Almost every account of abortion law in India begins with a sentence the statute does not support: that the Medical Termination of Pregnancy Act gives women a right to end a pregnancy. It does not, and the distance between that sentence and the text explains nearly everything that confuses readers, patients and practitioners about how termination of pregnancy actually works in India. Causing a miscarriage is an offence under the general criminal law of India, and it was an offence long before 1971. The 1971 statute did not repeal that offence, did not create a competing entitlement, and did not transfer any decision to the pregnant woman. What it did was narrower and, in its consequences, larger: it built a carefully fenced exception inside the offence, available to one class of person, the registered medical practitioner, who acts within defined conditions. The legal question the statute poses is therefore never whether a woman has a right. It is whether a practitioner has a defence.

That single design choice, a shield drafted for the doctor rather than an entitlement drafted for the woman, is the key to the whole field. It explains why the grounds for termination are framed as medical opinions rather than as reasons a woman may give. It explains why the statute speaks of pregnancies that “may be terminated by” practitioners rather than of pregnancies a woman may choose to end. It explains why the characteristic failures of the system are refusal, referral and delay rather than prohibition, because a defence that a practitioner is unsure of is a defence the practitioner may decline to rely on, and the statute gives the woman no lever with which to compel reliance. And it explains why, half a century after enactment, the litigation in this field consists so often of women petitioning constitutional courts for permission the statute itself contains no machinery to grant. This article sets out the statute as it was designed and as it operates: the offence it displaces, the committee whose public health reasoning produced it, the conditions in Section 3, the consent rules, the approved place machinery, the paperwork and secrecy regime, the emergency exception, the penalties that surround the shield, the constitutional reading the courts have layered over the text, and the neighbouring statutes whose enforcement chills the very conduct this Act protects. The gestational limits and the changes made in 2021 are stated here only in outline, because that ground belongs to the companion treatment of the amendment; the pillar’s subject is the architecture of the 1971 Act itself. The wider statutory landscape of which this Act forms part is mapped in the series guide to health legislation in India, and this article stays inside its own statute.

The offence the Act had to displace

The starting point is not 1971 but 1860. The Indian Penal Code, in a chapter grouping the causing of miscarriage with injuries to unborn children, the exposure of infants and the concealment of births, made the voluntary causing of a miscarriage a crime in itself. Section 312 punished whoever voluntarily causes a woman with child to miscarry, with imprisonment extending to three years, or fine, or both, rising to seven years and a mandatory liability to fine where the woman was quick with child, the older expression for the stage at which foetal movement is felt. The section carried two features that matter for everything that followed. The first is its single exception: no offence was committed where the miscarriage was caused in good faith for the purpose of saving the life of the woman. The second is its explanation, which provided expressly that a woman who causes herself to miscarry falls within the section, so the woman herself was a potential accused, not merely the person who assisted her.

Around that core the Code graded the surrounding conduct. Section 313 took the same act done without the consent of the woman and raised the punishment to imprisonment for life, or imprisonment extending to ten years, with fine. Section 314 addressed the attempt that kills: where an act done with intent to cause a miscarriage caused the death of the woman, the punishment extended to ten years with fine, and where the act was done without her consent it could extend to imprisonment for life. Section 315 covered an act done before the birth of a child with intent to prevent the child being born alive or to cause it to die after birth, where the act had that result, punishable with imprisonment extending to ten years. Section 316 treated the causing of death of a quick unborn child by an act amounting to culpable homicide as culpable homicide itself, again punishable with imprisonment extending to ten years. The architecture was complete and severe: the act was criminal at every point, the woman’s consent was no answer except through the single life-saving exception, and the penalties rose with the gravity of the outcome rather than with any assessment of the woman’s circumstances.

When the Penal Code was replaced by the Bharatiya Nyaya Sanhita in the criminal law revision of 2023, these provisions were carried across with their substance intact and their numbers changed. Sections 312 to 316 of the old Code correspond to Sections 88 to 92 of the Sanhita, renumbered and gathered into the chapter on offences against women and children, as the series comparison of the Penal Code and its successor code sets out in detail. The renumbering changed nothing about the structure this article describes: the offence remains the rule, and the 1971 statute remains the exception carved into it. The exception’s own text still speaks the language of the older Code, because Section 3 of the 1971 Act opens with the words “notwithstanding anything contained in the Indian Penal Code”, and the shield it raises is a shield against guilt under that Code “or under any other law for the time being in force”.

Two consequences of the offence-first structure deserve emphasis before the exception is examined, because they are the two most commonly missed. The first is that unlawfulness is the default. A termination that falls outside the 1971 Act’s conditions is not an unregulated medical act; it is the Section 312 offence, or its Sanhita equivalent, committed by whoever performed it, and the woman who procures her own miscarriage outside the statutory scheme remains exposed by the explanation to that section. The second is that the good faith exception inside Section 312 was the only shelter the criminal law itself offered before 1971, supplemented by the general exception in the Code protecting acts done in good faith for a person’s benefit with consent, the ground on which bona fide surgery stands. Those shelters protected a doctor who acted to save a life in a crisis. They did not protect elective practice, they did not protect the contraceptive failure case, they did not protect the rape survivor, and they gave a practitioner no certainty in advance, because good faith would be tested afterwards, in a prosecution, on facts the practitioner could not fully control. The demand for a statute came from exactly that gap between the medicine practitioners were being asked to perform and the law under which they would be judged for performing it.

The table below is the article’s findable artifact, and it states the architecture in one object: the penal provision on one side, the condition under the 1971 Act that displaces it on the other, and the person who carries the risk when a condition is not met.

The penal provision	The exception under the 1971 Act	Who bears the risk if a condition fails
Section 312 of the Penal Code (Section 88 of the Bharatiya Nyaya Sanhita): voluntarily causing a miscarriage, punishable up to three years, or seven where the woman is quick with child, with an exception only for good faith action to save the woman’s life	Section 3 of the 1971 Act: a registered medical practitioner who terminates a pregnancy in accordance with the Act is guilty of no offence under the Code or any other law	The practitioner, who must be able to show every condition was met
Section 313 (Section 89 of the Sanhita): causing a miscarriage without the woman’s consent, punishable with imprisonment for life or up to ten years	None. Section 3(4) makes the pregnant woman’s own consent the rule, and the Act protects no termination carried out without it	The person who acts without consent, for whom the Act provides no shelter at all
Section 314 (Section 90 of the Sanhita): death of the woman caused by an act done with intent to cause miscarriage, punishable up to ten years, or life where done without her consent	Section 8 of the 1971 Act bars proceedings for damage caused by anything done in good faith under the Act, but the criminal exposure for a death turns on whether the termination itself stood within the Act	The practitioner, whose protection depends on good faith and on the conditions having been satisfied
Section 315 (Section 91 of the Sanhita): an act intended to prevent a child being born alive or to cause it to die after birth, punishable up to ten years	None within the Act’s scheme, which authorises termination of a pregnancy and not any act directed at a child born or being born	Whoever performs the act, entirely outside the statutory shelter
Section 316 (Section 92 of the Sanhita): causing the death of a quick unborn child by an act amounting to culpable homicide, punishable up to ten years	None as such; the Act’s protection attaches to terminations performed within its conditions by a registered medical practitioner at an approved place	The practitioner, and, after the 2002 amendment, the owner of an unapproved place as well
The Act’s own perimeter: termination by a person who is not a registered medical practitioner, or at a place that is not approved	Section 5, as substituted in 2002, makes each of these an offence punishable with rigorous imprisonment of not less than two years and extending to seven years	The unqualified person, the person who terminates at an unapproved place, and the owner of that place

Read as a whole, the table makes the pillar’s central point visible. The exception never abolishes the offence; it routes around it, for one actor, on conditions. Every condition in the statute is therefore a load-bearing part of a criminal defence, and every failure of a condition returns the parties to the penal provisions in the first column. That is why the conditions are drafted with the precision of a penal statute rather than the generosity of a welfare statute, and it is why the sections that follow treat them one by one.

What existed before, and why it failed

Before 1971 the formal law was the Penal Code position described above, and the practical position was that terminations happened anyway, outside the law, in numbers nobody disputed and under conditions that killed and injured women in numbers the medical profession could see from inside its own hospitals. The gynaecological wards of the period received the consequences: women admitted with sepsis, haemorrhage and injury following procedures performed by untrained hands, in a legal climate in which neither the woman nor anyone treating her complications could speak freely about what had occurred. The criminal law did not stop the practice. It determined only where the practice happened, who performed it, and how dangerous it was. That distinction, between a law that suppresses conduct and a law that merely drives it into unsafe channels, is the analytical foundation on which the statute was built, and it was assembled not by a rights movement but by a committee of officials and doctors working inside the health ministry.

What was the law on abortion in India before the 1971 Act?

Before 1971, abortion in India was governed by the Indian Penal Code. Causing a miscarriage was an offence under Section 312, punishable with up to three years of imprisonment, or seven where the woman was quick with child, unless done in good faith to save her life. A woman who caused her own miscarriage was within the section.

The penal provisions themselves were an inheritance. The Code’s drafters worked within an English legal tradition that had criminalised abortion through the nineteenth century, and the structure they produced, a graded set of offences organised around consent, quickening and consequences for the woman, carried that tradition into Indian law without any Indian deliberation on the subject at all. No legislature in India had ever debated the question on its merits before the 1960s, and no Indian court had been asked to. The law on the subject was simply part of the received criminal code, enforced unevenly, discussed rarely, and sustained by the assumption that the practice it forbade was marginal. By the time the health ministry’s own institutions were reporting the wards full of its casualties, that assumption was unsustainable, and the question arrived on the government’s desk as a problem of hospitals and mortality statistics rather than as a question of rights.

The Government of India constituted the committee whose report became the statute in September 1964, on the recommendation of the Central Family Planning Board. The Committee to Examine the Question of Legalisation of Abortion, as it was formally styled, was chaired by Shantilal H. Shah, then the Health Minister of Maharashtra, and is known in the literature simply as the Shah Committee. It studied the legal, medical and social aspects of the question, took evidence, and submitted its report in December 1966. The report’s reasoning is worth stating precisely, because the statute follows from it. The committee proceeded on public health grounds: the existing law, it found, was failing to prevent terminations while ensuring that a large share of them were performed dangerously, and the cost was paid in maternal deaths and in lasting injury to health. It noted that the great majority of women seeking termination were married women, for whom concealment was not the object and for whom a lawful, medically supervised route would substitute safety for secrecy. It recommended that the law be liberalised so that termination could be performed by qualified medical practitioners on defined medical, humanitarian and social grounds. And, in a point often misreported, the committee’s case rested on health and humanitarian reasoning rather than on population control as its object, even though the exercise sat within a government preoccupied with family planning and even though one of the grounds it proposed, contraceptive failure, plainly carried a family planning logic within it.

The drafting consequence followed directly from the reasoning. A reform justified as a public health measure, designed to move an existing practice from unsafe hands to qualified ones, is naturally drafted as a regulation of practitioners: it defines who may perform the procedure, on what medical assessment, in what premises, with what records. A reform justified as the recognition of a woman’s right would be drafted the other way round, as an entitlement with duties laid on the state. India got the first kind of statute because India made the first kind of argument. The committee’s method deserves a paragraph, because it set the pattern the legislature followed. It worked from the service outward: it collected the evidence of what was happening in hospitals and in the informal market, examined the legal position in India and the reforming legislation of other countries, and framed its grounds by generalising the cases that the existing law was visibly failing, the endangered woman, the raped woman, the woman facing a damaged child, the woman whose contraceptive trust had been betrayed. It did not conduct, and did not pretend to conduct, a philosophical inquiry into the status of the foetus, and the absence of that inquiry from its report is the reason the statute it produced could be carried through the politics of 1971. A bill about a woman’s right would have divided the Parliament of the day on lines nobody could have managed. A bill about mortality wards, medical protection and certified grounds divided it hardly at all. The lesson is inscribed in the text: the Act’s silences are not oversights but the prices of its passage, and later reformers who propose to fill them are proposing, whether they say so or not, to reopen the negotiation the committee declined to start. The committee had recommended a separate statute rather than a simple relaxation of the Penal Code, and the bill that emerged, the Medical Termination of Pregnancy Bill, was introduced in the Rajya Sabha on 17 November 1969, examined by a Joint Committee of the two Houses which took evidence through 1970, and passed by Parliament in 1971. The President’s assent was given on 10 August 1971, and the statute is accordingly Act 34 of 1971. It was brought into force on 1 April 1972. As enacted, it extended to the whole of India except the State of Jammu and Kashmir, an exclusion that stood in the text until the reorganisation of that state in 2019, when Parliament extended the central statute book, this Act included, to the successor Union territories.

