Sex Discrimination and the Constitution – XI: The Justification of the Anti-Stereotyping Principle

We have seen that in Anuj Garg, the Supreme Court adopted the anti-stereotyping principle: sex-based classifications could not be saved under Article 15(1) if their only justification was to invoke stereotypes about women’s sexual or social roles in the community. What, however, is the basis of this principle? Since the Supreme Court borrowed it from American jurisprudence, we must take a brief detour, and examine the history of constitutional sex discrimination claims in the United States. That history throws up a surprising link: between sex equality, and the right to vote.

Until 1919, women in the United States did not have the right to vote. The denial of this right was justified – among other things – on a theory of virtual representation: that the interests of women were represented (before marriage) by their fathers and (after marriage) their sons, so there was no need for a separate vote. The idea of virtual representation was not restricted to the sphere of voting, but extended to an entire legal regime known as coverture: through which men disposed off property, entered into contracts and engaged in commercial relations on behalf of their wives or daughters. The social philosophy underlying the law of coverture is now called “the separate spheres” theory: i.e., it holds that men and women belong to naturally-ordained separate spheres – the public sphere for men, and the private sphere for women. To perform the functions required of one’s sphere is a natural obligation, and the two spheres are exclusive and non-overlapping.

The denial of the right to vote, therefore, rested upon the legal framework of coverture, which, in turn, was justified by the social theory of the separate spheres. So when the women’s suffrage movement in the United States’ concretely demanded the right to vote, it was not simply asking for access to the ballot box, but challenging the legal regime of coverture and the philosophy of the separate sphere itself. This is evident from the 1848 Seneca Falls Declaration, widely believed to herald the start of the suffrage movement. The Declaration accused man of “claiming it as his right to assign for [woman] a sphere of action, when that belongs to her conscience“, and attempting to “destroy her confidence in her own powers, to lessen herself-respect, and to make her willing to lead a dependent and abject life.” The twin ideas of dependency and an imposed restriction of social roles formed the heart of the claim for suffrage. This was understood by opponents of the movement as well, who linked the right to vote and the transformation of the separate spheres, askingif our ladies will insist on voting and legislating, where, gentlemen, will be our dinners and our elbows? where our domestic firesides and the holes in our stockings?” The American legal scholar, Reva Siegel, argues therefore that “the arguments of suffragists and their opponents tied the idea of women voting to the prospect of women’s emancipation from traditional roles in marriage and the market. Once the question of woman suffrage was infused with this social meaning – once the question of woman suffrage was known simply as the “woman question” – the nation’s debate about whether women should vote turned into a referendum on a whole range of gendered institutions and practices.”

Of course, between 1848 and 1919, the suffrage movement developed multiple currents, not all of which were in harmony. Around the turn of the century, for instance, another strand of the movement began to invoke the separate sphere to justify the claim for suffrage, arguing that because of women’s unique knowledge about issues related to welfare, the bringing up of children, sanitation and hygiene etc., they ought to be allowed the power of the ballot box in shaping policy. A decade later, yet another strand raised the spectre of the recently-enfranchised African-American community overwhelming the Whites at the polls, and asked for the vote to counteract this threat (See Alieen Kraditor, The Ideas of the Woman Suffrage Movement). As is the case with al social movements, it is impossible to tell which strand had the greatest contribution towards ultimate constitutional success. However, what is important to note is that the 19th Amendment, which granted the women the vote, was framed as a right. Neither the second, nor the third arguments for the vote, that we have outlined above, were framed in the language of constitutional principles or rights. It was only the first, and original suffragist argument against the theory of the separate spheres, that was framed in the vocabulary of rights. Consequently, whatever the intentions of the drafters of the Nineteenth Amendment, its very language reflects the constitutional acceptance of the anti-separate spheres movement. More importantly, this is how the Courts understood it – at least initially. In 1923, in Adkins vs Children’s Hospital, the Supreme Court struck down differential working hours of men and women. In so doing, it overruled the pre-Nineteenth Amendment case of Muller vs Oregon (which some of our courts have relied upon), noting that “… the ancient inequality of the sexes, otherwise than physical, as suggested in the Muller Case has continued ‘with diminishing intensity.’ In view of the… revolutionary changes which have taken place since that utterance, in the contractual, political, and civil status of women, culminating in the Nineteenth Amendment… these differences have now come almost, if not quite, to the vanishing point.” Fifty years later, in Frontiero vs Richardson, the judgment which kickstarted the modern American law of sex discrimination, Justice Brennan referred to “traditional belief that the “paramount destiny and mission of women are to fulfill the noble and divine offices of wife and mother”, before stating:

“As a result of notions such as these, our statute books gradually became laden with gross, stereotypical distinctions between the sexes and, indeed, throughout much of the 19th century the position of women was, in many respects, comparable to that of blacks under the pre-Civil War slave codes.”

Justice Brennan’s opinion makes the link between the anti-stereotyping principle and the separate spheres theory. The transformative moment that changed separate-sphere based stereotypes from accepted classificatory tools to unconstitutional, discriminatory ones, was the Nineteenth Amendment, which repudiated virtual representation and its underlying justifications by affirming the right of women to vote.

It is, of course, dangerous to draw connections between jurisdictions in too facile a manner. Two things ought to be noted, however. The first is that the link between the right to vote and the repudiation of separate spheres is a conceptual link, and not jurisdiction-specific. And the second is that a brief look at our pre-Constitutional history reveals some striking similarities. Scholars like Partha Chatterjee and Tanika Sarkar have demonstrated that the public/private divide in the form of ghar/bahir (although in a subtly different form) arose in India towards the end of the nineteenth century, with British efforts at social reform resisted on the grounds of interference with the “inner domain” of community life, which was often represented by the figure of the woman. Chatterjee notes, for instance:

“The world is [deemed to be] a treacherous terrain of the pursuit of material interests, where practical interests reign supreme. It is also typically the domain of the male. The home in its essence must remain unaffected by the profane activities of the material world – and woman is its representation. And so one gets an identification of social roles by gender to correspond with the separation of social space into ghar and bahir.”

This, as we can see, closely corresponds to the separate spheres theory (again, one must be careful of too much reductiveness in comparisons – for the purposes of this argument, however, a rough analogy will suffice). The separate spheres theory spilt over powerfully into the nascent demand for self-representation during the 1920s and 1930s phase of the freedom movement. Initially, during the first opening up of suffrage by the colonial government, separate electorates were proposed for women. As Wendy Singers points out, these “characterized a candidate as a stand-in for her constituency. In other words, separate electorates for women made manifest the idea of a women’s constituency that represented women’s issues and was embodied by the candidate.” (See Singers, A Constituency Suitable for Ladies 25) This was strongly resisted by the All India Women’s Conference (AIWC) and other organisations; on the other hand, as Gerladine Forbes points out, the proposition that only women could represent the interests of the “home” was endorsed by leading figures such as Sarojini Naidu, who urged “women to utilize their housekeeping skills to put the ‘national house’ in order.” This was also reflected – as Forbes notes – in initial demands to restrict suffrage to educated women, who were better placed to advocate social reform. The fact that suffrage was being demanded on two very different grounds, which were based on two incompatible visions of society, was clearly understood by the representatives of the women’s movement during the Second Round Table Conference. Mrinalini Sinha notes that “the representatives speaking on behalf of the Indian women’s movement had insisted that women were neither a “minority” nor a “special interest”, but an integral part of the people… Hence they demanded only universal adult suffrage and a declaration of fundamental rights in the new constitution that removed sex, along with caste, class and religion, as the grounds for any political disqualification.” (Mrinalini Sinha, Spectres of Mother India 223) Here, for the first time, we see the implicit connection between the right to vote, separate electorates, separate spheres, and equality and non-discrimination, being made explicit.

This is, admittedly, a sketchy history; what is worth pointing out, however, is that the Indian Constitution rejected both separate electorates for women and educational qualifications for suffrage. The intentions behind the framers’ decisions are complex, but what matters is that the text of the Constitution merely speaks about adult suffrage. This, in turn, would suggest – based upon our previous arguments – a rejection of the separate spheres theory, much along the lines of the Nineteenth Amendment in the United States.

The anti-stereotyping principle, therefore, is grounded in the transformative nature of the Constitution, which – in simultaneously guaranteeing women the unconditional right to vote along with a guarantee of non-discrimination, rejected separate spheres (and therefore, stereotypes) as justifications for sex-based classifications. Consequently, the line of High Court cases culminating in Anuj Garg was correctly decided, and should be followed in the future. Of course, as Reva Siegel points out, “anti-stereotyping” is an empty phrase without more; to decide whether or not classifications are based on stereotypes needs detailed historical enquiry, tracing the roots of the classifications and their evolution over time. It is an enquiry that the Court is yet to seriously embark upon, but Anuj Garg has, at least, laid the foundation for the future.

(A more detailed version of my argument drawing a link between sex equality, anti-stereotyping, and the right to vote, is available here)


Sex Discrimination and the Constitution – X: The Culmination of the Anti-Stereotyping Principle in Anuj Garg

In the last three essays, we discussed the complex intersections between labour and service laws, and sex discrimination. Let us now return to our original line of cases, which present discrimination claims in a simpler and starker background. In A.M. Shaila vs Chairman, Cochin Port Trust, decided by the Kerala High Court in 1994, the question was whether the Cochin Port Trust’s decision to exclude women from working as shed clerks violated Articles 14 and 15 of the Constitution. The Court held that it did not, noting that if women are excluded from employment of a particular category because of their physical structure and special susceptibilities, it means that women have been placed in a class by reason of the distinct circumstances. In such a case the denial of opportunity of employment, though it strikes at women ceases to be “solely” on the ground of sex.” In noting some of the differences in “physical structure” and “special susceptibilities”, the Court relied upon American cases that had referred to the “natural functions of motherhood” and “social and moral hazards, which had already been pointed out as having been overruled in Rajamma, twelve years before. Summing up, the Court held:

The continuous work while standing or moving and the movement at the shipping wharf amidst the menacing movement of cranes and forklifts demand protective restriction on the right to employment. The policy of the Port Trust indeed protects women from the hazardous effect of such work on their well being. Therefore the policy is not based only on sex. A woman working at the shipping wharf away from the main office, isolated and alone can be an object of violence on her person-especially at night. That is why Curt Muller v. The State of Oregon (supra) used the words” protect her from the greed and passion of man”. The policy of the Prot Trust impugned in this case does not violate Articles 14 and 15(1) of the Constitution of India for these reasons.”

