completed the Act’s territorial extent, recast its commencement and application provisions, added two weeks of leave for a tubectomy operation, and raised the medical bonus in stages, while the core entitlement of twelve weeks, the 80 day condition and the employer pays rule survived the half century unaltered.

The medical bonus shows the interval’s characteristic scale of reform. Section 8 as the Act operated for most of its life fixed a small sum, and the amendment of 2008 substituted the section so that the bonus stood at one thousand rupees, with power in the Central Government to increase the amount by notification up to a ceiling of twenty thousand rupees. By notification of August 2008 the amount was raised to two thousand five hundred rupees, and by a further notification of December 2011 to three thousand five hundred rupees, where it rested. A legislature minded to adjust a three figure sum by delegated notification across three years was not a legislature rethinking the scheme; it was administering it. Meanwhile the world the statute governed was changing around the static text. Women’s employment was shifting from factories and plantations toward offices, retail, education, health care and information services, most of it in establishments close to or below the ten person threshold, and the twelve week entitlement, designed against the industrial workforce of 1961, was increasingly a formal sector privilege enjoyed by a minority of working women. The pressure for a larger reform built from two directions: a health and child development argument that twelve weeks fell short of what infant care required, and an equality argument that India’s leave was short by the standards India compared itself against. Both arguments were aimed at the length of the leave. Neither was aimed, with any force, at the financing rule, and that asymmetry in the pressure explains the shape of what eventually passed.

The road to 2017: how twenty-four weeks became twenty-six

The policy anxiety underneath the reform was the recorded direction of women’s work. Across the decade before the amendment, the official survey record and the international series built on it showed the female share of the workforce in India falling, from a little over a third of the workforce in the middle of the first decade of the century to a little over a quarter in the years before the Bill, a decline that puzzled policy because it ran against rising female education and rising household incomes and that every government document on the subject treated as a problem to be reversed. The maternity amendment was framed inside that anxiety and should be read there: its object, as its defenders stated it, was retention, to stop the exit of women from formal employment at childbirth by making the exit temporary, protected and paid, and to convert the point in a working life at which women most often left the workforce into a point at which the law held their place in it. That framing explains features of the design that a health framing alone does not: the emphasis on the woman’s return, the nursing breaks and the crèche that serve the months after return, and the intimation duty that tells her at appointment that the statute expects her back. It also states the standard by which the amendment asked to be judged, retention of women in formal work, and the TeamLease evidence mattered politically precisely because it claimed the opposite movement on the hiring side of the same ledger. A trail that records only the weeks misses the wager: Parliament bet that doubling the protected absence would raise the number of women who remained in formal employment across a career by more than the hiring caution it provoked would reduce the number who entered, and the evidence available in the years after the amendment, survey based and partial in the ways already described, does not settle whether the bet was won.

The institutional origin of the 2017 amendment is unusually well documented, because the government recited it in the Statement of Objects and Reasons of the Bill. The Indian Labour Conference, the apex tripartite body of government, employers and unions, recommended at its 44th session that maternity leave under the Act be enhanced from twelve weeks to twenty-four weeks, and the recommendation was reiterated at the 45th and 46th sessions. The Ministry of Women and Child Development and other stakeholders pressed for enhancement as well, and the proposal went through tripartite consultation before the Bill was settled. The number in the Bill as introduced went beyond the conference’s recommendation: twenty-six weeks rather than twenty-four, a figure that aligned the entitlement with the six months of exclusive breastfeeding that public health guidance prescribes for an infant’s first half year, and placed India near the top of the international table for the duration of paid maternity leave. The choice of twenty-six over twenty-four matters for reading the amendment’s politics: this was not a grudging concession extracted from a reluctant executive, but a government bidding above the tripartite recommendation, on a measure with no organised opposition, where the credit for generosity was available precisely because its cost fell on employers rather than on the budget.

The pressure behind the Bill was not confined to the conference hall. The Ministry of Women and Child Development, then headed by Maneka Gandhi, pressed the enhancement publicly and repeatedly, and raised in public discussion the companion question of paternity leave, arguing that a settlement which loads the entire absence onto mothers entrenches the assumption that care is women’s work; the Labour Ministry’s answer, delivered through the Bill itself, took the leave and left the paternity question for another day, where it remains. Women’s organisations and the central trade unions supported the enhancement while recording, in the consultations that preceded the Bill, the reservation that runs through this whole trail: that a benefit charged to the individual employer in the informal margin of the formal sector would be resisted in practice, and that the state’s own contribution to the cost was the reform that would make the entitlement real in small establishments. The reservation was noted; it was not acted on. The Bill that emerged from tripartite consultation therefore represents a precise political equilibrium, in which every organised interest could support the headline, the unions and women’s groups because the entitlement grew, the government because the credit was its own and the budget untouched, and employers’ organisations because opposition to maternity leave was untenable and because the alternative models on the table internationally, an insurance fund financed by a payroll levy, would have charged all employers including those who employ few women. Employer liability, for a large employer with a balanced workforce, can be cheaper than a universal levy, and that arithmetic inside the employer camp is part of why the financing rule met no organised resistance in 2016. The trail’s most important decisions are sometimes the ones every interest found convenient, and this was one.

The Maternity Benefit (Amendment) Bill, 2016 was introduced in the Rajya Sabha by the Minister of Labour and Employment, Bandaru Dattatreya, and passed by that House on 11 August 2016, the day of its introduction. The Lok Sabha passed it on 9 March 2017. The President gave assent on 27 March 2017, and the Maternity Benefit (Amendment) Act, 2017, Act 6 of 2017, was published in the Gazette of India on 28 March 2017. The debates were brief and consensual by parliamentary standards; members across parties welcomed the enhancement, and the questions that were pressed concerned coverage, the position of women in the unorganised sector, and the absence of paternity leave, rather than the principle of the increase. Nothing in the passage record suggests that the financing design was seriously in contest. No amendment was moved to substitute an insurance or exchequer funded model for employer liability in the establishments outside the state insurance scheme, and the Bill’s financial memorandum could therefore present a reform of real generosity to a defined class of women at no cost to public funds. An amendment trail records what a legislature did; it should also record what the legislature was never asked to decide, and on this trail the road not taken, a social insurance maternity fund for the non insurance sector, was not so much rejected as never placed on the table in 2016 and 2017.

What the Maternity Benefit Amendment of 2017 changed, provision by provision

The amendment Act is a short instrument, and its method is surgical: a definition inserted, Section 5 rebuilt, and one new section added. Reading it closely repays the effort, because two of its most consequential features live in provisos that summaries routinely drop.

The first change is a definition. The amendment inserted clause (ba) into Section 3, providing that a commissioning mother means a biological mother who uses her egg to create an embryo implanted in any other woman. The definition is the gateway to a benefit the Act had never offered: the woman who becomes a mother through surrogacy, and who does not herself bear the child, was brought inside a statute organised around pregnancy and delivery. The entitlement it unlocks sits inside the wider statutory settlement on surrogacy, whose own restrictive statute is the subject of the surrogacy and assisted reproduction legislation, and the amendment should be read as the labour law limb of that settlement: whatever the conditions on which surrogacy itself is lawful, the commissioning mother’s employment protection was settled here.

The second and largest change rebuilt Section 5(3). The amendment substituted, for the words twelve weeks of which not more than six weeks, the words twenty-six weeks of which not more than eight weeks. The maximum period of maternity benefit for a woman with fewer than two surviving children thereby became 26 weeks, with the pre-delivery portion capped at eight weeks rather than six. Into the same sub-section the amendment inserted a new proviso, the provision most often lost in summary: the maximum period entitled to maternity benefit by a woman having two or more than two surviving children shall be twelve weeks, of which not more than six weeks shall precede the date of her expected delivery. The longer leave is therefore not the universal entitlement it is frequently described as being. It is an entitlement graded by family size, 26 weeks for the first two children and the old twelve weeks thereafter, and any statement of Indian maternity law that omits the condition misstates the law for precisely the women whose births occur in larger families.

The third change inserted Section 5(4). A woman who legally adopts a child below the age of three months, and a commissioning mother, each became entitled to maternity benefit for a period of twelve weeks from the date the child is handed over to her. The drafting repays attention. The benefit is not tied to a birth the woman undergoes but to the event of the child entering her care, and it runs from the handing over, so the leave is calibrated to the child’s arrival rather than to any medical event. Twelve weeks for an adoptive or commissioning mother is half the headline entitlement of a mother who delivers her first or second child, a grading that reflects the absence of physical recovery from childbirth while recognising the care work of the first months; whether that grading is right is contested, but its logic is visible on the face of the provision. The qualifying condition of 80 days of work in the preceding twelve months applies to these mothers as it does to others, since Section 5(4) confers benefit within the scheme of Section 5 as a whole.