The passage record also discloses what the reform was not. It was not a response to a court judgment, since no court had questioned the Penal Code position. It was not the product of street mobilisation on any scale comparable to the movements that produced other social legislation of the period. And it was not framed, in the debates or in the committee report, in the vocabulary of autonomy or choice that dominates the subject in other jurisdictions. The imported framing that describes the Indian statute as though it were a judicially recognised right, borrowed from the constitutional history of the United States, maps onto nothing in this record. The Indian statute is a legislative instrument of a particular and unusual kind: early by world standards, broad in its stated grounds, and yet built, from its first section to its last, around the protection of the practitioner rather than the entitlement of the woman. Grasping that the liberalisation and the doctor-centred design are the same fact, not two facts in tension, is what the committee’s report makes possible.

A shield, not an entitlement: how the Act is addressed

The operative section of the Act is Section 3, and its first sub-section performs the entire manoeuvre. It provides that, notwithstanding anything contained in the Indian Penal Code, a registered medical practitioner shall not be guilty of any offence under that Code or under any other law for the time being in force, if any pregnancy is terminated by him in accordance with the provisions of the Act. Every element of that sentence is doing work. The non obstante opening subordinates the penal provisions without amending them, leaving the offence intact for everyone outside the exception. The subject of the sentence is the practitioner, and the benefit conferred is framed negatively, as an absence of guilt, not positively, as a power or a right. The condition is total: the termination must be in accordance with the provisions of the Act, so the protection is exactly as wide as the conditions and no wider. A termination that satisfies nine of ten conditions is not nine-tenths protected. It is a termination outside the Act, judged under the first column of the table above.

Does the Act give a woman a legal right to a termination?

No. The Act confers no entitlement on a woman and creates no duty in any practitioner or hospital to provide a termination. Section 3 provides that a registered practitioner who terminates a pregnancy in accordance with the Act commits no offence. The woman holds a consent, not a claim: she can refuse, and her refusal binds, but she cannot compel.

The definitions in Section 2 confirm the design. The Act defines a guardian, a minor, and a registered medical practitioner, and in its original text it defined a lunatic by reference to the Indian Lunacy Act of 1912, a term the amendment of 2002 replaced with that of a mentally ill person, defined as a person in need of treatment by reason of any mental disorder other than mental retardation. What the definitions conspicuously do is locate every legal actor except the woman herself. The practitioner is defined by qualification and registration. The guardian is defined for the two classes of case where the woman’s own consent is not the operative consent. The woman at the centre of the procedure is not defined at all, because the statute needs nothing from her except, in the ordinary case, her consent. She is the occasion of the legal act, not its subject.

The registered medical practitioner definition deserves close reading, because it is the gate through which the whole exception passes. A practitioner qualifies only if three things coincide: possession of a recognised medical qualification within the meaning of the Indian Medical Council Act of 1956, entry of the practitioner’s name in a State Medical Register, and such experience or training in gynaecology and obstetrics as the rules prescribe. The first two limbs establish that the person is a doctor in the eyes of the medical registration system. The third limb, left by the Act to subordinate legislation, is where the Act decides what kind of doctor may do this work, and the rules made under it prescribe the routes: a practitioner registered before the Act commenced may rely on three years of experience in the practice of gynaecology and obstetrics; a practitioner registered later may qualify through six months of house surgency in gynaecology and obstetrics, through a year of hospital experience in the practice of obstetrics and gynaecology, through assisting a registered practitioner in twenty five cases of medical termination of pregnancy of which at least five were performed independently in an approved institution, or through a postgraduate degree or diploma in the speciality. The assisted-cases route, the rules specify, equips the practitioner for first trimester terminations only. The structure reveals the statute’s instinct throughout: competence is certified in advance, in categories, by reference to supervised experience, because the practitioner’s defence will one day be audited against exactly these categories.

Section 8 completes the protective scheme on the civil side. It provides that no suit or other legal proceeding shall lie against a registered medical practitioner for any damage caused or likely to be caused by anything done or intended to be done in good faith under the Act. Between Section 3(1) on criminal liability and Section 8 on civil proceedings, the statute builds the practitioner a two-walled shelter. But both walls are built of the same material, good faith and compliance, and neither wall protects a practitioner who steps outside the conditions, or who cannot afterwards demonstrate that the conditions were met. The paperwork requirements examined below are therefore not bureaucratic decoration. They are the evidence by which the shelter is proved, and a practitioner who treats them as decoration has misunderstood what the statute is for.

One further feature of the drafting should be fixed before the conditions are examined, because it recurs in the litigation. The statute nowhere obliges any practitioner to form the opinion the conditions require, and nowhere obliges any institution to make the procedure available. A practitioner who declines, whether from conscience, from caution, from lack of facilities or from unwillingness to accept the residual risk, breaks no provision of the Act. There is no offence of refusal, no duty to refer written into the statute, and no authority to which a woman can appeal a refusal within the Act’s own machinery. Whatever remedies refusal may attract lie elsewhere: in professional standards, in the constitutional litigation discussed below, or in politics. Within the four corners of the 1971 Act, the practitioner’s willingness is the hinge on which access turns, and it is a hinge the statute deliberately leaves free.

The conditions in Section 3: grounds, opinions and thresholds

Section 3(2) states when a pregnancy may be terminated by a registered medical practitioner. Where the length of the pregnancy does not exceed twelve weeks, the opinion of one practitioner suffices. Where the length exceeds twelve weeks but does not exceed twenty weeks, the concurring opinions of not less than two practitioners are required. In either case the opinion must be formed in good faith that one of two things is true: that the continuance of the pregnancy would involve a risk to the life of the pregnant woman or of grave injury to her physical or mental health, or that there is a substantial risk that if the child were born, it would suffer from such physical or mental abnormalities as to be seriously handicapped. Those are the grounds as enacted, and Article 136 owns the account of how the limits attached to them were recast in 2021; the pillar states the enacted scheme because the scheme’s logic, rather than any particular week’s number, is the thing to understand.

Three features of the ground structure repay attention. The first is that the grounds are stated as the content of a medical opinion. The statute does not say that a pregnancy may be terminated where a woman’s health is at risk; it says a pregnancy may be terminated where a practitioner is of opinion, formed in good faith, that continuance would involve such a risk. The legal object is the opinion, not the underlying fact. A court reviewing a termination therefore asks whether the opinion was genuinely formed within the statutory language, not whether the court would have assessed the risk the same way. This is the doctrinal root of the practitioner’s centrality, and it is also the source of a real protection for honest practice, since a good faith opinion honestly recorded is difficult to assail. The second feature is the breadth of the health ground read with its statutory aids. Section 3(3) provides that in determining whether continuance would involve the risk of injury to health, account may be taken of the pregnant woman’s actual or reasonably foreseeable environment. Health, in other words, is not confined to the clinical presentation; the woman’s circumstances, present and foreseeable, are admissible material for the opinion. The third feature is the pair of explanations that convert two recurring fact patterns into presumptions. Explanation I provides that where a pregnancy is alleged by the pregnant woman to have been caused by rape, the anguish caused by the pregnancy shall be presumed to constitute a grave injury to her mental health. Explanation II, as enacted, provided that where a pregnancy occurs as a result of failure of any device or method used by a married woman or her husband for the purpose of limiting the number of children, the anguish caused by the unwanted pregnancy may be presumed to constitute a grave injury to her mental health. The first explanation removes the need for a practitioner to conduct an inquiry into the truth of the allegation; the allegation itself triggers the presumption. The second, in its original form, reveals the era’s assumptions in a single phrase, since it extended the presumption to a married woman and her husband and said nothing of anyone else, an exclusion whose correction half a century later belongs to the amendment’s story rather than this one.

Why does the Act use twelve weeks and twenty weeks as its two thresholds?

The two thresholds organised the period’s medical practice. Twelve weeks marked the point beyond which termination ceased to be a simple procedure and intervention grew riskier, so the Act required a second concurring opinion and fuller practitioner experience. Twenty weeks, the midpoint of pregnancy and near foetal viability as then understood, marked the outer boundary the legislature would set.

The thresholds also allocate institutional effort. Within twelve weeks, one practitioner, duly qualified, may form the opinion and act on it, and the rules allow the narrower training route to suffice, which is what makes early termination capable of being a routine service in a primary setting. Between twelve and twenty weeks, two practitioners must concur, the fuller experience requirements apply, and the practical centre of gravity shifts to institutions. Beyond twenty weeks, the Act as enacted simply stops. There is no third tier, no committee, no board, no residual discretion vested in anyone to authorise a termination on the grounds alone. The only provision of the Act that operates beyond twenty weeks is the emergency provision in Section 5, and it operates on a different ground altogether, the immediate necessity of saving the woman’s life. The hard edge at twenty weeks is therefore not an oversight or an ambiguity. It is the legislature’s chosen boundary, drawn where the committee-era balance between the woman’s health and the state’s interest in the developing pregnancy was struck, and every later controversy about the boundary, from the Bombay litigation of 2008 to the amendment of 2021, is a controversy about moving a line Parliament drew deliberately.

The foetal ground in Section 3(2) requires separate handling, because it is the ground on which the statute is most often misdescribed. The text does not authorise termination for abnormality as such. It requires a substantial risk that the child, if born, would suffer from such physical or mental abnormalities as to be seriously handicapped. Three qualifiers do the work: the risk must be substantial, the abnormality must be of the physical or mental kind described, and the consequence must reach the gravity of serious handicap. A diagnosed condition that is treatable, or minor, or uncertain in its effects does not answer the description merely because it is a diagnosis. The practitioner forming the opinion must connect the clinical finding to the statutory consequence, and that connection, recorded in the certification, is what the protection rests on. It is also the feature of the ground that draws the sharpest ethical objection, examined in the disagreement section below, since a threshold expressed as serious handicap embodies a judgment about which lives are burdens that disability scholarship and disability activism have never accepted.

What the grounds do not contain is equally instructive. There is no ground of economic hardship as such, though the environment provision lets a woman’s circumstances enter the medical assessment of risk to health. There is no ground of the woman’s request as such, at any stage. There is no ground addressed to the pregnancy of a minor as a category; a minor’s pregnancy is handled through the ordinary grounds, with the consent machinery altered as the next section describes. And there is no ground of sex selection: the Act nowhere permits the sex of the foetus to figure in a lawful opinion, and the separate statute against sex determination, discussed below, makes the point with criminal force. The Act’s grounds are medical in form even where their reach is social in effect, and the distance between the form and the effect is where most of the argument about this statute lives.

Good faith, environment and the ground that disappeared

Two elements of Section 3 do quiet work that repays separate attention: the good faith standard, and the environment clause. Good faith is the statute’s universal solvent. The opinion under Section 3(2) must be formed in good faith; the emergency judgment under Section 5 must be formed in good faith; the civil immunity in Section 8 attaches to things done in good faith. In the general law of India, good faith in this setting imports honesty of purpose rather than correctness of conclusion, and its presence in the statute is a deliberate allocation of the benefit of the doubt. A practitioner who assesses a risk and gets the assessment wrong, in a case another practitioner would have assessed differently, is not thereby outside the Act, because the Act asks whether the opinion was genuinely formed on the statutory ground, not whether it was the best opinion available. Conversely, a practitioner who certifies a ground that was never genuinely assessed, who signs the form as a formality for a termination arranged on other reasons entirely, is outside the Act however impeccable the paperwork looks, because the substance the paperwork certifies never existed. The distinction matters in practice chiefly at the extremes. Prosecutions of practitioners acting within the scheme are rare, and the reported controversies tend to involve terminations with no genuine opinion behind them at all, which is to say terminations that were never candidates for the Act’s protection. The good faith standard is therefore best understood as the statute’s way of policing the boundary between judgment and pretext, leaving the first alone and abandoning the second.