 The problems with this line of reasoning have been examined at length before, and need not be repeated. In the 2000s, however, numerous High Courts were moved against provisions of the Factories Act that allowed governments to prohibit women from working in certain kinds of employment between 7 PM and 6 AM. In 2001, in Vasantha vs Union of India, this was challenged before the Madras High Court. Much like in A.M. Shaila’s case, this was justified on the ground that it was a “benign” measure designed to protect women, so that they could avoid “strenuous work”, and fulfill their household duties. Rejecting the argument, the Court noted that “it is not always so easy to verify whether discrimination that is claimed to be “affirmative action” or “benign” whether really is and at times it is demonstrably established that such a discrimination actually reinforces a negative and untrue stereotype of them.” In other words, it was not enough for the State to simply claim that it was enacting measures to benefit women under Article 15(3) (as the Supreme Court had held earlier, in its adultery decision). The measure itself would be subjected to judicial review, to confirm whether it was based on negative stereotypes. The Court struck down the provision, and also issued guidelines by requiring employers to take steps in order to prevent sexual harassment, provide secure working conditions, separate canteen and transportation facilities for women, etc. That same year, in Triveni vs Union of India, the Andhra Pradesh High Court followed suit, expressly agreeing with the Madras High Court and disagreeing with the Kerala High Court. The Court also observed:

“We have been told that there is a G.O. by which certain safeguards have been provided to the women who are working in Fish industry during the night hours. The same safeguards shall be given to women workers in other industries during the night time.”

Recall Catherine MacKinnon’s observation that the judgment upholding the prohibition of women from working in jails took the viewpoint of the “reasonable rapist.” The Madras High Court’s guidelines, and the Andhra Pradesh High Court’s observation are important, because they make it clear that the State cannot invoke social realities as grounds to burden women, but rather, has a positive obligation to change that reality in order that the need for discrimination disappear. In A.M. Shaila, the Court held that the dangers to women from “the greed and passion” of man formed part of the immutable background conditions (along with the “physical structure” of the sexes), which, if taken into account by the State in classifying the sexes for differential treatment, would be constitutionally valid. In this case, however, those “immutable background conditions” were treated as human-made social structures, which were allowed to flourish through State inaction – consequently, placing a duty upon the State to remedy them. The difference between A.M. Shaila and Vasantha and Triveni reflects, yet again, that the analytical baseline chosen by the Court, which is a deeply political choice, will end up having a profound effect upon the final judgment, even though the assumptions remains hidden.

Three years later, however, when the same provision was challenged before the Kerala High Court, the Court adopted the opposite reasoning, once again invoking the place and role of women in society: “the very nature of their commitment to the family and the social environment require that they cannot be entrusted with all those duties which men may be asked to perform… The place of women has been recognized in the Indian society since the hoary past. The Constitution has made a special provision in Article 15(3). It is calculated to protect and promote the interest of women, The impugned provision clearly falls within the protective umbrella of Article 15(3). It does not embody a principle of discrimination on sex, but is calculated to save women from the hazards of working during night in factories.”

 Examples need not be multiplied. But by now, the deep conflict in the basic understanding and interpretation of Article 15(1) should be evident, across time and place. In 2007, however, it would appear that the Supreme Court finally definitively settled the issue, in favour of the anti-stereotyping principle. In Anuj Garg vs Hotel Association of India, which we have discussed previously on this blog, the validity of Section 30 of the Punjab Excise Act, which prohibited the employment of women (and men under 25 years) in premises where liquor or other intoxicating drugs were consumed by the public, was challenged. The Delhi High Court struck down the statute, which was carried in appeal. Before the Supreme Court, it was defended by the State on the grounds of maintenance of security, akin to the arguments in A.M. Shaila and Triveni. The Court rejected the argument, holding – as in Triveni – that “new models of security must be developed, if necessary.” In a separate paragraph titled “Stereotype Roles and Right to Options”, it then quoted with approval the judgment in Frontiero vs Richardson, USA vs Virgnia, and Justice Marshall’s dissent in Dothard vs Rawlinson, all of which were based upon the anti-stereotyping analysis, and observed:

“The impugned legislation suffers from incurable fixations of stereotype morality and conception of sexual role. The perspective thus arrived at is outmoded in content and stifling in means.”

The High Court’s judgment in striking down the Section, therefore, was upheld.

There has been some controversy over whether the Supreme Court in Anuj Garg incorporated a standard of “strict scrutiny” in dealing with sex discrimination claims. In paragraph 44, the Court stated that “strict scrutiny should be employed” in cases assessing the validity of sex-based legislation. In paragraph 47, however, the Court seemed to adopt a proportionality test: “the legislative interference to the autonomy in employment opportunities for women is justified as a legitimate aim and proportionate to the aim pursued”, before going on to cite cases from the ECHR. It is important to note that strict scrutiny in sex discrimination claims is not the existing position of law, even in the United States. Such claims are adjudicated under a less exacting, “intermediate scrutiny” standard of review, which is fairly close to the proportionality review employed by the ECHR. It is therefore unlikely that the Supreme Court meant to adopt the standard of strict scrutiny as followed in the United States, given the rest of the tenor of its judgment.

It is also unlikely for another reason: the Supreme Court in Anuj Garg did not merely settle the controversy about whether or not stereotypes about women’s sexual and social roles could be invoked to justify a discriminatory law on “sex-plus” grounds. It also settled the controversy about whether an Article 15 enquiry was limited to the motive, or purpose of law, or whether it also included its effects. The Court held:

“Legislation should not be only assessed on its proposed aims but rather on the implications and the effects.”

And again:

“No law in its ultimate effect should end up perpetuating the oppression of women.”

In Anuj Garg, the law at issue was directly discriminatory – i.e., the law, in its very wording, created two categories (men and women), that were composed entirely and exclusively by the two sexes. However, note that the effects test necessarily includes both direct and indirect discrimination: a pregnancy-based classification, for instance, has the effect of disproportionately burdening women, even though the classification is not along the lines of sex. This is extremely important, since – as we have seen before in this series – it tracks an interpretation of the word “grounds” in Article 15 that qualifies not the motive/intent of the law (or lawmakers), but protected personal characteristics. Although the Court did not follow this analysis in Anuj Garg, it is clear that an effects-oriented interpretation of Article 15 must necessarily be based on the second meaning of “grounds”.

In most other jurisdictions, indirect discrimination is analysed within a proportionality framework, making it more likely that the Court’s invocation of strict scrutiny was not in its technical sense.

However, if the framework within which Article 15(1) is to be analysed is an effects-based one, then where does anti-stereotyping come in, which is concerned with motivations? In Anuj Garg, while responding to the State’s contention that the purpose of the law was to protect women, the Court noted that two conditions would have to be satisfied in such cases:

“… (a) the legislative interference (induced by sex discriminatory legalisation in the instant case) should be justified in principle, (b) the same should be proportionate in measure.”

The anti-stereotyping principle is relevant insofar as it rules out certain kinds of principled justifications (i.e., those based on stereotypes). In other words, at the first stage of enquiry, the effect of a statute will be analysed, to ascertain whether Article 15(1) is infringed. At the second stage, if the State then advances a justification (whether based on Article 15(3) or otherwise), it will be prohibited from relying upon stereotypes.

Anuj Garg, therefore, is authority for two crucial interpretive propositions: first, that both direct and indirect discrimination are covered under Article 15, within the framework of a broad, effects-based test; and secondly, the State may not rely upon stereotypes to justify prima facie discriminatory legislation. While there are many issues it does not deal with (and did not need to), such as whether different standards apply for direct and indirect discrimination, it nonetheless lays the foundation for a conceptually solid jurisprudence on sex equality. Whether and how it will be followed remains to be seen.


Guest Post: Sex, Sexual Orientation, and the Courts

(In this guest post, Vansh Gupta explores a recent ruling of the US Equal Opportunity Commission that speaks directly to the Indian debate over the constitutionality of s. 377)

The U.S. Equal Employment Opportunity Commission recently issued a ruling (available here) recognising discrimination on the basis of sexual orientation as discrimination on the basis of sex. Its immediate impact is that employees can invoke the Civil Rights Act (Title VII) when they are discriminated against on the basis of their sexual orientation, even though sexual orientation is not an explicitly prohibited basis for making employment decisions. As an interpretive exercise, however, the decision has wide-ranging ramifications. It affirms the position that ‘sex’ can be read to include ‘sexual orientation’ in statutory material.

The Commission considers that sexual orientation cannot be defined or understood without reference to sex (a gay man is gay precisely because he attracted to males). The Commission finds that there is ‘an inescapable link’ between sexual orientation discrimination and sex discrimination as discrimination on the basis of orientation is premised on ‘sex-based preferences, assumptions, expectations, stereotypes or norms’. It describes this link in the following 3 ways-

  • Sexual orientation discrimination is sex discrimination because it will require a person’s sex to be taken into account, for instance when a lesbian woman is reprimanded for speaking about her female partner in the workplace, but a straight man isn’t.

The treatment of the person is different precisely because the employee’s sex is different. The equivalence therefore that is drawn here is not between gay men and lesbian women (i.e. both would be reprimanded for demonstrating same-sex affection), but rather between a woman (who happens to be lesbian) and a man, for engaging in the same conduct (speaking about a female affectionately).

  • There is also the understanding that sexual orientation discrimination is associational discrimination on the basis of sex (sex must be taken into account when an employee is treated differently for associating with a person of the same sex). This follows from a line of rulings concerning interracial marriage or friendship. (A white man being punished for marrying a black woman, necessarily takes into account his race), but the EEOC expands the reasoning to apply to sex discrimination as well.
  • And finally, the third rationale is that sexual orientation discrimination is sex discrimination because it involves discrimination on the basis of gender stereotypes, which has already been held to prohibited by the term ‘sex’ in the Civil Rights Act. Courts have already held that claims against gender stereotyping can be brought by lesbian, gay and bisexual individuals if they are treated adversely for being inadequately ‘masculine’/’feminine’, based on their appearance, mannerisms or conduct. The EEOC also recognises however that discrimination of LGB individuals on the basis of gender stereotypes is about more than assumptions about what behaviour is masculine or feminine; there is also an element of enforcing ‘heterosexually defined gender norms’. Real men are not just masculine, but also straight.

India does not have an omnibus Civil Rights Act as in the US or an Equality Act as in the UK, but reference to the principle of sex(ual?) equality is found in statute (see S.4 and 5 of the Equal Remuneration Act) and A. 15 of the Constitution prohibits discrimination which is only on the basis of sex.

Readers will recall that the Delhi High Court had invoked A. 15 in its analysis while deciding the now-overturned Naz petition. Although the court read down S. 377 of the IPC on the basis of a web of reasoning concerning A. 14 (equal protection of laws), 15 and 21 (privacy), the intent and the result insofar as A. 15 is clear: the term ‘sex’ in A. 15 must be read to include ‘sexual orientation’.