The fourth change inserted Section 5(5), the work from home provision, and it is the provision most often misdescribed. After the expiry of the period of maternity benefit, where the nature of work assigned to a woman permits her to work from home, the employer may allow her to do so, on conditions mutually agreed between the employer and the woman. Every element of that sentence is permissive or conditional. The provision operates only after the benefit period ends; it operates only where the nature of the work permits; the employer may allow it, and is not obliged to; and the conditions are a matter of mutual agreement. Section 5(5) is an enabling provision that removes any doubt about the lawfulness of such an arrangement and gives the woman a statutory hook on which to open the conversation. It is not an entitlement to work from home, and writing that treats it as one hands the reader a right she does not hold.

The fifth change is the new Section 11A, and it carries two duties. Section 11A(1) requires every establishment having fifty or more employees to have the facility of a crèche within such distance as may be prescribed, either separately or along with common facilities, and it requires the employer to allow four visits a day to the crèche by the woman, which include the interval for rest allowed to her. The threshold is a headcount of employees, not of women employees, so a workforce of fifty men and one woman triggers the duty, and the common facilities route lets neighbouring establishments share a crèche rather than each building its own. Section 11A(2) imposes the intimation duty: every establishment shall intimate in writing and electronically to every woman, at the time of her initial appointment, regarding every benefit available under the Act. The intimation duty is the amendment’s quiet enforcement device. A benefit that a woman learns of only when she becomes pregnant arrives too late for planning and too late for her to know what a refusal means; the duty to state the benefits in writing and electronically at appointment converts an obscure statute into a term of the employment relationship from its first day.

The amendment’s drafting technique is itself part of the trail, because it shows how a legislature enlarges a statute without reopening it. The amending Act works entirely by insertion and substitution inside the architecture of 1961: a clause added to the definition section, sub-sections substituted and inserted inside Section 5 where the entitlement lives, and a new Section 11A placed immediately after the nursing breaks provision in Section 11, so that the childcare duties sit together in the text as they sit together in a working day. Nothing was renumbered, no chapter was disturbed, and the case law, the rules and the forms built on the old numbering survived the amendment without translation, which is a real administrative virtue in a statute administered by thousands of establishments and a thin inspectorate. The choice has a cost that trails of this kind should record. Placing the new mothers inside Section 5 as sub-section (4), beneath the delivery based entitlement and its provisos, made the adoptive and commissioning entitlements read as extensions of the birth entitlement rather than as a freestanding recognition of care, and it left the interactions between the sub-sections, the counting of adopted children toward the two surviving children among them, to inference. A legislature drafting the scheme afresh would have written a single entitlement to leave on the arrival of a child, graded by the mother’s recovery needs, and stated the counting rules once. The amendment of 2017, like most amendments that pass without opposition, preferred the graft to the rewrite, and the seams it left are the seams the clarifications and the courts have since been asked to stitch.

Two of the changes reward a closer reading than they usually receive. The raising of the pre-delivery cap from six weeks to eight is the amendment’s most clinically literate sentence. The original Act assumed a woman would work until six weeks before her expected delivery; the amended figure concedes that the final two months of pregnancy are, for a large class of working women, the period in which continued attendance is hardest and medically least advised, and it moves the boundary of the paid absence to meet that reality while keeping the total arithmetic anchored to the delivery date, so a woman who delivers early or late adjusts within the same 26 week envelope rather than gaining or losing by the accident of the date. Read with Section 4’s prohibition on arduous work in the closing period, the scheme after 2017 is coherent in a way the scheme of 1961 was not: the Act no longer merely forbids the worst work in late pregnancy while expecting attendance, it funds absence across the whole of the period it treats as protected. The intimation duty rewards attention for the opposite reason, because it is evidentiary rather than substantive. A woman who was told in writing and electronically at her appointment what the Act gives her holds, from her first day, the answer to the most common defence in maternity disputes, the suggestion that a benefit informally described or locally varied is the whole of her entitlement; and an establishment that has performed the duty has fixed the terms of the conversation in advance, which is why compliance professionals treated Section 11A(2), and not the crèche, as the amendment’s most immediately actionable sentence. Duties that cost nothing and prove everything are rare in this field, and the trail should mark that the cheapest provision of the 2017 amendment may be the one that moved the most claims.

What the amendment did not change is as definite a list as what it did. It did not change the 80 day qualifying condition. It did not lower the ten person threshold for shops and establishments or otherwise widen Section 2. It did not disturb the state insurance boundary or the bridging provisions in Sections 5A and 5B. It did not touch Section 12’s dismissal protection, Section 4’s prohibition on employment after delivery, the nursing breaks in Section 11, the miscarriage and tubectomy entitlements, the illness leave in Section 10, or the penalty and prosecution machinery. And it did not touch the financing rule: the benefit remained payable by the employer of the woman, at her average daily wage, for the whole period of absence, which at 26 weeks is a wage bill of roughly half a year, together with the cost of covering her work while she is away, borne by the firm that employs her and by no one else.

Commencement and transition: two dates and a clarification

When did the 2017 amendment commence, and which duty began three months later?

The amendment commenced on 1 April 2017 under a commencement notification issued on 31 March 2017, and the crèche duty in Section 11A was appointed a later date, 1 July 2017, so that establishments had a quarter in which to arrange facilities before the obligation bit.

The staged commencement followed the amendment Act’s own Section 1, which left commencement to notification and expressly allowed different dates for different provisions. The sequence in the transition deserves a reader’s care because it generated real disputes in workplaces during 2017. A woman whose expected delivery fell after 1 April 2017 claimed the enhanced benefit under the amended Section 5. A woman who had already completed and exhausted her twelve weeks before that date had no subsisting absence to which the longer period could attach. The Ministry of Labour and Employment issued clarifications in response to the queries that flooded in, and two of its answers define the transition. The enhanced benefit was available to women who were on maternity leave when the amendment commenced, in the sense that a subsisting period of benefit was measured under the amended text; a woman who had already availed the full twelve weeks before the amendment’s enforcement was not entitled to claim the extension. The Ministry also confirmed that the Act reaches women engaged on contractual terms, a confirmation consistent with the text, since nothing in Section 5 confines benefit to permanent employees, and with the judicial line examined below. The crèche duty’s three month deferral acknowledged a practical truth the rest of the amendment could ignore: leave can be granted by a letter, but a crèche has to be found, fitted out and staffed, and a duty impossible to perform on the day it commences teaches establishments to treat duties as aspirational. Legislatures rarely concede that openly, and the concession here is a small mark of drafting realism inside an otherwise maximalist instrument.

The crèche duty also carried a delegation problem that the transition exposed. Section 11A prescribes the threshold and the visits but leaves the operative content, the distance within which the crèche must sit and the standards it must meet, to rules to be prescribed, and the rule making in the states proceeded at uneven speeds and with uneven content, so that for a period the duty existed in the Act while the specification of what compliance required existed only in draft rules, circulars and the practice of larger employers who built to their own standards. Establishments were thus asked to comply with a duty whose dimensions the state had not finished stating, and the predictable result was a two tier compliance record mirroring the rest of the trail: professionally managed establishments built or contracted crèche facilities and advertised them, while the broad middle of covered establishments waited for rules, inspections or a claim to tell them what was required. The Code on Social Security carried the crèche obligation forward in its maternity chapter, and its full operation rides on the same rule making, which is one of the concrete senses in which the trail’s story is unfinished rather than merely unrepealed.

Administration: the state the amendment legislated for

An amendment is also a bet on the administrative machinery that will carry it, and the 2017 amendment bet on the machinery of 1961: the same Inspectors, the same state rule books, the same exhibition of abstracts and maintenance of registers, asked to deliver a benefit twice the size with two entirely new duties attached. The subordinate legislation had to be moved first. The rules governing mines and circus establishments are central rules; the rules for factories, shops and establishments are state rules, and the amendment’s commencement therefore set off a round of rule revision across the states, amending forms, registers and return formats to the amended text. Where a state’s rules lagged the Act, the entitlement was unaffected in law, since the Act prevails over its own rules, but the paperwork of compliance, the notice forms an employer is told to accept, the register an Inspector asks to see, spoke in the vocabulary of the old Act, and small establishments experienced the amendment through that lag as uncertainty about what exactly to file and keep. The episode is a standard feature of Indian labour law reform and belongs in the trail because it disciplines any claim about what the amendment did on a given date: it changed the rights on 1 April 2017, and it changed the administered reality across the following years at the pace of state rule making.

The crèche duty presented the bet in its purest form, because it is the one duty in the amendment that requires capital rather than paperwork. An establishment crossing fifty employees must find space within the prescribed distance, fit it for infants and small children, staff it, and keep it open across the working day, or contract with a common facility neighbours share. For a hospital, a large factory or a technology campus, the duty is a facilities line item and was frequently already met voluntarily as a retention measure. For the establishment in the band just above the threshold, a mid sized warehouse, a regional office, a private school, the duty is a step cost of a different order, and the common facilities route in Section 11A, which lets the obligation be discharged through a shared crèche, exists because the drafters saw the problem. The threshold also creates a cliff at the boundary of the count: the establishment with forty-nine employees owes no crèche and the establishment with fifty owes a complete one, a structure that invites the same boundary behaviour at fifty that the ten person threshold of Section 2 invites at ten, in an economy where establishment size is partly a choice. None of this makes the duty wrong; childcare at the workplace is the single support most likely to convert a maternity return into a maternity retention. It makes the duty the amendment’s most administration dependent provision, and its record the hardest to read from the statute book alone.