The environment clause in Section 3(3) is the residue of a larger ambition. The Shah Committee’s proposed grounds were conventionally grouped as medical, eugenic, humanitarian and social, and the social ground contemplated termination where the woman’s socio-economic circumstances made the pregnancy and the child an injury the law should recognise. The enacted statute contains no free standing social ground. What it contains instead is the direction that, in determining whether continuance of a pregnancy would involve the risk of injury to health that the first ground describes, account may be taken of the pregnant woman’s actual or reasonably foreseeable environment. The drafting is a compromise preserved in amber. The social reality the committee wanted to reach, the household that cannot feed another child, the woman whose circumstances make the pregnancy a catastrophe for her health in any realistic sense, is admissible, but only as material for a medical judgment about health, and only through the practitioner’s opinion. The foreseeable environment limb extends the same logic forward: a practitioner may weigh not merely the woman’s present condition but the conditions into which the pregnancy will run and the child would be born. Read together with the breadth of grave injury to mental health and the two presumptions, the clause gives the Indian grounds a practical reach comparable to the social ground statutes of other jurisdictions, while preserving the form that the decision is medical. Whether that preservation is a virtue, keeping the decision clinical and the practitioner protected, or a fiction, disguising a social judgment in medical dress and charging the disguise to the woman’s dependence on the gatekeeper, is one of the oldest arguments about the Act, and it is an argument about the Act’s honesty rather than its content. The pillar records it without resolving it, because the resolution depends on the moral question reserved for the disagreement section below.

One more feature of the grounds deserves record, because it structures a great deal of litigation and clinical difficulty: the grounds are asymmetrical in time. The health ground speaks of risk from continuance, which can be assessed at any presentation. The foetal ground depends on diagnosis, and the diagnostic calendar does not respect the statutory one. The anomaly scan through which many serious abnormalities are detected is performed in the middle of pregnancy, and some conditions, including cardiac abnormalities of the kind in the Bombay case, can be characterised reliably only at or beyond the point where the statutory limit falls. A ground that can be known with confidence only after the period for acting on it has closed is a ground with a built in failure mode, and the failure mode is not accidental in the sense of being unforeseen: it is the predictable consequence of fixing a limit by weeks in a field where knowledge arrives on its own schedule. The 2021 amendment’s Medical Board mechanism was Parliament’s eventual answer to that failure mode, and the permission litigation of the intervening decades was the courts’ answer. The pillar’s concern is the design point: the mismatch between the diagnostic calendar and the statutory calendar is a property of the 1971 scheme itself, visible from its text, and any assessment of the scheme’s working that omits it will misdiagnose the litigation as an aberration rather than as an output.

Consent: the one decision the Act does give the woman

If the grounds belong to the practitioner, the consent belongs to the woman, and Section 3(4) draws that allocation in two clauses. Clause (b) states the general rule: save as otherwise provided, no pregnancy shall be terminated except with the consent of the pregnant woman. Clause (a) states the exceptions: no pregnancy of a woman who has not attained the age of eighteen years, or who having attained eighteen years is a mentally ill person within the Act’s definition, shall be terminated except with the consent in writing of her guardian. The general rule is unqualified by any other person’s interest. The statute does not require the consent of a husband. It does not require the consent of parents in the case of an adult woman. It does not require the consent of the family, the in-laws, an employer, or any authority. A married woman’s termination requires her consent and the practitioner’s opinion, and no third signature appears anywhere in the scheme.

The persistence of the contrary belief, that a husband’s consent is needed, is one of the most consequential misunderstandings in Indian health law, because it operates at the clinic door. It draws strength from social practice, from institutions that ask for a spouse’s signature as a matter of internal routine, and from the general habit of treating medical decisions within a family as family decisions. In law it has no foundation in the Act, and the point has been tested in the adjacent litigation: in the Bihar case that reached the Supreme Court in 2017, a woman’s access was obstructed by institutional demands for third party consent she was never legally required to produce, the delay carried her beyond the statutory limit, and the Court condemned the obstruction and awarded compensation. The statute’s allocation is not accidental. The committee-era legislature, for all its practitioner-centred design, treated consent as the woman’s own act because the procedure is performed on her body and the penal consequences of its absence are the gravest in the group: a termination carried out without the woman’s consent is not a defective termination under the Act at all, but the Section 313 offence, the causing of a miscarriage without consent, punishable with imprisonment for life. Consent is thus the one place where the Act’s logic and the woman’s autonomy coincide completely, and the statute protects the coincidence with its heaviest penalty.

The exceptions in clause (a) require care, because they are routinely stated too broadly. The guardian’s written consent is required where the woman is under eighteen, and where she is an adult who is a mentally ill person. The second category is a defined and narrow one. Following the amendment of 2002, a mentally ill person for this purpose is a person in need of treatment by reason of a mental disorder other than mental retardation. Intellectual disability, described in the statute’s older vocabulary as mental retardation, is therefore outside the guardian consent category altogether. The Supreme Court gave that distinction constitutional weight in Suchita Srivastava v. Chandigarh Administration in 2009. A woman with mild mental retardation, living in a government institution in Chandigarh, became pregnant following rape. The Punjab and Haryana High Court directed that the pregnancy be terminated, reasoning from her circumstances and her supposed inability to care for a child. The Supreme Court reversed the direction. The woman’s right to make reproductive choices, the Court held, is a dimension of personal liberty under Article 21 of the Constitution; mental retardation is not mental illness; the guardian consent provision did not displace her own choice; and her expressed wish to continue the pregnancy was to be respected notwithstanding official assessments of her capacity to parent. The case is examined again in the courts section below, but its consent holding belongs here: the exception in Section 3(4)(a) is confined to mental illness properly so called, and no administrator, guardian or court may enlarge it into a general power over the reproductive decisions of women with intellectual disabilities.

For minors, the guardian consent rule interacts with the rest of the legal system in ways that generate genuine difficulty, and the pillar should state the difficulty plainly. A pregnancy in a girl under eighteen is, under the Protection of Children from Sexual Offences Act of 2012, evidence of an offence which that statute obliges specified persons, including medical practitioners, to report. The practitioner treating a pregnant minor therefore stands at the intersection of two statutes pulling in opposite directions: the 1971 Act, which routes the termination decision through a guardian’s written consent and surrounds the record with secrecy, and the 2012 Act, which treats the pregnancy itself as a matter for the criminal process and demands disclosure. The tension is not theoretical; it shapes whether minors are brought to lawful services at all, since a family that fears the reporting machinery may avoid the approved place altogether. The Supreme Court addressed the sharpest edge of this collision in its 2022 decision on the statute and its rules, which is part of the amendment’s story and is treated there. What belongs to the pillar is the structural observation: the consent exception for minors was drafted in 1971 for a legal world in which no mandatory reporting regime existed, and the strain visible in the case law is the strain of that older allocation meeting a newer one.

The correlative proposition, that no one else can forbid what the woman has consented to, has had to be defended from the other direction. Husbands and families who oppose a termination have occasionally sought the courts’ help to stop one, and the consistent answer, grounded in the same clause, is that the statute knows no veto: the practitioner who has the woman’s consent and has formed the opinion needs nothing more, and a third party’s objection, however deeply felt, is not a legal event under the Act. The symmetry is worth dwelling on, because it shows the clause working as a rule of allocation rather than as a courtesy. Consent in this statute decides who is the decision maker of record in both directions. It excludes the family’s affirmative demand, and it excludes the family’s negative demand, and the institution that honours the first exclusion while breaching the second, turning away the unsupported woman while accepting the accompanied one, has understood the custom of its corridor better than the text of its statute.

Place, paperwork and secrecy: the machinery of the exception

A termination under the Act must happen in the right place, generate the right record, and keep that record secret, and each of the three requirements is a condition of the practitioner’s protection. Section 4 as enacted provided that no termination of pregnancy shall be made in accordance with the Act at any place other than a hospital established or maintained by Government, or a place approved for the purpose of the Act by Government. The amendment of 2002 substituted a new Section 4 which kept the government hospital limb and widened the approval limb: a place may be approved by Government or by a District Level Committee constituted by that Government, chaired by the Chief Medical Officer or the District Health Officer and consisting of not less than three and not more than five members. The rules attach facility standards to approval, distinguishing what a place must have to undertake terminations up to twelve weeks from the fuller surgical and anaesthetic support required for the later band. The decentralisation of 2002 was an access measure dressed as an administrative one: approval at state capitals had been a bottleneck, and moving the power to the district was meant to multiply the places where the exception could lawfully be exercised.

The paperwork begins with the opinion itself. Section 7 empowers state governments to make regulations requiring the opinion referred to in Section 3 to be certified by the practitioner or practitioners concerned, requiring a practitioner who terminates a pregnancy to give intimation of the termination and such other information as the regulations specify, and prohibiting the disclosure of intimations and information except to the persons and for the purposes the regulations allow. Contravention of the regulations is punishable with a fine extending to one thousand rupees, a small sum that understates the provision’s real significance, which is not punitive but evidentiary and protective. Under the regulations in force, the certified opinion is recorded on the prescribed form, the woman’s consent is recorded on the prescribed consent form, intimations travel to the Chief Medical Officer of the state, and the head of the hospital or the owner of the approved place maintains an admission register, in the prescribed form, in which the details of each woman admitted for termination are entered serially. The register is retained for five years from the end of the calendar year to which it relates. It is, by the express words of the regulations, a secret document: the information in it as to the name and other particulars of the pregnant woman shall not be disclosed to any person, it is kept in safe custody, and it is not open to inspection except under the authority of law. Where a termination is performed under the emergency provision, the documentation travels to the Chief Medical Officer in a sealed envelope marked secret. The design problem the register solves, a record strong enough to prove legality yet sealed against the harm its disclosure would do, is one other professions meet in their own material, a parallel traced in the case of journalists and the records they keep on their sources in this study of source protection practice.

What did the 2002 amendment actually change?

The Medical Termination of Pregnancy (Amendment) Act of 2002, brought into force in 2003, modernised the statute’s language and hardened its perimeter. It replaced lunatic with mentally ill person, moved approval of places to District Level Committees chaired by the Chief Medical Officer, and attached imprisonment of two to seven years to terminations by unqualified persons and at unapproved places.

Confidentiality in the 1971 scheme is therefore real but indirect, and the pillar should describe it at its true strength rather than in the inflated form in which it is often reported. The Act itself contains no free standing confidentiality section in its original text; what it contains is the regulation making power in Section 7(1)(c), under which disclosure of the intimations and information generated by the process may be prohibited except as specified, and the regulations made under the scheme have used that power to seal the register and the record. The protection is strongest exactly where the process is most formal, and weakest where a woman’s identity becomes known through channels the regulations never governed: the referral conversation, the family negotiation, the police report generated by a neighbouring statute, the court petition in which her case is argued. The amendment of 2021 later wrote an express confidentiality provision into the Act itself, with a penal sanction for disclosure, and the detail of that provision belongs to the treatment of the amendment. The pillar’s point is the older position: for the Act’s first five decades, confidentiality was an artefact of its paperwork rules, carefully constructed around the admission register, and a practitioner or institution that treated identity as gossip to be shared was breaking the regulations that made the record, even though the Act’s own text had not yet said so in terms.

The machinery section also holds the explanation for a paradox practitioners report consistently: a statute can be complied with only by generating records, yet the records are what practitioners fear. The certification proves the defence, but it also creates the file an inspector, a prosecutor or a hostile complainant will later read. The statute’s answer to the paradox is the secrecy regime just described, which attempts to make the record complete for the practitioner’s protection and invisible for everyone else’s purposes. Where that attempt succeeds, the paperwork is a shield. Where a practitioner doubts it will succeed, because inspection practice under the neighbouring sex selection statute has taught the profession that files invite trouble, the paperwork becomes a reason to refuse the case. That interaction is the subject of the section on neighbouring statutes below, and it is one of the places where the Act’s design and its environment work against each other.

The second layer: rules, forms and the life of an approved place

The Act is eight sections long, and nobody can operate it from its eight sections alone. Beneath the statute sits a second layer of subordinate legislation that carries much of the working scheme: rules made by the Central Government under Section 6, which prescribe among other things the experience and training a practitioner must hold, and regulations made by state governments under Section 7, which govern certification, intimation and the custody of records, together with the prescribed forms through which the whole process is documented. The practitioner encounters the Act almost entirely through this layer. The opinion is formed on the prescribed certification form. The consent is taken on the prescribed consent form. The termination is reported through the prescribed intimations, including periodic returns from institutions to the Chief Medical Officer, so that the state’s picture of the service is built from the same paperwork that protects the practitioner. The training routes in Rule 4, the facility standards attached to approval, the constitution and procedure of the District Level Committees: all of it sits in the second layer, amendable by notification without returning to Parliament, which is why the working law of termination in India can shift in detail, on training, on approval, on medical methods, while the eight sections above it stand still for decades.