“104. We hold that sexual orientation is a ground analogous to sex and that discrimination on the basis of sexual orientation is not permitted by Article 15. Further, Article 15(2) incorporates the notion of horizontal application of rights.”

While the Court has dealt with A. 14 and 21 extensively, its A. 15 analysis is limited to the a few paragraphs, relying primarily only on the conclusions of a number of foreign judgements. The High Court does however make an effort at explaining how an unenumerated ground can be the basis for holding discrimination. (“There will be discrimination on an unspecified ground if it is based on an attributes or characteristics which have the potential to impair the fundamental dignity of persons as human beings, or to affect them adversely in a comparably serious manner”, following from the South African constitutional Court’s decision in Harksen v. Lane.”)

The difference between the treatment of the matter by the Delhi High Court and the EEOC is that the EEOC relies on an understanding and interpretation of ‘sex’ to determine that sexual orientation discrimination is sex discrimination, whereas the High Court considers discrimination on the basis of sexual orientation to be analogous to sex discrimination.

This difference is not trivial. A. 15 specifically enumerates prohibited categories of discrimination. By including an analogous ground, the Court opened the way for A. 15 to be read expansively vis-à-vis other grounds as well. But the Civil Rights Act can only be used for the grounds that it specifies (necessitating Americans with Disabilities Act to be enacted to provide similar protections for persons with disabilities).

There is also the consideration that A. 15 is a constitutional provision, and is more amenable to expansive interpretation than a statutory term. EEOC-like reasoning would perhaps be necessary for interpreting ‘sex’ in statutory provisions such as S. 4 and 5 of the Indian Equal Remuneration Act.

Naz offered a promising start to a new way of dealing with A. 15 altogether in terms of both its method and result. Since its total reversal by Suresh Kumar Koushal, this possibility is now gone as Naz losing all its precedential value. It is another matter that in doing so the Supreme Court did not rule or comment on the High Court’s treatment of A. 15 at all, choosing to notice only that the High Court struck down S. 377 inter alia on the basis of 15, recording the contentions of the parties on its applicability, and after a (shoddy) analysis of only 14, concluding that the High Court was unjustified in finding S. 377 to be ultra vires both A. 14 and 15. The curatives pending before the Supreme Court highlight this glaring omission, but for the moment, sexual minorities remain a ‘miniscule fraction of the country’s population’ undeserving of the protection of their so-called rights.

(Vansh is a law researcher with Justice Ravindra Bhat at the High Court of Delhi)

Edward Said, Alessandro Portelli, Law’s Representing Power, and Judicial Humility

In a piece called Death and the Sovereign (the reference is unmistakable), Pratap Bhanu Mehta has an important critique of the Rajasthan High Court’s Santhara judgment. He argues that legal categories (such as “suicide” and “attempted suicide”) are insufficient and inaccurate placeholders for understanding the range and complexity of religious practices, as well as them meaning and significance they carry for their adherents.

Mehta’s argument reminds me of something Edward Said wrote about in Culture and Imperialism: that representation is an act of power, and one way to recognize an unequal relationship is by observing how it acts to silence, exclude and marginalize the experiences of the subjects of representation. Said writes:

Power even in casual conversation to represent what is beyond metropolitan borders derives from the power of an imperial society, and that power takes the discursive form of a reshaping or reordering of ‘raw’ or primitive data into the local conventions of European narrative and formal utterance, or in the case of France, the systematics of disciplinary order. And these were under no obligation to please or persuade a ‘native’ African, Indian, or Islamic audience: indeed, they were in most influential instances premised upon the silence of the native. When it came to what lay beyond metropolitan Europe, the arts and the disciplines of representation – on the one hand, ficton, history and travel writing, philology, racial theory – depended on the powers of Europe to bring the non-European world into representations, the better to be able to see it, to master it, and, above all, to hold it.

Said’s point resonates. In the Santhara judgment, the Rajasthan High Court takes the “raw” or “primitive” data (i.e., the fact that some Jains are undergoing a fast unto death). The Court shoehorns it into the “local conventions” of legal “narrative and formal utterance” (which can equally well be classified as a disciplinary order), i.e., the prohibition against abetment to “suicide”. This, of course, is premised upon the “silence of the… [Jains]”. The questions whether they regard it as “suicide”, and whether there a divergence between how they experience Santhara, and how the law understands suicide, are not addressed.

In a similar vein, Alessandro Portelli, the oral legal historian, writes about the anatomy of an Italian terror trial in the 1970s. In defining a political movement as a criminal conspiracy, Portelli argues that the magistrates were thus involved in reconstructing the past, redefining its meaning, and attempting an overall interpretation. These are historical tasks, and it is appropriate to examine the way they were performed from the point of view of the theory and method of history, oral history specifically, given the nature of most sources used.” A trial, according to Portelli, always involves reconstructing (or constructing history), through documents and – where there are gaps in the documents – through oral testimonies. This, of course, is another form of representation, and another form of power: through its final judgment, the Court will declare one version of history to be true, which may well be contrary to the experiences of its participants (note, for instance: any finding of guilt must, by necessity, affix whole and complete responsibility upon individuals, rather than structural or social causes. Portelli extracts the evidence of a prosecution witness, before observing that “prosecution witness Romito favors political over social history: mass struggles or insurrections do not depend upon the masses or on broad social causes, but on the secret dealings of leaderswhose influence on the working class was never more than marginal anyway.”)

The judge’s task, therefore, repeatedly requires her to answer questions that are centrally anthropological (what is the character of Santhara?) or historical (what was the cause of the terror attack?), without obligating her to adhere to the rigorous standards of method and scholarship required of an anthropologist or a historian. Paradoxically, however, the judge has far more power than the anthropologist or historian. While the latter’s conclusions are provisional, tentative and always open, the judge’s findings assume the crystalline immutability of “judicial truth” one they are ensconced in a judgment. More than that, of course, they have very real consequences. As Robert Cover points out, “legal interpretation takes place in a field of pain and death.” In the Santhara case, that is literally true.

All this points to an urgent need for judicial humility. Hard questions should be approached with an awareness of the judge’s privileged subject position, the power that she wields to impose a representation of the “truth”, the further power to convert that representation into actual facts on the ground, and the consequent need to be wary of sweeping, assured positions. True “demosprudence” would require the judge to listen to the experiences of those who will soon be represented in judgments, especially when their experiences seem not to fit easily into pre-defined, a priori, universalising legal categories. It is a trait that, as Mehta points out, is unfortunately completely missing in the Rajasthan High Court’s judgment.

Judicial humility, I suspect, will be a recurring theme when the historians of the future, finally unshackled by contempt of court laws, sit down to write the history of the present Court. In particular, I think they will marvel at how easily our judges liberated themselves from the gravitational pull of doubt, which keeps most of us earthbound, and escaped into the stratosphere of diamond-bright moral certainty. Recently, in confirming a death sentence (Edit: V. Venkatesan has kindly pointed out that it was an abetment to suicide case, and not a death penalty confirmation. My thanks to Mr. Venkatesan), Justice Dipak Misra stated that judges have a duty to respond to the “collective cry of society”. We have heard such words before – in the mouths of prophets and madmen, who have considered themselves to have privileged communion with the Voice of God (vox dei or vox populi, the clarity of the moral vision never dulls). Prophets and madmen, however, have no need of explaining themselves. Judges do, because they live under the Constitution, like the rest of us, and have as much a duty of fidelity to the Constitution as the rest of us do.

What is the collective? How can you know that there is only one collective, which speaks with one voice (one “cry”)? How can you know that the collective is right? And above all else, how can you be so sure that you, among all citizens, have correctly heard and interpreted this “collective cry”? Such pretensions would provoke laughter, did they not have such real and tragic consequences.

Almost five hundred years ago, Oliver Cromwell wrote to the synod of the Church of Scotland, stating thus: “I beseech you… think it possible that you might be mistaken.” It is a plea that one wishes would be heard by the present judges of the Court.

“Essential Religious Practices” and the Rajasthan High Court’s Santhara Judgment: Tracking the History of a Phrase

This week, the Rajasthan High Court held that the Jain practice of santhara – a ritual of “voluntary and systematic fasting to death” was illegal, since it amounted to abetment to suicide (criminalised under Section 306 of the Indian Penal Code).

There are many issues that arise out of this judgment. This was a PIL filed by a ‘lawyer’ who had no connection with, and was not affected in any way by santhara. The Court’s decision to issue notice and hear the case on merits demonstrates yet again that PIL has been entirely cast off from its moorings: the loosening of standing rules was intended to ensure the representation of those who could not represent themselves. By now, it is used to  transform the Court into a super-legislature, where any social question might be agitated by any person (something similar is ongoing in the Supreme Court, in the Kamlesh Vaswani anti-porn petition). The Court’s analysis of whether santhara is equivalent to suicide is fraught with problems as well. In this essay, however, I will focus on another part of the Court’s judgment: its finding that santhara did not constitute an “essential religious practice” for the Jains, and consequently, was not protected by Article 25 of the Constitution. The Court held:

“We do not find that in any of the scriptures, preachings, articles or the practices followed by the Jain ascetics, the Santhara or Sallekhana has been treated as an essential religious practice, nor is necessarily required for the pursuit of immortality or moksha.”

The essential practices test has been used consistently by the courts at least since 1957. The test allows the Court to initiate a judicial enquiry into whether or not an impugned religious practice is an “essential practice”, independent of what the religion’s adherents themselves say about it. Commenting upon the Supreme Court’s use of the test, Jacobsohn has insightfully noted that it has become “an internal level of reform”: by holding that certain regressive practices do not constitute “essential” parts of a religion, the Court not only denies them constitutional protection, but also takes upon itself the task of recharacterising the religion in a more progressive light, and, in a sense, create new social facts through its holdings. Naturally, for these very reasons, the test has met with fierce criticism. The judiciary, it is argued, possesses neither the competence nor the legitimacy to decide what constitutes an “essential practice”; it is not, after all, “the Supreme Court of Hinduism” (Galanter)  These criticisms are powerful ones, but in this essay, I want to ask a different question. The Constitution does not mention the term “essential religious practice”: it grants protection to the right to practice, profess and propagate one’s religion, not just to engage in the “essential practices” of religion. So where does this concept comes from?

We can find a clue in the Constituent Assembly Debates. On the 2nd of December, 1948, Ambedkar delivered a speech in the Constituent Assembly where, among other things, he observed:

“The religious conceptions in this country are so vast that they cover every aspect of life, from birth to death. There is nothing which is not religion and if personal law is to be saved, I am sure about it that in social matters we will come to a standstill. I do not think it is possible to accept a position of that sort. There is nothing extraordinary in saying that we ought to strive hereafter to limit the definition of religion in such a manner that we shall not extend beyond beliefs and such rituals as may be connected with ceremonials which are essentially religious. It is not necessary that the sort of laws, for instance, laws relating to tenancy or laws relating to succession, should be governed by religion.”