The intimation duty, by contrast, was absorbed into ordinary paperwork within a single hiring cycle, and the manner of its absorption is instructive. Establishments discharged it through the appointment letter itself, through a schedule of statutory benefits annexed to the offer, or through the employee handbook whose receipt the joining form records, and the electronic limb of the duty was met by the same email that carried the offer. Human resources practice converted the duty into a template, and templates are how obligations survive contact with scale: once the paragraph exists in the standard offer letter, it is performed for every woman appointed, by every branch and franchise that uses the letter, without any manager needing to know that Section 11A exists. The trail should record the asymmetry this creates inside the amendment. The duty that was fully templated is fully performed; the duty that required construction, staffing and recurring expenditure is performed unevenly; and the entitlement that required only a payroll instruction is performed wherever the payroll is professional. A legislature choosing among instruments for the next reform has the evidence of this one before it: duties discharged by documents outperform duties discharged by facilities, and both outperform duties whose performance depends on an employer volunteering money at the moment the statute’s beneficiary can least afford a dispute. Design for the template, the facility and the payroll in that order of confidence.

The work from home provision underwent the opposite experience: drafted as a cautious, permissive experiment, it was overtaken by events that tested its premise at national scale. The years of enforced and then normalised remote work during and after the pandemic demonstrated, across precisely the office sectors where the provision matters, that a large share of the work women return to can be performed from home without loss, which was the very contingency Section 5(5) left to the nature of the work and to mutual agreement. The demonstration changed the facts and not the law: feasibility, which an employer could once contest as a matter of assertion, became a matter of the establishment’s own recent practice, and a refusal to agree to home working for a returning mother in a role the establishment itself had run remotely is harder to sustain as a position, although the provision gives her no right to insist and a tribunal no standard by which to override the refusal. The trail here records a recurring pattern in labour legislation, in which a permissive provision’s real content is fixed less by its text than by the world that grows around it, and in which the amendment’s most modest sentence may prove, in the long run, to have been its most adaptive.

Who gets what and who pays

The amendment’s settlement is easiest to hold in one view as a table. The final row is the article’s argument, because the distance between the first row and the last is the distance between the two Indias the maternity law serves.

Category of woman	Entitlement	Source of the obligation	Who bears the cost	The practical position
Employee in a covered establishment, fewer than two surviving children	26 weeks of paid absence, not more than 8 before the expected delivery, at the average daily wage, with dismissal protection and nursing breaks on return	Maternity Benefit Act, 1961 as amended in 2017, Sections 5, 11 and 12; the content carried into the maternity benefit chapter of the Code on Social Security, 2020	Her employer, in full, for the whole period	The strongest statutory maternity entitlement in Indian law, dependent on the establishment being covered and the employer complying
Employee in a covered establishment, two or more surviving children	12 weeks of paid absence, not more than 6 before the expected delivery, on the same protective terms	The proviso inserted into Section 5(3) by the amendment of 2017	Her employer, in full	The pre-2017 position preserved by an express condition; summaries that state 26 weeks without the condition misstate her law
Adoptive mother of a child below three months	12 weeks of paid absence from the date the child is handed over	Section 5(4), inserted in 2017	Her employer, in full	Recognition that maternity is care work and not only recovery from childbirth; half the headline period, because no delivery is undergone
Commissioning mother	12 weeks of paid absence from the date the child is handed over	Section 3(ba) and Section 5(4), inserted in 2017	Her employer, in full	The employment limb of the surrogacy settlement; entitlement depends on the statutory definition of a commissioning mother
Woman in an establishment covered by the state insurance scheme	Maternity benefit under the Employees’ State Insurance scheme for the insured period, drawn from the insurance fund	Employees’ State Insurance Act, 1948; the Maternity Benefit Act stands aside subject to Sections 5A and 5B	The insurance fund, financed by employer and employee contributions	The social insurance model: the cost is pooled across employers instead of resting on the firm that employs the woman
Woman in the unorganised sector, outside covered establishments	A cash maternity benefit of not less than six thousand rupees under the food security statute, paid in instalments under the governing scheme	Section 4 of the National Food Security Act, 2013	The exchequer, through the scheme	A sum smaller than a single month of the formal entitlement; this row, not the first, describes the position of the large majority of working women

The two-child line

Why does the longer leave stop at two surviving children?

The condition grades the entitlement by family size: 26 weeks for a woman with fewer than two surviving children and twelve weeks where she has two or more. It carries the population policy assumption of the two child family into a welfare statute, rewarding the smaller family with the longer leave.

The condition deserves scrutiny from both directions, because each side of it is genuinely arguable. The defence runs as follows. Public provision graded by family size is a familiar instrument of Indian policy; the state’s interest in a smaller family norm is legitimate; the benefit is a wage replacement subsidy delivered through employers, and confining the most expensive tier to the first two children contains the cost of the scheme and the burden on employers; and no woman is left without benefit, since twelve weeks remains a substantial entitlement, equal to what the whole workforce received before 2017. The criticism answers at the level of the child rather than the parent. A third child’s need of maternal care in the first months of life is identical to a first child’s; the infant did not choose its birth order; and a provision justified throughout its passage by the health of mothers and infants withdraws half its protection from the infants whose mothers, by the pattern of fertility in India, are disproportionately poorer, more rural and less able to buy care in the market. The grading also sits awkwardly with the amendment’s own logic elsewhere: the adoptive and commissioning entitlements recognise that the statute protects care, and care does not diminish with the number of siblings. The Supreme Court has met the two child condition in the adjacent territory of service rules, in Deepika Singh v. Central Administrative Tribunal, decided in 2022, where a woman was denied child care leave because her husband had two children from an earlier marriage although the child for whom she claimed was her own first; the Court read the governing rules to sustain her claim, reasoning that the object of such leave is the care of the child before the mother, and that a construction defeating that object on a counting of children she had not borne could not stand. The case concerned leave rules rather than Section 5(3), and it does not unsettle the statutory condition, which is express. It shows the direction of judicial travel when a counting rule collides with the purpose of a care entitlement, and it marks the two-child line as the provision in the amended Act most exposed to a future challenge argued on the statute’s own stated objects.

The drafting of the condition carries its own consequences, which a trail should notice because they fall on cases the passage debate never mentioned. Parliament graded the benefit by surviving children, not by pregnancies undergone or children borne, and the choice of the survivor test over a birth count cuts in two directions. In the woman’s favour, a child who does not survive does not consume the longer entitlement: a woman whose first child died in infancy and who has one living child stands, on the words of the proviso, as a woman with fewer than two surviving children, and takes 26 weeks for her next delivery, which is the humane reading and the textual one. Against her, a single first delivery of twins exhausts the longer tier as completely as two separate deliveries: the mother of twins who is delivered of a third child takes twelve weeks, although she has been pregnant twice and although the care burden that justified 26 weeks, an infant’s first months, is undiminished. A birth count and a survivor count are different instruments, and the statute’s selection between them was a real drafting decision with distributive effects at the margins, of the kind that careful amendment drafting surfaces and hurried amendment drafting leaves for the courts and the clarifications to absorb. The adoptive entitlement interacts with the same count in ways the text does not expressly resolve: whether children adopted count toward the two surviving children for a later delivery is a question the provision’s language, which speaks of surviving children without distinguishing the route by which they entered the family, answers more readily than the administration has always assumed, since a child in the family is a child in the family for the purposes of a care based entitlement, and the parity reasoning of the courts points the same way. These are the cases in which the two child line will be litigated if it is litigated at all, and they show the condition working not as a bright population signal but as a counting rule at the edge of real families.

The population policy pedigree of the condition should be recorded fairly, because it is neither accidental nor unique to this statute. Indian public law has carried a two child norm in several fields for decades: a number of states enacted laws disqualifying persons with more than two children from contesting or holding panchayat office, and the Supreme Court upheld the Haryana version of that disqualification in Javed v. State of Haryana, decided in 2003, accepting the promotion of family planning as a legitimate object and the disqualification as a proportionate instrument of it. A legislature working inside that settled landscape did not experience the grading of maternity benefit by family size as an innovation requiring special justification; it experienced it as the routine application of a familiar policy grammar. The counter-argument has also developed since those cases were decided, and it is specific to benefit statutes as opposed to eligibility rules for office. A disqualification from office burdens the person who chose the larger family in pursuit of a civic privilege; a grading of maternity benefit burdens, in the first instance, an infant, by halving the maternal care the state itself has declared necessary in the first months of life, and it burdens precisely the class of mothers least placed to purchase substitute care. Whether that distinction would move a court asked to test the proviso against the statute’s own health objects is an open question, and the honest statement of the trail is that the proviso is express, politically orthodox in origin, and analytically the least defended provision of the 2017 settlement, resting on a population rationale its own passage debate barely articulated because the debate was busy celebrating the weeks.

