The Vishaka guidelines present a problem that no drafting technique can dissolve, because the problem is constitutional rather than textual. In August 1997 the Supreme Court of India confronted a subject on which Parliament had enacted nothing at all: there was no statute defining sexual harassment at a workplace, no statute obliging an employer to prevent it, no forum in which a working woman could complain of it as a wrong against her working conditions rather than as a crime to be reported to the police, and no machinery anywhere in Indian law through which such a complaint could be examined. The Indian Penal Code punished discrete acts, assault or criminal force to outrage a woman’s modesty and words or gestures intended to insult it, but a penal code addresses an offender after an act; it does not organise a workplace, and it offered nothing to a woman whose complaint was that her place of work had become a place she could not safely attend. The question the Court had to answer was therefore not what a statute meant. It was whether the Constitution, unaided by any statute, could supply enforceable obligations of this operational kind, and if it could, how far a court may go in writing them. The answer the Court gave governed every workplace in India for sixteen years, and the argument about whether it should have been given has never closed.
This article owns that judicial history. The statute that eventually occupied the field, the Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal) Act, 2013, is analysed provision by provision in the pillar article on the Act, and this article does not re-derive its machinery. What belongs here is everything that happened before the statute and explains it: the petition and the constitutional reasoning, the precise source of the Court’s authority to declare its guidelines to be law, the international convention the Court read into the fundamental rights, the content of the guidelines themselves, the two later decisions that tested and then audited them, the sixteen years in which an interim judicial regime was the only regime there was, and the standing objection that a court which writes committee compositions and reporting duties has ceased to interpret law and begun to legislate. A reader who finishes should be able to state, without hesitation, what the Vishaka bench held, on what provisions it rested the holding, what each guideline required, what the 2013 Act kept and what it discarded, and where the legitimate criticism of the whole enterprise lies.
The petition and the question it actually posed
The judgment arose from an incident of grave sexual violence against a woman working on a state social development programme in Rajasthan, and the petition that followed it was not a claim for compensation in her case but a class action brought by social activists and non-governmental organisations, Vishaka among them, for the enforcement of the fundamental rights of working women generally. The petitions, Writ Petition (Criminal) Numbers 666 to 70 of 1992, were filed under Article 32 of the Constitution, the provision that lets a person move the Supreme Court directly for the enforcement of a fundamental right, and that choice of vehicle shaped everything that followed. A class action under Article 32 does not ask a court to construe a statute or to review an administrative order. It asks the court to say what the fundamental rights require in a field where the ordinary instruments of government have produced nothing, and to fashion the order that makes those rights real. The petitioners’ framing, which the Court adopted in its first paragraph, was that the petition sought to focus attention on the violation and to find suitable methods for the realisation of gender equality, and to prevent sexual harassment of working women in all workplaces through the judicial process, expressly in order to fill the vacuum in existing legislation.
The bench that heard the matter was presided over by Chief Justice J. S. Verma, sitting with Justice Sujata V. Manohar and Justice B. N. Kirpal, and the judgment of 13 August 1997, reported at (1997) 6 SCC 241, was delivered by the Chief Justice for the bench. Fali S. Nariman appeared as amicus curiae, Meenakshi Arora and Naina Kapur assisted the Court, and the Solicitor General, T. R. Andhyarujina, appeared for the Union of India. The composition of the argument matters to the later debate about legitimacy, because the guidelines were not framed in an adversarial vacuum: paragraph 9 of the judgment records that the formulation of the guidelines had the consent of the Union of India, given through the Solicitor General, who indicated that these should be the guidelines and norms declared by the Court to govern the behaviour of employers and all others at workplaces. A regime later criticised as imposed by an unelected court on an unconsulted country was, at the point of its creation, consented to by the elected government’s own senior law officer, who rendered what the judgment calls valuable assistance in the true spirit of a law officer to help find a proper solution to a social problem of considerable magnitude. That fact does not answer the constitutional objection, but no honest account of the case can omit it.
The question the Court posed to itself was narrower than the rhetoric the case has attracted. It was not whether sexual harassment is wrong, and it was not whether the Constitution disapproves of it. It was a question about remedy and institutional competence: given that the conduct violated fundamental rights, given that no enacted law provided for the effective enforcement of those rights against this conduct, and given that a writ of mandamus which merely declared the violation would leave the violation recurring daily in workplaces across the country, what may a court acting under Article 32 lawfully direct? Chief Justice Verma’s formulation in paragraph 3 is the hinge of the whole judgment. A writ of mandamus in such a situation, if it is to be effective, needs to be accompanied by directions for prevention, because the violation of fundamental rights of this kind is a recurring phenomenon; the fundamental right to carry on any occupation, trade or profession depends on the availability of a safe working environment; and when instances of sexual harassment resulting in violation of the fundamental rights of women workers are brought before the Court for redress under Article 32, an effective redressal requires that some guidelines should be laid down for the protection of these rights to fill the legislative vacuum. Every element of the later controversy is present in that paragraph: the insistence that a remedy must be effective rather than declaratory, the characterisation of the field as a vacuum rather than as a policy choice Parliament had made, and the move from adjudication to direction-writing that the objection fastens on.
The rights held to be violated
The first step in the reasoning was to locate sexual harassment inside the fundamental rights, and the Court located it in four places at once. Sexual harassment at a workplace was held to violate Article 14, the guarantee of equality before the law and equal protection of the laws, and Article 15, the prohibition of discrimination by the state on grounds that include sex, because conduct that drives women out of workplaces, or prices their continued presence there at the cost of submission to sexual demands, denies them the equality the Constitution promises in the one sphere where most adults spend most of their waking lives. It was held to violate Article 19(1)(g), the right to practise any profession or to carry on any occupation, trade or business, because that right is hollow if its exercise is conditioned on enduring conduct that no male colleague is required to endure; the judgment’s sentence that the fundamental right to carry on any occupation depends on the availability of a safe working environment converts a liberty usually read against state interference into a claim about the conditions under which the liberty can be exercised at all. And it was held to violate Article 21, the right to life and personal liberty, read as the judgment reads it: right to life means life with dignity, and gender equality includes protection from sexual harassment and the right to work with dignity, which the Court described as a universally recognised basic human right.
Around those four provisions the Court placed supporting text. Article 15(3), which permits the state to make special provision for women and children, removed any doubt that a regime addressed specifically to the protection of women sat comfortably within the constitutional scheme rather than offending the equality guarantee it invoked. Article 42, a directive principle requiring the state to make provision for securing just and humane conditions of work, and Article 51A(e), the fundamental duty of every citizen to renounce practices derogatory to the dignity of women, were cited as provisions envisaging intervention for the eradication of the evil. None of these supporting provisions carries the load; the load is carried by the conjunction of equality, occupation and dignity. But the arrangement repays attention, because it shows the Court building a holding that a later legislature could enact without constitutional embarrassment and that a later bench could defend without strain. When Parliament came to legislate in 2013 it inherited a field whose constitutional geography had already been surveyed: the dignity reading of Article 21 that does the deepest work in Vishaka is the same reading that later constitutional adjudication expanded, and the treatment of dignity and personal liberty in the Puttaswamy privacy and Aadhaar judgment descends from the line of cases to which Vishaka belongs, in which Article 21 is read as protecting the conditions of a dignified human existence rather than a bare survival.
How long did the Vishaka guidelines operate as the law?
The guidelines were framed to operate until suitable legislation was enacted, and they operated from 13 August 1997 until the POSH Act commenced on 9 December 2013, a span of sixteen years and four months in which they were, by the judgment’s own declaration, binding and enforceable law in every workplace in India, public and private.
Sixteen years is the fact that organises this entire subject, and it deserves to be stated without decoration before it is analysed. The Court did not present its guidelines as a permanent settlement. It presented them as a bridge, built because the far bank had not been constructed, and it said in terms that the bridge would stand until Parliament built the road. Parliament took sixteen years. In that interval the guidelines were not a proposal, a recommendation or a model code. They were, by force of Article 141, the law declared by the Supreme Court, binding on all courts within the territory of India, and an employer who ignored them was not declining a suggestion but disobeying the law of the land as declared by its highest court. The series of which this article forms part has traced statutes that were passed and never commenced, and rules that were required and never made. Vishaka is the mirror image of that pathology: an obligation that existed in full force for sixteen years with no statute behind it at all, created by the one organ of the state that is not supposed to create obligations of general application, and kept alive far beyond any plausible meaning of the word interim by the inaction of the organ that is.
Where a court found the authority to declare law
The holding that sexual harassment violates fundamental rights was, by 1997, well within the Court’s ordinary competence; courts declare the content of constitutional rights every working day. The extraordinary step was the second one, and the judgment is careful and explicit about its sources, because the bench knew exactly how extraordinary it was. Four distinct constitutional provisions do the work, and they must be kept separate, because later accounts habitually blur them into a general claim that the Court acted under Article 141, which is only one strand of the four.
The first strand is Article 32 itself. The petitions were brought under Article 32, and the guidelines were laid down, in the judgment’s words at paragraph 16, in exercise of the power available under Article 32 for the enforcement of the fundamental rights. Article 32 empowers the Supreme Court to issue directions or orders or writs for the enforcement of any of the rights conferred by Part III, and the judgment’s move is to treat the word directions at its full width: if the Court may direct, and if a bare declaration cannot enforce a right against a recurring pattern of conduct in millions of workplaces, then directions that specify the preventive and redressal machinery are within the power, provided they are directed at enforcement and framed to last only until the legislature acts. The second strand is Article 141, which provides that the law declared by the Supreme Court shall be binding on all courts within the territory of India. Paragraph 16 emphasises that the guidelines would be treated as the law declared by the Court under Article 141, and the concluding direction repeats that they would be binding and enforceable in law until suitable legislation is enacted to occupy the field. Article 141 does not itself confer a power to make norms; it confers precedential force on the law the Court declares. Its function in the architecture is to convert the guidelines from orders in a single case into law of general application, so that a High Court entertaining a complaint about a workplace with no committee would apply the guidelines as law rather than as persuasion.
The third strand is the international-law bridge, built from Article 51(c) and Article 253 read with Entry 14 of the Union List. Article 51(c), among the directive principles of state policy, requires the state to endeavour to foster respect for international law and treaty obligations. Article 253 empowers Parliament, notwithstanding the ordinary distribution of legislative competence, to make any law for implementing any treaty, agreement or convention, and Entry 14 of the Union List covers entering into and implementing treaties. The judgment’s use of these provisions at paragraph 7 is precise and limited: they are not said to make the convention self-executing, and they are not said to give the Court a legislative power. They are said to make the contents of international conventions and norms significant for the interpretation of the constitutional guarantees, so that any international convention not inconsistent with the fundamental rights and in harmony with their spirit must be read into those provisions to enlarge their meaning and content and to promote the object of the constitutional guarantee. The fourth strand is Article 73, the executive power of the Union, which extends to matters on which Parliament has power to legislate: the judgment notes that executive power was therefore available until Parliament enacted legislation, aligning the Court’s interim regime with an executive competence that the Constitution itself contemplates as gap-filling. Around these four strands the Court placed the Beijing Statement of Principles of the Independence of the Judiciary, accepted by the Chief Justices of the Asia and Pacific region at Beijing in 1995, whose statement of the objectives of the judiciary includes ensuring that all persons are able to live securely under the rule of law and promoting, within the proper limits of the judicial function, the observance and attainment of human rights. The phrase within the proper limits of the judicial function is the judgment’s own chosen standard, and the whole of the later debate is an argument about whether detailed committee-design guidelines fall within those limits or beyond them.