Ambedkar’s use of the term “essentially religious”, therefore, was in response to a very specific concern. He was worried that unlike in the West, with its seemingly clear demarcation between the City of God and the City of Man, there was no aspect of Indian life which was untouched by religion. Consequently, insofar as the Constitution protected religion and personal laws, there was a very real risk that it would entirely hamstring the State’s power to pass social legislation. He was, therefore, adamant that there must be a separation between religious activities, and secular activities tinged with religion. The latter could have no constitutional immunity from legislation. In Ambedkar’s formulation, it is clear the word “essentially” qualified “religious”, and was designed to separate the religious from the secular.

The wording of Article 25 responds to Ambedkar’s concern. Unlike Article 19, where the main Article lists out the fundamental freedoms (Article 19(1)), followed by the scope of reasonable restrictions (Articles 19(2) – 19(6)), Article 25 starts off with limitations: “Subject to public order, morality and health, and to the other provisions of this Part, all persons are equally entitled to freedom of conscience and the right freely to profess, practise and propagate religion.”

At this point, the following question may be raised: with the limitations built into the right to freedom of religion, is there any further need for Amedkar’s proposed demarcation between that which is “essentially religious” and that which is not? Surely, social and welfare legislation could be justified under the “public order, morality and health” categories. In fact, the very structure of the Article – specifying both the right and its restrictions – ought to preclude threshold enquiries separating the religious from the non-religious (this is the form of rights-adjudication practiced in South Africa, for instance).

Notwithstanding the presence of limitations, however, there might still be a role for the threshold enquiry. To start with, as Ambedkar pointed out, if every sphere of existence has religious significance, then there seems little point in a constitutionally guaranteed right to the freedom of religion in the first place. Secondly, prima facie constitutional protection places a heavy burden of justification upon the State, and if every regulatory law has to run the gauntlet of the “public order, morality and health” tests, many might not survive. And thirdly, the expressive significance of holding something to be a fundamental right protected by the Constitution might well require a threshold enquiry to ascertain whether the reason why the Constitution protects religion in the first place, ought to extend to the practice under question.

The first few judgments after the coming into force of the Constitution did use “essentially religious” in the sense that Ambedkar had used it. In Lakshmindra Swamiar (1954), the Supreme Court held that “what constitutes the essential part of a religion is primarily to be ascertained with reference to the doctrines of that religion itself. If the tenets of any religious sect of the Hindus prescribe that offerings of food should be given to the idol at particular hours of the day, that periodical ceremonies should be performed in a certain way at certain periods of the year or that there should be daily recital of sacred texts or oblations to the sacred fire, all these would be regarded as parts of religion and the mere fact that they involve expenditure of money or employment of priests and servants or the use of marketable commodities would not make them secular activities partaking of a commercial or economic character; all of them are religious practices and should be regarded as matters of religion within the meaning of article 26(b).”

For the Court, therefore, “essential” marked the border between the religious and the secular. These observations were repeated that same year in Ratilal vs State of Bombay, where the Court added that “no outside authorities has any right to say that these are not essential parts of religion and it is not open to the secular authority of the State to restrict or prohibit them in any manner they like under the guise of administering the trust estate.”

Three years later, however, in Ram Prasad Seth vs State of UP, the Allahabad High Court put a very different gloss on things. UP Government regulations, which prohibited bigamous marriages to those in public employment, were challenged on the grounds of Article 25. It was argued that the Hindu religion allowed certain funeral rites for a deceased individual to be performed only by sons. Consequently, it was imperative for a Hindu individual to have a son, and sometimes, bigamy was the only way of achieving this. In response, the Court analysed certain important Hindu religious texts, and on the basis of analysis, held that “[bigamy] cannot be regarded as an integral part of a Hindu religion… the acts done in pursuance of a particular belief are as much a part of the religion as belief itself but that to my mind does not lay down that polygamy in the circumstances such as of the present case is an essential part of the Hindu religion.”

Here is the key shift: the word “essential” has gone from qualifying the nature of the practice (i.e., whether it is religious or secular), to qualifying its importance (within the religion) – i.e., from whether something is essentially religious to whether it is essential to the religion. It is a minor grammatical shift, but with significant consequences, because it allows the Court to define questions that are internal to religion in a judicial enquiry, and thereby define the nature of the religion itself.

The Supreme Court adopted this interpretation one year later, in Qureshi vs State of Bihar, holding that the sacrifice of a cow on the occasion of Id was not an essential religious practice for Muslims: “We have… no material on the record before us which will enable us to say, in the face of the foregoing facts, that the sacrifice of a cow on that day is an obligatory overt act for a Mussalman to exhibit his religious belief and idea. In the premises, it is not possible for us to uphold this claim of the petitioners.” The Court further entrenched this position in 1962, in Syedna Saifuddin, while striking down a law that prohibited excommunications. The Court held that Article 25(2), which allowed the State to pass reform legislation, “is intended to save the validity only of those laws which do not invade the basic and essential practices of religion which are guaranteed by the operative portion of Art. 25(1).” Four years later, this attitude seemed to have become such an undisputed part of judicial wisdom, that Chief Justice Gajendragadkar was able to devote an entire judgment to answering the central questions: who is a Hindu, and what constitutes Hinduism?

We can see, therefore, that the essential practices test did not originally mean what it has come to mean now. At this point, however, another objection might be raised: does not the question of whether a particular practice is religious or secular involve as much judicial interference as the question of whether it is essential to a religion? The answer is: not necessarily. In its early judgments, the Court held that this question could only be settled by the tenets of the religion itself, which does not necessarily mean judicial enquiry into what those tenets say. In other jurisdictions, the Court only asks whether a particular practice is “sincerely held” by its adherent, a question that requires it to go into the adherent’s past behaviour and conduct, but not into the substantive nature of the practice itself. A purely subjective test, however, might simply smuggle back in Ambedkar’s fundamental worry: that religion could now be invoked to cover every aspect of a person’s life. A possible answer to this was suggested by Justice Sinha in his dissenting judgment in Syedna Saifuddin, holding that practices that directly impacted a person’s enjoyment of his civil rights that were guaranteed by law (as excommunication did), would not be given constitutional protection.

I cannot here go into a full analysis of Justice Sinha’s fascinating proposal. What I hope to have established is that in its present form, the essential practices test is based on an interpretive mistake: it misinterprets what Ambedkar said, as well as the early judgments of the Supreme Court itself. This, coupled with the institutional problems that it creates, should be enough for a fundamental reappraisal of this test within the scheme of Indian constitutional jurisprudence.


Sex Discrimination and the Constitution – IX: Rajendra Grover, the Culmination of the Air Hostess Cases, and Thoughts on the Supreme Court’s Institutional Role

After two decades of litigation, Air Hostesses, Air Flight Pursers and the Union of India met again, in one final battle before the Courts. After the decision of the Supreme Court in Yeshaswinee Merchant, refusing the merger of cadres, the struggle for equalisation was taken (again) outside the judiciary. In 2003, Air India allowed its female cabin crew members (who had been recruited after the first equalisation in 1997) to undertake flying duties until the age of 58, bringing them on par with the male cabin crew. Two years later, in 2005, Air India also decided that henceforth, executive female crew members would be eligible for the post of “In Flight Supervisor” (IFS), thus effectively merging a promotional category that had only been open to men. In Rajendra Grover vs Union of India, this was challenged by the male cabin crew that had been recruited before 1997.

Recall, once again, the tortuous history. In 1995, for the first time, the management and cabin crew of Air India had entered into an understanding, clarifying that for new recruits, cabin crew functions would be interchangeable, but without affecting promotional avenues. Promotional avenues up until the first executive level were themselves merged in 1997, when one cadre – that of “cabin crew” was created. It was the 1997 settlement that had been challenged by the Air Hostesses who were part of the executive class, leading to the Bombay High Court order creating absolute parity between male and female cabin crew – which, in turn, had been struck down in Yeshaswinee Merchant. But now, what Nargesh Mirza and Yeshwasinee Merchant had refused to do via the Constitution, had been done by Air India through Office Orders: complete equalisation of male and female cabin crew.

Before the Delhi High Court, the Petitioners argued that Air India was not entitled to disturb the status that had been guaranteed to them by Nargesh Mirza and Yeshaswinee Merchant, as well as by numerous prior settlements. They contended that the “In-Flight Supervisor” was a promotional post, specifically part of the male cabin crew cadre, which had clearly been held to be separate and different from the Air Hostess cadre in both Nargesh Mirza and Yeshaswinee Merchant, as well as by the Settlements. Consequently, “the impugned administrative order inasmuch as it seeks to grant parity between two unequal classes and separate and distinct cadres is a violation of Article 14 of the Constitution of India.” The Respondent Air Hostesses, on the other hand, argued that the IFS was an entirely functional post (and not a promotional one), and that it was therefore not bound to any one particular cadre. In any event, the equalisation did not affect either the Nargesh Mirza judgment or the settlements, since IFS was a supervisory or executive post, whereas the controversy in the prior cases had been between cabin crew who constituted “workmen” within the meaning of labour law. The Air Hostesses countered the male cabin crew’s invocation of the Constitution by making constitutional claims of their own: “denying a woman the functions of the IFS and maintaining it as a male preserve is violative of Articles 14, 15 and 16 of the Constitution.”

The Delhi High Court found that, on a close reading of Nargesh Mirza and Yeshaswinee Merchant, those cases had only held that the existing status quo did not constitute hostile discrimination. They had not, however, barred the government from changing the status quo by exercising its executive prerogative. This, indeed, was what had happened: “Post 1997, there has been a merger of the cadres of the male and female members of the cabin crew. This is quite different from what prevailed at the time of the judgment in Nergesh Meerza. It is nobody’s argument that such a merger is unconstitutional or invalid or that it runs contrary to what was held in Nergesh Meerza. In fact, in Yeshaswinee Merchant also, the Supreme Court recognized that the employer can take a policy decision to re-organise its organizational structure to remove elements of discrimination.

The last line is particularly interesting, because of course it was the Court’s opinion in Nargesh Mirza and Yeshaswinee Merchant that unequal service conditions of male and female cabin crews did not amount to discrimination. Taken literally, the Delhi High Court’s sentence would lead to the odd result that the Supreme Court had acknowledged that discrimination existed, but contrary to the requirements of the Constitution, left it to the State to deal with them (of course, our argument throughout this series has been that that was what the Court effectively did, but that is neither here not there).

In any event, the Court concluded the matter by saying that after the merger of cadres in 1997, “in the matter of either flying duties or in the matter of seeking avenues of promotion, the rights of one cadre need not be to the exclusion of the rights of the other. In fact what has happened is that in terms of the impugned order both the pre-1997 male and female cabin crew have an equal chance of being considered for performing the function of IFS in accordance with their seniority.”