The provision the amendment never touched: who pays

The economics of the amended Act reduce to one fact. In every establishment outside the state insurance scheme, the cost of a maternity absence is the full wages of the absent woman for the period of leave, plus the cost of the replacement labour or lost output during her absence, and the whole of it is charged to the establishment that employs her. Doubling the leave therefore doubled, and more than doubled, the expected cost to an employer of employing a woman who might become pregnant, relative to employing a man of the same age and skill, while leaving the legal prohibition on acting on that calculation, in recruitment and in dismissal, exactly as strong on paper as before. The statute thus created a priced incentive against its own equality object and relied on enforcement and employer goodwill to defeat it. This is the sense in which the financing rule, and not the leave period, is the amendment’s most consequential feature, and it is why an assessment of the trail that stops at the generosity of the entitlement has not begun.

The case for employer liability, stated at its strongest, is not contemptible and should not be caricatured. Its defenders, in the tripartite tradition from which Indian labour law grew, argue that the enterprise which takes the benefit of a woman’s labour should bear the ordinary human costs of that labour, of which childbirth is the most predictable; that a benefit funded by the individual employer is a benefit delivered with certainty, without a fund to be designed, contributions to be collected or a bureaucracy to interpose between the woman and her wages; that social insurance in India has historically reached only establishments above thresholds and workers below wage ceilings, so an insurance model risked excluding the better paid women the Act covers; and that employers of scale had in many cases already moved voluntarily toward longer leave as a retention measure, evidence that the cost is absorbable where the establishment is large. The retention argument deserves its weight: a statute that keeps an experienced woman attached to her employer through childbirth protects her seniority, her provident fund continuity and her career in a way that a cash benefit from a distant fund does not, and employer liability is the mechanism that fuses the wage payment to the job protection in Section 12. On this view, the 2017 design is coherent: the firm pays, the job waits, and the woman returns.

The case against, stated at its strongest, begins by accepting every word of the entitlement’s value and attacking only its incidence. A cost that falls on the specific firm that hires the specific woman is a cost that rational firms will try not to incur, and the margin on which they can act is hiring, because dismissal is policed by Section 12 and hiring decisions are effectively unreviewable. The burden is also regressive as between firms: a large company spreads the expected cost across thousands of employees and treats it as a predictable overhead, while a small establishment with a handful of women employees faces the whole cost of one absence, wages and replacement together, as a discrete shock, so the design charges the obligation most heavily exactly where compliance capacity is thinnest. And the burden is arbitrary as between women: two women doing identical work draw radically different security depending on whether their establishments fall inside the insurance scheme, where the cost is pooled, or outside it, where their employer’s private calculation about their childbearing becomes part of their employment prospects. The remedies proposed by the critics are all ways of keeping the entitlement while moving its incidence: extending the social insurance model so that a fund pays, reimbursing employers from the exchequer in whole or part, or enacting parental leave that distributes expected absence across fathers as well, so that the cost of reproduction stops being a sex specific charge on the women a firm employs. Each remedy has been urged in the debate that followed the amendment, by trade unions, by women’s organisations and by employer bodies themselves, which is a mark of how widely the diagnosis is shared even among those who disagree about everything else in labour law.

The evidence most often cited in this debate is the study published by TeamLease Services in June 2018, titled The Impact of Maternity Benefits on Business and Employment. TeamLease is one of India’s largest staffing companies, and its report was built on a survey of 300 employers across ten sectors, including aviation, information technology and technology enabled services, banking and financial services, manufacturing, retail, education, e-commerce, real estate and tourism. The report projected a net loss of jobs for women of between eleven lakh and eighteen lakh across those ten sectors in the financial year 2018-19, attributable to the amendment, over and above ordinary attrition; it reported that large companies and medium sized public sector employers were positive about the amendment and expected to hire as before or more, because for them the cost of replacing a trained woman exceeded the cost of retaining her; and it reported that start-ups, small and medium enterprises, medium sized multinational companies and closely held family run businesses expected to reduce their intake of women. Rituparna Chakraborty, the company’s co-founder and executive vice president, attributed the projected loss to the amendment itself and pointed to the British mechanism as the contrast, one in which a woman’s entitlement runs longer than India’s and the cost of paying for it does not rest entirely on the employing organisation. The same survey carried the other half of the story, which is quoted less often: employers surveyed expected post maternity attrition, reported at 56 percent, to fall to 33 percent over the following four years as the longer leave took effect, which is the retention dividend the amendment’s defenders predicted.

A reader should know what kind of evidence that is. The TeamLease figures are projections of employer hiring intention gathered in a survey conducted a year after commencement, not measurements of what employment of women actually did; they state what 300 employers said they expected to do, and the conversion of such statements into a count of jobs that would not exist involves modelling choices the headline number conceals. No official series isolates the amendment’s effect on women’s employment from the general movement of the labour market, and this article asserts no national figure. What the report does establish, and what no serious participant disputes, is the direction of the incentive and the identity of the firms that feel it: the smaller and less formal the establishment, the sharper the perceived cost of the entitlement, and the greater the temptation to comply on paper, to prefer male candidates at the margin, or to keep women’s engagement informal and outside the Act altogether. Human resources teams that modelled the exposure for their boards worked through precisely these computations, establishment size by establishment size, and a reader following the same exercise can open the staffing workings in a PPT viewer and trace how the cost lands differently on a firm of forty employees and a firm of four thousand. The analytical verdict the evidence supports is narrower than either camp’s slogan. The amendment did not need to destroy women’s jobs to do damage, and it did not need to be costless to be worth enacting. It concentrated a social cost on the actor with the strongest incentive to avoid it, and the predictable consequences, hiring caution in small firms, informalisation at the margin, uneven compliance, were the price of financing a public good through a private charge. A legislature that had pooled the cost would have bought the same 26 weeks without that price, and the trail shows it was never asked to choose.

The insurance alternative deserves a precise statement, because it exists inside the Indian system already and is not a foreign import. Under the Employees’ State Insurance scheme, maternity benefit for an insured woman is paid from a fund built out of contributions levied as a percentage of wages on employers and employees across every covered establishment, so the cost of any one woman’s absence is spread across the entire insured workforce and no individual firm’s bill moves when one of its employees has a child. The incentive structure is the mirror image of employer liability: hiring a woman of childbearing age carries no identifiable marginal cost, because the contribution is owed for the workforce as it is and does not rise with the sex composition of new recruits. The limits of the model are equally structural. The insurance scheme reaches establishments above its own thresholds and employees below a wage ceiling, which is why it has historically covered the lower paid half of the formal workforce and left the better paid half to the Act; a universal insurance model for maternity would need either a levy without a ceiling, which employers’ organisations have resisted as a payroll tax, or exchequer subvention, which returns the question to the budget the 2017 reform was designed to avoid. The British arrangement to which the TeamLease leadership pointed in the debate works on the reimbursement principle: the employer pays the statutory maternity pay and recovers it, in whole or in large part, from public funds, so the administrative convenience of payment through the payroll is retained while the incidence is moved to the tax base. Each model was available to the drafters of 2016 in working form, two of them inside Indian law or the law of the country whose parliamentary practice India follows. The choice of employer liability was therefore a choice among known alternatives, made under a political equilibrium in which it was the only model that charged the reform to no organised constituency that was in the room, and that, rather than ignorance of the alternatives, is the explanation the trail supports.

One further feature of the debate connects this statute to its neighbours. The objection that a woman’s employment is made expensive by her biology is the same objection, in another register, that the equal remuneration legislation answered on the wage side: a labour market left to price sex differences will price them, and the law’s response has been to forbid the pricing rather than to subsidise it. The wages limb of that settlement, carried from the Equal Remuneration Act into the Code on Wages, is examined in the Equal Remuneration Act, 1976 explained; the maternity limb shows the limit of the strategy, because a prohibition can govern a payslip in a way it cannot govern a shortlist. The honest conclusion is that the financing question remains open in Indian law, carried forward unresolved into the Code on Social Security, whose maternity chapter preserves employer liability for the establishments outside the insurance scheme just as the Act of 1961 did.

The international frame

How does the Indian entitlement compare with the international standard?

Generously, on duration. The Maternity Protection Convention, 2000 (No. 183) of the International Labour Organization sets a floor of fourteen weeks of maternity leave, and the 26 weeks enacted in 2017 is close to double that floor, placing India’s formal entitlement among the longest in the world for the first two children.

The historical line sharpens the point. The Act of 1961 was drafted against the Maternity Protection Convention (Revised), 1952 (No. 103), whose floor was twelve weeks of leave, and the original Section 5 adopted that floor exactly: the twelve weeks that later generations would treat as meagre were, at enactment, the international standard itself, written into Indian law within a decade of its adoption. Each revision of the international standard thereafter left the Indian figure behind, not because Parliament reduced anything but because it never returned to the provision, until the amendment of 2017 carried India past the standard of 2000 in a single step. A trail that moves in one jump of this size tells its own story about the cadence of Indian social legislation in this field: long stasis at a figure that was orthodox when chosen, followed by a correction so large that it reset the comparison entirely, with the intermediate positions, fourteen weeks, eighteen weeks, twenty-four weeks, debated by conferences and committees across the decades and enacted by no one.