Two authorities were marshalled for the reading-in technique itself. In Nilabati Behera v State of Orissa, the Court had drawn on the International Covenant on Civil and Political Rights to support an enforceable right to compensation as a public law remedy under Article 32, and Vishaka treats that case as establishing that international conventions may be used for construing the fundamental rights expressly guaranteed in the Constitution, which embody the basic concept of gender equality in all spheres of human activity. From the High Court of Australia’s decision in Minister for Immigration and Ethnic Affairs v Teoh, the judgment took the concept that ratification of a convention can generate a legitimate expectation of its observance in the absence of a contrary legislative provision, a concept the judgment describes as an accepted rule of judicial construction where there is no inconsistency and there is a void in domestic law. The Australian borrowing matters less for its content than for what it reveals about the bench’s method: the Court was conscious that it was performing an operation other common law courts had performed, in systems with and without a bill of rights, and it assembled the comparative cover deliberately. The doctrine that results, stated at paragraph 14, is the proposition for which Vishaka is cited far beyond the law of workplace harassment: the international conventions and norms are to be read into the fundamental rights in the absence of enacted domestic law occupying the field, when there is no inconsistency between them. Both conditions are load-bearing. Where domestic law occupies the field, or where the convention conflicts with it, the technique has no application, and the judgment does not suggest otherwise.
This gap-filling technique is not unique to this field, and the wider doctrine of courts constructing obligations in statutory silence, of which the environmental jurisprudence of the same era is the largest example, is traced in the account of environmental jurisprudence in the Supreme Court, which owns that wider story. What distinguishes Vishaka within it is the completeness of the regime produced: not a standard, not a principle awaiting application, but a working code of employer duties, committee composition and reporting obligations, declared to be law and expressed to endure until a legislature displaced it. Judicial vacuum-filling of that completeness is rare enough in any legal system to require the justification the judgment took such care to assemble, and rare enough that its sixteen-year life, rather than any defect in its reasoning, is what ultimately made it controversial. An interim regime that lasts sixteen years tests the premise on which it was justified, which is that the legislature will shortly act; the justification does not decay because the Court reasoned badly but because the condition it assumed failed to occur.
The convention the Court read in: CEDAW
The international instrument on which the judgment rests is the Convention on the Elimination of All Forms of Discrimination against Women, adopted by the United Nations General Assembly in 1979 and ratified by India in 1993; the judgment records the ratification, with reservations it describes as not material in the context, and supplements the treaty text with the Indian government’s commitments at the Fourth World Conference on Women at Beijing in 1995. The use the judgment makes of the Convention is worth stating exactly, because it is routinely overstated in both directions. The Court did not hold that CEDAW was enforceable in India as a statute is enforceable. India is a dualist system in which a treaty, without legislation under the power the Constitution confers for that purpose in the provision on legislation to implement international agreements, does not by itself create rights a court can apply against private parties. What the Court held is that in the absence of domestic law occupying the field, the Convention’s content may be used to construe the amplitude of rights the Constitution itself guarantees, so that the obligation enforced remains a constitutional obligation, sourced in Articles 14, 15, 19(1)(g) and 21, whose content is informed by the standard India had already accepted internationally. The distinction between enforcing a treaty and construing a constitution in a treaty’s light is the distinction on which the judgment’s legality stands or falls, and the bench observed it scrupulously.
India’s ratification of the Convention was not unqualified, and the judgment’s handling of the qualification is part of its discipline. India accompanied ratification with declarations concerning the Convention’s provisions on modifying social and cultural patterns of conduct and on marriage and family relations, matters entangled with the personal laws of religious communities, in which it declared that it would abide by those provisions in conformity with its policy of non-interference in the personal affairs of any community without its initiative and consent, together with a reservation on the Convention’s inter-state dispute settlement clause. The judgment recorded the ratification, noted the reservations, and stated that they were not material in the present context, which was exactly right: the provisions the bench drew on, equality in employment and the national measures undertaking, stood outside the declared qualifications, and a construction of the fundamental rights informed by them therefore built on ground the Indian state had accepted without reserve. The point matters beyond this case, because the reading-in technique’s legitimacy in any given application depends on the same check: the convention relied on must be one the state has accepted, in the relevant part, without a reservation that withdraws the very content the court proposes to read in.
The General Recommendation quoted in the judgment, General Recommendation No. 19 of 1992, did three things the treaty text alone could not do. It named sexual harassment as a form of gender-specific violence that seriously impairs equality in employment. It supplied a definition in terms that a legal instrument can operate: unwelcome sexually determined behaviour such as physical contacts and advances, sexually coloured remarks, showing pornography and sexual demands, whether by words or actions, which can be humiliating, may constitute a health and safety problem, and is discriminatory when the woman has reasonable grounds to believe that her objection would disadvantage her in connection with her employment, including recruiting or promotion, or when it creates a hostile working environment. And it stated that effective complaints procedures and remedies, including compensation, should be provided. The definition in guideline 2 of the Vishaka guidelines tracks that recommendation nearly word for word, and the complaints machinery of guidelines 6 and 7 is the institutional form of its closing sentence. The compensation element of the recommendation, alone among its elements, found no counterpart in the guidelines, a gap whose significance is examined below, because the statute of 2013 repaired it and the repair is one of the clearest demonstrations of what a legislature could add that a court framing interim guidelines did not attempt.
The guidelines themselves, head by head
The operative part of the judgment is a code of twelve numbered heads, prefaced by two recitals that fix its character: the Court took note of the definition of human rights in Section 2(d) of the Protection of Human Rights Act, 1993, and of the fact that the existing civil and penal laws did not adequately provide for the specific protection of women from sexual harassment at workplaces and that enactment of such legislation would take considerable time. The guidelines were then declared necessary and expedient for employers in workplaces as well as other responsible persons or institutions to observe. Their content repays a slow reading, because the 2013 Act is unintelligible as a drafting exercise until the reader knows exactly what the drafters had in front of them, and because the mapping between guideline and section, set out in the table that follows this section, is the article’s central evidence.
The first head states the duty in its general form: it shall be the duty of the employer or other responsible persons in workplaces or other institutions to prevent or deter the commission of acts of sexual harassment and to provide the procedures for the resolution, settlement or prosecution of acts of sexual harassment by taking all steps required. Three features of that sentence carried the regime. The duty is preventive before it is redressive; an employer who waits for complaints has already breached it. The duty attaches to the employer and to other responsible persons, not to the state alone, which is what made the guidelines a workplace code rather than a direction to governments. And the procedures to be provided run through the full arc of resolution, settlement and prosecution, so the internal mechanism was never intended to displace the criminal law but to sit alongside it.
The second head is the definition, and it has two limbs. The first limb enumerates the conduct: sexual harassment includes such unwelcome sexually determined behaviour, whether directly or by implication, as physical contact and advances, a demand or request for sexual favours, sexually coloured remarks, showing pornography, and any other unwelcome physical, verbal or non-verbal conduct of a sexual nature. The second limb, too often dropped in summary accounts, supplies the circumstances in which such conduct is discriminatory in the equality sense: where the victim has a reasonable apprehension that the conduct can be humiliating and may constitute a health and safety problem in relation to her employment or work, whether she draws salary or honorarium or works voluntarily, in government, public or private enterprise; it is discriminatory, for instance, when the woman has reasonable grounds to believe that her objection would disadvantage her in connection with her employment or work, including recruiting or promotion, or when it creates a hostile work environment, and adverse consequences might be visited if the victim does not consent to the conduct or raises any objection to it. That second limb is the analytical core of the definition, because it explains why the enumerated conduct is a rights violation rather than mere impropriety: the conduct weaponises the employment relationship itself, so that objection carries a professional price. When the 2013 Act came to be drafted, the two limbs were separated: the first became the definition in Section 2(n), and the second became, in altered form, the free-standing list of circumstances in Section 3(2). The separation is a drafting improvement with a doctrinal cost, examined in the mapping below.
The third head prescribes preventive steps for all employers or persons in charge of a workplace, in the public or private sector. The express prohibition of sexual harassment as defined should be notified, published and circulated in appropriate ways. The rules and regulations of government and public sector bodies relating to conduct and discipline should include rules prohibiting sexual harassment and provide for appropriate penalties against the offender. Private employers should take steps to include the prohibitions in the standing orders under the Industrial Employment (Standing Orders) Act, 1946. And appropriate work conditions should be provided in respect of work, leisure, health and hygiene, to further ensure that there is no hostile environment towards women at workplaces and that no woman employee has reasonable grounds to believe she is disadvantaged in connection with her employment. The fourth head deals with criminal proceedings: where the conduct amounts to a specific offence under the Indian Penal Code or under any other law, the employer shall initiate appropriate action in accordance with law by making a complaint with the appropriate authority; victims and witnesses are not to be victimised or discriminated against while complaints are dealt with; and the victim should have the option to seek the transfer of the perpetrator or her own transfer. The fifth head requires disciplinary action where the conduct amounts to misconduct in employment as defined by the relevant service rules.
The sixth head requires the complaint mechanism itself: whether or not the conduct constitutes an offence under law or a breach of the service rules, an appropriate complaint mechanism should be created in the employer’s organisation for redress of the complaint made by the victim, and such mechanism should ensure time-bound treatment of complaints. The seventh head constitutes the Complaints Committee and is the most consequential design decision in the judgment. The complaint mechanism should be adequate to provide, where necessary, a Complaints Committee, a special counsellor or other support service, including the maintenance of confidentiality. The Committee should be headed by a woman, and not less than half of its members should be women. To prevent the possibility of any undue pressure or influence from senior levels, the Committee should involve a third party, either a non-governmental organisation or another body, who is familiar with the issue of sexual harassment. The eighth head protects the space in which grievances can be voiced collectively: employees should be allowed to raise issues of sexual harassment at workers’ meetings and in other appropriate forums, and the subject should be affirmatively discussed in employer-employee meetings. The ninth head requires awareness: awareness of the rights of female employees should be created in particular by prominently notifying the guidelines, and appropriate legislation when enacted, in a suitable manner. The tenth head addresses third party harassment: where sexual harassment occurs as a result of an act or omission by any third party or outsider, the employer and the person in charge will take all steps necessary and reasonable to assist the affected person in terms of support and preventive action. The eleventh head is addressed to governments rather than employers: the Central and State Governments are requested to consider adopting suitable measures, including legislation, to ensure that the guidelines are also observed by employers in the private sector. The twelfth head is a savings clause: the guidelines will not prejudice any rights available under the Protection of Human Rights Act, 1993.
What did the Complaints Committee have to report, and to whom?
The Complaints Committee had to make an annual report to the government department concerned, setting out the complaints it had received and the action taken on them, and employers and persons in charge had separately to report their own compliance with the guidelines, including the Committee’s reports, to the same government department.
That reporting duty is the least remembered of the guidelines and one of the most instructive, because it shows the bench reaching for the only supervisory architecture available to it. A court cannot inspect workplaces. It cannot audit committee constitutions across a subcontinent. What it can do is require paper to flow upwards to a government department, creating a record against which a later court, entertaining a later petition, can measure compliance. That is precisely the use to which the record was put fifteen years later in Medha Kotwal Lele, where the failure of the regime was proved largely out of the absence of the very structures the reporting duty presupposed. The 2013 Act retained the annual report but redirected it, sending the Internal Committee’s report to the employer and the District Officer under Section 21 rather than to a government department at large, a change that looks administrative and is in fact a change of supervisory theory, from vertical reporting to a named local officer: the detail belongs to the pillar article, and it is flagged here because the mapping table records it as a modification, not an adoption.
Why did the guidelines insist on a third party inside the Committee?
The third party requirement answers the structural defect of every internal forum, which is that an internal body sits inside the hierarchy whose conduct it may have to judge, and a complainant against a senior officer faces the risk that the committee will protect the institution rather than examine the complaint; an outsider familiar with the issue breaks that closure.
The reasoning in the guideline is compressed to a single phrase, to prevent the possibility of any undue pressure or influence from senior levels, but the design intelligence in it should not be missed. The bench did not respond to the closure problem by taking the forum outside the workplace altogether, which would have required a tribunal system no court could conjure into existence with staff, budgets and appellate routes. It responded by contaminating the internal body with one member whose position, reputation and livelihood did not depend on the employer at all. The statute of 2013 adopted the device and remains attached to it, requiring an external member on every Internal Committee, and the compliance litigation of the following decade found, again and again, committees that existed without the external member or with an external member in name alone. The guideline and the section share a diagnosis: the point at which workplace redress fails is not the absence of a norm but the capture of the forum, and both instruments legislate against capture with the same modest instrument, one outsider in the room.