The Delhi High Court’s decision was carried in appeal to the Supreme Court which, in 2011, dismissed the appeal with a few perfunctory lines that need not concern us. Thus, the decades-long battle ended. But the thirty-year long litigation compels us to ask a few important questions about the Supreme Court’s institutional role.

As we have seen, the basis of the dispute was the undeniable fact that male cabin crew and female cabin crew were treated differently. In the first round of litigation before the Supreme Court, the treatment of the female cabin crew was clearly inferior: compulsory retirement upon first pregnancy or within four years of marriage, or on attaining the age of 35. Despite the fact that the entire basis of division was the sex of an employee, the Court managed to find that there was no discrimination under Articles 14, 15 and 16 of the Constitution. The Air Hostesses took their case to the Parliament, and won important concessions, such as an increase in retirement age (although still not on par with men) – and, in 1997 – an equalisation of functions. Partly based upon this, the Bombay High Court then ordered complete parity and a merger of the cadres. The Supreme Court reversed this decision, and reinstated the old, unequal system. Parity was finally made complete through executive action, which – ultimately – was upheld by the Court.

Notably, all three institutions of the State were involved in various stages: the Supreme Court, Parliament and the Government. Out of these, the Air Hostesses failed to find relief in only one of the three: the Court. Equalisation was repeatedly rejected by the Courts, which refused to find unconstitutional discrimination, but was won before the Parliament and the Government. If you agree with the basic analysis that constitution of cadres (with unequal service conditions) is clearly sex-based discrimination (as even the Delhi High Court seemed to do in Rajendra Grover), then this entire litigation is strangely counter-intuitive: civil rights were, ultimately, protected not by the institution that is constitutionally mandated to do so, but by the institutions that they are supposed to be protected from.

The Air Hostess litigation, and the area of non-discrimination law, is not an isolated example. In the domain of free speech, the Supreme Court in 1952 upheld the constitutionality of the draconian Press (Emergency) Powers Act, which had been passed by the colonial British government in 1931, with sweeping powers of censorship designed at controlling newspapers. The Act was repealed five years later by Parliament. In the domain of personal liberty, the Court upheld the stringent provisions of the Terrorist and Disruptive Activities (Prevention) Act, which systematically denuded criminal procedure safeguards. The TADA was repealed by Parliament soon after, in the face of sweeping criticism. More recently, the draft women’s rights bill put out by the Delhi government has the first mention of non-discrimination on the grounds of sexual orientation, two years after the Supreme Court upheld the constitutionality of Section 377.

While these are merely anecdotal references, there is something amiss when whatever protection is accorded to civil rights, is won in majoritarian institutional fora, and lost in the constitutional court. I think this points to a need to fundamentally reappraise the institutional role of the Supreme Court within the Indian constitutional scheme. Over the last three decades, a narrative has developed that has characterised the Court as activist and interventionist, aggressively going beyond the text of the Constitution to find and enforce new rights, and coaxing or goading a moribund executive into performing its mandated functions. This narrative is based entirely on the Court’s PIL jurisprudence. Supporters have praised the Court for fulfilling the vacuum left by a non-functioning executive and fractious coalition politics, and for interpreting the Constitution in a way that is “pro-people”; critics have accused the Court of violating the separation of powers and encroaching into the domain of elected representatives. But in all this debate over whether the Supreme Court has gone beyond its mandated functions and whether that is justified, the question is rarely asked now whether the Court is effectively doing what it is uncontroversially required to do: protect civil rights under Part III of the Constitution. Does the fact that the Court has, more and more, begun to resemble the executive in its sweeping directions and its rhetoric, bear any connection with the gradual erosion of its counter-majoritarian role in protecting civil rights? We need to ask not only what (if anything) we have gained through “social action litigaton” (what one scholar describes – in my opinion with extreme inaccuracy – as “demosprudence”), but what we have lost with a Court that has become – as Lord Atkin once said – “more executive-minded than the executive.” And my sense is that with an honest analysis, we might find that what we have lost has not been worth losing after all.

Sex Discrimination and the Constitution – VIII: The further discontents of Air India Cabin Crew vs Yeshaswinee Merchant

Air India vs Nargesh Mirza was only the first salvo in a legal battle that was to last thirty more years. After losing in the Supreme Court, the air hostesses took their battle to the political arena: in 1989, they petitioned the Lok Sabha against the discrimination in retirement ages and other service conditions. In response to the recommendations of the Petition Committee, the Central Government issued a direction to the Air india, requiring them to allow both male and female cabin crew to serve until the age of 58. After some resistance from Air India, the government clarified that while the retirement age for both men and women was now 58, air hostesses could be given “suitable alternate jobs on the ground” after they had crossed the age of 35. Accordingly, in 1990, Air India issued a circular stating that air hostesses who had attained the age of 45 would be given ground employment. In 1993, this was extended to 50, subject to tests for medical fitness for flying duties.

Soon afterwards, through legislation, Air India and Indian Airlines became two separate entities. Air India (now a new company) entered into an agreement with the Air India Cabin Crew Association in March 1995. According to this agreement, there would be complete interchangeability of functions between male and female cabin crew members for new entrants to the service; however, for all recruits who had joined before 1997, the old policies would apply. This agreement was enshrined in a formal settlement in June 1997, where it was also clarified that all previous settlements and agreements, entered into between Air India in is previous avatar and its employees, would continue to hold where applicable.

In accordance with the new settlement, a revised promotion policy for cabin crew was brought into force the same month. At this point, fifty-three Air Hostesses, most of whom were nearing the age of 50, formed an Association (Air India Air Hostesses Association), and filed a petition before the National Industrial Tribunal, asking for a merger of the cadres, including for recruits who had joined before 1997. Interestingly, this claim was resisted by other Air Hostesses (a majority, in fact), who stated that they wished to retain their right to early retirement (and, consequently, accelerated seniority), and therefore, to continue to be treated as a separate cadre. The Tribunal rejected the petition for merger. Accordingly, a writ petition was filed in the Bombay High Court.

Before the High Court, the Air India Cabin Crew Association asked to be impleaded as a party, but the High Court only allowed them to file an intervention application (this fact is important, because it seemed to weigh significantly with the Supreme Court; we shall return to it later in this essay). It was argued by Air India and the AICCA (as intervenor), that the interchangeability of jobs between male and female cabin crew was applicable only after 1997; it was a policy decision, and did not imply the pre-existing position was discriminatory. The High Court, however, was not convinced; after some prodding from Justice Shah, Air India submitted a proposal for the merging of cadres. According to the proposal, the male and female cadres would be merged, and service conditions would be equalised. A corollary of this was that female air hostesses would lose the accelerated promotions and higher allowances that they were entitled to by virtue of their early retirement.

The Air India Cabin Crew Association approached the Supreme Court. It claimed to be the only recognised union entitled to represent the cabin crew (with 684 air hostesses out of a total of 1138 being a part of it), and argued that the 53 air hostesses who had claimed parity (and were working in executive posts) were “unmindful of the interest[s] of the larger number of air hostesses who are of the workmen category and have agreed for an early retirement age.”

The Supreme Court handed down its judgment in 2003, in a case called Air India Cabin Crew Association vs Yashaswinee Merchant, agreeing with the claims of the Association. It began by noting that the correct test for determining whether there was discrimination “on grounds” of sex was the “but-for” test: i.e., the question that needed to be asked was whether, but for his/her sex, an individual had been disadvantaged. Notice that the but-for test makes no mention of the discriminator’s reason – but in the very same paragraph, the Court undid its own analysis by stating that “women [cannot] be deliberately selected for less favourable treatment because of their sex.” The use of the word “deliberately” was entirely unnecessary, because it fixes the locus of discrimination within the conscious mental state of the discriminator, which is something that is in no way required by the but-for test.

In any event, the Court then reiterated the Nargesh Mirza holding, stating that on the “but-for” test, it had been found that because air hostesses and air flight pursers formed separate cadres, there was no discrimination on grounds of sex. The problems with this analysis have been explained in detail in a previous essay, and need no repetition. The Court then went on to observe, however:

“The constitutional prohibition to the State not to discriminate citizens only on sex, however, does not prohibit a special treatment to the women in employment on their own demand. The terms and conditions of their service have been fixed through negotiations and resultant agreements, settlement and awards made from time to time in the course of industrial adjudication. Where terms and conditions are fixed through collective bargaining as a comprehensive package deal in the course of industrial adjudication and terms of service and retirement age are fixed under agreements, settlements or awards, the same cannot be termed as unfavourable treatment meted out to the women workers only on basis of their sex and one or the other alone tinkered so as to retain the beneficial terms dehors other offered as part of a package deal. The twin Articles 15 and 16 prohibit a discriminatory treatment but not preferential or special treatment of women, which is a positive measure in their favour.”

First of all, the Court’s recourse to Article 15(3) is strange. If the Court had already decided to follow Nargesh Mirza on the point that Articles 15 and 16 were entirely inapplicable to the dispute, because the discrimination was not on the basis of sex, then the constitutional questions ended right there – there was no need to additionally justify the classification under Article 15(3). Secondly, if Article 15(3) was invoked, then clearly there was an element of discrimination under Article 15 and 16. But if that was the case, then how could the Court invoke Article 15(3) to override the barrier on discrimination in public employment under Article 16(2)? The scheme of Article 16 has no provision equivalent to Article 15(3); in fact, Article 16(4) makes it clear that affirmative action under the Article 16 scheme is limited to socially and educationally backward classes. In a previous essay, we have seen how High Courts were split over the issue of the relationship between Article 15(3) and 16(2). In blithely invoking preferential treatment for women, the Supreme Court simply ignored this fraught textual question.

Lastly – and most problematically – is the Court’s assumption that the Settlement was a positive measure in favour of women. To buttress this opinion, it observed:

“We have taken a resume of several agreements, settlements and awards made after negotiations from time to time and periodically, between Air India and the AICCA being the recognised association with majority of male and female cabin crew members. In all the demands, it insisted on maintaining two separate cadres for pre-1997 recruits and agreed for early retirement age to air hostesses compared to males from flying duties with option to go for ground duty between 50 to 58 years of age. In the course of industrial adjudication through conciliation and negotiation the employer could legitimately acknowledge women’s perspective, their life experience and view point. After giving consideration to the same, the employer could agree for terms and conditions which suited the air hostesses.”