The comparison cuts both ways, and the trail is incomplete without its other edge. The Convention’s design assumes that benefits are paid through compulsory social insurance or public funds, and states the principle expressly that an employer is not individually liable for the direct cost of the benefit except where a national settlement so provides with the agreement of the social partners; the architecture exists precisely to prevent the cost of maternity from becoming a charge on the woman or on the firm that hires her. India’s settlement, 26 weeks of benefit with individual employer liability across the non insurance sector, therefore combines the longest duration in the comparison with the financing method the international standard treats as the exception. The six months of exclusive breastfeeding recommended in public health guidance explains the choice of 26 weeks as a duration; nothing in health guidance speaks to who should pay, and the amendment’s silence on financing cannot borrow authority from the health case for its length. India has not ratified Convention No. 183, and the comparison is one of standards rather than obligations, but as a standard it frames the trail’s central question exactly: a country may choose the duration and the financing independently, and the amendment of 2017 exercised only half of that choice.

The parallel trail in the courts

The courts worked on the Act along a line the legislature did not draw, the line between forms of employment. The text of Section 5 speaks of a woman employed in an establishment and draws no distinction between the permanent employee, the contractual employee, the casual worker and the daily wage worker; establishments drew the distinction anyway, treating maternity benefit as an incident of regular service. The leading decision is Municipal Corporation of Delhi v. Female Workers (Muster Roll), decided by the Supreme Court in 2000, reported at (2000) 3 SCC 224. The women were engaged by the Municipal Corporation of Delhi on muster roll, on daily wages, doing the same work as regular employees, and were refused the maternity benefit that regular women received; their union’s claim succeeded before the Industrial Tribunal and the Delhi High Court, and the Corporation carried the matter to the Supreme Court. A bench of Saghir Ahmad and Wadhwa JJ. dismissed the appeal. The Court held that the provisions of the Act are wholly in consonance with the directive principles in Articles 39 and 42, that a woman employee at the time of advanced pregnancy cannot be compelled to undertake hard labour detrimental to her health and to the health of the foetus, and, in the passage on which the later case law rests, that the Court had scanned the different provisions of the Act and found nothing in them which entitles only regular women employees to the benefit of maternity leave and not those engaged on casual basis or on muster roll on daily wage basis. The decision also drew on the Convention on the Elimination of All Forms of Discrimination against Women, and its durable proposition is that the Act classifies by establishment and by days worked, and not by the label on the engagement.

Later courts applied the same purposive method at the Act’s newer edges. Before the amendment of 2017 admitted adoptive and commissioning mothers to the statute, High Courts were asked to place them within leave frameworks built for birth mothers. The Madras High Court in K. Kalaiselvi v. Chennai Port Trust, decided in 2013, treated a commissioning mother as entitled to maternity leave on par with an adoptive mother under the applicable leave rules, and the Bombay High Court in Hema Vijay Menon v. State of Maharashtra, decided in 2015, held that maternity related leave could not be restricted to biological mothers, since the entitlement serves the rearing of the child and not only recovery from delivery. These decisions operated on service rules rather than on the 1961 Act, and should not be cited as constructions of it; their place in the trail is that they built the parity reasoning, motherhood as care rather than motherhood as parturition, which the amendment of 2017 then wrote into Section 5(4). The Delhi High Court’s treatment of contractual employees follows the same line: in a line of decisions of which its ruling on an ad hoc employee’s claim is representative, the court held that the Act does not differentiate between a permanent employee and a contractual employee or a daily wage worker, and that the two limiting factors on a claim are the 80 days of work in the preceding twelve months and the ceiling on the period, not the tenure printed on the appointment letter.

The Supreme Court returned to the point in Dr. Kavita Yadav v. Ministry of Health and Family Welfare, decided in 2023. The appellant was a contractual employee whose pregnancy fell within her engagement and whose contract ended before the maternity period ran out; the employing institution resisted the benefit for the period beyond the contract’s term. The Court held that the entitlement to maternity benefit under the 1961 Act is not coterminous with the tenure of employment: where the pregnancy occurs during the subsistence of the engagement and the qualifying condition is met, the benefit runs for its full statutory period even though the contract has ended in the meantime. Read with the muster roll case, the position across three decades of case law is consistent and worth stating as a rule of the trail: the Act’s boundaries are its own, the covered establishment and the 80 days, and courts have refused every attempt to write in a third boundary based on the form of the woman’s engagement. The practical importance of that refusal grows as formal employment itself becomes more contractual, because each shift toward fixed term and outsourced staffing would otherwise have silently contracted the Act’s coverage without any amendment at all.

The dismissal jurisprudence under Section 12 has its own discipline, which the trail should state because it defines the one door through which an employer may lawfully act. The protection is not a protection against dismissal for cause unconnected with the absence; it is a protection against dismissal during or on account of the absence, and against the variation of service conditions to the woman’s disadvantage in that window. The gross misconduct exception is hedged with form: the deprivation of benefit must be by an order in writing, the misconduct must be of the kind prescribed, and the order is appealable within sixty days, so an employer who acts on an unrecorded allegation, or who converts a performance process opened before the pregnancy into a dismissal delivered during the leave, acts unlawfully and exposes the dismissal itself, not merely the benefit, to challenge. Courts and tribunals reading Section 12 have treated timing and sequence as evidence: a dismissal that follows hard upon the announcement of pregnancy or the filing of the Section 6 notice invites the inference the section exists to defeat, and the employer who wishes to sustain such a decision must show a record that predates the protected event. The practical consequence for the trail is that the dismissal shield, unlike the financing rule, has been strengthened by its enforcement culture: the amendment of 2017 did not need to touch Section 12 because forty years of case law had already given it the meaning its text promises, which is the clearest contrast inside the statute between a provision the courts could complete and a design choice only the legislature could revisit.

Compliance as the second trail: how establishments answered

The amendment’s second trail runs through establishments rather than through Parliament, and it is written in policies, clarifications and disputes rather than in sections. Its first chapter was alignment. Large employers, many of which already offered leave at or near the new figure as a matter of competitive policy, revised their maternity policies to the statutory text, extended their return support, and treated the crèche duty as a facilities programme; for this tier the amendment functioned as a floor beneath an existing practice and its main effect was to remove the competitive disadvantage of generosity, since a firm could no longer be undercut on maternity cost by a rival offering the old statutory twelve weeks. The second chapter was interpretation. The months after commencement produced a traffic in questions that the Ministry’s clarifications only partly answered: the treatment of women whose leave straddled 1 April 2017, the position of fixed term and agency staff on the establishment’s premises, the counting of employees for the crèche threshold where a workforce is split across sites and contractors, and the form the electronic intimation should take. Each question is a place where the amended text runs out and administration begins, and the pattern of the questions confirms the design analysis: they cluster at the coverage boundaries and the financing points, and almost none of them concern the length of the leave, which the amendment had made the one unambiguous thing in the scheme.

The third chapter is the disputed one, and the trail should record it at the strength the evidence allows and no further. In the debate that followed the TeamLease report, unions and women’s organisations reported cases of the practices the design predicts: women asked to resign before leave and rehired after it, engagements converted to consultancy or short contracts around the qualifying period, recruitment conversations that priced the risk the statute forbids pricing, and small establishments pleading inability as a reason for informal settlements below the statutory figure. Employer bodies answered that such cases were exceptions in a record of general compliance, that the survey projections had been misread as outcomes, and that retention gains were visible in their own workforces. Both bodies of claim are reports of experience rather than measurements, and the honest verdict is the one already given on the study itself: the incentive is structural, the practices it predicts were reported by the organisations closest to the women affected, no official measurement exists that could weigh them against the compliant majority, and a legal system that leaves the measurement of its own statute’s central risk to a staffing company’s survey has an evidence problem that is itself part of the implementation record. The courts, in the years after the amendment, continued to decide the cases the machinery produced, and their holdings, on contractual tenure, on the counting of days and on dismissal in the protected window, have been collected in the section on the judicial trail; what the litigation map shows is that disputes reach court at the Act’s boundaries of status and coverage, which is where an amendment aimed at working women generally, rather than at employees of a defined kind, would have legislated. The amendment of 2017 legislated for the employee the Act already knew, at twice the length and the same address, and the compliance trail is the record of that choice working itself out in establishments the statute reaches unevenly.