One consequence of the two level structure is that the working law is not uniform across India. The Act is central and uniform; the rules on training and approval are central and uniform; but the regulations on certification, intimation and records are made by each state government under Section 7, and the forms, the intimation intervals and the administrative routing therefore differ from state to state, as does the energy of the inspecting machinery. A multi state provider works to several paperwork regimes under one statute, and a practitioner who moves states carries the Act unchanged and leaves the regulations behind. The variation is easy to miss in commentary, which speaks of the MTP framework as a single thing, and it matters in enforcement, because the technical defaults on which prosecutions and disciplinary proceedings turn, a late intimation, an unsigned certification, a register defect, are defaults against a state text. It also matters to reform: several of the scheme’s working improvements over the decades, in forms, in reporting, in the streamlining of approvals, were achieved at the regulation level by states and at the rule level by the Centre, below the threshold of parliamentary attention, which is one reason the eight sections above them could remain untouched for so long. The statute is the skeleton the country argues about; the second layer is where the service is actually administered, and the two are amended on different clocks by different hands.

The life of an approved place illustrates the layer at work. Approval is not a permanent status conferred once; it is a continuing conformity to prescribed standards, and the scheme provides for inspection of approved places and for the suspension or cancellation of approval where the standards are not maintained, so that the place limb of Section 4 is policed administratively in the intervals between criminal cases. The approving authority examines the facilities against the band for which approval is sought: a place approved for early terminations requires the equipment and support appropriate to that work, while approval extending to the later band demands the fuller surgical, anaesthetic and emergency arrangements, including access to blood and to referral facilities, that later procedures require. The certificate of approval, the inspections, the returns and the registers together make the approved place a known point in an administrative map. That map is the state’s answer to the question the Shah Committee posed: if the practice cannot be suppressed, it must at least be located, staffed and counted. The private sector’s place within the map grew steadily as approval was decentralised, and the greater part of facility based provision in the incidence figures comes from private and non government facilities, which means the statutory service in India is, in its majority, a regulated private service operating inside a public legal frame.

The second layer also holds the scheme’s treatment of medical methods, which belongs here because it is a creature of rules rather than of the Act. The 1971 text was written for a surgical world and says nothing about method at all; it regulates the actor, the ground, the place and the record, and leaves technique to medicine. When medication abortion, the combination of mifepristone and misoprostol, entered Indian practice, it was absorbed into the scheme through the drugs framework and through the rules: the training rules were adapted to recognise a route for conducting terminations by medical methods in the earliest weeks, prescribing a shorter experience requirement for that work, three months of experience in obstetrics and gynaecology or ten cases performed independently under supervision in an approved setting. The absorption was characteristic of the statute’s design. Because the Act regulates the practitioner’s status and opinion rather than the instruments used, a new technique could be brought inside the fence by adjusting the second layer, without reopening the settlement of 1971. The same feature, however, fixed the boundary problem that dominates the next section: a method that can be used by the woman herself, outside any facility, fits a practitioner centred scheme hardly at all.

The machinery applied: three passages through the scheme

The scheme is best tested by walking three ordinary cases through it, in the form in which they present at a facility. The cases are composites drawn from the statute’s own categories, and their purpose is to show where, in each, the legal event occurs and where the friction sits.

The first case is the one the statute was built around. A married woman whose contraceptive method has failed attends an approved private clinic at nine weeks. Explanation II supplies the presumption: the anguish of the unwanted pregnancy may be presumed to be a grave injury to her mental health, so the practitioner is not required to construct a psychiatric case, only to form the opinion in good faith on the ground as the explanation frames it. One practitioner suffices at this gestation, and the practitioner who holds the narrower assisted cases qualification may act. The woman signs the consent form; no one else signs anything. The opinion is certified, the termination performed, the entry made in the admission register under its serial number, and the intimation sent through the prescribed channel. The legal event, from beginning to end, is the practitioner’s opinion and its record; the woman’s part in the legality of her own procedure is exhausted by her consent and her history. The case is administratively unremarkable, and it is worth pausing on how completely the design achieves, for this case, what the Shah Committee wanted: a dangerous informal transaction converted into a safe, recorded, professional one.

The second case is a survivor of rape who presents at a public hospital at sixteen weeks, having spent weeks in the processes that follow such an assault. Explanation I supplies her ground without inquiry: the pregnancy being alleged to have been caused by rape, the anguish is presumed to be a grave injury to her mental health, and no practitioner is required, or entitled under the Act, to test the allegation as a condition of acting. Two practitioners must concur at this gestation, which in a district facility can mean waiting for a second qualified doctor to be available, the first point of friction. The Act nowhere makes a police report, a first information report, or any step in the criminal process a condition of the termination; the medical and the criminal tracks are legally independent, and an institution that suspends the medical track until the criminal track moves is adding a condition the statute does not contain. Her consent is the operative consent, and if she is an adult it is hers alone. The documentation proceeds as in the first case, under the same secrecy regime. The friction in this case is not in the grounds, which the legislature deliberately made frictionless, but in the machinery around them: the second opinion, the institution’s caution, the parallel criminal process and its demands on the same woman at the same time.

The third case is the one the scheme handles worst, and it is the case that generated the litigation. A serious foetal abnormality is suspected at the anomaly scan and characterised with confidence only at nineteen or twenty weeks, or just after. Within the limit, the ground requires the practitioners to connect the diagnosis to the statute’s consequence, a substantial risk that the child, if born, would suffer such physical or mental abnormalities as to be seriously handicapped, and two must concur while the clock runs. Every day spent obtaining the confirmatory imaging, the specialist opinion and the second signature is a day off a fixed allowance, and a case that begins inside the limit can be pushed outside it by the process of establishing its own ground. Beyond the limit, under the Act as enacted, the machinery simply stops, and the case converts into the writ petition anatomised above. The design flaw is precise: the ground most dependent on late arriving knowledge is governed by the limit least tolerant of delay, and no provision of the 1971 text, not the environment clause, not the emergency section with its life saving threshold, reaches it. When the histories of this statute’s reform are written, they are written around this case.

The emergency exception and the perimeter penalties

Section 5 is the Act’s release valve and its fence, in that order. Sub-section (1) provides that the provisions of Section 4, and so much of Section 3(2) as relates to the length of the pregnancy and to the opinion of not less than two registered medical practitioners, shall not apply to a termination by a registered medical practitioner where the practitioner is of opinion, formed in good faith, that the termination is immediately necessary to save the life of the pregnant woman. Read the provision with precision. It disapplies the place requirement entirely, so the emergency termination need not occur at a government hospital or an approved place. It disapplies the length limit, so it operates at any gestation, including beyond twenty weeks. It disapplies the two opinion requirement in the later band, so one practitioner’s good faith opinion carries the case. It also relaxes, for its purposes, the experience and training limb of the practitioner definition, so the emergency does not fail because the doctor at hand holds the narrower qualification. What it does not disapply is the ground. The opinion must be that termination is immediately necessary to save the woman’s life. Not desirable, not beneficial to health, not justified by grave injury: immediately necessary to save life. The provision is the statutory descendant of the single exception the Penal Code itself always carried, restated in workable form, and its narrowness is the price of its breadth everywhere else. A practitioner who stretches the emergency provision to cover a hard case that is not a life saving case has stepped outside it, and the shelter of Section 3(1) does not follow.

The rest of Section 5 faces outwards. As substituted by the amendment of 2002, sub-section (2) provides that, notwithstanding anything in the Penal Code, the termination of a pregnancy by a person who is not a registered medical practitioner is an offence punishable with rigorous imprisonment for a term of not less than two years and extending to seven years, and that the Code stands modified to that extent. Sub-section (3) applies the same punishment to whoever terminates a pregnancy at a place other than a place Section 4 permits. Sub-section (4) applies it to the owner of a place that is not approved, owner being defined to reach the person who is the administrative head of the institution or otherwise responsible for its working. The perimeter penalties are severe by the standards of the Act, and deliberately so. The legislature of 2002 had before it the evidence of terminations continuing to be performed by unqualified persons in unapproved premises, the precise harm the Shah Committee had set out to extinguish, and it answered by making the boundary of the exception a criminal boundary with a mandatory minimum. The design is coherent: inside the fence, protection; on the fence, the practitioner’s own conditions; outside it, the penal law at its most emphatic.

Between the emergency provision and the perimeter penalties sits the question of what happens to the practitioner whose termination, honestly performed, produces a bad outcome. Section 8, already noticed, bars suit or other legal proceeding for damage caused by anything done in good faith under the Act. Good faith is the operative standard throughout the statute, in the forming of opinions, in the emergency judgment, and in the civil immunity, and it is the standard the whole scheme is calibrated to reward. The practitioner the Act imagines is not the practitioner who is always right; it is the practitioner who forms opinions genuinely, records them properly, works in approved premises, and stays within the grounds. Such a practitioner the Act protects against both prosecution and suit. The practitioner it declines to protect is the one who treats the conditions as elastic, and the elasticity, once discovered, is judged under the penal provisions with which this article began.

A boundary remains to be marked, between the Act’s protection and the ordinary law of medical negligence. A termination performed within the conditions, which ends in the woman’s death through negligence, is not converted into the Section 314 offence merely by the outcome, because the Act’s shield covers guilt under the Code for a termination in accordance with its provisions, and the question the criminal law then asks is the question it asks of any medical treatment that ends badly: whether the practitioner’s conduct crossed from error into the gross negligence that the Supreme Court’s medical negligence jurisprudence, stated for criminal liability in Jacob Mathew v. State of Punjab in 2005, requires before a doctor may be prosecuted for causing death by negligence. The distinction preserves the coherence of the whole arrangement. The Act decides whether the termination was lawful. The negligence standard decides whether the treatment was criminally careless. A practitioner may satisfy the first and fail the second, or fail the first and never reach the second, and confusing the two inquiries is a recurring error in commentary on prosecutions in this field. The 2002 perimeter offences, for their part, have a prehistory worth recording, since they answer the objection that the original Act had fences without sanctions: an expert group constituted in 1997 to review the Act’s working, and suggestions from the National Commission for Women directed at provisions it regarded as discriminatory, fed the review that produced the amending Bill, and the punishments that emerged in 2002 were aimed at the supply side of the unsafe market, the unqualified practitioner and the unapproved premises, rather than at women.

Medical methods and the self managed majority

The single largest fact in the modern operation of this statute is that most terminations in India do not pass through it. The incidence study’s estimate for 2015, set out in the practice section below, is that close to three quarters of all abortions were medication abortions occurring outside health facilities, overwhelmingly through combipacks of mifepristone and misoprostol obtained from chemists or informal sellers and used by the woman herself. The legal classification of that majority practice deserves exact statement, because public discussion habitually blurs it. The 1971 Act protects a defined course of conduct: a termination by a registered medical practitioner, on a certified opinion, at an approved place, with the record made. A woman who obtains the drugs from a chemist and uses them at home has not been terminated by a practitioner within the Act at any point. The transaction sits outside the statutory channel, and the penal background against which it occurs is unchanged: the explanation to Section 312 of the Penal Code, carried into Section 88 of the Bharatiya Nyaya Sanhita, provides that a woman who causes herself to miscarry is within the offence, subject always to the section’s own good faith exception for saving her life. The chemist who sells, and any person who assists, stand in their own positions under the penal and drugs law. That prosecutions of women for self managed medication abortion are not a visible feature of the enforcement landscape does not alter the classification; it records enforcement practice, not law.

The drugs themselves sit under a second legal regime that does some of the boundary work. Mifepristone and misoprostol are prescription medicines under the drugs law, and the combipack in which they are marketed carries conditions of supply directed at the statutory scheme: it is intended for use under medical supervision, within the gestational limit the drugs approval specifies, and dispensing without prescription offends the drugs framework independently of anything in the 1971 Act. State drug control administrations periodically enforce those conditions against chemists, and the enforcement has its own perverse geography, since pressure on urban chemists displaces sales to informal channels rather than into facilities. The result is a layered illegality that public discussion chronically flattens. The woman who self manages commits, in strict law, the penal offence the explanation describes; the chemist who sells without prescription offends the drugs law; the practitioner who certifies nothing has no involvement at all; and the statute that the public believes governs the transaction governs none of it. Against that analysis must be set the enforcement record, which is permissive in practice, and the public health record, which associates the spread of the drugs with falling mortality. A legal system can hold a practice unlawful in classification, tolerated in enforcement and beneficial in outcome for a long time, and India has. The cost of the arrangement is that nobody inside it, woman, chemist or doctor, can know in advance which of the three registers will be applied to their case.