What did the judgment take from the Protection of Human Rights Act, 1993?
The judgment took note of the definition of human rights in Section 2(d) of the Protection of Human Rights Act, 1993, rights relating to life, liberty, equality and dignity guaranteed by the Constitution and embodied in the international covenants, and its twelfth head saved all rights under that Act from prejudice by the guidelines.
Guideline to section: the mapping the statute never published
The table below performs the exercise the brief for this article identifies as its contribution: every head of the guidelines is set against its counterpart in the Act of 2013, with a column recording whether the statute adopted the guideline, modified it, or introduced something the guidelines did not contain, and a final column recording what the guidelines contained that the statute left behind. Readers who keep their own provision maps and committee papers in document files for reference can open and check Word, Excel and PowerPoint files directly with the ReportMedic office file viewer while working through the comparison. The last column is the one no summary of the Act supplies, because accounts of the statute describe what it contains, and only a mapping against the guidelines reveals what a decade and a half of judicial law had that Parliament chose not to carry forward.
| Vishaka guideline | Counterpart in the POSH Act 2013 | Adopted, modified or new | In the guidelines, absent from the Act |
|---|---|---|---|
| Head 1: the employer’s duty to prevent and deter, and to provide procedures for resolution, settlement or prosecution | Section 3 prohibition and Section 19 duties of the employer | Adopted and elaborated into a list of specified duties | Nothing material; the general duty survives with its content enumerated |
| Head 2, first limb: the five enumerated forms of unwelcome sexually determined behaviour | Section 2(n) definition of sexual harassment | Adopted in substance, near verbatim | Nothing; the enumeration is the most faithfully copied text in the Act |
| Head 2, second limb: conduct discriminatory where objection would disadvantage the woman in employment, recruitment or promotion, or create a hostile environment | Section 3(2) list of circumstances that may amount to sexual harassment | Modified: a definitional gloss converted into a free-standing provision | The guidelines treated the circumstances as part of what the conduct is; the Act treats them as circumstances that may amount to it, a softer connector |
| Head 3(a): express prohibition to be notified, published and circulated | Section 19 duties to display the prohibition and organise orientation | Adopted | Nothing material |
| Head 3(b): government and public sector conduct rules to prohibit sexual harassment and prescribe penalties | Disciplinary action under the Act proceeds through the inquiry report and the service rules under Section 13 | Modified: the Act relies on its own inquiry and on action in accordance with service rules rather than mandating rule amendments | The free-standing direction that every conduct rulebook be amended to name the prohibition and its penalties |
| Head 3(c): private employers to write the prohibition into standing orders under the Industrial Employment (Standing Orders) Act, 1946 | No counterpart; the Act applies directly to employers it defines | Dropped as a mechanism, because a statute needs no standing-orders bridge | The standing-orders route itself, and with it the guideline’s technique of colonising existing labour instruments |
| Head 3(d): appropriate work conditions in work, leisure, health and hygiene, and no hostile environment | Section 19 duty to provide a safe working environment | Adopted in compressed form | The specification of leisure, health and hygiene as the content of the working-conditions duty |
| Head 4: employer to initiate criminal proceedings where the conduct is an offence; no victimisation of victims or witnesses; victim’s option of transfer | Section 19 duty to assist a woman who chooses to file a complaint under the penal law; Section 12 interim relief including transfer; Section 28 saving of other laws | Modified: the initiative moved from the employer to the woman, with the employer assisting | The employer’s own duty to set the criminal law in motion, and the express bar on victimising witnesses as a guideline obligation |
| Head 5: disciplinary action where the conduct is misconduct under the service rules | Section 13 action on the inquiry report, including deduction from salary where service rules do not govern | Adopted, with a monetary route the guidelines did not contain | Nothing material |
| Head 6: a complaint mechanism in every organisation, ensuring time-bound treatment | Sections 9 and 11 complaint and inquiry, with a three-month complaint window and a ninety-day inquiry | Adopted and given clocks the guidelines never fixed | Nothing; the guidelines’ time-bound aspiration became enforceable periods |
| Head 7: Complaints Committee headed by a woman, not less than half women members, a third party from an NGO or other body familiar with the issue, a special counsellor and confidentiality | Section 4 Internal Committee and Section 6 Local Committee, each with an external member and a majority framework of women members | Adopted and split into two tiers by workplace size and sector | The special counsellor as a named figure in the mechanism, and the Committee’s constitution in every workplace regardless of size |
| Head 7: the Committee’s annual report to the government department concerned, and employer compliance reporting to the same department | Section 21 and Section 22 annual reports to the employer and the District Officer | Modified: the addressee changed from a government department to a named district officer | Reporting to a government department as the supervisory destination |
| Head 8: workers’ initiative; employees to raise the issue at workers’ meetings and employer-employee meetings | No counterpart | Dropped | The only guideline addressed to employees themselves, and the only collective, voice-based element in the regime |
| Head 9: awareness, by prominently notifying the guidelines | Section 19 duties of awareness and orientation | Adopted as employer duties | Nothing material |
| Head 10: third party harassment; employer to assist the affected person with support and preventive action | The workplace definition in Section 2(o) and the employer’s duty to assist | Adopted indirectly and partially | A free-standing employer duty framed for harassment by outsiders as a distinct head |
| Head 11: Central and State Governments requested to consider legislation for the private sector | The Act itself, applying to the private and unorganised sectors | Spent: answered by enactment | The request form itself, which recorded that in 1997 the Court could request of legislatures what it could not compel |
| Head 12: savings for rights under the Protection of Human Rights Act, 1993 | Section 28: the Act is in addition to and not in derogation of any other law | Adopted in generalised form | The specific reference to the human rights statute |
| No guideline counterpart: the ten-employee threshold for an Internal Committee | Section 4 | New | The guidelines knew no threshold; the Committee was required wherever women worked |
| No guideline counterpart: the Local Committee and District Officer machinery | Sections 5 to 8 | New | Nothing in the guidelines built district administration |
| No guideline counterpart: conciliation, with a bar on monetary settlement as its basis | Section 10 | New | The guidelines provided procedures for resolution and settlement without a conciliation stage or its safeguard |
| No guideline counterpart: interim relief, including leave of up to three months | Section 12 | New in this form; the transfer element descends from Head 4 | The guidelines gave the transfer option without the leave entitlement |
| No guideline counterpart: compensation determined on stated factors | Section 15 | New | The guidelines omitted the compensation that CEDAW’s General Recommendation No. 19 had said should be provided |
| No guideline counterpart: penalty on the employer, and action against a false or malicious complaint | Sections 26 and 14 | New | The guidelines created duties with no penalty for the employer who ignored them, the regime’s deepest structural weakness |
Three conclusions fall out of the last column, and they are conclusions about legislation rather than about the guidelines. The first is that what the statute dropped was, with one exception, machinery of transmission rather than substance: the standing-orders bridge, the conduct-rule amendment direction, the report to a government department. A statute does not need transmission machinery of that kind because it binds directly, and the drops are the predictable consequence of converting judge-made norms, which had to colonise existing instruments to reach private workplaces, into enacted law that reaches them by its own force. The second conclusion is that the one substantive drop is head 8, the workers’ initiative. Every other guideline obliged an employer or a government; head 8 alone obliged nobody and empowered the employees themselves, entitling them to raise sexual harassment at workers’ meetings and requiring the subject to be affirmatively discussed in employer-employee meetings. Nothing in the Act reproduces it. The statute’s awareness duties are things an employer does to and for employees; the guideline contemplated something employees do, in a forum the employer must open. A reader can judge whether the loss is large, but it is the only place where the enacted regime is narrower in kind than the judicial one, and it passes unnoticed in accounts that treat the Act as the guidelines plus penalties.
The third conclusion runs the other way and is the table’s real surprise. The guidelines’ most serious omission was not Parliament’s discovery; it was visible against the judgment’s own source. General Recommendation No. 19, which the bench quoted, stated that effective complaints procedures and remedies, including compensation, should be provided. The guidelines built the procedures and omitted the compensation. For sixteen years a woman who proved harassment before a Complaints Committee could obtain disciplinary consequences for the respondent and nothing for herself beyond what a transfer or a service-rule penalty incidentally produced. Section 15 of the Act, directing the committee to determine compensation on factors that include the mental trauma, pain, suffering and emotional distress caused, the loss of career opportunity, the respondent’s income and financial status and the feasibility of payment in a lump sum or in instalments, is therefore not an ornament on the judicial scheme. It is the repair of a gap the judicial scheme carried from its first day, and it illustrates the general proposition this article presses throughout: the guidelines were a remarkable instrument for what a court could build unaided, and their limits were exactly the limits of the judicial tool kit, which contains duties and forums but not budgets, penalties, compensation funds or district administrations.
The first test: the Apparel Export Promotion Council case
The guidelines were eighteen months old when the Court was asked to apply their underlying definition in a contested disciplinary case, and its answer in Apparel Export Promotion Council v A. K. Chopra, decided on 20 January 1999 and reported at (1999) 1 SCC 759, is the first of the two later decisions the brief for this article requires to be stated precisely. A. K. Chopra was private secretary to the chairman of the Apparel Export Promotion Council. A woman clerk-cum-typist complained that he had pressed her to accompany him to a hotel business centre to take dictation she was not trained to take, and there attempted to molest her. A departmental inquiry found the charge proved and he was removed from service. The Delhi High Court, examining his challenge, reasoned that he had tried to molest her but had not molested her, since no actual physical contact was established, and it softened the outcome accordingly. The Supreme Court, in a judgment delivered by Chief Justice A. S. Anand, set that reasoning aside and restored the dismissal.
The questions the case framed are stated at the head of the judgment, and they are the questions the guidelines had left open. Does an action of a superior against a female employee which is against moral sanctions and does not withstand the test of decency and modesty amount to sexual harassment? Is physical contact with the female employee an essential ingredient of such a charge? The Court’s answers did two things at once. On the definition, it held that physical contact is not an essential ingredient: the Vishaka enumeration had always included conduct with no contact at all, sexually coloured remarks and the showing of pornography among them, and the catch-all of unwelcome physical, verbal or non-verbal conduct of a sexual nature, so a reading that required contact would have deleted most of the definition the guidelines had laid down. The case is therefore best understood not as an extension of Vishaka but as a refusal to let a High Court contract it, delivered at the first moment contraction was attempted. On the disciplinary standard, the Court held that in cases of this character the courts are required to examine the broader probabilities of the case and not to be swayed by insignificant discrepancies or narrow technicalities, and that the High Court’s distinction between trying to molest and molesting, deployed to save the respondent’s employment, inverted the inquiry: an attempt by a superior, abusing his position over a subordinate sent where she did not belong on work she could not do, was itself the misconduct.
The case also fixed the appellate discipline for the regime’s first era. Punishment, the Court held, is for the competent authority to decide, and a court will not interfere with a penalty unless it is so disproportionate to the proved misconduct as to shock the conscience; a reduction of sentence by a High Court in such a case would have a demoralising effect and be a retrograde step; and a last-stage apology does not call for sympathy or mercy. Read together with Vishaka, the two decisions define the operating posture of the guidelines period. The substantive law was the guidelines’ definition, construed against narrowing. The procedural law was the employer’s own disciplinary process, reviewed with deference. The forum was internal, the standard of proof was the civil standard of the departmental inquiry, and the court’s role was to police the boundaries of both. That settlement worked, where it worked, for a class of employers with rulebooks and inquiry officers: government departments, public sector bodies, organised councils and companies. Its limits were the limits of that class, and the case that exposed them took thirteen more years to arrive.