The third system: maternity leave in government service

No account of the trail is complete without the system that sits outside the Act entirely and set the benchmark the amendment chased. Women in central government service do not draw maternity benefit under the Maternity Benefit Act; they take maternity leave under the Central Civil Services (Leave) Rules, and the generosity of that system is the standard against which the statutory entitlement was publicly measured in 2016 and 2017. Under those rules a woman government servant is entitled to maternity leave of 180 days, a period raised from 135 days on the recommendation of the Sixth Central Pay Commission with effect from September 2008, and the same rules carry a child care leave of up to 730 days across the whole of a woman’s service for the care of her children, the provision whose construction reached the Supreme Court in Deepika Singh. The political arithmetic of the amendment is visible in the figures: 26 weeks is 182 days, which is the 180 days of the government system translated into the weekly units of the Act, and the reform was presented, accurately, as carrying the private formal sector to parity with the public sector. Two lessons for the trail follow. The first is that the state, when it is itself the employer, had settled the duration question nine years earlier without a financing crisis, because the cost of a government servant’s leave falls on a consolidated fund and a cadre large enough to absorb absence as a routine establishment expense; the employer’s identity, and not the length of the leave, decides whether the entitlement is experienced as a burden. The second is that the parity framing, powerful as it was, quietly defined the amendment’s constituency: parity with government servants is an argument about the formal workforce, and it supplied no language in which the unorganised majority’s six thousand rupees could be debated in the same breath, which is one reason the two tracks were never debated together although they answer the same constitutional sentence. State government servants stand under state leave rules that broadly track the central pattern, so the third system is a family of rule books rather than a single text, but its direction is uniform and its position in the trail is fixed: it led on duration, it finances by absorption, and it is the standing demonstration that the obstacle to a long maternity entitlement in India was never administrative capacity but the incidence of cost in the private establishment.

The unorganised majority

What separates the formal sector entitlement from the unorganised sector’s cash benefit?

Almost everything: amount, form and payer. The formal entitlement replaces a woman’s full wages for 26 weeks, paid by her employer; the unorganised sector’s entitlement is a cash benefit of not less than six thousand rupees under the food security statute, paid by the state in instalments, below one month of the formal entitlement at most wage levels.

An honest account of the trail has to end its analysis of coverage here, because the amendment debate was conducted almost entirely about a minority. The overwhelming majority of women who work in India work outside covered establishments: in agriculture, in domestic work, in home based piece work, in construction, in small family enterprises and in the vast informal economy where no establishment of ten or more persons stands between the woman and her livelihood in the form the Act requires. For them the maternity law of India is not the Maternity Benefit Act at all. It is Section 4 of the National Food Security Act, 2013, which entitles every pregnant woman and lactating mother, subject to the scheme made under it, to a maternity benefit of not less than six thousand rupees, paid in instalments; the scheme that operationalises the entitlement, the Pradhan Mantri Matru Vandana Yojana, pays five thousand rupees for the first living child. The structure of that entitlement, its place in the food security statute and its delivery record are the subject of the National Food Security Act 2013 guide; its place in this trail is comparative. Six thousand rupees is the statutory recognition that maternity has a cost for every woman; 26 weeks of full wages is the recognition of what that cost actually is for a woman with an employer. The two figures belong to the same constitutional promise in Article 42, and the distance between them measures how far the promise has been kept for the minority inside the establishment gate and how thinly it has been spread outside it.

The scheme lineage behind the cash entitlement shows the state answering the same constitutional direction by a different technique. A conditional maternity benefit scheme was piloted from 2010 in selected districts, paying a cash amount in instalments tied to registration of the pregnancy, institutional delivery and immunisation of the child, and the National Food Security Act, 2013 converted that scheme logic into a statutory floor by fixing the entitlement at not less than six thousand rupees for every pregnant woman and lactating mother. The national scheme that succeeded the pilot, the Pradhan Mantri Matru Vandana Yojana, pays five thousand rupees for the first living child, a figure below the statutory floor, which implementing orders reconcile with the Act by treating the balance as met through the cash incentive for institutional delivery paid under the health mission, the Janani Suraksha Yojana. However the arithmetic is assembled, the contrast with the formal track is structural and not merely numerical. The formal entitlement is wage replacement, calculated from the woman’s own earnings and designed to hold her standard of living and her job together across the absence; the unorganised entitlement is a contribution toward the expenses of childbirth, calculated from the state’s fiscal capacity and designed to encourage institutional delivery, and it replaces no wages at all, because the woman it addresses has no single employer from whom wages could be replaced. The two instruments share a name, maternity benefit, and almost nothing else, and a reader who meets the phrase in a headline about the formal amendment should not carry its meaning across to the majority who stand under the other statute.

The comparison should not be pushed into a cheap verdict against the 2017 amendment, which improved the position of the women it reached and is not responsible for the women it did not. The amendment’s terms of reference were the 1961 Act, and the 1961 Act’s terms of reference were establishments. But a reader assessing the trail should hold both facts at once: the most generous maternity leave in the statute book and the smallest maternity cash benefit in it coexist in the same legal system, divided by the accident of whether a woman’s work is performed inside a covered establishment, and no amendment in the trail, from 1963 to the Code, has treated that division itself as the problem to be solved. The unorganised workers’ chapter of the Code on Social Security continues the scheme based approach for this majority, under which benefits exist as schemes framed by government rather than as entitlements claimed against an employer, so the two track structure of 1961 survives the consolidation intact, with a third track, the insurance fund, running beside them.

Subsumption: the trail enters the Code on Social Security

The final event in the trail is a repeal. The Code on Social Security, 2020, Act 36 of 2020, consolidated nine welfare statutes into a single code of fourteen chapters, and the Maternity Benefit Act, 1961 was one of the nine. The Code carries the maternity settlement into its own maternity benefit chapter, Chapter VI, and the content it carries is the content as the amendment of 2017 left it: the 26 weeks of paid leave for a woman with fewer than two surviving children at the centre, the graded twelve weeks beyond, the provisions for adopting and commissioning mothers, the crèche obligation graded by establishment size, and employer liability as the financing rule for establishments outside the insurance scheme, whose architecture the Code restates in its own state insurance chapter. The statute-by-statute account of that consolidation, and the reasons the Code restated some chapters at length and sketched others, belongs to the Social Security Code 2020 explained, which owns the Code as a whole; this article records only what the repeal did to the maternity trail.

What it did, in the event, was delayed for five years. The Code was enacted in 2020 with commencement left to notification, and the four labour codes were brought into force together on 21 November 2025, from which date the twenty-nine subsumed statutes of the labour code project stood repealed and the maternity entitlement continued under the Code’s chapter rather than under the Act of 1961. The long interval between enactment and commencement, during which the Act of 1961 governed every actual claim while its repeal sat printed in a code awaiting notification, and the incomplete state of the rules on which the Code’s operation depends, are traced in the labour codes implementation delay; for the maternity trail the operative points are three. First, the repeal extinguished no entitlement: the savings structure of the Code deems action taken under the repealed Act to have been taken under the corresponding provisions of the Code, so claims, notices and proceedings cross over rather than restart. Second, the repeal moved the address of every provision: a reader citing Section 5 of the Act of 1961 for a claim arising under the Code is citing a repealed text, and the discipline of checking which instrument governs her workplace on the date of her claim became, after November 2025, the first step of any maternity claim. Third, the consolidation changed nothing in the trail’s argument. The Code had the opportunity, once in two generations, to redesign the financing of maternity benefit for the whole workforce, and it chose restatement: the entitlement architecture of 2017 was carried forward intact, employer, insurance and scheme tracks preserved in their old proportions. The mechanics of the transfer repay a paragraph, because they determine what a claim looks like under the Code. The Code works with the unified definitions of the labour code project, so the wage base from which maternity benefit is computed is the base defined for the codes as a family, the definition carried in the Code on Wages under which basic pay, dearness allowance and retaining allowance are wages and specified heads are excluded subject to the fifty percent ceiling on exclusions; a maternity computation under the Code therefore begins from the same figure as a provident fund computation, which closes the old possibility of divergent bases inside one payroll. The Code restates the inspectorate in the modern form of the inspector cum facilitator, preserves the claims and appeal structure in substance, and continues the offences and penalties architecture in its general chapters. What a claimant experiences is continuity of substance with a change of citation: the entitlement, its conditions, its grading by surviving children and its financing all survive, while the section numbers, the forms and the authority to whom the claim is addressed are the Code’s. Advisers and human resources departments that had built their maternity documentation on the Act of 1961 were required, at commencement, to re-point every reference, and establishments that failed to do so hold paperwork citing a repealed statute for a live entitlement, a small but telling instance of how a consolidation taxes the administered world even when it changes nothing of substance.

The trail therefore does not end in 2025. It pauses there, with the question it has carried since 1961 transferred to a new statute book in the same words.

The entitlement at work: three situations the trail produces

The trail’s distinctions are easiest to test against the situations a claim actually presents. Take first a woman employed for four years as a software engineer in a technology company employing several thousand persons, expecting her first child. Her establishment is covered, her service exceeds the 80 days in the preceding twelve months by years, and she gives notice under Section 6 stating the date from which she will absent herself. She takes eight weeks before the expected delivery and eighteen after it, twenty-six weeks at her average daily wage computed on the three preceding months, paid in advance for the first period and within 48 hours of proof of delivery for the second. Her post cannot be filled permanently in her absence in any way that displaces her, her conditions cannot be varied to her disadvantage, and on return she has two nursing breaks a day until the child is fifteen months old and four daily visits to the establishment’s crèche, which its headcount obliged it to provide. If the nature of her work permits, she may ask, after the benefit period expires, to work from home on mutually agreed conditions, and the employer may agree or decline, because Section 5(5) gives her the conversation and not the outcome. Her claim is the amendment working exactly as its passage debate imagined, and it depends on nothing but coverage, service and a compliant employer.