The gap between the classification and the practice is the deepest structural tension in the field, deeper than the gestational limit, because it concerns not the edge of the service but its centre of mass. Three observations discipline any judgment about it. The first is about safety: medication abortion in the early weeks, properly dosed, is a low risk intervention by the standards of the procedures it displaced, and the mortality collapse that the Shah Committee sought has been delivered in large part by this channel rather than by the formal one, a fact that complicates any simple story in which the statute alone made abortion safe in India. The second is about information: the self managed route carries no certified opinion, no screening for the contraindications that matter, no certain knowledge of gestation, and no guaranteed path to care when the process is incomplete or goes wrong, so its risks, while far below the backstreet surgery of 1970, are distributed to precisely the women least equipped to manage them. The third is about the statute’s premise: a scheme built on the practitioner’s opinion as the legal event cannot recognise the woman’s own administration as the same event without ceasing to be itself, which is why reform argument in this area is really argument about replacing the scheme’s premise, conducted by both sides in the vocabulary of amendment. Reformers who would bring self managed abortion within the law must explain what happens to the opinion, the place and the record; defenders of the scheme must explain why the majority of actual practice should remain outside the only framework the law provides. The pillar states the tension as structural, because that is what it is: the Act governs termination as a professional act, and the country increasingly practises it as a pharmaceutical one.

The 2021 changes, in outline

Half a century after enactment, Parliament returned to the statute. The Medical Termination of Pregnancy (Amendment) Act of 2021 recast the gestational architecture: the opinion of one practitioner was extended across a longer early period, the upper limit was raised for prescribed categories of women to twenty four weeks, a mechanism of state level Medical Boards was created for cases of substantial foetal abnormality beyond that limit, the contraceptive failure explanation was freed of its original restriction to married women, and an express confidentiality provision was written into the Act. The following year, the Supreme Court read the statute and its rules together so as to remove the distinction between married and unmarried women for the purposes of the extended limit. Each of those propositions is the subject matter of the companion treatment of the 2021 amendment and the limits, and this pillar does not compete with it. What the pillar records is only the structural observation that matters for understanding the 1971 design: the amendment altered the numbers, the categories and the machinery, but it did not alter the address of the statute. The amended Act still speaks to practitioners, still frames its grounds as opinions, still confers no entitlement on the woman, and still leaves the beyond the limit case to a gatekeeping body rather than to her decision. The renovation was substantial; the architecture was kept.

The courts and the constitutional overlay

The statute’s text gives the woman a consent and no claim. The constitutional case law has built, alongside the text, a doctrine of reproductive choice that speaks a different language, and the relationship between the two is the most intellectually interesting feature of this field. The courts have not struck down the 1971 Act or any part of its conditions. What they have done is to hold, in a line of authority running through the privacy and personal liberty jurisprudence, that reproductive choice is constitutionally protected, and then to apply that protection at the edges of the statute: to consent, where the statute’s own logic already pointed the same way, and to the beyond the limit case, where the statute gives no answer and the courts have supplied one case by case.

The starting point is Suchita Srivastava in 2009, whose consent holding has been set out above. Its constitutional holding was larger. A woman’s right to make reproductive choices, the Supreme Court said, is a dimension of personal liberty as understood under Article 21 of the Constitution, and reproductive choices can be exercised to procreate as well as to abstain from procreating; the woman’s right to privacy, dignity and bodily integrity was to be respected. The direction of the case is worth noticing, because it is the opposite of the usual one. The constitutional right was invoked not to obtain a termination the statute withheld, but to resist a termination the state sought to impose, and it prevailed. The doctrine, that is to say, protects the woman’s decision as such, in whichever direction she makes it, which is exactly what a consent based reading of Section 3(4) would predict and what a purely medical reading of the statute might have missed.

Eight years later, the nine judge bench in Justice K. S. Puttaswamy v. Union of India held privacy to be a fundamental right, and the plurality reasoning placed decisions of the most personal kind, including decisions about reproduction and the intimacies of family life, within the protected core of that right, citing the reproductive choice line of authority with approval. The full treatment of that jurisprudence belongs to the series article on the Puttaswamy judgment and the privacy doctrine; here it is enough to mark the consequence. After Puttaswamy, the constitutional atmosphere in which the 1971 Act operates is one in which the woman’s choice is a rights bearing act, while the statute through which the choice must usually be exercised continues to address itself to the practitioner. That dissonance, between the constitutional subject and the statutory addressee, is the unresolved structural fact of Indian abortion law, and both the amendment debates and the court permission cases are best understood as attempts to manage it without confronting it directly.

Did the Supreme Court strike down any part of the MTP Act?

No part of the Act has been struck down. The courts have instead worked at its edges: reading the consent provision so that mental retardation does not displace a woman’s own choice, affirming reproductive choice as an aspect of personal liberty and privacy, and, in individual petitions beyond the twenty week limit, granting or refusing permission on medical board reports.

The permission cases are the visible tip of that management. Because the Act as enacted provided no route beyond twenty weeks except the life saving emergency, a woman whose pregnancy crossed the limit, through late detection of a foetal abnormality, through delayed access, through rape, through the failure of every institution she approached, had nowhere to go except a constitutional court. The courts became, in effect, the third tier the legislature had declined to build. The Bombay proceedings of 2008 showed the system’s hardness at its most exposed. Niketa Mehta, whose pregnancy was found at about twenty four weeks to involve a congenital heart block in the foetus, petitioned the Bombay High Court with her husband and her treating doctor for permission to terminate beyond the limit. Expert opinion placed before the court divided and then resolved against the prognosis the petitioners relied on; the panel’s assessment was that there was no medical evidence the child would be seriously handicapped after birth. The High Court declined permission, holding that the petitioners had not made out an exceptional case for the exercise of its discretion and that altering the statutory boundary was Parliament’s work. The pregnancy ended in a miscarriage some weeks later. The case entered the public memory as the emblem of the limit’s rigidity, and the ethical debate it provoked, including a sustained exchange in the Indian Journal of Medical Ethics on whether the foetal ground itself discriminates against disabled children, shaped the reform argument for a decade.

The Supreme Court’s later practice moved, by increments, in the other direction, without disowning the statute. In X v. Union of India in 2016, the Court permitted a termination at about twenty four weeks in a rape case, acting on a medical board’s assessment of the risk to the woman. In Meera Santosh Pal v. Union of India, decided on 16 January 2017, the petitioner’s foetus was diagnosed with anencephaly, a condition in which the skull does not form and survival outside the womb is impossible. A medical board’s report supported the conclusion that continuance involved risk to the woman’s life and possible grave injury to her physical and mental health within Section 3 itself, and the Court, speaking of bodily integrity and of the woman’s right to take the steps necessary to preserve her own life on an informed choice, permitted the termination at twenty four weeks. In the same period the Court permitted a termination beyond the limit for a thirteen year old survivor of rape on the particular facts before it, notwithstanding a board’s caution about comparative risks. The pattern across these orders is consistent and should be stated exactly. The courts did not hold the twenty week limit unconstitutional. They did not rewrite Section 3. They treated each petition as an individual case, obtained a medical board’s report, and granted or withheld permission on the conjunction of the medical evidence, the ground the case could be fitted to, and the woman’s own stated choice. Refusals occurred as well, most starkly where boards reported that termination itself had become the more dangerous course. The doctrine supplied the vocabulary; the boards supplied the answers; the limit supplied the reason a court was being asked at all.

The writ route anatomised

Because the permission cases recur in reporting as though the courts were a standing appellate tier of the statute, the mechanics deserve a section of their own. A petition in this line typically begins after the statutory route has closed: an abnormality detected at or after the limit, a survivor whose case surfaced late, a pregnancy discovered at an advanced stage. The woman, or those supporting her, move the High Court under its writ jurisdiction or the Supreme Court directly, and the court’s first act is almost invariably to constitute or convene a medical board at a public hospital and direct it to examine the woman and report on the medical questions: the condition of the foetus, the risk to the woman of continuance, and the risk to her of termination at the stage reached. The board’s report then does the decisive work. Where it supports the statutory language, risk to life or possible grave injury to physical or mental health, or certifies an abnormality incompatible with survival, permission tends to follow, expressed in the vocabulary of the Act itself. Where it reports that termination has become comparably dangerous, or that the foetal condition is survivable and treatable, permission tends to be refused, and the court’s sympathy expresses itself, if at all, in directions for care and support. Between the filing and the order, days and weeks pass in which the pregnancy advances and the medical position itself changes, which is why the cases so often read as races against the very limit they are testing.

The anatomy explains the pattern of outcomes better than any theory of judicial attitude. The courts in these cases behave less like constitutional tribunals revising a statute than like the missing tier of the statute itself, applying its grounds with its methods, the medical opinion, on material the boards supply. That is also why the jurisprudence, for all its Article 21 vocabulary, has never generalised into a ruling that the limit is unconstitutional or that a woman may elect termination beyond it: the courts have been deciding cases, not redesigning the scheme, and they have said so. The cost of the arrangement falls where the pillar has already located it. A remedy that requires a writ petition, counsel, a board examination and a hearing is a remedy rationed by resources, and the reported cases skew towards women in metropolitan centres with institutional support, while the same facts in a district town end, ordinarily, in continuance or in the unrecorded channel. It also falls on the woman in a second currency, publicity: the petitioner litigates the most private facts of her life in open court records, protected by anonymised initials in the reports but exposed in the process itself. The anonymising convention, the petitioner who appears in the law reports only as X, is the courts’ own acknowledgment of the price, and it interacts awkwardly with the statute’s secrecy machinery: a woman whose termination the regulations would have recorded under seal in a district register finds the facts of her pregnancy summarised in a reported judgment, readable by anyone, whenever her case crosses the limit and enters a courtroom. The confidentiality the scheme promises is therefore conditional on the case never needing the remedy of last resort, which is to say it is least available in the cases of greatest hardship. No court created that result deliberately. It is the compound interest of two reasonable arrangements, a secret register for the statutory route and open justice for the constitutional one, and it stands in the record as one more instance of the article’s recurring pattern: the protections of this Act attach to the channel, and lapse at the points where the channel fails. The 2021 amendment’s Medical Boards were Parliament’s attempt to administrative what the courts had been doing by writ, moving the function from the courtroom to a standing body; whether the attempt succeeded belongs to the amendment’s assessment. The pillar’s point is anterior: for fifty years the Act had no beyond the limit institution at all, and the writ route was the improvisation by which the legal system supplied one, at a price.

For the practitioner and the advocate, the practical teaching of the case law is twofold. Within the limit, the statute governs and the courts’ role is the ordinary one of policing consent, good faith and the conditions. Beyond the limit, until the 2021 amendment built its Medical Board mechanism, the only lawful door was a writ petition, with all the cost, publicity and delay that door implies, and the cases that came through it were disproportionately those of women with the resources and support to litigate. The constitutional overlay dignified the woman’s choice in principle while the statutory structure continued to price its exercise in practice, and that gap between principle and price is the honest summary of the judicial record.

The neighbouring statutes that frighten practitioners

No statute operates alone, and the 1971 Act operates in the company of two others whose enforcement reaches into the same consulting rooms. The first is the law against sex selection: the Pre-Conception and Pre-Natal Diagnostic Techniques (Prohibition of Sex Selection) Act of 1994, enacted as the Pre-Natal Diagnostic Techniques (Regulation and Prevention of Misuse) Act and renamed and strengthened at the start of the following decade. Its object is distinct and legitimate: to prohibit the use of diagnostic techniques to determine the sex of a foetus and the selective elimination that follows. Its machinery, however, overlaps the terrain of the 1971 Act. It regulates the clinics where ultrasound is performed, which are frequently the same premises where pregnancies are assessed and terminated. It imposes its own dense record keeping, centred on prescribed forms for every scan of a pregnant woman. It arms inspectors with powers of search and seizure, and it attaches criminal penalties to record keeping failures as well as to the substantive offence. A practitioner can therefore stand in a single room, on a single day, fully compliant with the 1971 Act’s conditions for a lawful termination and exposed under the 1994 Act for a defect in a form that has nothing to do with the termination’s lawfulness.

The Federation of Obstetric and Gynaecological Societies of India has argued for years, in representations to government and in the reform debates, that the practical consequence is a conflation in the minds of practitioners and inspectors alike between sex determination, which is prohibited, and second trimester termination, which the 1971 Act permits on conditions, and that the conflation makes practitioners defensive about the whole field. Facility studies and provider interviews collected by health researchers record the same anxiety from the other side: practitioners declining lawful later terminations, insisting on documentation beyond what either statute requires, or leaving the field entirely, because the personal cost of an inspection or a prosecution is high and the professional reward for providing the service is low. The position of the two statutes in law is clear and has been stated in the case law: the 1994 Act does not amend the 1971 Act, a termination within the 1971 Act’s conditions is lawful, and sex selection is not among its grounds. The position in practice is murkier, and the murk is itself a finding about how this area of law works. A defence based statute delivers the access its conditions describe only if the class it protects believes the defence will hold when tested, and belief is formed by the whole enforcement environment, not by the statute book alone.