The definition’s second limb: apprehension, disadvantage and the hostile environment
Most litigation under the guidelines was not about the enumerated acts, which are concrete enough, but about the definition’s second limb, and the limb deserves the slow treatment the brief for this article requires, because it is where the judgment’s equality reasoning does its operative work. The limb provides that the enumerated conduct falls within the definition where it is committed in circumstances whereunder the victim has a reasonable apprehension that, in relation to her employment or work, whether she is drawing salary or honorarium or works voluntarily, in government, public or private enterprise, the conduct can be humiliating and may constitute a health and safety problem. It then states when such conduct is discriminatory: when the woman has reasonable grounds to believe that her objection would disadvantage her in connection with her employment or work, including recruiting or promotion, or when it creates a hostile work environment, adverse consequences being liable to follow if she does not consent to the conduct or raises an objection to it.
Three drafting choices inside that passage controlled the regime’s practical reach. The first is the move from the actor’s intention to the victim’s reasonable apprehension and reasonable grounds. The wrong is not defined by what the respondent meant but by what the complainant could reasonably apprehend about the conduct’s effect on her working life, which relieved committees of an inquiry into motive they were ill-equipped to conduct and aligned the test with the discrimination analysis the limb announces: harassment is discrimination because of the price it attaches to objection within an unequal relationship. The second choice is the enumeration of the sites of disadvantage, employment, recruitment and promotion, which extends the definition backwards to the woman not yet hired and forwards to the woman whose advancement is conditioned, and which is why the guidelines were never confined to the treatment of existing employees by their colleagues: a selection process is a workplace event for this purpose. The third choice is the hostile environment clause, which detaches the wrong from any transaction at all. Conduct need not demand anything or threaten anything; a working environment made hostile by unwelcome sexual conduct is itself the violation, and the woman who objects to the atmosphere rather than to a proposition is within the definition. That clause is the reason the Vishaka definition, alone among the legal instruments available in India in its era, could reach the ambient case, the office saturated with pornography and sexually coloured comment in which no individual demand is ever made of the complainant, and it is the element of the guidelines with the strongest claim to have changed Indian legal culture, because it named a harm that the Penal Code’s act-by-act structure could not see.
The statute’s handling of the limb is the most debated drafting decision in the succession, and it belongs to this article rather than to the pillar because it is a decision about the judgment’s text. Section 2(n) reproduces the enumerated acts and closes with the same catch-all, any other unwelcome physical, verbal or non-verbal conduct of sexual nature, but the apprehension-and-disadvantage passage does not appear in the definition. It reappears, transformed, in Section 3(2), which provides that a set of circumstances may amount to sexual harassment if they occur or are present in relation to or connected with any act or behaviour of sexual harassment: an implied or explicit promise of preferential treatment in employment, an implied or explicit threat of detrimental treatment, an implied or explicit threat about present or future employment status, interference with work or the creation of an intimidating or offensive or hostile work environment, and humiliating treatment likely to affect health or safety. Two changes ride on the relocation. The circumstances list is expanded and particularised, gaining the employment-status threat and the health-and-safety formulation in express terms. And the connector is softened: where the guidelines made the circumstances part of what rendered the conduct discriminatory, and so part of the wrong itself, the Act makes them circumstances that may amount to sexual harassment in connection with the enumerated behaviour, leaving committees and courts a discretion the guideline text did not obviously confer. Defenders of the statutory arrangement treat the relocation as clarification, separating the act from its aggravating context so that either can be proved cleanly. Critics treat it as dilution, because a committee minded to dismiss a complaint can reason that the conduct occurred but the circumstances did not amount to harassment, a two-step the guidelines foreclosed by building the circumstances into the definition. The pillar article states the section as it operates; the judgment call between the two readings belongs here, and this article’s verdict is that the critics have the better of the textual argument and the defenders the better of the practical one: the definition’s centre of gravity moved, but the hostile environment survived the move in express words, and the committees that wished to find harassment under the guidelines found it under the Act on materially the same facts.
Sixteen years: the interval the guidelines had to govern
The judgment closed with directions expressed to operate until suitable legislation is enacted to occupy the field, and with head 11’s request that governments consider legislation. The interval that followed is the strongest evidence in the entire debate about judicial legislation, and it must be narrated accurately, because both sides of that debate quote it. The defence quotes it to show that the Court had no choice: the legislature, invited to act, did nothing for over a decade, and a court that had vacated the field in deference to legislative primacy would have left the fundamental right it had declared unenforced for the whole of that period. The critics quote the same interval to show what interim judicial law becomes in practice: a regime nobody owns, which no legislature feels pressure to replace precisely because the court has supplied something, and which employers treat as hortatory because its enforcement machinery is a committee inside their own offices and a report to a department that may never read it.
The legislative record of the interval is thin enough to state completely. For ten years after the judgment there was no bill at all. A first bill on the subject was introduced in 2007 and lapsed. The measure that became law, the Protection of Women Against Sexual Harassment at Workplace Bill, was introduced in the Lok Sabha on 7 December 2010, thirteen years after the guidelines, and was referred to the Parliamentary Standing Committee on Human Resource Development, which reported in 2011. In 2012 the government amended the bill to bring domestic workers within its protection, a constituency the guidelines’ workplace generality had arguably covered all along but whom an unamended bill would have left to the criminal law. The Lok Sabha passed the bill on 3 September 2012 and the Rajya Sabha on 26 February 2013. Between the two passages, the interval acquired its final and strangest feature. The Delhi gang rape of December 2012 produced a committee headed by Justice J. S. Verma himself, the author of the guidelines, and its report of January 2013 addressed the pending bill directly and recommended that the committee system the guidelines had designed be replaced by an employment tribunal, and that the bill’s provision penalising false complaints be reconsidered. Parliament retained the committees and retained the false-complaint provision in qualified form. The guidelines’ own author, given one formal opportunity to revise his creation fifteen years on, proposed to replace its central institution; the legislature, legislating at last, kept the institution and declined the revision. Any account of the sixteen years that presents them as a simple story of judicial creation followed by legislative ratification has to accommodate that exchange, in which the court-side actor argued for redesign and the legislature conserved.
Measured against the guidelines’ performance rather than its passage history, the interval resolves into three phases. In the first, running roughly to the turn of the decade, the guidelines did the work the Court had intended in the organised sector: service rules were amended, committees were constituted in government and in much of the organised private sector, and the definition entered the working vocabulary of Indian employment practice. In the second phase, the regime’s dependence on voluntary institutional energy showed. A committee is a continuing institution, not an event; it must be reconstituted as members leave, its external member must be found and retained, its reports must be written and sent, and none of these acts is performed by the judgment that ordered them. The guidelines carried no penalty for the employer who let the machinery lapse, no inspectorate to detect the lapse and no fund from which any of it was paid. Compliance therefore decayed unevenly, invisibly and without legal consequence, and the decay was worst exactly where the need was least visible to metropolitan professional life: smaller establishments, educational institutions outside the large universities, district-level public bodies and the unorganised sector, which the guidelines in principle covered and in practice barely reached. In the third phase, the Court was drawn back, not by a fresh scandal about the definition, which nobody any longer disputed, but by documentation of the decay itself, assembled state by state by petitioners who had spent years extracting compliance information from governments. That documentation produced the second of the two later decisions, and it is the hinge between the guidelines era and the statute.
What did the Court ask governments to do for the private sector?
Head 11 of the guidelines requested the Central and State Governments to consider adopting suitable measures, including legislation, to ensure that the guidelines were observed by employers in the private sector, in the form of a request because the Court could direct compliance with law it had declared but chose to invite, rather than command, the legislative act itself.
The wording of head 11 is a small masterclass in the judgment’s self-discipline, and it is worth pausing on why the bench used the language of request in the one head addressed to legislatures. The guidelines bound employers because the Court declared them to be law under Article 141; no corresponding device could bind a legislature to legislate, and the Court did not pretend otherwise. What it could do, and did, was place on the record, in a judgment every law ministry in the country would read, the identification of the gap and the identity of the organ responsible for closing it. Sixteen years later the request was answered by a statute that covered the private sector comprehensively, and the delay in answering it is the legislature’s fact to carry, not the Court’s. But the request form also exposes the asymmetry at the regime’s heart, which the objection section below presses: a court can create obligations for millions of private employers by declaration, while the corresponding act of the democratic legislature required thirteen years, a standing committee, a cabinet amendment and a national crisis of confidence in the state’s protection of women to accomplish.
The guidelines enter the rulebooks: the organised sector, 1998 to 2004
The speed with which the central government moved inside its own employment is the interval’s most underreported fact, and it complicates the simple story of a regime universally ignored. Within six months of the judgment, the Department of Personnel and Training circulated the guidelines to all ministries and departments by an office memorandum of 13 February 1998, directing strict compliance, and by the same instrument the Central Civil Services (Conduct) Rules, 1964 were amended to insert Rule 3-C, declaring sexual harassment of working women to be misconduct on the part of a government servant, published in the Gazette of India as G.S.R. 49 on 7 March 1998. The rule’s two limbs track the guidelines exactly: no government servant shall indulge in any act of sexual harassment of any woman at her workplace, and every government servant in charge of a workplace shall take appropriate steps to prevent it; and the explanation to the rule reproduces the Vishaka definition, physical contact and advances, demand or request for a sexual favour, sexually coloured remarks, showing pornography, and any other unwelcome physical, verbal or non-verbal conduct of a sexual nature. Head 3(b) of the guidelines, the direction that government conduct rules should prohibit the conduct and provide penalties, had been executed by its largest addressee inside a single administrative season.
The second instalment took longer and mattered as much. A conduct rule makes the conduct misconduct; it does not decide who inquires. In 2004 the Central Civil Services (Classification, Control and Appeal) Rules, 1965 were amended, by a notification of 1 July 2004 published as G.S.R. 225 on 10 July 2004, to provide that the Complaints Committee contemplated by the guidelines would be treated as the inquiring authority for the purposes of the disciplinary rules and would follow the procedure those rules lay down for holding an inquiry into a complaint of sexual harassment. That amendment is the organised sector’s answer to the question Medha Kotwal Lele would later put to the states in 2012, when the Supreme Court directed every state to amend its conduct rules so that the committee’s report would be deemed an inquiry report rather than a preliminary fact-finding: the Centre had built that bridge for its own servants eight years earlier. The sequence is a study in how the guidelines regime actually propagated: not by the force of Article 141 operating on millions of employers at once, but instrument by instrument, rulebook by rulebook, where an employer existed with a rulebook to amend and a personnel department to amend it.
The contrast with the rest of the field is the analytical point. In central government employment, the three transmission steps the guidelines required, circulation of the norms, amendment of the conduct rules, and integration of the committee into the disciplinary procedure, were each performed by identifiable instruments on identifiable dates. In the states, the same steps were performed late, partially or not at all, which is why the 2012 audit is a state-by-state document. In the private sector, head 3(c)’s standing-orders route depended on each employer’s certified standing orders being amended establishment by establishment, a process no single instrument could complete, and in the unorganised sector there were, in the main, no standing orders, no conduct rules and no personnel departments at all: the transmission machinery the judgment had designed presupposed the organised employment relation at every joint. The guidelines were therefore strongest where their beneficiary was a government servant or an employee of a large formal employer, and weakest where work was casual, dispersed and undocumented. That gradient of protection, tracking the formality of the employment relation rather than the exposure to the wrong, is the guidelines era’s deepest structural criticism, and it is a criticism of coverage rather than of content. The statute’s district machinery, its Local Committees and its application to workplaces with fewer than ten employees and to domestic work, is best read as the legislature’s attempt to flatten exactly that gradient, fifteen years after the judgment had drawn it.