Take next a woman expecting her third child, employed in a retail establishment that employs eleven persons, two of whom are women. The establishment is covered, because the threshold for shops and establishments is ten, and she satisfies the 80 day condition. Her entitlement is twelve weeks, not twenty-six, because two of her children survive: six weeks before the expected delivery at most, and six after, at her average daily wage, with the medical bonus and the Section 12 shield. The employer’s arithmetic is the trail’s other face: a wage bill of a quarter of a year for the absent woman, a temporary replacement to be hired and trained for a shop that runs on a staff of eleven, and no fund, insurer or exchequer to which any part of it can be sent. If he complies, he has borne the whole cost of a national policy on a small balance sheet; if he does not, her remedies are the Inspector’s direction under Section 17 and the penal provisions behind it, set in motion by her complaint. Every pressure the TeamLease survey recorded is present in this single case in miniature, which is why compliance in this segment is the trail’s weakest record and why the financing question, and not the leave period, is the question her case poses.

Take finally a commissioning mother, a bank officer whose child is born to a surrogate under the arrangements the surrogacy legislation permits, the embryo created from her egg. Before 2017 the Act had nothing for her: no delivery, no benefit, and her claim to leave depended on her employer’s leave rules and the parity reasoning the High Courts had begun to accept in the service cases. Under Section 5(4) she is entitled to twelve weeks of maternity benefit from the date the child is handed over to her, at her average daily wage, with the dismissal shield applying to that absence as it applies to any other, and she learned of the entitlement, if her employer obeyed Section 11A(2), in writing and electronically on the day of her appointment rather than in the middle of the surrogacy process. Her case shows the amendment at its most considered: a category of mother the original legislature could not have foreseen, admitted by definition, given a benefit calibrated to care rather than recovery, and fitted into machinery built for another purpose in 1961 without disturbing it. The three situations together are the trail in miniature: an entitlement that works completely where the establishment is large, strains where it is small, and extends with real ingenuity to cases its drafters never contemplated, while the rule that allocates its cost sits underneath all three, unchanged.

What the trail predicts

A trail read closely supports a disciplined prediction, and this one supports three. The first concerns the forum of the next reform. The maternity question has migrated, across the trail, from a standalone Act to a chapter of a consolidated code, and the next material change will be made inside the Code’s architecture rather than by reviving the Act: through the Code’s rule making on the crèche obligation, through the scheme chapters that govern unorganised, gig and platform workers, and through the insurance chapter’s thresholds and ceilings, each of which can be moved by notification or rule with a fraction of the parliamentary effort a fresh amendment requires. Reformers who continue to draft amendments to a repealed Act will miss the venue where the entitlement’s real boundaries are set. The second prediction concerns the order of the questions. The duration question is settled for a generation: no future government will shorten 26 weeks, and the two child grading, though analytically exposed, is politically orthodox and unlikely to be the first item revisited. The live questions are the ones the trail leaves open, financing and coverage, and they will be taken in the order of their administrative ease: coverage of the unorganised majority through schemes, because schemes need no employer’s consent and can be expanded by budget lines, before any pooling of the formal sector’s cost, because pooling requires either a levy or an appropriation and a decision about which employers pay for which women. The third prediction concerns the paternity question, which the trail has deferred three times over, in the consultations of 2016, in the debates of 2017 and in the consolidation of 2020. Each deferral has made the eventual reform larger, because a parental leave enacted into a system built on 26 weeks of maternity leave must either add a father’s entitlement alongside it, at visible cost, or redivide the existing period between the parents, at visible political risk; the systems that solved this problem cheaply solved it when the total period was short. The trail therefore ends in a stable and unstable position at once: the entitlement itself is beyond political contest, and every question that determines what the entitlement is worth in the labour market, who funds it, whom it reaches and whether fathers stand inside it, remains exactly where the amendment of 2017 left it.

Verdict

The amendment of 2017 was a genuine and substantial advance, and this article has not qualified that judgment, because the evidence does not ask for qualification on the entitlement side: 26 weeks of wage protected absence, recognition of adoptive and commissioning mothers, a crèche duty with a defined threshold, an intimation duty that makes the benefit visible from the first day of employment, and a dismissal shield that the courts have read at full strength. For a woman in a covered establishment with a compliant employer, Indian law after the amendment is among the most protective maternity settlements anywhere. The trail’s lesson lies in the assessment rule it teaches. An amendment is read correctly only when what it left alone is read with the same care as what it changed, and the provision the amendment of 2017 left alone, the rule that the individual employer pays, explains more of the amendment’s recorded consequences, the hiring caution of small firms, the formalism of compliance, the persistence of the informal margin, than any provision it enacted. The statute doubled the leave and kept the payer. The first fact is the reform’s achievement; the second is its explanation; and the reform that finally addresses the second, by pooling the cost as the insurance track already pools it for its own sector, is the amendment this trail is still waiting for.

The rule of assessment this article has applied is worth stating once, plainly, because it travels to every other trail in this series. An amendment tells the truth about a legislature’s intentions not in the provisions it enacts, which are written for the audience that celebrates them, but in the provisions it examines and leaves, which are written for the audience that pays for them. The amendment of 2017 examined the length of maternity leave and found it wanting, and doubled it; it examined the classes of mother the statute recognised and found them incomplete, and completed them; it examined the duties of the establishment and added two it had never owed. It did not examine the incidence of the cost, the boundary of the covered establishment, or the father’s absence from the scheme, and in each case the unexamined provision, not the amended one, has governed the record since. To read an amendment trail correctly is to read the silences with the text, and to ask of every generous figure the question this statute teaches: who was asked to pay for it, and what will they rationally do.

Frequently Asked Questions

Q: How much maternity leave does Indian law provide?

For a woman employed in a covered establishment, the entitlement is 26 weeks of paid maternity benefit, of which not more than eight weeks may precede the expected date of delivery, provided she has fewer than two surviving children. A woman who has two or more surviving children is entitled to twelve weeks, of which not more than six may precede the expected delivery. The benefit is paid at the rate of her average daily wage for the period of actual absence. These periods were fixed by the amendment of 2017 to the Maternity Benefit Act, 1961, and the content stands carried into the maternity benefit chapter of the Code on Social Security, 2020.

Q: Which establishments must provide maternity benefit?

The obligation falls on establishments within the Act’s application provision: every factory, mine and plantation, including government establishments; establishments where persons are employed for the exhibition of equestrian, acrobatic and other performances; and every shop or establishment in which ten or more persons are employed or were employed on any day of the twelve preceding months. The appropriate government may extend the Act to other classes of establishment with the approval of the Central Government after two months’ notice. Where the Employees’ State Insurance scheme covers an establishment, maternity benefit is drawn under that scheme instead, subject to bridging provisions that protect a woman when insurance coverage begins or ends.

Q: What did the 2017 maternity amendment change?

The Maternity Benefit (Amendment) Act, 2017, Act 6 of 2017, raised the maximum period of paid maternity benefit in Section 5 from twelve weeks to 26 weeks for a woman with fewer than two surviving children, raising the pre-delivery cap from six weeks to eight, and retained twelve weeks for a woman with two or more surviving children. It inserted a definition of commissioning mother and gave adoptive mothers of children below three months and commissioning mothers twelve weeks of benefit from the date the child is handed over. It added a permissive work from home provision in Section 5(5), a crèche duty for establishments with fifty or more employees, and a duty to intimate every woman in writing and electronically of the benefits at her initial appointment.

Q: Do adoptive and commissioning mothers get maternity leave?

Yes. Section 5(4), inserted by the amendment of 2017, entitles a woman who legally adopts a child below the age of three months, and a commissioning mother, to maternity benefit for a period of twelve weeks from the date the child is handed over to her. A commissioning mother is defined by Section 3(ba) as a biological mother who uses her egg to create an embryo implanted in any other woman. The benefit runs from the handing over rather than from a delivery, because its object is the care of the child in the first months. The general qualifying condition of 80 days of work in the preceding twelve months applies within the scheme of Section 5.

Q: When must an employer provide a creche?

Section 11A(1) requires every establishment having fifty or more employees to have the facility of a crèche within such distance as may be prescribed, either separately or along with common facilities shared with other establishments. The threshold counts employees, not women employees, so the duty is triggered by the size of the workforce as a whole. The employer must allow a woman four visits a day to the crèche, and those visits include the interval for rest allowed to her. The duty was inserted by the amendment of 2017 and was commenced on 1 July 2017, three months after the rest of the amendment, to give establishments time to arrange the facility.

Q: Can a woman be dismissed during maternity leave?

No, save for one narrow exception. Section 12 makes it unlawful for an employer to discharge or dismiss a woman during or on account of her absence in accordance with the Act, or to vary the conditions of her service to her disadvantage, and a dismissal notice timed to expire during her absence does not defeat the benefit. The exception is gross misconduct as prescribed: an employer may, by an order in writing, deprive a woman of maternity benefit or medical bonus on that ground, and the woman may appeal against the order to the prescribed authority within sixty days. A dismissal dressed up as something else, where the real reason is the maternity absence, falls within the prohibition.