The second neighbour is the child protection statute noticed in the consent section, the Protection of Children from Sexual Offences Act of 2012, whose mandatory reporting obligation converts a pregnant minor’s attendance at a clinic into the starting point of a criminal process whether or not the minor or her family wishes it. Between the 1994 Act’s inspection regime and the 2012 Act’s reporting regime, the practitioner treating the cases at the margin of the 1971 Act, the later termination, the minor, the rape survivor, faces legal exposure from three directions at once, of which the 1971 Act’s own conditions are only one. It is not a sufficient explanation of uneven access, but it is a necessary part of any explanation, and it is entirely absent from accounts that describe the statute’s liberal grounds and profess surprise at the service’s scarcity.

The gatekeeper in practice

Between the statute and the woman stands the gatekeeper, and the gatekeeper’s working habits, rather than the text, determine the service most women actually meet. The habits are not lawless; they are the predictable product of the design. A practitioner whose protection depends on a record will want the record complete, and will often want more than the record: corroboration of the history, a spouse present, a family member informed, an ultrasound filed, a second opinion where the statute is satisfied with one. Some of these wants are clinical prudence. Some are institutional routine. Some are the practitioner’s insurance against the inspection regimes of the neighbouring statutes. From the woman’s side of the desk they are indistinguishable: each is a condition she must satisfy to reach a service the law is popularly supposed to have given her, and each is imposed by an actor she cannot appeal. The case law’s clearest illustration has already been noticed: the Bihar litigation, in which institutional insistence on third party consent that the Act does not require delayed a survivor’s case past the statutory limit and drew compensation from the Supreme Court. The significance of that case is not that one institution erred. It is that the error is structurally available at every desk where the Act is administered, because the Act polices the practitioner’s compliance and contains no corresponding machinery to police the practitioner’s additions.

Conscience operates in the same unregulated space. The Act contains no conscience clause, in either direction: it neither protects a practitioner who declines on grounds of conscientious objection nor obliges any practitioner to participate. Declining is simply outside the statute’s field of vision, as the design section explained. Professional ethics fills part of the space, since a practitioner who declines is expected in ordinary medical practice to make the patient’s further care possible rather than to obstruct it, but the expectation is ethical rather than statutory, and its enforcement is the slow machinery of professional discipline rather than any remedy the woman can invoke at the time it matters. In institutions the pattern compounds: a public facility in which the one trained practitioner declines, or is absent, or is unwilling to take the later band cases, is a facility that does not provide the service whatever its approval certificate says, and the incidence literature’s repeated finding that many approved public facilities do not in practice offer terminations is, in substantial part, the sum of these individual positions. None of this requires bad faith anywhere in the system. It requires only a defence based statute, a risk bearing gatekeeper, and a service nobody is under a duty to supply. The referral chain that results, from the chemist to the clinic to the district hospital to the medical college to, at the limit, the High Court, is the lived form of the Act for a large class of women, and each link in it consumes the one resource the statute’s own thresholds make precious, which is time.

How the Act has worked in practice

The most reliable single measurement of the distance between the statute’s scheme and the reality it governs is the national incidence study carried out by the Guttmacher Institute with the International Institute for Population Sciences and the Population Council, published by Singh and colleagues in The Lancet Global Health with estimates for the year 2015. The study estimated that 15.6 million abortions took place in India in 2015, a rate of 47 abortions per thousand women aged fifteen to forty nine. Of that total, it estimated that 3.4 million, about twenty two per cent, were provided in health facilities. About 11.5 million, seventy three per cent, were medication abortions that took place outside health facilities, the great majority using the combipacks of mifepristone and misoprostol sold through chemists and informal channels. The remaining 0.8 million, five per cent, were estimated to be abortions outside facilities by methods other than medication, the category the study treated as probably unsafe. Within facilities, the private sector accounted for about seventy three per cent of procedures and the public sector about twenty four per cent, although the public sector is the main source of care for poor and rural women. Set against these estimates, the health ministry’s own service statistics, which count terminations reported through the recorded system, captured well under a million in a year, a discrepancy the study’s authors treated as a measure of how much of the practice lay outside the recorded statutory channel altogether.

Household survey evidence tells the same story from the woman’s side. The fifth National Family Health Survey, covering 2019 to 2021, recorded that about three per cent of pregnancies ended in abortion, and that of the abortions women reported, more than half were performed in the private sector, about a fifth in the public sector, and more than a quarter were performed by the woman herself at home. The distribution matters more than the shares. The public sector is the only channel that is free at the point of use and the only one with any presence in much of rural India, yet it accounts for the smallest facility share and figures in the survey as the channel of the minority; the private sector, which the approval machinery of the Act brought into the lawful map, carries the service; and the home, which the Act does not reach at all, carries more than the public system does. A statute cannot be assessed apart from the delivery system beneath it, and beneath this statute the delivery system is a thin public layer, a thick private layer available at a price, and a self managed layer available at a chemist. The grounds in Section 3 are the same in all three layers. Almost nothing else is.

The mechanism that converts those layers into outcomes is cost, in its several currencies. A facility termination costs money in the private layer, where most provision sits, and the price rises steeply with gestation, so that the women delayed longest by the referral chain face the highest price at the end of it. It costs time in every layer: the confirmatory scan, the second opinion, the repeat visit, the day of travel to the district facility and the day lost when the practitioner is absent, and time under this statute is not a neutral currency, because the thresholds convert delay into illegality on a fixed date. It costs wages, childcare, explanations at home and at work, and, where secrecy fails, it can cost the woman her household position, which is why the confidentiality machinery is not a courtesy but a condition of the service being usable at all. None of these costs appears in the statute, and all of them are distributed by it, because a design that routes every lawful termination through scarce professional time at approved places in a country of India’s distances will price its service in exactly these currencies whatever its text says about grounds. The access record is the design, read from the paying end.

The figures reward careful reading, because they cut in two directions and the article will not force them into one. On the reading favourable to the statute, the 1971 Act has presided over a transformation in safety: the dangerous core of the pre-1971 problem, the septic backstreet procedure performed by untrained hands, has shrunk to a small residue of the total, because medication abortion, whatever its legal classification when self managed, is pharmacologically far safer than the methods it displaced, and the Registrar General’s Sample Registration System bulletins record the direction of travel in maternal deaths, with the special bulletin for 2001 to 2003 attributing eight per cent of maternal deaths to abortion, a cause that had bulked far larger in the Shah Committee’s world. On the reading unfavourable to the statute, the formal channel it built, one practitioner, approved place, certified opinion, secret register, accounts for barely a fifth of what actually happens, and the Act’s real function in most women’s experience is not to enable their termination but to stand unused beside it. Both readings are supported by the same table of numbers. What they jointly establish is that the statute succeeded at the problem the Shah Committee set it, the safety of the procedure as a medical matter, far more completely than at the problem it did not set itself, the delivery of termination as an organised legal service.

The reasons for the formal channel’s thinness are structural and recur across the implementation literature. The stock of practitioners who satisfy the definition and hold the prescribed training is concentrated in urban and private practice, because the training routes run through obstetric and gynaecological experience that rural public facilities struggle to staff. A large share of public facilities that are approved on paper do not in practice offer the service, for want of a trained provider, of equipment, of supplies of the drugs, or of institutional will. The approval of private places multiplied the lawful map without curing its geography. Above these supply facts sit the demand side facts: stigma within families and communities, the preference for the anonymity of a chemist over the register of an institution, however secret the register is said to be, and the simple arithmetic of cost and travel. The characteristic operational problems named at the start of this article, refusal, referral and delay, operate inside this structure. A practitioner who refers a woman onward because the second opinion is unavailable in the facility, or because the case sits in the band the practitioner is not trained for, or because the paperwork or the neighbouring statutes make the case feel dangerous, is not defying the statute. The practitioner is exhibiting exactly the behaviour a defence based statute predicts from a risk bearing actor with no duty to act. The woman experiences the referral as the system declining her. The statute experiences it as the system working as drafted. Both experiences are accurate, and the distance between them is the measure of the design.

Why does access remain uneven under a statute with broad grounds?

Because the Act distributes a defence, not a service. Its grounds are broad, but only a practitioner who holds the prescribed training, works at an approved place, and is willing to form and record the required opinion can use them. Where practitioners are scarce, unwilling or deterred by neighbouring enforcement, the grounds describe a service that does not exist locally.

How provision organised itself

A lawful channel does not supply itself, and the half century of the Act is also the history of the institutions that grew up inside its permissions. The public system was the intended backbone: the approval of government hospitals by status put a lawful provider, in principle, in every district, and state programmes built training for medical officers into the family welfare and later the reproductive and child health frameworks, with comprehensive abortion care training standardised and delivered with technical support from organisations working in the field, among them the Ipas Development Foundation, which has partnered with state governments on training and service quality for many years. The Federation of Obstetric and Gynaecological Societies of India, the professional body, has been a continuous institutional presence in the field, in the standard setting for practice, in the debates over the later band and the Medical Boards, and in the profession’s longstanding complaint, recorded above, about the enforcement of the sex selection law falling on lawful providers. Alongside both, a non government sector specialised in exactly this service: organisations such as Parivar Seva Sanstha have run clinic networks providing terminations within the Act for decades, and their existence is a comment on the design, since a statute that creates lawful demand without a duty of supply invites precisely this kind of provision to fill the space between the public backbone and the private market.

The organisation of provision explains several features of the practice record that would otherwise look accidental. The concentration of the service in private facilities reflects not only approval but economics: the private clinic monetises the statutory channel, trains its practitioners through the assisted cases route, and locates where paying demand is densest, which is urban India. The thinness of the public channel reflects the opposite arithmetic: approval without staffing, staffing without the confidence to take later band cases, and no institutional reward for provision, produce the approved facility that does not provide. The pharmacist, finally, organised the channel the statute never contemplated, because the drugs were effective, the demand was there, and the enforcement of prescription discipline for a medicine in the prescription schedules of the drugs law was unequal to the market. Each layer of the delivery system is, in this sense, a rational response to the 1971 design and its silences. The statute regulated legality and left supply to incentives, and the supply that arrived is the supply those incentives described.

The disagreement the Act cannot settle

Underneath every argument about this statute’s machinery lies a disagreement the machinery cannot resolve, and the pillar’s obligation is to state it fairly rather than to smuggle in a resolution. The disagreement concerns the moral status of the developing foetus and the weight its interests should carry against the pregnant woman’s. It is genuinely held on both sides, by people who understand the law and the medicine, and no drafting change dissolves it.

The position that the law should protect the developing foetus treats human life as beginning before birth and regards the foetus as a bearer of interests, at the limit a rights bearer, whom the state may legitimately protect against the wishes of others, including at some point the wishes of the woman carrying it. On this view, expressed in the teaching of the Roman Catholic Church, which holds direct abortion to be impermissible, and in the advocacy of pro life organisations in India and elsewhere, a permissive statute is not a neutral framework but a decision by the state to withdraw protection from one class of human life, and the correct direction of reform is towards restriction, support for women to carry pregnancies to term, and the treatment of abortion as a social failure rather than a medical service. Supporters of this position point to the statute’s own foetal ground as its most objectionable feature, since in their reading it licenses the ending of pregnancies on the basis of predicted disability, a judgment they regard as discriminating against disabled people and as inconsistent with the commitments the same legal system makes to disability equality elsewhere. That objection is not confined to religious voices; it was pressed by disability commentators in the debate following the Niketa Mehta case and surfaces in the Indian Journal of Medical Ethics exchange that case provoked.

The opposing position treats decisions about one’s own body and reproductive life as belonging to the person whose body it is. On this view, associated in the legal arena with the constitutional jurisprudence set out above, with the Supreme Court’s own language in Suchita Srivastava and in Puttaswamy, and in the policy arena with networks such as the Pratigya Campaign for Gender Equality and Safe Abortion, the state’s proper role is minimal: pregnancy is a condition of the woman’s body, its continuance imposes the burdens and risks on her, and a legal structure that makes her access depend on a professional’s permission mistakes a gatekeeper for a safeguard. Supporters of this position read the 1971 Act as a statute of its era, liberal by the standards of 1971 and paternalist by construction, and argue that a framework honest about its own premises would start from the woman’s decision and regulate only the safety of the procedure, as the law regulates other medical procedures she may choose. They point to the access record set out above as evidence that gatekeeping does not prevent terminations but merely sorts them, into the lawful and recorded for women with money and proximity, and the unrecorded for women without.