But there are two key questions that the Court entirely ignored. First, is discrimination no longer discrimination, if I have agreed to it? And if the answer to the question is no, if discrimination is at least partially an objective fact, then can I contract out of my right not to be discriminated against? The Court’s stress on the negotiated settlement between the Employer and the Association as solving the problem of discrimination relied upon the answer to at least one of those questions being in the affirmative. But these are crucial questions of constitutional philosophy, that require a detailed analysis – not blanket assumptions. Philosophical thought experiments such as that of the happy slave warn us about the dangers of a subjectivist approach to discrimination; and the question of whether it is permissible to waive one’s fundamental rights is controversial even within Indian jurisprudence (waiver was expressly rejected for Article 14 in Basheshar Nath’s Case).

Admittedly, the fact situation in Yashaswinee Merchant was nowhere near as straightforward or obvious as the happy slave hypothetical. Arguably, the service conditions of air hostesses (accelerated promotions and allowances in return for early retirement) were not significantly worse than that of air flight pursers. Consequently, the negotiated settlement certainly ought to have played a part in the Court’s reasoning. However, it could not have replaced  the overriding requirement of a constitutional analysis, which would call upon Air India to justify why there existed all-male and all-female cadres with different service conditions in the first place. Notice that the enquiry would not have ended at that point, even if, all things considered, it had been found that the air hostesses’ service conditions were more beneficial. As we have seen earlier, Article 15(3) cannot be interpreted to present a blank cheque to the State to impose any policy that advantages women over men. The purpose of Article 15(3) is to redress long centuries of structural disadvantage faced by women, which cannot be wiped out through formal a non-discrimination guarantee. Consequently, measures based upon the same stereotypes and prejudices that contributed to this structural disadvantage surely cannot be justified by taking recourse to Article 15(3). However, towards the end of the judgment, it became clear that this was precisely what motivated the opinion of the Court:

“Duties on flight demand of air hostesses physical fitness, agility and alertness. Duties in air are full of tension and sometimes hazardous. They have, therefore, agreed for comparatively early age of retirement with option to accept duties on the ground. There is nothing objectionable for air hostesses to wish for a peaceful and tension-free life at home with their families in the middle age and avoid remaining away for long durations on international flights.”

The implication (as we have seen innumerable times by now) is, of course, that women somehow lose the physical fitness, agility, alertness and ability to deal with tension after crossing the age of 50, while men retain it until the age of 58; and that the fulfillment of family obligations is the sole preserve of women. And what’s crucial to notice here is that the Court’s framing of the issue as one of consent and choice was a sleight of hand: after crossing the age of 50, it was mandatory for an air-hostess to accept ground employment. In fact, this, precisely, was why the policy had been challenged – i.e., that it actively denied women eight year of in-flight employment that was available to men, for no other reason that they were women.

It is this aspect, indeed, that constitutes the most disturbing part of the judgment; because what seemed to sway the Court was not simply the fact of a settlement, but that it was entered into by a majority of the employees, including air hostesses. The Court noted: “A small number of air hostesses nearing the age of 50 years and who are now in executive cadre cannot wriggle out of the binding agreements and settlements to which they were parties through the association.”

But why should “small numbers” matter in a constitutional claim based upon the non-discrimination guarantee of Article 15? Civil rights would be meaningless if their enforcement was dependent upon a showing of numbers; in fact, the very purpose of civil rights is to serve as a counter-majoritarian bulwark for the protection of basic liberties. And in this case, the Court effectively held that a majority of air hostesses could impose their preferences upon a minority, regardless of whether or not they were discriminatory. Thus, the question in Yashaswinee Merchant was not merely one of waiver, because not every air hostess had waived her right against non-discrimination. But not only did the Court assume the legitimacy of waiver, it also effectively held that some could waive their rights on the behalf of others. Needless to say, this is deeply problematic.

Attentive readers will, of course, have noted that the “small numbers” of Yashaswinee Merchant would, twelve years later, be transmitted into the “minuscule minority” of Koushal vs Naz Foundation. The pathology remains the same.

Yashaswinee Merchant, therefore, suffers from many of the discontents that plague Nargesh Mirza. Apart from the stereotypes that play both an overt and a covert role in the judgments, there is something else that unites them: a clear-headed constitutional analysis is muddied and obscured by the technicalities of service law (cadres, service conditions, settlements) and statutory provisions (in both cases, those of the Equal Remuneration Act). It seems that the scope and ambit of constitutional rights flows from service law and the Equal Remuneration Act, instead of the latter being held to the standards of the former. The broader question that crops up repeatedly in these cases is the relationship between specific domains of law and the Constitution. This is not limited to service law: recall that the competition regime raises questions about free speech (in the case of media holdings), the intellectual property rights regime intersects with free speech (copyright) and access to health (patents); and so on. The Air Hostess cases reflect a deeper problem with the Court’s approach to constitutional adjudication: according primacy to these specific domains, deciding cases with references to the internal norms that govern them, and – consequently – subordinating the Constitution to those norms. Unlike the South African Constitutional Court, which examined trademarks in the context of free speech, or the Canadian Supreme Court, which analysed copyright in the context of user’s rights to educational materials, the Indian Supreme Court is yet to create a constitutional framework within which it can analyse these laws. That time, we can only hope, comes sooner rather than later.


Sex Discrimination and the Constitution – VII: The Retreat from Nargesh Mirza in MacKinnon Mackenzie

In the previous essay, we noted the analytical problems with the Supreme Court’s judgment in Air India vs Nargesh Mirza. However, six years later, in 1987, the Supreme Court handed down another judgment that substantially undercut the ratio in Nargesh Mirza. In MacKinnon Mackenzie vs Audrey d’Costa, a company offered different remuneration to male stenographers and female stenographers (the latter were called “Confidential Lady Stenographers”, because they performed a range of functions such as attending to correspondence etc. The Supreme Court held that paying a lesser wage to female stenographers violated the Equal Remuneration Act. It observed:

“Wherever sex discrimination is alleged, there should be a proper job evaluation before any further enquiry is made. If the two jobs in an establishment are accorded an equal value by the application of those criteria which are themselves non-discriminatory (i.e. those criteria which look directly to the nature and extent of the demands made by the job) as distinct from criteria which set out different values for men and women on the same demand and it is found that a man and a woman employed on these two jobs are paid differently, then sex discrimination clearly arises.”

Notice the underlined part: “by the application of criteria which are by themselves non-discriminatory.” This effectively forestalls a Nargesh Mirza-type situation, where sex is the criterion that constitutes the differentiation in the job in the first place. The Court in MacKinnon Mackenzie emphasised that in adjudicating the validity of an unequal classification, it would consider the nature of the work involved, factoring out those criteria that could not themselves be explained without resorting to discriminatory justifications. Again, for instance, this would prevent a situation such as Nargesh Mirza, where the Court relied upon the fact that the qualification conditions for Air Hostesses required them not to be married, and that the absence of a similar condition for Air Flight Pursers was proof of how the two jobs were different! Indeed, the Court specifically held that “deliberately create such conditions of work only with the object of driving away women from a particular type of work which they can otherwise perform with the object of paying them less remuneration elsewhere in its establishment.”

This observation is crucial, not merely because of its repudiation of Nargesh Mirza’s logic, but also because of its affirmation of the proposition that discrimination occurs not merely through direct exclusion of women, but also from structuring the workplace in a manner that disproportionately burdens women. Of course, the Court is cautious: it limits its observation to “deliberate” creation of conditions, which, in the context of the case seems to be merely about paying women less. As we have seen before, though, unequal remuneration is not the only way in which the workplace might operate to the detriment of women. The issues are structural, and involve a host of conditions, such as a minimal number of working days in a year, a perception of maternity leave as a benefit, and not a right, etc. In a seminal article written in 1989, the feminist scholar Joan Williams observed:

“Western wage labor is premised on an ideal worker with no child care responsibilities. In this system men and women workers are allocated very different roles. Men are raised to believe they have the right and the responsibility to perform as ideal workers. Husbands as a group therefore do far less child care, and earn far more, than their wives. Women are raised with complementary assumptions. They generally feel that they are entitled to the pleasure of spending time with their children while they are small. Moreover, even upon their return to work, the near-universal tendency is to assume that women’s work commitment must be defined to accommodate continuing child-care responsibilities. This gender system results in the impoverishment of women, since it leads mothers systematically to “choose” against performing as ideal workers in order to ensure that their children receive high-quality care. The phenomena that comprise the gender system today are often noted, but the way the system functions as a coherent whole remains largely hidden.”

The key insight of feminist scholars, therefore, was to expand the understanding of discrimination from individual acts of exclusion or subordination (which naturally map on to a reason-based model of discrimination law), to interrogating the systems and structures within which discrimination is made possible. This, of course, is a much more difficult enquiry, especially because it requires questioning the baselines that we intuitively assume as “normal” (such as the default structure of the workplace). And however hesitant and cautious, MacKinnon Mackenzie’s focus on “conditions of work” allows for the beginnings of a more holistic and structural understanding of sex discrimination, than what the Courts have otherwise admitted.

It is important to note that MacKinnon Mackenzie was decided by a bench of two judges, and consequently, was bound by Nargesh Mirza. This led to an odd contradictions towards the end of the judgment, where the Court was constrained to observe: “We do not agree that the work of the Confidential lady Stenographers is a sex based one like the work of air hostesses. There is no custom or rule that only ladies can be Confidential Stenographers.” But surely there was no “rule” that required a division of cabin crew into (female) Air Hostesses and (male) Air Flight Pursers, and if there was a custom, surely it could not be made immune from constitutional scrutiny. As the Court itself had pointed out earlier in its judgment, the key issue was whether the two jobs were accorded equal value using criteria that were in themselves non-discriminatory, and in Nargesh Mirza, there had been a judicial finding to the effect that there was no legally significant difference in the work performed by Air Hostesses and Air Flight Pursers. Whichever way one spins it, the logic of MacKinnon Mackenzie is directly at odds with Nargesh Mirza. Nonetheless, in its (understandable) attempt to distinguish Nargesh Mirza, the Court got itself into all kinds of contortions, arguably undermining its own argument.

On an optimistic reading of MacKinnon Mackenzie, however, there’s no conceptual conflict with Nargesh Mirza. After all, even in Nargesh Mirza, the Court stated that “attributes” would be a part of determining whether the two cadres could be treated separately. It then went on to ignore that condition in adjudicating the concrete issue, but it was merely this proposition of law that MacKinnon Mackenzie was following when it held that equal remuneration is required “if the two jobs in an establishment are accorded an equal value by the application of those criteria which are themselves non-discriminatory (i.e. those criteria which look directly to the nature and extent of the demands made by the job).” On this reading, MacKinnon Mackenzie correctly applied Nargesh Mirza’s logic, where Nargish Mirza had failed to correctly apply its own logic. The core reasoning – that sex-based classifications cannot be sustained simply by invoking service law and creating separate cadres – remains good law.