The audit: Medha Kotwal Lele and the return of the Court
Medha Kotwal Lele v Union of India, decided on 19 October 2012 and reported at (2013) 1 SCC 297, is the judgment that closed the guidelines era by auditing it. The petitioners had spent years documenting, state by state and institution by institution, what the Vishaka directions looked like fifteen years after they were declared to be law, and the record they placed before the Court is the single most important enforcement document of the period, because it converted an assumption every participant in the debate had been making into findings a court recorded. Several states had not amended their civil services conduct rules to treat sexual harassment as misconduct and to give the Complaints Committee’s report the status the guidelines presupposed. Complaints Committees had not been constituted across large classes of public bodies. The third-party member, the design feature on which the guidelines rested their answer to the capture problem, was frequently missing. The Court’s observation, in the passage by which the case is remembered, was that lip service, hollow statements and inert and inadequate laws with sloppy enforcement are not enough for the genuine upliftment of women, and that the guidelines in Vishaka should not remain symbolic.
The directions that followed were supervisory in form and administrative in content, and their specificity is the measure of how elementary the defaults were. States and union territories that had not amended their Civil Services Conduct Rules were directed to do so within two months, providing that the report of the Complaints Committee shall be deemed to be an inquiry report in a disciplinary action, so that a disciplinary authority would treat the committee’s findings as the findings in an inquiry into the misconduct of the delinquent employee rather than as a preliminary investigation on which a second inquiry would have to be built. The Industrial Employment (Standing Orders) Rules were to be amended on the same lines within the same period, discharging at last, fifteen years on, the transmission work head 3 of the guidelines had assigned to them. States were directed to form an adequate number of Complaints Committees so as to ensure that they function at taluka level, district level and state level, each headed by a woman and, as far as possible, with an independent member associated, replacing the single state-level committee some states had treated as sufficient. State functionaries and public and private sector undertakings were directed to put in place sufficient mechanism to ensure full implementation, including provision that a complainant found to have been harassed should not be forced to work with or under the harasser, that transfer of the harasser be considered, and that harassment and intimidation of witnesses and complainants be met with severe disciplinary action. And the professional regulators were brought inside the regime: the Bar Council of India was directed to ensure that bar associations and enrolled advocates followed the guidelines, and the Medical Council of India, the Council of Architecture, the Institute of Chartered Accountants, the Institute of Company Secretaries and other statutory institutes were directed to ensure compliance by the organisations and persons registered with or affiliated to them.
Two features of Medha Kotwal Lele fix its place in the doctrine. The first is what the Court did not do. Confronted with proof that its own fifteen-year-old regime had been widely ignored, it did not redesign the regime, did not attach penalties it had no power to fund or enforce, and did not convert the guidelines into a contempt-driven audit of named defaulters. It repeated the original directions with deadlines and with a widened circle of addressees, and it did so expressly as a holding operation, in terms that the further directions were necessary until legislative enactment on the subject is in place. The judgment is thus the Court’s own verdict on the limits of its 1997 instrument: declaration plus exhortation had achieved a norm and failed to achieve a system, and the missing ingredients, sanctions, administration and money, were ingredients only a legislature could supply. The second feature is timing. The decision was delivered in October 2012, after the Lok Sabha had passed the bill and while it awaited the Rajya Sabha, so the Court was auditing a regime whose statutory replacement was already halfway through Parliament. Read in sequence, Vishaka creates the obligation, Apparel Export defends its content against narrowing, and Medha Kotwal Lele certifies the failure of its enforcement machinery at the moment the legislature finally moves. The three decisions together are the complete judicial biography of the subject before the statute, and no account of the statute’s design is adequate that does not read its penalties, its district machinery and its reporting duties as answers to the specific defaults the 2012 audit recorded.
A complaint under the guidelines: the route as a complainant met it
The regime is usually described from the institution’s side, as a set of duties owed by employers. Its character is clearer from the side of the woman who had to use it, because the guidelines allocated every practical burden of the system to her initiative, and the statute’s later additions read, one by one, as reliefs from those burdens. A woman harassed at work in, say, 2003, in a country where the POSH Act did not exist, held the following legal assets, and only the following. She had a definition, declared to be law, which she could quote to her employer. She had a right to a complaint mechanism in her employer’s organisation ensuring time-bound treatment of complaints, and, where necessary, to a Complaints Committee headed by a woman, at least half of whose members were women, with a third party from a non-governmental organisation or another body familiar with the issue participating to guard against senior-level pressure, with a special counsellor or other support service and with confidentiality maintained. She had a right that, if the conduct amounted to an offence under the Penal Code, her employer would initiate appropriate action in accordance with law by complaining to the appropriate authority, and a right not to be victimised, and to have witnesses protected from victimisation, while her complaint was dealt with, together with the option to seek her own transfer or the perpetrator’s. And she had the disciplinary route: if the conduct was misconduct under the applicable service rules, her employer was to initiate disciplinary action in accordance with those rules, in which, in government service after 2004, the committee itself would be the inquiring authority.
What she did not have defines the regime as sharply. She had no statutory clock of her own: the guidelines required time-bound treatment without fixing a period within which she must complain or the committee must conclude, so a complaint about conduct long past met no limitation bar, and an inquiry that drifted met no deadline either. She had no interim relief as an entitlement: the transfer option was hers to seek, but there was no leave provision, no restraint on the respondent’s contact with her framed as a right, and no direction restraining publication or retaliation beyond the guideline’s own words. She had no compensation: a successful complaint produced consequences for the respondent, a penalty under the service rules, a transfer, a warning, a dismissal in the gravest cases, but the guidelines routed no money to the woman whom the process had vindicated, notwithstanding that the CEDAW recommendation the judgment quoted had said remedies including compensation should be provided. She had no appeal structure: the committee reported within her employer’s own disciplinary universe, and her remedies against a committee that exonerated the respondent, or an employer who shelved the report, lay in the ordinary public law of the writ courts, a remedy priced and paced for institutions rather than for individual employees. And if her workplace had simply never constituted a committee, her asset was the judgment itself: she could invoke Article 141 in a High Court and ask it to direct her employer to obey the law the Supreme Court had declared, which is a real remedy and was used, but one whose cost and duration placed it beyond the ordinary run of complainants.
The comparison this section invites is not between a bad regime and a good one. It is between a regime that a court could create unaided and the additions that required a legislature, and the statute’s text becomes legible as a schedule of the guidelines’ deficits when read against the complainant’s route. The three-month complaint window with its recorded extension answers the absence of clocks, from the side of certainty rather than generosity. The ninety-day inquiry, the sixty-day employer duty to act on the report, and the appeal the Act provides answer the drift and the absent appeal structure. Section 12’s interim relief, transfer of either party and leave of up to three months in addition to the leave otherwise due, answers the missing entitlements. Section 15’s compensation answers the omission the judgment’s own source had flagged in 1992 and the judgment itself had not repaired in 1997. And the Local Committee with its District Officer answers the complainant whose workplace had no committee at all, the case the guidelines left to the writ courts. A legal-information point belongs here rather than in a footnote: the routes described in this article are historical in part, the guidelines having been displaced for workplaces the Act covers, and the position on any live matter depends on the statute and the rules under it, not on the guidelines, a distinction committee members and employers still blur when they cite Vishaka as though it remained the operative text.
The deemed-inquiry-report direction deserves its own emphasis, because it is the most technically consequential of the Medha directions and the least understood. A Complaints Committee that inquires and reports occupies an uncertain place in a disciplinary system built around the inquiry officer: if its report is a mere preliminary fact-finding, the employer must run a second, full inquiry before acting, duplicating the complainant’s ordeal and halving the committee’s authority; if its report is the inquiry, the disciplinary authority proceeds directly to penalty on its findings. Vishaka had presupposed the second position without building it, and the Centre had built it for its own servants in the 2004 amendment of the appeal rules. Medha’s direction that every state amend its conduct rules to the same effect within two months was therefore the audit’s practical core: it converted a fifteen-year ambiguity, exploitable by every employer who wished to re-open a committee’s findings, into a rule. The Act of 2013 settled the question at the statutory level for its own committees by providing for action on the inquiry report in accordance with the service rules, and the episode stands as the clearest instance in the whole history of a proposition this article has pressed from several directions: the guidelines repeatedly identified the right destination and just as repeatedly lacked the transmission machinery to arrive there, and the history of the subject is the history of that machinery being built, late, by other hands.
What a breach cost: enforcement in the guidelines era
The enforcement architecture of the guidelines can be stated in one sentence, and its brevity is the finding: the regime created duties of the most detailed kind and attached to their breach, for the employer who simply ignored them, no penalty at all. A respondent found guilty faced the disciplinary consequences of his service rules, up to dismissal, as the Apparel Export case demonstrated. A victim or a witness faced a guideline prohibition on victimisation. But an employer who never constituted a committee, never circulated the prohibition, never filed the annual report and never amended a standing order faced, in the text of the guidelines themselves, no fine, no prosecution and no named consequence. The duties bound as law under Article 141, and the theoretical instruments for vindicating declared law existed: a High Court could be moved under Article 226 to direct a defaulting employer to comply, and disobedience of a court’s direction in a given case could be met in contempt. But both instruments require an applicant, a filing and a hearing, and they convert every act of routine non-compliance into litigation. A regime whose enforcement depends on each default being separately litigated by the class of persons the regime exists to protect has, in structural terms, no enforcement system; it has a cause of action.
The contrast usually drawn is with the criminal law, which the guidelines engaged only obliquely through head 4, but the sharper contrast is internal to employment law itself. The standing-orders and conduct-rule instruments into which the guidelines were transcribed carried their own disciplinary logic for the individual delinquent, and none for the institution. Rule 3-C of the Central Civil Services (Conduct) Rules illustrates the asymmetry precisely: it made the harassing government servant guilty of misconduct, and it obliged the government servant in charge of a workplace to take appropriate steps to prevent harassment, but the consequence of the second limb’s breach was itself a disciplinary matter inside the same employer’s discretion. The regime’s sanctions, at every joint, ran against individuals for acts of harassment and stopped at the threshold of institutional default, which is the default that determines whether any individual complaint will ever be heard fairly. Medha Kotwal Lele’s state-by-state record is, on analysis, a catalogue of institutional defaults, unamended rulebooks, unconstituted committees, missing external members, and the directions it issued were framed as directions because no penalty provision existed under which a single defaulting institution could have been proceeded against in the ordinary way.
This is the structural fact that the judicial legislation debate habitually misses, because the debate is conducted at the level of authority rather than of machinery. The objection says the Court legislated; the defence says the Court enforced rights. Both descriptions pass over the point that what the Court produced was legislation with the sanctions chapter missing, for reasons internal to the judicial office: a court declaring interim norms under Article 32 can direct conduct, but the creation of a penal or fiscal consequence for breach, a fine payable to the state, a licence consequence, an inspectorate to detect defaults, belongs to the class of acts that require a legislature’s authority and a budget’s support, and the Vishaka bench, to its credit, did not attempt them. The result was a regime whose normative completeness disguised its coercive incompleteness for sixteen years, and whose history is therefore the cleanest available demonstration that the difference between a judgment and a statute is not primarily a difference of democratic pedigree. It is a difference of equipment. The statute’s Section 26, fining the defaulting employer fifty thousand rupees and visiting repetition with enhanced penalty and the possible cancellation of the licence or registration on which the business depends, is a modest provision, and the pillar article records how unevenly even it has been used. But its presence in the Act and its necessary absence from the guidelines mark the exact boundary of what the judicial process could build in 1997, and they explain why the guidelines, for all their authority, were always going to be a bridge: a bridge is judged by whether traffic can cross it, and the traffic of enforcement could cross this one only in the organised sector, on the rulebooks, at the pace litigation allowed.