Q: Who pays for maternity benefit in India?

In establishments outside the state insurance scheme, the employer of the woman pays, in full, at the rate of her average daily wage for the whole period of her absence. The amendment of 2017 doubled the period without changing the payer, which is the design feature from which most of the debate about the amendment follows. Where the Employees’ State Insurance scheme covers the establishment and the woman is insured, the benefit is paid from the insurance fund, which is financed by contributions from employers and employees, so the cost is pooled. For women outside covered establishments altogether, the state pays a cash maternity benefit under the food security statute through a government scheme.

Q: Is the Maternity Benefit Act still in force?

The Act of 1961 was repealed when the labour codes commenced. The Code on Social Security, 2020, Act 36 of 2020, consolidated nine welfare statutes including the Maternity Benefit Act, and when the four codes were brought into force together on 21 November 2025 the repeal took effect and the maternity entitlement continued in the maternity benefit chapter of the Code. The content carried across is the entitlement as the amendment of 2017 left it: 26 weeks for a woman with fewer than two surviving children, the graded twelve weeks beyond, the adopting and commissioning mother provisions and the crèche obligation. Claims arising under the Code are made under its chapter rather than under the repealed Act, with savings for action already taken.

Q: How many days must a woman have worked to qualify for maternity benefit?

Eighty days. Section 5(2) conditions the entitlement on the woman having actually worked in an establishment of the employer from whom she claims benefit for a period of not less than 80 days in the twelve months immediately preceding the date of her expected delivery. The count is generous in one respect: the explanation to the provision treats days on which she has been laid off and holidays with wages as days she has actually worked, so interruptions of a kind the establishment itself imposed do not break the qualifying period. The condition measures attachment to the establishment, and the courts have applied it as such, refusing to add further conditions based on the form of the woman’s engagement.

Q: How much maternity leave does a woman get for a third child?

Twelve weeks. The 26 week period inserted in 2017 is confined by a proviso to women with fewer than two surviving children; the maximum period for a woman having two or more than two surviving children is twelve weeks, of which not more than six weeks may precede the expected date of delivery. The count is of surviving children, so the question is how many of her children are living when the entitlement falls to be measured, not how many pregnancies she has had. For a woman expecting a third child whose two earlier children survive, the entitlement is the twelve weeks that the whole workforce received before 2017, together with the same dismissal protection, medical bonus and nursing breaks.

Q: What maternity benefit is payable after a miscarriage or medical termination of pregnancy?

Section 9 entitles a woman, on production of prescribed proof of a miscarriage or a medical termination of pregnancy, to leave with wages at the rate of maternity benefit for a period of six weeks immediately following the day of the event. The entitlement is separate from the main maternity benefit and does not depend on the stage of pregnancy reached. Where the event leads to illness, Section 10 adds a further entitlement: leave with wages at the maternity benefit rate for up to one month for illness arising out of pregnancy, delivery, premature birth of a child, miscarriage or medical termination, over and above the other periods of absence the Act allows.

Q: Does the Act give leave for a tubectomy operation or for illness arising out of pregnancy?

Yes, both. Section 9A entitles a woman to leave with wages at the rate of maternity benefit for a period of two weeks immediately following a tubectomy operation, on production of prescribed proof; the provision was added to the Act during the decades between its enactment and the 2017 amendment, in recognition that a sterilisation operation undertaken by the woman should not cost her wages. Section 10 covers illness: a woman suffering from illness arising out of pregnancy, delivery, premature birth of a child, miscarriage, medical termination of pregnancy or tubectomy is entitled, on proof, to additional leave with wages at the maternity benefit rate for a maximum period of one month.

Q: What is the medical bonus, and when is it payable?

The medical bonus under Section 8 is a lump sum payable by the employer to every woman entitled to maternity benefit, where the employer does not provide pre-natal confinement and post-natal care free of charge. The amount stood at one thousand rupees under the section as substituted by the amendment of 2008, which also empowered the Central Government to raise it by notification up to twenty thousand rupees; it was raised to two thousand five hundred rupees by notification in 2008 and to three thousand five hundred rupees by a further notification in 2011. It is a supplement to the wage replacement, not a substitute for it, and it is forfeitable, with the maternity benefit itself, only through the written order procedure for gross misconduct under Section 12.

Q: What nursing breaks must an employer allow after a mother returns to work?

Section 11 entitles every woman who returns to work after delivery to two breaks of the prescribed duration in the course of her daily work for nursing the child, in addition to the interval for rest allowed to her, until the child attains the age of fifteen months. The breaks are part of the working day and are not deducted from it. Where the establishment has fifty or more employees, the crèche duty in Section 11A sits alongside the nursing breaks: the woman must be allowed four visits a day to the crèche, including her rest interval, so a mother whose child is kept in the workplace crèche has a defined right of access across the day rather than a favour to request.

Q: Can a woman employed on contract or on daily wages claim maternity benefit?

Yes, if her establishment is covered and she satisfies the 80 day condition. The Act classifies by establishment and by days worked; it nowhere confines benefit to regular or permanent employees. The Supreme Court settled the point in Municipal Corporation of Delhi v. Female Workers (Muster Roll), (2000) 3 SCC 224, holding that nothing in the Act entitles only regular women employees to maternity benefit and excludes women engaged on casual basis or on muster roll on daily wages. The Court returned to the theme in Dr. Kavita Yadav v. Ministry of Health and Family Welfare in 2023, holding that the benefit is not coterminous with the tenure of a contract and runs its full period where the pregnancy occurred during the engagement. The Ministry of Labour and Employment’s clarifications on the 2017 amendment likewise confirmed its application to contractual employees.

Q: How does a woman claim maternity benefit, and when must it be paid?

Section 6 sets the procedure. The woman gives notice to her employer that she will absent herself under the Act, states the date from which she will be absent, and may nominate a person to receive the benefit on her behalf. The amount for the period preceding the expected delivery is paid in advance, on production of prescribed proof that she is pregnant; the amount for the period after delivery must be paid within 48 hours of production of prescribed proof that she has been delivered of a child. Notice is not a trap: the failure to give notice does not disentitle a woman who is otherwise entitled, and an Inspector may direct payment where the machinery has not been set in motion by the employer.

Q: What can a woman do if her employer withholds maternity benefit?

The Act gives her an inspectorate route rather than leaving her to a civil suit. An Inspector appointed under the Act may, on the woman’s application or of the Inspector’s own motion, direct the employer to pay benefit or medical bonus wrongfully withheld, and the direction is appealable by either side under Section 17 and the rules. A woman deprived of benefit by an order citing gross misconduct may appeal to the prescribed authority within sixty days under Section 12. Contraventions of the Act are offences punishable with imprisonment and fine under Section 21, although a prosecution can be instituted only by or with the previous sanction of an Inspector and within a year of the offence under Section 23. The general position here is information about the statutory machinery; an individual claim is fact specific and benefits from advice on its own record.

Q: Did the 26 week maternity law reduce women’s hiring?

The most cited evidence is a projection, not a measurement. A report by TeamLease Services published in June 2018, The Impact of Maternity Benefits on Business and Employment, surveyed 300 employers across ten sectors and projected a net loss of jobs for women of between eleven lakh and eighteen lakh in the financial year 2018-19 attributable to the amendment, with small and medium enterprises, start-ups and family run businesses reporting an intention to reduce intake, while large employers expected to hire as before. The same survey expected post maternity attrition to fall sharply as longer leave improved retention. No official statistical series isolates the amendment’s effect on women’s employment from wider labour market movements, so the defensible conclusion is about the incentive the design creates and the firms that reported feeling it, not a national count of jobs lost.

Q: Is there any statutory paternity leave for private sector employees in India?

No. Indian statute law contains no paternity or general parental leave entitlement for employees in the private sector; the maternity scheme assigns the whole of the paid absence around childbirth to the mother, which is one reason the cost of reproduction falls on the employment of women alone. Fathers employed by the Central Government are in a different position, because the Central Civil Services (Leave) Rules give a male government servant fifteen days of paternity leave, a service rule benefit rather than a statutory entitlement extending across employment. Private members’ bills proposing a statutory paternity benefit have been introduced in Parliament and have lapsed without enactment. The absence is the most frequently proposed correction to the maternity design, since leave available to both parents would spread the expected cost of a child across the workforce rather than concentrating it on women.

Q: Does the maternity law apply where the state insurance scheme covers the establishment?

The two statutes divide the field between them. Where an establishment is covered by the Employees’ State Insurance Act, 1948 and the woman is an insured employee, maternity benefit is payable under the insurance scheme, for the period and on the conditions that scheme prescribes, and the Maternity Benefit Act stands aside. The division is managed by two bridging provisions. Section 5A preserves continuity where a woman’s position moves between the two regimes, and Section 5B entitles a woman to claim under the Maternity Benefit Act where the insurance scheme ceases to apply to her, on the conditions those sections state, so that a change in coverage does not strip her of maternity protection altogether. Which regime governs a particular claim is therefore the first question to settle, before the length of leave is reached.