The logic of MacKinnon Mackenzie was taken to its conclusion by the Supreme Court in Uttarakhand Mahila Kalyan Parishad vs State of UP, decided by the Supreme Court in 1992. In that case, the Supreme Court found that the Uttar Pradesh education department’s creation of cadres of male and female teachers, and of paying female teachers less than male teachers, and according them inferior promotional avenues, was illegal. Consequently, it held: “While we direct by mandamus that the State of Uttar Pradesh shall equate the pay scales with effect from 1-10-1991, so far as the creation of promotional avenues is concerned, we direct the State of Uttar Pradesh to have this examined by an appropriate Committee within two months and give additional promotional avenues to the lady teachers and other employees in the educational line doing administrative business so that their grievance as raised in this petition is properly met.” Consequently, the remedy for violation was not simply invalidation of the cadres (which would leave an administrative vacuum), but at least a partial merging of the cadres (as far as pay scales were concerned), with directions to the State to equalise promotional avenues in a time-bound manner.


Sex Discrimination and the Constitution – VI: The Discontents of Air India v Nargesh Mirza

With the exception of its 1954 judgment, Yusuf Abdul Aziz, which cursorily upheld the constitutionality of adultery, the Supreme Court did not have occasion to seriously deal with sex discrimination under Article 15(1) in the first three decades after the Constitution. The cases that we have discussed so far were all decided by various High Courts. That changed in 1981, with the decision of a three judge bench in Air India v Nargesh Mirzawhich remains one of the landmark Supreme Court judgments on the issue. It is also, as I shall argue, an analytically unsatisfactory decision, which fails to take into account the nuanced and reflective sex discrimination jurisprudence which, as we have seen, was in the process of being developed by various High Courts across the country.

The facts and history that led to the case are extremely complex, and involve two entities (Air India and Indian Airlines Corporation), and multiple rounds of litigation before two tribunals, and then the courts. The following is a highly simplified summary. Briefly, Regulations 46 and 47 of the Air India Employees Service Regulations were challenged. These Service Regulations had created a significant amount of disparity between the pay and promotional avenues of male and female in-flight cabin crew (in accordance with Air India’s designations, the male cabin crew shall be referred to as “Air Flight Pursers” [“AFPs”], and the female cabin crew as Air Hostesses [“AH”]).  For instance, under Regulation 46, while the retirement age for Flight Pursers was 58, Air Hostesses were required to retire at 35, or on marriage (if they married within four years of joining the service), or on their first pregnancy, whichever occurred earlier. Under Regulation 47, this period could be extended, subject to the absolute discretion of the Managing Director.

The first round of litigation took place before two Tribunals, which had successively upheld the Regulations, making observations such as the necessity of having “young and attractive” air hostesses to deal with temperamental passengers. The case finally came to the Supreme Court, which upheld the regulations in part, modified them in part, and struck them down in part.

Before the Supreme Court, the constitutional provisions at issue were Articles 14 (equality before law), 15(1) (non-discrimination on grounds of sex), and 16(2) (non-discrimination on grounds of sex in public employment). In order to have a clear understanding of the Court’s reasoning, it will be important to consider them separately.

Article 14 (the classification test): The Court began by observing that “Art. 14 forbids hostile discrimination but not reasonable classification. Thus, where persons belonging to a particular class in view of their special attributes, qualities, mode of recruitment and the like, are differently treated in public interest to advance and boost members belonging to backward classes, such a classification would not amount to discrimination.” It then noted that Air Flight Pursers and Air Hostesses formed different classes (in service law terminology, separate “cadres”): they had different recruitment conditions and different promotional avenues. Consequently, the “inescapable conclusion that follows is that … there are two separate and different classes having different conditions of service and different incidents, [and] the question of discrimination does not arise.”

But there is something extremely puzzling about this reasoning. Admittedly, in service law, there are different cadres within the same department, with their own separate recruitment rules, promotional avenues and service conditions. This, in itself, does not raise any Article 14 issue. These cadres are constituted in order that persons doing one kind of work are streamlined into a single category when it comes to the rules governing them. In Nargesh Mirza, however, the constitution of the cadre was itself on the basis of sex. That is to say, by definition, only women could become air hostesses, and only men could become Air Flight Pursers. Instead of starting off by examining whether this initial sex-based classification was in conformity with Articles 14, 15 and 16, the Court instead began by noting that because the service conditions and promotional avenues of the two cadres were different, therefore they formed separate classes, and could legitimately be treated separately! This is strange reasoning. It effectively means that the guarantees under Articles 15 and 16 are useless; in order to get around them, all I need to do is to divide my workforce along sex, caste or religious lines, give them different names, treat one class in a manner far inferior to the other, and then justify it by invoking that very separateness of treatment to argue that the two form separate cadres. To put it another way, the Court used the fact that women were being treated less favourably than men to hold that women and men formed separate classes, and that therefore inferior treatment was justified. The vicious circularity of his reasoning is exhibited by the fact that in order to prove that AFPs and AHs formed different cadres, the Court observed that one of the recruitment conditions for AHs was that they must be unmarried, whereas there was no such condition for the AFPs. As we have seen before, imposing marriage as a disqualification upon women but not upon men is itself discriminatory; here, the Court uses that as proof that AFPs and AHs form different classes, for different treatment!

What is stranger, however, is that in its initial formulation, the Court seemed to be aware of this. It noted that “where persons belonging to a particular class in view of their special attributes, qualities, mode of recruitment and the like, are differently treated in public interest…”, there would be no discrimination. In its actual analysis of the flight pursers/air hostess distinction, though, the “special attributes” and “qualities” somehow dropped out of the analysis, and the Court only focused on the recruitment, service conditions and promotional avenues. Indeed, one of the major issues before the Court was whether AFPs and AHs performed different functions. The Court emphatically rejected Fali Nariman’s submissions that the functions of the two sets of cabin crew were different; it held that “a perusal of the job functions which have been detailed in the affidavit, clearly shows that the functions of the two, though obviously different overlap on some points but the difference, if any, is one of degree rather than of kind.” And yet, despite this finding, the Court nonetheless was able to hold that there was no Article 14 violation.

Articles 15 and 16: In comparison with its exposition of Article 14, the Court’s analysis of Articles 15 and 16 was threadbare. It dismissed the challenge in a couple of paragraphs. It noted, first, that under Section 16 of the Equal Remuneration Act, it was stated that “where the appropriate Government is, on a consideration of all the circumstances of the case, satisfied that the differences in regard to the remuneration, or a particular species of remuneration, of men and women workers in any establishments or employment is based on a factor other than sex, it may, by notification, make a declaration to that effect, and any act of the employer attributable to such a difference shall not be deemed to be a contravention of any provision of this Act.” In this case, the government had passed a notification under Section 16. The Court held that “the declaration by the Central Government, therefore, completely concludes the matter.”

The matter, however, is not quite as simply as that. There are three reasons why the Section 16 notification should not have concluded the issue in this case. First, as the text of the Section makes clear, the deeming fiction is limited in nature: the Government’s notification only exempts a classification from liability under the Equal Remuneration Act, and doesn’t give it immunity from a constitutional challenge. Secondly – and most importantly – Section 16 deals with situations where the classification is deemed by the government to be based on a factor other than sex. This, as we have seen before, is an approach to discrimination law that focuses on the discriminator’s reasons, and not on the law’s effects upon protected constituencies. But, as we discussed in the previous essay, there are overwhelming jurisprudential reasons for holding the effects-based test to be the correct one for adjudicating constitutional claims against discrimination. Consequently, Section 16 could have no effect upon the Article 15(1) or 16(2) enquiries. And lastly, even if we take Article 15(1) to incorporate a reason-based model, S. 16 clearly cannot be read to foreclose an independent constitutional enquiry by the Court! If that were to be the case, then S. 16 effectively becomes the following: “where the Appropriate Government makes a declaration that a classification does not violate Article 15(1) of the Constitution, any such classification will be deemed not to violate Article 15(1).” At the very least, such a reading would make Section 16 suffer from the vice of excessive delegation. Therefore, it seems evident that the Supreme Court was incorrect in holding that “the declaration by the Central Government… completely concludes the matter.”

After noting this, the Supreme Court observed that “even otherwise, what Articles IS (l) and 16 (2) prohibit is that discrimination should not be made only and only on the ground of sex. These Articles of the Constitution do not prohibit the State from making discrimination on the ground of sex coupled with other considerations.” It then cited excerpts from Yusuf Abdul Aziz and M.C. Muthamma, before observing “for these reasons, therefore, the argument of Mr. Setalvad that the conditions of service with regard to retirement, etc., amount to discrimination on the ground of sex only is overruled and it is held that the conditions of service indicated above are not violative of Art. 16 on this ground.”

Yet there is absolutely no reasoning by the Court to link the premise to the conclusion. It made no effort to adduce the “grounds” apart from sex that were at play in the impugned classification. There are very good reasons for this: it would have required the Court to ask the principal question that it had avoided throughout its judgment: what was the basis for the initial classification into Air Flight Pursers (male) and Air Hostesses (female), upon which the different conditions of service were superimposed? Especially after finding that the work performed overlapped, the Court would have been left with no option but to hold that the classification was only on the basis of sex. Instead, and unfortunately, it evaded the question.

Article 14 and Arbitrariness: With the classification test under Article 14, and the discrimination claims under Articles 15 and 16 having been rejected, there was one last arrow left in the bow of the challengers: to impugn the Regulations as “arbitrary” under Article 14. As we all know, there are two parallel tests under Article 14: first, the classification test, which requires there to be an intelligible differentia between the two groups that are created by a classification, and a rational nexus with a State objective; and secondly, the judicially-invented “arbitrariness” test: legislation will fail under Article 14 if it is “manifestly unreasonable” or “absolutely arbitrary”. In the history of the Court, the “arbitrariness test” has never really been provided strong conceptual or definitional foundations. Its manipulability was in evidence again, in Nargesh Mirza. On the condition that the Air Hostess’ employment would be terminated if she married within four years of joining, the Court noted:

“So far as the question of marriage within four years is concerned, we do not think that the provisions suffer from any constitutional infirmity. According to the regulations an AH starts her career between the age of 19 to 26 years. Most of the AHs are not only SSC which is the minimum qualification but possess even higher qualifications and there are very few who decide to marry immediately after entering the service. Thus, the Regulation permits an AH to marry at the age of 23 if she has joined the service at the age of 19 which is by all standards a very sound and salutary provision. Apart from improving the health of the employee, it helps a good in the promotion and boosing up of our family planning programme. Secondly, if a woman marries near about the age of 20 to 23 years, she becomes fully mature and there is every chance of such a marriage proving a success, all things being equal. Thirdly, it has been rightly pointed out to us by the Corporation that if the bar of marriage within four years of service is removed then the Corporation will have to incur huge expenditure in recruiting additional AHs either on a temporary or on ad hoc basis to replace the working AHs if they conceive and any period short of four years would be too little a time for the Corporation to phase out such an ambitious plan.”