From guideline to bill: the drafting decade inside the interval
The legislative half of the interval rewards a closer account than the bare dates given above, because the bill that finally emerged was not drafted from first principles and its drafting choices are unintelligible except as decisions about the guidelines’ text. The organised campaign for a statute ran through the whole period: women’s organisations that had litigated and served on the committees the guidelines created, working with the National Commission for Women, pressed governments for an enactment that would give the regime the penalties and machinery the judgment could not. The first bill, introduced in 2007, lapsed without passage. The second, the Protection of Women Against Sexual Harassment at Workplace Bill introduced on 7 December 2010, was referred to the Parliamentary Standing Committee on Human Resource Development, whose examination produced a report in 2011, and the bill as eventually passed incorporated many of the amendments the Committee recommended. A drafting decision taken in 2012 illustrates the process at work: the bill as introduced had excluded domestic workers, notwithstanding that a dwelling-house is a workplace to the woman employed in it and that the guidelines’ own coverage, framed for workplaces and other institutions without a sectoral carve-out, reached her in principle; in May 2012 the Union Cabinet approved an amendment to include domestic workers, and the enacted statute serves them through the Local Committee, since a household cannot constitute an internal one. The sequence is a small parable of the succession: a judicial text of general application, a bill that narrowed it by oversight, a correction extracted through the committee and cabinet process, and a statutory solution, the district forum, that the judicial text had never imagined.
The contest over the bill’s design, where it is documented, ran on three lines, and each line is a decision about the guidelines. The first was the forum. The guidelines’ internal committee was attacked from two directions that rarely agreed with each other: from the side of complainants, on the ground that an internal body, however composed, answers to the employer whose reputation the complaint damages, and from the side of employers, on the ground that the committee system imposed an adjudicative burden on bodies without adjudicative training. The Justice Verma Committee’s report of January 2013, delivered after the Lok Sabha had passed the bill and while the Rajya Sabha considered it, gave the first line its most authoritative statement, recommending an employment tribunal in place of the committee system. Parliament’s refusal to follow it is the single clearest evidence that the legislature of 2013 legislated independently rather than transcribing either the Court of 1997 or the committee of 2013: it examined the judicial design, heard its own author propose its replacement, and kept the design, on the stated logic of the bill’s defenders that a forum inside the workplace, reformed by an external member and statutory clocks, would be used by women whom a distant tribunal would deter. The second line was the false-complaint provision, which the Verma Committee also asked Parliament to reconsider and which the Act retained in a qualified form, penalising a complaint found malicious rather than merely unproved, with an express protection that inability to substantiate a complaint need not attract action. The third line was coverage, and here the critique was pressed in public by the All India Democratic Women’s Association, whose statement on the Lok Sabha’s passage of the bill on 3 September 2012 welcomed a comprehensive definition and the mandatory internal mechanism while recording the gaps that remained: women agricultural workers, forming a large segment of the unorganised labour force, were not securely within the bill; women in the armed forces were left outside; and the ten-employee threshold for the internal mechanism bore hardest on the unorganised sector where protection was thinnest. Those three gaps map, almost exactly, onto the gradient this article identified in the guidelines era itself, protection tracking the formality of employment, which suggests that the gradient was never solely a judicial artefact: two different lawmaking processes, fifteen years apart, reproduced it, because both built their machinery on the organised workplace and reached beyond it only by special extension.
Set against that record, the description of the 2013 Act as the codification of Vishaka is accurate about its core and misleading about its margins, and the distinction is the one the mapping table in this article was built to police. The definition, the committee, the duties and the savings came across from the judgment. The penalties, the compensation, the clocks, the district tier and the conciliation bar had no source in the judgment and answered defects the judgment’s regime had displayed in operation. The workers’ initiative went the other way, present in the judgment and absent from the Act, and the coverage gradient survived both instruments. A legislative history that calls the Act a ratification flatters the legislature’s originality where it conserved and understates it where it repaired; a judicial history that calls the guidelines a complete code flatters the Court where it improvised transmission machinery and understates the two things, compensation and sanctions, that it left out from the beginning. The sixteen years are best read as a single, discontinuous drafting process conducted by two institutions in sequence, each supplying what the other’s tools allowed, and the seams between their contributions are exactly where the table in this article puts them.
The statute arrives, and what kind of arrival it was
The Sexual Harassment of Women at Workplace (Prevention, Prohibition and Redressal) Act, 2013, Act 14 of 2013, received the President’s assent on 22 April 2013 and commenced on 9 December 2013, together with the Rules framed under it, sixteen years and four months after the judgment whose guidelines it displaced. Its full anatomy, its committee architecture, its clocks, its penalties and its implementation record belong to the POSH Act 2013 pillar article, which owns the statute; what belongs to this article is the character of the succession. Parliament did not design a workplace harassment law and then discover the guidelines. It enacted, in the main, the guidelines themselves: the definition carried over near verbatim, the committee architecture carried over with its woman chair, its majority of women and its external member, the employer’s preventive and awareness duties carried over, and the parallel survival of the criminal law carried over through a general savings provision. Onto that inherited frame the legislature bolted exactly the components the judicial kit could not contain: a penalty of fifty thousand rupees on the employer who fails to constitute the committee or to act on its report, with enhanced consequences for repetition extending to cancellation of licence or registration; a district administration of District Officers and Local Committees to reach the workplaces an internal committee cannot serve; a conciliation stage with a proviso that no monetary settlement shall be made the basis of it; interim relief including leave; and, repairing the omission the guidelines had carried from their own CEDAW source, compensation determined on stated factors. The succession was therefore neither a ratification nor a repudiation. It was a completion, in which the legislature accepted the judiciary’s substantive settlement of the subject and supplied the enforcement apparatus the judiciary had spent fifteen years demonstrating it could not supply for itself. In the wider map of the field this statute sits inside the guide to women’s rights legislation in India, and its position there, as the one major protective regime in the field that began life as a judgment, is the anomaly this article exists to explain.
Later amendments adjusted the inheritance at the margins: the committee nomenclature was standardised by Act 23 of 2016, which renamed the Internal Complaints Committee as the Internal Committee and the Local Complaints Committee as the Local Committee throughout the Act. The criminal law moved in parallel in the same season as the statute, when the Criminal Law (Amendment) Act, 2013 inserted Section 354A into the Indian Penal Code and made sexual harassment itself a named offence, so that the two instruments the guidelines had contemplated in heads 4 and 5, the criminal proceeding and the workplace proceeding, both acquired statutory form within a single year. What did not follow was the quiet retirement of the judiciary from the field that the succession seemed to promise. Within a decade of commencement, the Supreme Court was again issuing directions in this subject, and the question those directions pose, whether judicial supervision of workplace harassment has ended, is one of the seeded questions of this article’s own territory, answered in the audit that follows rather than asserted here: the author of the obligation had become, under the statute, the auditor of the machinery, and the audit has proved to be a continuing one.
Supervision after the statute: the auditor’s return
The post-statute chapter opens, fittingly, with a case about the fairness of a Vishaka-era committee. In Aureliano Fernandes v State of Goa, decided on 12 May 2023 by a bench of Justice A. S. Bopanna and Justice Hima Kohli, the Supreme Court examined the termination of a Goa University professor after a committee inquiry conducted under the guidelines framework in 2009, found the proceedings had been conducted in a tearing hurry in defiance of natural justice, quashed the termination and remanded the inquiry, and then turned from the case to the field, describing the enforcement of the Act a decade after its commencement as a sorry state of affairs. The directions that followed were directed at the existence of the statutory machinery itself: the Union, the states and the union territories were to undertake a time-bound exercise verifying that ministries, departments, public and statutory bodies, universities, hospitals and institutions had constituted their committees with the composition the Act requires, and to publish committee details and complaint procedures. The Court kept the matter alive, and in orders of 22 October 2024 and 3 December 2024, and continuing through May 2025 and 12 August 2025, it pressed the states to appoint District Officers, constitute Local Committees against stated deadlines, arrange legal services assistance for complainants, conduct district-wise surveys of public and private establishments and upload committee data to the SHe-Box portal, under continuing judicial supervision.
For this article’s purposes the significance of that sequence is doctrinal before it is administrative. In 1997 the Court supervised the content of the obligation because no enacted content existed; it wrote the norms. In 2012 it supervised the transmission of those norms into rulebooks and committee constitutions. After 2023 it supervises the existence of enacted machinery, because the content is no longer in dispute and what fails is construction. The object of supervision has migrated from substance to structure, but the posture is continuous, and the continuity carries a lesson about the whole subject that neither the guidelines’ defenders nor their critics predicted in 1997: the enactment of the statute did not terminate judicial involvement, because the failure mode of this field was never primarily a failure of norms. It was, in 1997, in 2012 and again in 2023, a failure to build and maintain small, unglamorous institutions at an enormous number of sites, and that is a failure courts can expose and shame but cannot themselves repair. The reader who wants the detail of the current compliance directions will find the statute’s own record in the pillar article; the point owned here is that the supervision question and the legislation question turned out to be different questions, and answering the second did not answer the first.
The objection: judicial legislation, stated fairly
The serious criticism of Vishaka has never been that its cause was unworthy or its definition unsound. It is an institutional criticism, and the brief for this article requires it to be given in its strongest form, because it identifies the genuine cost of the technique the judgment normalised. In that strongest form, the objection runs as follows. The guidelines are legislation in everything but the identity of their author. They define a wrong in enumerative detail, impose duties on millions of private parties who were never consulted, design an institution with composition rules and a gender quota, mandate reporting to government, and direct the amendment of service rules and standing orders across two sectors of the economy. Legislation of that character is supposed to emerge from a process with features a courtroom lacks: public consultation, costing, the testing of administrative feasibility by the departments that must run the scheme, amendment in committee, and the political accountability of its authors to those it binds. The Vishaka regime had none of these. Its substantive content was drawn substantially from a treaty that India had ratified but that Parliament, exercising the very power Article 253 confers, had chosen not to enact into domestic law, and the technique of reading the treaty into the fundamental rights achieved indirectly the incorporation the political branches had not performed. The guidelines were framed as interim, but interim law made by a court has no sunset mechanism a legislature can be forced to respect, and it lasted sixteen years, during which its existence arguably reduced the pressure on Parliament to legislate at all: the classic displacement effect, in which judicial gap-filling relieves the legislature of the political cost of the gap. And the precedent generalises. A court equipped with Vishaka’s technique, a rights violation, a legislative void and a sympathetic international instrument, can in principle manufacture a code on any subject the legislature has neglected, restrained only by its own sense of the proper limits of the judicial function, which is the very phrase the judgment borrowed from the Beijing Statement: a limit the constrained institution states for itself.
The defence, stated at equal strength, begins by denying the objection’s premise about the alternative. The choice before the bench in 1997 was not between judicial guidelines and parliamentary legislation. It was between judicial guidelines and nothing, for an interval whose length nobody could then know and which proved to be sixteen years. The fundamental rights were not in dispute: the Union of India, appearing through its Solicitor General, consented to the guidelines and assisted in their formulation, so the regime carries a species of executive concurrence that pure judicial impositions lack. The power exercised was not free-standing: Article 32 is a fundamental right itself and empowers the Court to issue directions for the enforcement of the rights in Part III, and a mandamus that declares a recurring violation and directs nothing to prevent its recurrence is a remedy designed to fail. The guidelines were expressly interim, expressly deferred to legislation, and expressly framed to be displaced the moment Parliament acted, which is the opposite of the judicial entrenchment the objection fears: the Court built the legislature a bridge and left the toll in the legislature’s hands. Parliament remained free throughout to legislate differently, and when it did legislate it changed the design where it wished, separating the definition’s two limbs, redirecting the annual report, adding conciliation and compensation, and rejecting, in the Verma Committee’s tribunal proposal, a redesign urged from the judicial side of the argument. A legislature that can do all that has not been displaced; it has been waited for. On the treaty point, the defence is that the judgment enforced no treaty: it construed the Constitution, observing the dualist discipline, and used the ratified convention only to fix the content of rights whose existence and supremacy are domestic facts. Finally, the defence turns the sixteen years from an embarrassment into the argument’s foundation. The objection assumes a legislature standing ready to act, delayed only by judicial pre-emption. The record shows a legislature that required thirteen years to introduce the bill and sixteen to enact it, in a field where the Court’s norm was already available for copying. Without Vishaka, the women who used Complaints Committees between 1997 and 2013 would have had the Penal Code and nothing else, and the counterfactual the objection must defend is that their rights should have remained unenforced for a generation in order to preserve an institutional boundary the Constitution itself, in Article 32, deliberately qualifies.