A third critique cuts across both positions and should be distinguished from each, because it is about the statute’s origins rather than its morality. Organisations working on women’s health in India, among them Sama, the Delhi based resource group for women and health, have long argued that the 1971 Act is best read as an instrument of the population programme rather than of women’s welfare: that its timing within the family planning era, its contraceptive failure ground, its silence about the woman’s own reasons, and its investment of authority in the medical profession all mark it as a statute designed to make termination available for the state’s purposes while keeping it under professional and administrative control. On this reading, the doctor centred design is not a technical choice but the point, and the access failures are not malfunctions but the system declining to serve women whose needs fall outside the purposes it was built for. The critique has force as an account of emphasis, and the pillar has already recorded the committee’s own disavowal of population control as its object, which is the strongest documentary answer to it. The two can be held together without contradiction: a committee may reason in public health terms while a state enacts its recommendations inside a population programme, and the statute that results will serve, and be seen to serve, both masters. What the critique usefully prevents is the sentimental reading of 1971 as a charter of women’s liberation manqué. It was a health measure of a paternalist state, and its text has never pretended otherwise.

Between the two positions sit the practical criticisms that holders of either can and do share, and the pillar should record them as common ground rather than as either side’s property. The first is that a framework resting on practitioner opinion produces access that tracks the distribution and disposition of practitioners, which is to say it produces inequality by design, between city and village, between private means and public dependence, between the married woman the system expected and the unmarried woman it did not. The second is that the enforcement climate of the sex selection law frightens practitioners away from lawful work, so that a statute aimed at one grave social evil operates in practice as a tax on a different and lawful service. The third is that a system with a hard statutory edge and no administrative route beyond it delivers the hardest cases, the late detected abnormality, the child survivor, the woman failed by every earlier institution, into constitutional courts, where the price of the remedy is money, publicity and time, paid by precisely the women least able to pay it. A reader who holds the first moral position can accept all three criticisms as reasons to build the boundary differently; a reader who holds the second can accept them as reasons the boundary should not be a gate at all. The criticisms are about craft. The disagreement beneath them is about value, and it will survive any amendment Parliament enacts.

The pillar’s own verdict is confined to the ground the evidence governs, which is the ground of design. Judged as a piece of legislative craft addressed to the problem its authors set, the Act is a coherent and durable instrument: it identified the actor whose behaviour had to change, the practitioner working outside the law, and it changed that behaviour by the only lever the criminal law gives, the removal of guilt on conditions. Judged as an instrument of access, its central feature is also its central limitation, because a statute that confers a defence and imposes no duty cannot guarantee a service, and the history of its operation is the working out of that sentence. The authors of 1971 knew they were protecting doctors. The committee said so in substance, the preamble says so in terms, speaking of termination of certain pregnancies by registered medical practitioners, and the operative section says so in its grammar. Later argument that treats the Act as a failed rights statute is therefore aiming at a target the statute never claimed to be. The serious question the record poses is not why the Act failed to be something else, but whether Parliament, having understood in 1971 how much turns on the practitioner’s willingness, should at some point have added the duties that turn a defence into a service, and that question, unlike the moral one, is capable of an answer from evidence. The access record summarised above is that evidence, and it points one way.

Assessment: reading the Act whole

Read whole, the Medical Termination of Pregnancy Act of 1971 is best understood as three instruments compressed into eight sections. It is a criminal law instrument, because its operative act is the removal of guilt from conduct the Penal Code forbids, and its perimeter is policed with the mandatory minimum sentences added in 2002. It is a medical regulation, because everything within the perimeter, who may act, on what opinion, in what premises, with what training, with what records, is specified with the particularity of a licensing regime. And it is a document of its decade’s assumptions about women, in which a liberalising legislature extended protection to married women managing family size, to survivors of rape, and to women whose health pregnancy endangered, while routing every one of those protections through a professional’s opinion and stopping short of the woman’s own reasons as a ground in themselves. The statutes of the same era that addressed women’s position in Indian law shared the pattern of advance through a protective frame, whether the subject was termination of pregnancy or equal remuneration for equal work, and reading this Act beside them is a corrective to accounts that treat it as either a feminist landmark or an aberration. It is neither. It is a public health statute that worked on the problem it named.

For the practitioner, the payload of the whole analysis is short enough to state in one paragraph. The defence is complete where the conditions are complete: a practitioner within the definition, an opinion formed in good faith on a statutory ground and certified as the regulations require, the woman’s consent or, in the two defined cases, the guardian’s written consent, premises within Section 4, and a record made and kept as the secrecy regime prescribes. The defence is absent where any one of those elements is absent, and the Act offers no doctrine of substantial compliance. For the advocate, the corresponding payload is that argument in this field is almost never about whether termination is lawful in the abstract; it is about which element of a particular case is said to fail, about consent, about the opinion’s formation and record, about the limit and the route beyond it, and increasingly about the constitutional doctrine that frames the woman’s side of the case. For the student and the general reader, the payload is the One Test answer with which this article began, restated as a rule of reading: whenever a report describes what the Act allows a woman to do, translate it into what the Act permits a practitioner to do without guilt, and the report’s accuracy can be judged at once.

Three confusions account for most published error on this statute, and the pillar closes by naming them for the record. The first is the rights confusion: the Act creates no right to a termination, in its original or its amended form, and descriptions of Indian law as conferring one misstate the law in the direction of generosity, which is still a misstatement. The second is the consent confusion: no husband, parent or family member holds a consent over an adult woman’s termination, and institutions that demand one are adding a requirement the statute does not contain, with consequences the Supreme Court has already had to correct in litigation. The third is the geography confusion: the Act is a central statute of uniform text whose service map is anything but uniform, because its conditions can be satisfied only where trained practitioners and approved places exist, and the distance between the text’s uniformity and the service’s patchiness is the single largest fact about its operation. A reader who holds those three corrections has the working model the bare text alone would not supply. What this article has deliberately not attempted is advice on any individual situation: the account here is an explanation of a statutory scheme, and questions about a particular pregnancy are medical and legal questions of fact on which the general position is no substitute for the practitioner and the adviser seized of the case.

It sharpens the assessment to set the Indian design against the two ideal types between which it sits, because most confusion about the Act is confusion about which type it belongs to. The first type is the rights model, in which the law recognises the woman’s decision as hers, lays the corresponding duties on the state and the provider system, and treats gestational limits as regulations of an entitlement rather than conditions of somebody else’s immunity. The second type is the pure prohibition model, in which termination is a crime relieved only by a necessity defence tested after the fact in a criminal court. India is neither. The Indian scheme is a licensing model for professional conduct: the state’s instrument is the register of practitioners and the approval of places, its unit of legality is the certified opinion, and its characteristic question is never whether the woman was entitled but always whether the practitioner was covered. Judged as a rights statute it will always appear a failure, since it confers none; judged as a prohibition it will always appear a fraud, since it licenses the majority of the practice it nominally forbids; judged as what it is, a device for converting an unpoliceable crime into a regulated medical service, its strengths and its pathologies are the same set of facts. The gatekeeper is the device working. The gatekeeper is also the device failing the woman the gate was not built for. Reform debates in India recurrently stall because the participants are grading the statute against different types, and the first service an analysis can perform, as this pillar has tried to perform it, is to fix the type before the grading begins.

The pattern of its amendment is itself part of the record. In the half century to 2021 the Act was substantively amended only once, in 2002, which decentralised approval and built the perimeter offences, and not at all otherwise, in a period when the surrounding law of medicine, of privacy, of sex selection and of child protection was repeatedly remade. The stasis is instructive. A statute whose scheme is a settled bargain among interests, the profession protected, the state informed, the woman served through a gatekeeper, generates few internal pressures for revision, because each interest holds its position by holding the text still; and the pressures that did accumulate, the diagnostic calendar, the exclusion of the unmarried woman from the contraceptive failure presumption, the self managed majority, arrived from outside the bargain and were answered together, in 2021, rather than one at a time. Whether the next revision comes faster, and whether it revises the premise or only the parameters, is the question the pillar leaves with its successors. The companion study of the 2021 Act takes the story to the point where the parameters moved. The premise, on the evidence of the delivered amendment, remains where Parliament left it in 1971.

Frequently Asked Questions

Q: Is abortion legal in India and on what conditions?

Abortion in India is lawful only within the Medical Termination of Pregnancy Act of 1971, because causing a miscarriage remains an offence under the general criminal law. A termination is lawful where a registered medical practitioner, holding the prescribed experience or training in gynaecology and obstetrics, forms the opinion in good faith that continuance of the pregnancy would risk the woman’s life or cause grave injury to her physical or mental health, or that there is a substantial risk the child would be born seriously handicapped. One practitioner’s opinion suffices up to twelve weeks and two are required between twelve and twenty weeks under the Act as enacted. The woman’s consent is required, the procedure must take place at a government hospital or an approved place, and the prescribed records must be made. Outside those conditions, the penal provisions apply in full.

Q: What does the Medical Termination of Pregnancy Act permit?

The Act permits a registered medical practitioner to terminate a pregnancy without incurring guilt under the Penal Code or any other law, provided the termination accords with the Act. It specifies the grounds on which the practitioner’s opinion may be formed, the number of opinions required at each stage, whose consent is necessary, the places where the procedure may be performed, the emergency case in which several requirements fall away, and the records to be kept. It also protects the practitioner against civil proceedings for anything done in good faith under it. What it permits is therefore best understood as a protected course of professional conduct rather than a general legalisation: the same procedure performed by an unqualified person, at an unapproved place, or outside the grounds remains a criminal offence, punishable after the 2002 amendment with rigorous imprisonment of two to seven years.

Q: Who can lawfully perform a termination of pregnancy?

Only a registered medical practitioner within the Act’s definition may lawfully perform a termination. The practitioner must hold a recognised medical qualification under the Indian Medical Council Act of 1956, be entered in a State Medical Register, and possess the experience or training in gynaecology and obstetrics prescribed by the rules. The rules recognise several routes: three years of practice in the speciality for those registered before the Act commenced, six months of house surgency in gynaecology and obstetrics, a year of hospital experience in obstetrics and gynaecology, assisting in twenty five cases of medical termination of which at least five were performed independently in an approved institution, or a postgraduate degree or diploma in the speciality. The assisted cases route qualifies the practitioner for first trimester terminations only. Anyone else who terminates a pregnancy commits an offence under Section 5 of the Act, punishable with rigorous imprisonment of not less than two years and extending to seven years.

Q: Is the consent of a husband or family required for a termination?

No. Section 3(4) of the Act provides that, apart from the cases of a minor and of a mentally ill person, no pregnancy shall be terminated except with the consent of the pregnant woman herself. The statute names no other consent holder. A husband has no veto and no signature to give; parents have no role in the termination of an adult daughter’s pregnancy; no family council, employer or authority is consulted by the Act. Institutions that ask for a spouse’s consent are imposing an internal practice, not a legal requirement, and the Supreme Court has condemned the obstruction that results when such demands are treated as law, awarding compensation where a woman lost her access because third party consent was wrongly insisted upon. The consent must be the woman’s, and a termination carried out without it is the separate and grave offence of causing a miscarriage without consent under the penal law.

Q: What happens if a pregnancy exceeds the gestational limit?

Under the Act as enacted, the practitioner’s protection ends at twenty weeks. Beyond that point the statute contains no committee, board or discretion to which a case can be referred on the grounds alone, and the only provision that operates is the emergency exception, which requires that termination be immediately necessary to save the woman’s life. A woman whose case fell outside the limit, through late detection of a serious foetal abnormality, rape, or delay in reaching services, therefore had to petition a High Court or the Supreme Court for permission, and the constitutional courts built a case by case practice on medical board reports, granting permission in some cases and refusing it in others. That route was slow, public and expensive, and the hardship it produced was a principal argument for reform. The 2021 amendment later built a Medical Board mechanism for substantial foetal abnormalities beyond the extended limit, a change treated in the companion article on the amendment.

Q: Does a minor need guardian consent for termination of pregnancy?