A verdict is owed, and this article gives one. The objection succeeds as a warning and fails as a condemnation. It fails as a condemnation of Vishaka itself, because every condition the judgment set for its own technique was satisfied on the facts: a genuine legislative void rather than a legislative choice the Court disliked, a rights violation conceded in substance by the state itself, consent of the Union to the norms framed, an express interim character, and no inconsistency with any enacted law. A technique used under those conditions, with that discipline, is enforcement of the Constitution and not usurpation of Parliament. But the objection succeeds as a warning about the technique’s afterlife, because the conditions are easier to recite than to police, and later invocations of gap-filling rarely reproduce all of them. The sixteen years supply the warning’s sharpest edge from the opposite direction: interim judicial law places the burden of inertia on the legislature, and a legislature content to free-ride on a court’s interim regime will do so indefinitely, converting the exception into the arrangement. The discipline Vishaka actually teaches is therefore double. Courts should fill voids only on the Vishaka conditions. Legislatures should treat a court’s interim regime as an accusation with a date on it, not as a settlement. On the evidence of this field, the first discipline held and the second did not, and the reputation of the judgment has carried the cost of the legislature’s failure ever since.
Vishaka among the guideline judgments: the technique in its family
Vishaka did not invent the guideline judgment, and its authority is strengthened rather than weakened by placing it in its family. Months earlier, in D. K. Basu v State of West Bengal, reported at (1997) 1 SCC 416, the Court had laid down a set of basic requirements to be followed in all cases of arrest and detention until legal provisions were made on that behalf, as a measure to prevent custodial violence: identification of arresting officers, a memo of arrest attested by a witness, intimation to a friend or relative, medical examination, and like safeguards, again sourced in the fundamental rights and again expressed to operate in a legislative void. The two judgments share a grammar: a recurring rights violation, an absent statute, a set of operational requirements declared binding in the interim, and an open invitation to the legislature to occupy the field. The family resemblance explains why constitutional lawyers treat the late nineteen nineties as the high season of the guideline technique, and it also explains the discipline the better examples observe. In each case the guidelines regulate the conduct of the state or of parties the state can reach through existing instruments, they are tied to the enforcement of an identified fundamental right, and they disclaim permanence.
Where Vishaka stands alone in the family is in the identity of its addressees and the completeness of its institutional design. The D. K. Basu requirements bound the police, an arm of the state, in the exercise of a power only the state holds. The Vishaka guidelines bound every employer in India, including purely private parties with no public function, and they did not merely regulate conduct; they constituted a forum, prescribed its composition, assigned it a reporting duty and connected it to the disciplinary and criminal systems on either side of it. That is the step from regulating the state’s own agents to organising civil society’s internal institutions, and it is the step on which the judicial legislation objection fastens with most force. It is also the step that made the guidelines unusable as a template for quick imitation in unrelated fields: a court contemplating a Vishaka in another subject must contemplate designing the forum as well as declaring the right, and the sixteen-year record of the forum in this field, its decay, its audit and its eventual statutory reconstruction, is the standard exhibit in the argument that forum design is where judicial competence runs out. The doctrine that travelled most successfully from the case was not the committee but the canon of construction: the proposition that a ratified international convention may be read into the fundamental rights in the absence of inconsistent domestic law has been cited wherever Indian courts construe statutes and rights in the light of international obligations, including in the interpretation of later welfare legislation, where benches have treated consonance with international law as the preferred reading between two available constructions. The judgment’s most portable product, in the end, was a rule about how to read, not the institution it built.
The afterlife in citation: a judgment that outlived its own regime
A peculiar feature of this subject, noted in the brief for this article and visible in any training room where the law is taught, is that the guidelines continue to be cited as the operative law by people who do not realise that a statute replaced them. Committee members speak of the Vishaka committee, policies reproduce the guidelines’ text beneath a statutory heading, and examination answers describe the guidelines in the present tense a decade after commencement of the Act. The error is understandable, because the Act inherited the guidelines’ content so completely that a document faithful to the judgment is faithful to most of the statute, and it is instructive, because it shows what sixteen years of operation does to a legal instrument’s place in professional memory: the guidelines were the law for the whole of the working formation of a generation of practitioners, administrators and teachers, and instruments learned as law in one’s formation are displaced slowly even by instruments that bind in law. The doctrinal consequence is small but real. Where the guideline and the section differ, in the annual report’s destination, in the conciliation stage, in the complaint window, in the compensation the guidelines lacked, the practitioner working from memory of the judgment will state the law wrongly while believing it correctly cited, and the correction requires exactly the mapping exercise this article’s table performs.
The case’s place in legal education is the other half of the afterlife. Vishaka is among the most examined decisions in Indian constitutional law, set in syllabi not primarily for its subject but for its method: it is the standard teaching vehicle for the use of international law in constitutional interpretation, for the guidelines technique, for the Article 32 remedial jurisdiction and for the separation-of-powers argument in its most concrete form, because every element of the controversy is present in a short judgment that a student can read entire. That pedagogic centrality has a distorting effect the reader should discount. Taught as a method case, Vishaka is remembered for its boldest sentence, the declaration that the guidelines would be treated as law under Article 141, and its sixteen years of administrative grind, the rulebook amendments, the committee constitutions, the annual reports and the 2012 audit, fall away. The judgment becomes a moment. The record this article has assembled is the corrective: a moment that had to be administered for a decade and a half, audited for its failures, and finally completed by the institution it had waited for, and whose meaning lies as much in the administration and the audit as in the declaration.
What remains open, and the practical consequence
Three questions survive the statute, and a judicial interpretation article should close by stating them in their durable form. The first is the residue question: the guidelines were expressed to operate until legislation occupied the field, and the Act of 2013 occupies the field of sexual harassment of women at workplaces, but the reading-in doctrine Vishaka established is not exhausted by the succession, because it states a general rule of construction for any field where domestic law is silent and a ratified convention speaks without inconsistency; the doctrine’s future applications, and its limits, remain live questions of constitutional method that no statute can settle. The second is the supervision question: the post-2023 proceedings have kept workplace harassment under continuing judicial oversight of a structural kind, and the terms on which such supervision ends, what a court should require before it certifies a statutory scheme as built, is a question the orders themselves are still working out. The third is the gender question, which the statute inherited from the guidelines without re-examination: both regimes protect women, on the foundation of Article 15(3), and the claims of complainants outside that boundary rest on employer policy and the general law, a boundary Parliament chose in full knowledge of the neutral alternative and which the guidelines, framed on a petition about working women, never had occasion to test.
For the practitioner and the committee member, the practical consequence of the whole history is compact. The obligations under which a workplace operates descend in a direct line from the 1997 judgment, and the text a committee member is applying is, in its definition and its architecture, the text the Verma bench wrote; the statute added the clocks, the penalties, the district machinery and the compensation that the judgment could not supply. The general position described in this article is legal information and analysis, not advice on an individual case, and anyone acting on a live complaint should take advice on its facts, because limitation periods and forum rules are unforgiving and the committee route and the criminal route run on different clocks from the same events. For the student of the Constitution, the consequence is the one with which this article opened. Vishaka is the leading Indian authority for two propositions at once: that an unincorporated convention may be read into the fundamental rights where domestic law is silent and not inconsistent, and that a court may lay down enforceable norms to fill a legislative vacuum pending legislation. Both propositions were stated by the Court itself, both are contested, and both survived the arrival of the legislation they anticipated, because the statute displaced the guidelines and left the doctrine untouched. Sixteen years of interim law is the name this series gives the episode: the Court framed the guidelines to govern until suitable legislation was enacted, they governed every Indian workplace for sixteen years, and a further decade after the legislation arrived the same Court was supervising whether the committees existed. On this subject, judicially created obligation has been the operative law for far longer than the statute has, and the field will not be understood until that inversion is faced squarely.
Frequently Asked Questions
Q: What did the Vishaka judgment require employers to do?
The judgment required every employer, and every other person responsible at a workplace or institution, to prevent and deter sexual harassment and to provide procedures for the resolution, settlement or prosecution of such acts. In practice that meant notifying, publishing and circulating an express prohibition of sexual harassment; writing the prohibition and penalties into government and public sector conduct rules, and into private employers’ standing orders under the Industrial Employment (Standing Orders) Act, 1946; providing work conditions in work, leisure, health and hygiene that created no hostile environment; initiating criminal proceedings where the conduct was an offence; taking disciplinary action where it was misconduct; and maintaining a complaint mechanism with a Complaints Committee, headed by a woman, with not less than half women members and a third party from a non-governmental organisation or another body familiar with the issue.
Q: How could the Vishaka guidelines be binding without legislation?
The bench rested the guidelines on two constitutional provisions working together. The petitions were brought under Article 32, which empowers the Supreme Court to issue directions for the enforcement of fundamental rights, and the guidelines were laid down in exercise of that power, because a declaration that rights were violated, without directions to prevent a recurring violation, would not enforce them. The Court then directed that the guidelines be treated as the law declared by it under Article 141, the provision making the law declared by the Supreme Court binding on all courts in India. On that footing the guidelines were not advice or a model code but binding and enforceable law, expressed to operate only until suitable legislation was enacted to occupy the field, at which point the statute, and not the judgment, would govern.
Q: How did international conventions shape the Vishaka judgment?
The Court used the Convention on the Elimination of All Forms of Discrimination against Women, which India had ratified in 1993, to fix the content of the fundamental rights it was enforcing. It set out Article 11 of the Convention, on equality in employment including safety in working conditions, and Article 24, the undertaking to adopt all national measures to realise the Convention’s rights, and it drew the working definition of sexual harassment from the CEDAW Committee’s General Recommendation No. 19, which named sexual harassment as gender-specific violence impairing equality in employment. The technique was construction, not incorporation: relying on Article 51(c) and Article 253, the Court held that an international convention not inconsistent with the fundamental rights must be read into them where domestic law is silent, so the obligation enforced remained constitutional, with the Convention supplying its content.
Q: What did the Apparel Export Promotion Council case add to Vishaka?
Decided on 20 January 1999 and reported at (1999) 1 SCC 759, the case was the first application of the Vishaka framework to a contested dismissal, and it added two holdings. A private secretary to the Council’s chairman had been removed after an inquiry found he had attempted to molest a woman clerk-cum-typist; the Delhi High Court reasoned that he had only tried to molest her, since no physical contact was proved. The Supreme Court, in a judgment delivered by Chief Justice A. S. Anand, held that physical contact is not an essential ingredient of sexual harassment, restored the dismissal, and directed that courts examine the broader probabilities of such cases rather than insignificant discrepancies or narrow technicalities. It also held that punishment is for the competent authority, and that a court interferes only where a penalty is so disproportionate as to shock the conscience.
Q: Why did it take sixteen years to legislate on workplace harassment?
No single cause explains the interval, and the record supports several operating together. For the first ten years no bill was introduced at all; a first bill came in 2007 and lapsed, and the measure that became law was introduced only on 7 December 2010, then spent time with the Standing Committee on Human Resource Development, which reported in 2011, and was amended in 2012 to include domestic workers. The guidelines themselves reduced the pressure to legislate, because a functioning judicial regime, however incomplete, removed the visible emergency that forces a subject up the parliamentary agenda. The subject also sat between ministries and constituencies, a labour question, a women’s question and a criminal law question at once, owned fully by none. Passage finally came in the altered atmosphere after December 2012, with the Lok Sabha passing the bill on 3 September 2012 and the Rajya Sabha on 26 February 2013.
Q: Did the POSH Act adopt the Vishaka definition?
In substance, yes, with one structural change. Section 2(n) of the Act reproduces the guidelines’ enumeration almost word for word: physical contact and advances, a demand or request for sexual favours, sexually coloured remarks, showing pornography, and any other unwelcome physical, verbal or non-verbal conduct of a sexual nature. The guidelines, however, carried a second limb inside the definition, treating the conduct as discriminatory where the woman reasonably apprehended disadvantage in her employment, recruitment or promotion if she objected, or where it created a hostile work environment. The Act lifted that limb out of the definition and re-enacted it, expanded, as the free-standing list of circumstances in Section 3(2) that may amount to sexual harassment. The content survived the succession; the connector changed, and commentators dispute whether that relocation diluted the definition or clarified it.
Q: What did Medha Kotwal Lele add?
Medha Kotwal Lele v Union of India, decided on 19 October 2012 and reported at (2013) 1 SCC 297, added an audit and a set of supervisory directions rather than new doctrine. Fifteen years after Vishaka, the petitioners documented state by state that conduct rules had not been amended, Complaints Committees had not been constituted across public bodies and the third-party member was frequently missing, and the Court, holding that the guidelines should not remain symbolic, directed states to amend their Civil Services Conduct Rules and standing orders rules within two months so that a Complaints Committee’s report would be deemed an inquiry report in disciplinary proceedings, to constitute committees at taluka, district and state levels headed by women with an independent member, and directed the Bar Council of India and the professional councils to secure compliance in their fields, all expressly until legislative enactment was in place.
Q: Has judicial supervision of workplace harassment ended?
No. Enactment changed what the Court supervises, not whether it supervises. Under the guidelines the Court supervised the content of the obligation, because it had written the norms; after the Act it supervises the existence of the machinery the Act requires. In Aureliano Fernandes v State of Goa, decided on 12 May 2023, the Supreme Court described the Act’s enforcement a decade after commencement as a sorry state of affairs and directed a time-bound verification that bodies had constituted their committees with the required composition. In later orders in the same proceedings, on 22 October 2024 and 3 December 2024 and continuing through May 2025 and 12 August 2025, it pressed for the appointment of District Officers, the constitution of Local Committees against deadlines, district-wise surveys and the uploading of committee data to the SHe-Box portal, keeping the matter under continuing supervision.
Q: What was the Vishaka case about?
Vishaka v State of Rajasthan was a class action brought under Article 32 of the Constitution by social activists and non-governmental organisations for the enforcement of the fundamental rights of working women. It arose after an incident of grave sexual violence against a woman working on a state social development programme in Rajasthan, and it asked the Supreme Court to address the complete absence of any law protecting women against sexual harassment at workplaces. The Court held that sexual harassment violates the fundamental rights of women, and, finding no enacted law to enforce those rights, laid down guidelines for all workplaces that were to operate as law until Parliament legislated. The case is therefore about two things at once: the rights of working women, and the power of a constitutional court to create binding obligations in a legislative vacuum.
Q: Who decided Vishaka, and when was the judgment delivered?
The judgment was delivered on 13 August 1997 by a three-judge bench of the Supreme Court presided over by the Chief Justice of India, J. S. Verma, sitting with Justice Sujata V. Manohar and Justice B. N. Kirpal. The judgment, reported at (1997) 6 SCC 241 and also at AIR 1997 SC 3011, was delivered by Chief Justice Verma for the bench, on writ petitions filed in 1992. Fali S. Nariman appeared as amicus curiae, Meenakshi Arora and Naina Kapur assisted the Court, and the Solicitor General, T. R. Andhyarujina, appeared for the Union of India, which consented, through him, to the guidelines the Court framed. The bench’s identity matters to the case’s later history: the same Chief Justice, retired, headed the committee whose report of January 2013 recommended replacing the guidelines’ committee system with an employment tribunal, a recommendation Parliament declined.
Q: Which fundamental rights did the Vishaka judgment hold were violated by workplace sexual harassment?
The Court held that sexual harassment at a workplace violates four fundamental rights together. Article 14, equality before the law, and Article 15, the prohibition of discrimination on grounds including sex, are violated because the conduct denies women equal conditions in the sphere where they work. Article 19(1)(g), the right to practise any profession or carry on any occupation, trade or business, is violated because that right depends on the availability of a safe working environment, and a woman priced out of her occupation by harassment is denied the right in substance. Article 21, the right to life and personal liberty, is violated because the right to life means life with dignity, and gender equality includes protection from sexual harassment and the right to work with dignity. The Court also cited Article 15(3), Article 42 and the fundamental duty in Article 51A(e) in support.
Q: What is the difference between a Complaints Committee under the Vishaka guidelines and an Internal Committee under the POSH Act?
The two bodies share a design: each is an internal forum, each is headed by a woman, each must have not less than half its members women under the guidelines and a majority structure of women under the Act, and each includes an outsider, a third party familiar with the issue under the guidelines and an external member from a non-governmental organisation or a person familiar with issues of sexual harassment under Section 4 of the Act. The differences are the statute’s additions. The Complaints Committee was required in every workplace without a size threshold, reported annually to the government department concerned, worked without fixed clocks, and its report’s status in disciplinary proceedings had to be secured by separate rule amendments. The Internal Committee is required only where ten or more persons are employed, smaller workplaces going to the Local Committee; it works to statutory deadlines, its report goes to the employer and the District Officer, and the inquiry carries the powers the Act confers.
Q: Did the Vishaka guidelines apply to private sector employers?
Yes. The guidelines were addressed to employers in workplaces and other institutions without a public sector limitation: the definition itself speaks of work whether in government, public or private enterprise, the preventive steps are prescribed for all employers or persons in charge of a workplace whether in the public or private sector, and private employers were directed to take steps to include the prohibition in their standing orders under the Industrial Employment (Standing Orders) Act, 1946. The eleventh guideline additionally requested the Central and State Governments to consider suitable measures, including legislation, to ensure the guidelines were observed by private sector employers, which reflects a doubt about reach rather than about obligation: the duty bound private employers from 1997, but the transmission instrument for the private sector, the standing order, depended on establishment-by-establishment amendment, and compliance there was correspondingly uneven until the Act applied to the private sector by its own force.
Q: Are the Vishaka guidelines still in force after the POSH Act?
For workplaces the Act covers, the guidelines have been displaced, because they were framed from the beginning to operate only until suitable legislation was enacted to occupy the field, and the Act, which commenced on 9 December 2013, occupies the field of sexual harassment of women at workplaces. A committee member or employer who applies the guidelines as the operative text will state the law wrongly where the two differ: the guidelines fixed no complaint window, no inquiry deadline and no compensation, and sent the annual report to a government department, where the Act prescribes a three-month complaint window, a ninety-day inquiry, compensation under Section 15 and reports to the employer and District Officer. The judgment itself remains fully in force as precedent: its holdings on the fundamental rights violated, and its doctrine that a ratified international convention may be read into the fundamental rights where domestic law is silent and not inconsistent, continue to govern and are cited across constitutional law.
Q: Why is the Vishaka judgment criticised as judicial legislation?
The criticism is institutional, and in its strongest form it says that the guidelines are legislation in everything but their author. They define a wrong in detail, bind millions of private parties who were never consulted, design an institution with composition rules, direct reporting to government and require the amendment of service rules and standing orders, all without the consultation, costing, feasibility testing or amendment process that legislation undergoes, and they rest substantially on a treaty Parliament had not enacted into domestic law. The guidelines were framed as interim yet lasted sixteen years, and their existence arguably reduced the pressure on Parliament to act. The defence answers that the choice in 1997 was between guidelines and no enforcement at all, that Article 32 expressly empowers directions for the enforcement of fundamental rights, that the Union of India consented to the norms, that they were expressly interim and displaced the moment Parliament legislated, and that Parliament, when it came, freely kept, altered and rejected parts of the design.
Q: What did the Justice Verma Committee recommend about the workplace harassment bill in 2013?
The committee headed by Justice J. S. Verma, the retired Chief Justice who had authored the Vishaka judgment, was appointed after the Delhi gang rape of December 2012 and reported in January 2013, while the workplace harassment bill was between the two Houses. On this bill it recommended that the committee system the guidelines had designed and the bill retained be replaced by an employment tribunal, and that the provision penalising false complaints be reconsidered. The Rajya Sabha passed the bill on 26 February 2013 without adopting the tribunal model, retaining the Internal and Local Committees as the forum and retaining the false-complaint provision in qualified form. The episode is singular: the one formal opportunity to revise the guidelines’ design was taken by their own author to propose its replacement, and the legislature, legislating at last, conserved the design instead, a fact both sides of the judicial legislation debate have to accommodate.
Q: Can an international convention be enforced in India without an Act of Parliament after Vishaka?
Not directly, and Vishaka does not hold otherwise. India follows a dualist position in which a treaty does not by itself create enforceable domestic rights; Article 253 empowers Parliament to legislate to implement treaties, and until it does, the treaty is not a statute. What Vishaka establishes is a rule of construction: in the absence of enacted domestic law occupying the field, and where there is no inconsistency with the fundamental rights, the contents of an international convention India has ratified must be read into the fundamental rights to enlarge their meaning and content, and may be used to construe domestic law. The right enforced is then a constitutional or statutory right, with the convention supplying its interpretation, which is why the Court could draw on the Convention on the Elimination of All Forms of Discrimination against Women without treating the Convention itself as Indian law. Where domestic law occupies the field, or the convention conflicts with it, the technique has no application.
Q: Did the Vishaka guidelines give a complainant any compensation?
No, and the omission was visible against the judgment’s own source. The CEDAW Committee’s General Recommendation No. 19, which the judgment quoted, stated that effective complaints procedures and remedies, including compensation, should be provided. The guidelines built the procedures and routed their outcomes through the disciplinary system: a proved complaint could produce a service-rule penalty against the respondent, disciplinary action up to dismissal, a transfer, and the initiation of criminal proceedings, but no payment to the woman who had been harassed, and no mechanism by which her own losses, to health, earnings or career, could be assessed or recovered through the committee. The defect persisted for the whole sixteen years of the regime. The POSH Act repaired it in Section 15, which directs the committee, where it recommends deduction or payment, to determine compensation on stated factors including mental trauma, pain, suffering and emotional distress, loss of career opportunity, medical expenses, and the respondent’s income and financial status.
Q: What could a woman do if her employer ignored the Vishaka guidelines?
Her remedies were real but priced for institutions rather than individuals. The guidelines were law declared under Article 141, so an employer who had not constituted a Complaints Committee, or who refused to act on one, was disobeying binding law, and she could move the High Court under Article 226 for a direction compelling compliance, as petitioners did in the line of proceedings that culminated in the Supreme Court’s audit in Medha Kotwal Lele in 2012. Within her employment, she could invoke the conduct rules and standing orders into which the guidelines had been transcribed, pursue the disciplinary route against the respondent, and insist on the guideline protections against victimisation and on her transfer option. She could also set the criminal law in motion for conduct amounting to an offence. What she could not do was invoke a penalty against the defaulting employer, claim compensation through the committee, or rely on any inspectorate to detect the default: enforcement depended on her own initiative and, in the last resort, on litigation, which is the structural weakness the 2013 Act’s penalties and district machinery were designed to answer.
Q: What did the Vishaka judgment say about the role of the judiciary when the legislature has not acted?
The judgment framed the judicial role as enforcement under constraint. It recorded that the primary responsibility for ensuring safety and dignity at work through suitable legislation, and for creating a mechanism for its enforcement, is of the legislature and the executive, and it treated its own intervention as arising only because instances of violation were brought before it under Article 32, where an effective redressal required guidelines to fill the legislative vacuum. For the legitimacy of that intervention it invoked the Beijing Statement of Principles of the Independence of the Judiciary, accepted by the Chief Justices of the Asia and Pacific region in 1995, whose objectives of the judiciary include ensuring that all persons can live securely under the rule of law and promoting, within the proper limits of the judicial function, the observance and attainment of human rights. The qualifying phrase is the judgment’s own chosen limit: the judiciary acts where rights are violated and the legislature is silent, within the proper limits of its function, and only until the legislature occupies the field.