Yes. Section 3(4)(a) provides that the pregnancy of a woman who has not attained the age of eighteen years shall not be terminated except with the consent in writing of her guardian, and the same rule applies to an adult woman who is a mentally ill person within the Act’s definition. For a minor, the guardian’s written consent replaces the woman’s own consent as the operative legal act, although practitioners record the minor’s position as a matter of practice and the grounds in Section 3 must still be satisfied by the practitioner’s opinion. The rule interacts with the Protection of Children from Sexual Offences Act of 2012, under which a minor’s pregnancy is evidence of an offence subject to mandatory reporting, so a practitioner treating a pregnant minor faces a reporting obligation alongside the consent and confidentiality machinery of the 1971 Act. That collision between the two statutes has generated substantial litigation and judicial mitigation, particularly in cases involving consensual adolescent relationships.

Q: How does the Act interact with the law against sex determination?

The two statutes govern different conduct and neither amends the other. The Medical Termination of Pregnancy Act permits termination on its stated grounds and nowhere treats the sex of the foetus as a ground or as material to a lawful opinion. The Pre-Conception and Pre-Natal Diagnostic Techniques (Prohibition of Sex Selection) Act of 1994 prohibits the determination and communication of the sex of a foetus and the use of diagnostic techniques for sex selection, regulating the clinics, machines and records through which prenatal diagnosis occurs. A termination within the 1971 Act’s conditions remains lawful even where the same clinic is regulated under the 1994 Act. In practice, however, the second statute’s inspection and record keeping regime operates on the same premises and the same profession, and practitioners and their federation have long argued that enforcement practice blurs the line between prohibited sex determination and permitted later termination, deterring lawful services. That practical interaction, rather than any conflict in the texts, is where the two laws meet.

Q: Is confidentiality protected in termination cases?

Yes, through the machinery of the Act’s regulations rather than through a free standing promise in the original text. Section 7 of the Act empowers regulations prohibiting the disclosure of the intimations and information the process generates, except to specified persons for specified purposes. Under the regulations, the practitioner’s certified opinion is recorded on the prescribed form, intimations travel to the Chief Medical Officer, and the admission register recording each woman’s particulars is declared a secret document, kept in safe custody, retained for five years, and open to inspection only under the authority of law; documentation of an emergency termination is sent to the Chief Medical Officer in a sealed envelope marked secret. The protection attaches to the record rather than to everything a woman may experience in reaching the service, and the amendment of 2021 later wrote an express confidentiality duty into the Act itself with a penalty for disclosure, strengthening a protection the original scheme had built out of its paperwork rules.

Q: Why was the Medical Termination of Pregnancy Act passed in 1971?

The Act followed the report of the Shah Committee, constituted by the Government of India in September 1964 under the chairmanship of Shantilal H. Shah to examine the question of legalising abortion. The committee reported in December 1966 and reasoned principally on public health grounds: the Penal Code prohibition was not preventing terminations but was driving them into the hands of untrained practitioners, at heavy cost in maternal deaths and injuries, while most women seeking termination were married women who would use a lawful medical service if one existed. It recommended a separate statute allowing qualified practitioners to terminate pregnancies on defined medical, humanitarian and social grounds. A bill was introduced in the Rajya Sabha in November 1969, examined by a Joint Committee of both Houses, passed by Parliament in 1971 and assented to on 10 August 1971, coming into force on 1 April 1972. The reform was a health measure, not the product of a rights judgment or a mass movement.

Q: How many doctors must approve a termination of pregnancy?

Under the Act as enacted, the number depends on the length of the pregnancy. Where the pregnancy does not exceed twelve weeks, the opinion of one registered medical practitioner, formed in good faith on a statutory ground, is sufficient. Where the pregnancy exceeds twelve weeks but does not exceed twenty weeks, the opinions of not less than two registered medical practitioners must concur on the same ground. The requirement is about the opinion, not about a committee or an approval from any authority: the practitioners themselves form and certify the opinion, and no official sanction is needed. In the emergency case, where a practitioner forms the opinion in good faith that termination is immediately necessary to save the woman’s life, the two opinion requirement falls away with the place requirement and the length limit. The 2021 amendment later redistributed the numbers across a longer period, one practitioner up to twenty weeks and two between twenty and twenty four weeks for prescribed categories.

Q: Where can a termination of pregnancy legally be performed?

A termination under the Act may be performed only at a hospital established or maintained by Government, or at a place approved for the purpose of the Act. As enacted, approval lay with the Government; the amendment of 2002 widened the approving authority to include a District Level Committee constituted by the Government and chaired by the Chief Medical Officer or District Health Officer, with three to five members, in order to decentralise a process that had bottlenecked at state level. Approval is tied to facility standards prescribed by the rules, which distinguish the requirements for early terminations from the fuller surgical and anaesthetic support required for later ones. A termination at any other place is outside the Act, and after 2002 the person performing it, and the owner of the unapproved place, are each punishable with rigorous imprisonment of two to seven years. The single exception is the genuine emergency under Section 5, where the place requirement does not apply.

Q: What happens in an emergency when delay would endanger the woman’s life?

Section 5(1) of the Act disapplies the machinery where a registered medical practitioner forms the opinion in good faith that termination of the pregnancy is immediately necessary to save the life of the pregnant woman. In that case the place requirement in Section 4 does not apply, so the termination may occur wherever the emergency finds the patient; the length limits and the requirement of two concurring opinions in the later band do not apply, so gestation is no bar and one practitioner’s judgment carries the case; and the experience and training limb of the practitioner definition is relaxed for the purpose. The documentation is adjusted rather than abandoned: the termination is recorded through the sealed envelope procedure to the Chief Medical Officer marked secret. What the provision does not relax is the ground itself. The opinion must be one of immediate necessity to save life, and a hard case that does not reach that standard cannot be brought within Section 5 by calling it an emergency.

Q: What does grave injury to mental health mean under the Act?

Grave injury to mental health is one limb of the health ground in Section 3(2), and the Act builds three aids around it. First, Section 3(3) allows account to be taken of the pregnant woman’s actual or reasonably foreseeable environment when assessing whether continuance risks injury to her health, so the assessment is about her life circumstances and not only her clinical presentation. Second, Explanation I presumes that the anguish caused by a pregnancy alleged to have been caused by rape constitutes a grave injury to mental health, without any inquiry into the allegation. Third, Explanation II, as enacted for a married woman, allowed the anguish of an unwanted pregnancy resulting from failure of a contraceptive device or method to be presumed to be such an injury. The standard is applied through the practitioner’s good faith opinion, and the phrase covers the anguish, distress and psychiatric risk of an unwanted or dangerous continuation; it is the principal route by which social circumstances enter a ground drafted in medical language.

Q: What training or experience must a doctor have to perform terminations?

The practitioner must first be a registered medical practitioner: holder of a recognised medical qualification under the Indian Medical Council Act of 1956, with a name entered in a State Medical Register. Beyond registration, Section 2 requires such experience or training in gynaecology and obstetrics as the rules prescribe, and Rule 4 of the Medical Termination of Pregnancy Rules supplies the routes. A practitioner registered before the Act commenced may rely on three years of experience in the practice of gynaecology and obstetrics. A later registered practitioner may qualify through six months of house surgency in the speciality, or a year of hospital experience in obstetrics and gynaecology, or a postgraduate degree or diploma in the speciality. The remaining route is practical: assisting a registered practitioner in twenty five cases of medical termination of pregnancy, of which at least five were performed independently, in an approved hospital or training institute; that route qualifies the practitioner for first trimester terminations only.

Q: Can a termination be carried out without the pregnant woman’s consent?

No, and the attempt is among the gravest offences in this area of the penal law. Section 3(4)(b) states the rule inside the Act: save for the two defined exceptions, no pregnancy shall be terminated except with the consent of the pregnant woman. A termination performed on a woman who has not consented is therefore never within the Act’s protection, whatever the grounds or the practitioner’s opinion. It falls instead under Section 313 of the Indian Penal Code, Section 89 of the Bharatiya Nyaya Sanhita, which punishes causing a miscarriage without the woman’s consent with imprisonment for life, or imprisonment extending to ten years, and fine. Consent given under coercion or by deception is not the consent the statute contemplates, and consent cannot be supplied by a husband, a parent or an institution in the case of an adult woman of sound mind. The only substitutions the law allows are the guardian’s written consent for a minor and for a mentally ill person within the Act’s narrow definition.

Q: What is the legal position when the woman has a mental illness?

Section 3(4)(a) provides that the pregnancy of a woman who has attained eighteen years but is a mentally ill person shall not be terminated except with the consent in writing of her guardian. The category is defined and narrow: after the amendment of 2002, a mentally ill person is a person in need of treatment by reason of any mental disorder other than mental retardation. Intellectual disability is therefore expressly outside the provision, a distinction the Supreme Court enforced in Suchita Srivastava v. Chandigarh Administration in 2009. There, the pregnancy of a woman with mild mental retardation in a state institution had been directed to be terminated by the High Court; the Supreme Court held that she was not mentally ill within the Act, that her reproductive choices were protected by Article 21, and that her own expressed wish to continue the pregnancy governed. For a woman who is mentally ill within the definition, the guardian’s written consent is the operative consent, and the grounds in Section 3 must still be satisfied.

Q: Is a doctor protected from prosecution for a termination done under the Act?

Yes, on two levels, provided the termination stands within the Act. Section 3(1) provides that, notwithstanding anything in the Penal Code, a registered medical practitioner shall not be guilty of any offence under that Code or under any other law for the time being in force if a pregnancy is terminated in accordance with the provisions of the Act. Section 8 adds the civil protection: no suit or other legal proceeding lies against a practitioner for any damage caused or likely to be caused by anything done or intended to be done in good faith under the Act. Both protections are conditioned on the same things, genuine formation of the statutory opinion, satisfaction of the conditions, and compliance that can be demonstrated from the record. A practitioner who cannot show the conditions were met, or who acted outside them, has no shelter under either provision, and the exposure is then the ordinary penal exposure for causing a miscarriage, aggravated where the case falls within the perimeter offences added in 2002.

Q: When did the MTP Act come into force, and did it extend to the whole country?

The Act is Act 34 of 1971. It received the President’s assent on 10 August 1971 and was brought into force on 1 April 1972 by notification. As enacted, Section 1 provided that it extends to the whole of India except the State of Jammu and Kashmir, so the statute book position for most of its life was that the Act governed the rest of the country while the excluded state stood outside it. The position changed when Parliament reorganised Jammu and Kashmir in 2019 and extended the central statutes, this Act among them, to the successor Union territories of Jammu and Kashmir and Ladakh, completing the Act’s territorial extent by legislation rather than by any amendment of the Act itself. Its short title, the Medical Termination of Pregnancy Act, 1971, takes its year from the year of enactment, not from the year of commencement, a distinction that accounts for the frequent misstatement that the Act has operated since 1971 rather than since April 1972.

Q: How have the courts linked termination of pregnancy to personal liberty under Article 21?

The courts have built the link in two steps without disturbing the statute. In Suchita Srivastava v. Chandigarh Administration in 2009, the Supreme Court held that a woman’s right to make reproductive choices is a dimension of personal liberty under Article 21, that reproductive choices may be exercised to procreate as well as to abstain from procreating, and that a woman’s privacy, dignity and bodily integrity must be respected; the holding protected a woman’s decision to continue a pregnancy against a state directed termination. In Justice K. S. Puttaswamy v. Union of India in 2017, the nine judge bench recognised privacy as a fundamental right, and the plurality reasoning counted reproductive decisions among the personal choices the right protects, approving the earlier line of authority. The doctrine therefore treats the woman’s choice as constitutionally significant in both directions, while the statute through which the choice is usually exercised continues to address the practitioner, a dissonance the case law manages rather than resolves.

Q: Who carries the legal risk if a condition of the Act is not met?

The practitioner carries it, and after the amendment of 2002 the institution’s owner shares part of it. The Act is drafted as a defence, so every condition is an element of that defence: a ground within Section 3, the correct number of concurring opinions, the prescribed qualification, the woman’s consent or the guardian’s written consent where required, and an approved place. If an element fails, the termination is judged under the penal provisions for causing miscarriage, and the practitioner cannot answer by pointing to the elements that were satisfied. Where the failure is of the perimeter kind, termination by a person who is not a registered medical practitioner, termination at an unapproved place, or ownership of an unapproved place where terminations occur, Section 5 attaches rigorous imprisonment of not less than two years and extending to seven years. The woman is not the Act’s risk bearer in respect of a practitioner’s defaults, although the Penal Code’s explanation to Section 312 leaves her exposed where she causes her own miscarriage outside the statutory scheme altogether.