One need not dwell too long on the outright stereotyping that the Court engaged in here, apart from noting, yet again, that the elephant in the room – i.e., that men were not subjected to the same standards with respect to marriage which, according to the Court, were crucial for “health” and “family planning”. While upholding the marriage requirement, the Court nonetheless found the requirement of termination on the first pregnancy to be unconstitutional, observing: “It seems to us that the termination of the services of an AH under such circumstances is not only a callous and cruel act but an open insult to Indian womanhood the most sacrosanct and cherised institution.” In response, Fali Nariman proposed replacing “first pregnancy” with “third pregnancy”, and subjecting Air Hostesses to a series of requirements, such as taking leave without pay during the pregnancy period and having to undergo an annual fitness examination “in the interests of maintenance of efficiency”. These Amendments were described by the Court as “quite reasonable“. Bizarrely, even as it upheld termination on third pregnancy, the Court also accepted the dissenting judgments in the American case of General Electric Company vs Martha Gilbert, which had held that a classification on the basis of pregnancy amounted to sex discrimination! The two standpoints are impossible to reconcile, and the Court’s attempts to do so perhaps highlight the intellectual aridity of the “arbitrariness” doctrine:

“In the first place, the provision preventing third pregnancy with two existing children would be in the larger interest of the health of the AH concerned as also for the good upbringing of the children. Secondly, as indicated above while dealing with the rule regarding prohibition of marriage within four years, same considerations would apply to a bar of third pregnancy where two children are already there because when the entire world is faced with the problem of population explosion it will not only be desirable but absolutely essential for every country to see that the family planning programme is not only whipped up but maintained at sufficient levels so as to meet the danger of over population which, if not controlled, may lead to serious social and economic problems throughout the world.”

It would, by now, be multiplying repetitions to point out, yet again, that at the heart of the Court’s reasoning are stereotypes: here, it is the stereotype that the responsibility for bringing up children lies with women, and the assumption the burden of controlling population explosion (!) may also be imposed only upon women.

Lastly, the Court struck down the absolute discretion vested in the Director to terminate the employment of Air Hostesses after they reached the age of 35, holding that the uncanalised discretion amounted to excessive delegation of power.

The above analysis should be sufficient to bring out the discontents of Air India vs Nargesh Mirza. To recap:

(1) In holding that AFPs and AHs constituted separate classes, and therefore different service conditions were valid, the Court ignored the fact that the classes themselves were constituted along the lines of sex. In effect, the Court used the fact that women’s service conditions were inferior to men’s to hold that the two constituted separate cadres in service law, and that therefore the difference in service conditions were justified. This argument is viciously circular.

(2) In holding that the Government’s notification under Section 16 of the Equal Remuneration Act was dispositive of the question of whether the Regulations discriminated on the grounds of sex, the Court made three mistakes: first, extending the scope of Section 16 to the Constitution; secondly, treating the government’s statement on a question of fundamental rights as conclusive; and thirdly, ignoring the difference between the phrases “on the basis of” (ERA), and “on grounds… of“. (Article 15)

(3) The Court made no attempt to show why, independent of the Government’s notification under Section 16 of the ERA, Articles 15(1) and 16(2) were not applicable to the facts at hand.

(4) The Court’s arbitrariness enquiry ended up perpetuating and endorsing the exact stereotypes that discrimination law is designed to obliterate. These included the role of women as care-givers and vessels of “family planning”.

All in all, Nargesh Mirza is a highly disappointing judgment in the annals of Indian sex discrimination law. As we shall see in subsequent essays, its harm was not only in the reasoning that it employed, but its role as precedent. Some of the most important sex discrimination cases in India (and elsewhere) have been brought to the Court in the context of service and labour law. Nargesh Mirza initiated a constitutional jurisprudence where the rules of service law overshadowed rigorous constitutional analysis, and were even giving primacy over the latter. This, as we shall see, would lead to a progressive undermining of constitutional analysis under the non-discrimination guarantee of Articles 15 and 16.

The Right to Privacy and the Supreme Court’s Referral: Two Constitutional Questions

Today, in a detailed order in the ongoing Aadhar litigation, a three-judge bench of the Supreme Court referred the question of whether there exists a fundamental right to privacy under the Constitution, to a five-judge bench. In its order, the Court explains that the Attorney-General referred to the judgments in M.P. Sharma vs Satish Chandra (eight judges) and Kharak Singh vs State of UP (six judges), both of which seem to hold that there is no right to privacy under the Constitution. The Court then observes:

“The institutional integrity and judicial discipline require that pronouncement made by larger Benches of this Court cannot be ignored by the smaller Benches without appropriately explaining the reasons for not following the pronouncements made by such larger Benches. With due respect to all the learned Judges who rendered the subsequent judgments – where right to privacy is asserted or referred to their Lordships concern for the liberty of human beings, we are of the humble opinion that there appears to be certain amount of apparent unresolved contradiction in the law declared by this Court.”

One can have no quarrel with this reasoning. Nonetheless, I think that two important constitutional questions arise, which need to be addressed.

(1) (Edit: It was correctly pointed out to me by Suhrith Parthasarthy that the Supreme Court order does not specify that it is making a referral under Article 145(3). Taking that point on board, I rest the following discussion on the first sentence of paragraph 12 of the Order, where the Court says, in language that mirrors that of Article 145(3): “We are of the opinion that the cases on hand raise far reaching questions of importance involving interpretation of the Constitution.” At the end of the same paragraph, however, it also suggests that the reason for referral is to clear up an inconsistency between different benches. My thanks to Suhrith for bringing this to my attention)

The referral has been made under Article 145(3) of the Constitution, which states that:

“The minimum number of Judges who are to sit for the purpose of deciding any case involving a substantial question of law as to the interpretation of this Constitution or for the purpose of hearing any reference under Article 143 shall be five: Provided that, where the Court hearing an appeal under any of the provisions of this chapter other than Article 132 consists of less than five Judges and in the course of the hearing of the appeal the Court is satisfied that the appeal involves a substantial question of law as to the interpretation of this Constitution the determination of which is necessary for the disposal of the appeal, such Court shall refer the question for opinion to a Court constituted as required by this clause for the purpose of deciding any case involving such a question and shall on receipt of the opinion dispose of the appeal in conformity with such opinion.”

What, precisely, constitutes “a substantial question of law as to the interpretation of this Constitution”? Before today’s order, there have been two constitutional challenges where the Court has refused to make a referral. The first is the constitutional challenge to Section 377 of the IPC, which was decided by a two-judge bench in Koushal vs Naz. The second is the constitutional challenge to Section 499 of the IPC (Subramanian Swamy vs Union of India), which is being heard by a two-judge bench at present. Both these cases raised questions of constitutional interpretation: Koushal vs Naz asked whether “sexual orientation” was a protected ground under Article 15(1) (whether arising out of “sex”, or as an analogous ground read into the provision), and whether sexual orientation was protected under Article 21’s right to privacy. The criminal defamation challenge asks whether a 155-year old legal provision under the Penal Code violates the freedom of speech and expression; it also asks the Court to harmonise the constitutional free speech standards that apply to civil defamation and criminal defamation.

What differentiates today’s order from the previous two? The only relevant difference seems to be the existence of conflicting judgments of varying bench strengths. But surely whether or not something involves a substantial question of law as to the interpretation of the Constitution cannot depend upon whether previous judgments have been consistent on the point! To a large extent, the fragmentation of constitutional jurisprudence in recent years has been due to two or three-judge benches hearing important constitutional matters (R. Rajagopal vs Union of India is a case in point). In the referral, the Supreme Court has a chance to clarify the criteria for future references; at the least, a challenge to the constitutional validity of a legal provision should certainly merit a hearing by a Constitution Bench!

(For an excellent discussion, see Suchindran Narayan’s essay on Law and Other Things)

(2) In the logical course of things, a three-judge bench will refer the matter to a five-judge bench, which will then decide whether or not to refer the matter to an even greater bench. The question in this case is: what should be the strength of the bench that decides the question of whether or not there exists a fundamental right to privacy? One argument is that since the decision in M.P. Sharma vs Satish Chandra has expressly been relied upon by the Attorney-General, the bench must consist of at least nine judges, since they can, if required, overrule that case.

With respect, I think this is incorrect. In my opinion, five judges (a Constitution Bench) are enough to decide this issue, for the reason that neither M.P. Sharma (eight judges) nor Kharak Singh (six judges) are binding on the specific issue of whether there exists a constitutional right to privacy. M.P. Sharma was a case involving Article 20(3) of the Constitution (the right against self-incrimination), and the Court’s observations on privacy are entirely incidental to the issue. Kharak Singh is a more complex case, that involved a constitutional challenge to certain forms of surveillance carried out by the police upon a “history-sheeter’s” residence, which included tracking his movements. With respect to the right to privacy, the Court noted:

“Does then the fact that an enquiry is made as regards the movements of the suspect and the facts ascertained by such enquiry are verified and the true facts sifted constitute an infringement of the freedom to move? Having given the matter our best consideration we are clearly of the opinion that the freedom guaranteed by Art. 19 (1) (d) is not infringed by a watch being kept over the movements of the suspect. Nor do we consider that Art. 21 has any relevance in the context as was sought to be suggested by learned Counsel for the petitioner. As already pointed out, the right of privacy is not a guaranteed right under our Constitution and therefore the attempt to ascertain the movements of an individual which is merely a manner in which privacy is invaded is not an infringement of a fundamental right guaranteed by Part III.”

What is crucial to note here is that the observations on privacy arise out of a specific and concrete question: whether the tracking of someone’s movements violates his fundamental rights. The Court’s reasoning, which is contained in a previous part of the judgment, is as follows:

“The question that has next to be considered is whether the intrusion into the residence of a citizen and the knocking at his door with the disturbance to his sleep and ordinary comfort which such action must necessarily involve, constitute a violation of the freedom guaranteed by Art. 19 (1) (d) or “a deprivation” of the “personal liberty” guaranteed by Art. 21. Taking first Art. 19 (1) (d) the “freedom” here guaranteed is a right “to move freely” throughout the territory of India. Omitting as immaterial for the present purpose the last words defining the geographical area of the guaranteed movement, we agree that the right to “‘move” denotes nothing more than a right of locomotion, and that in the context the adverb “‘freely” would only connote that the freedom to move is without restriction and is absolute, i. e., to move wherever one likes, whenever one likes and however one likes subject to any valid law enacted or made under cl. 5. It is manifest that by the knock at the door, or by the man being roused from his sleep, his locomotion is not impeded or prejudiced in any manner. Learned Counsel suggested that the knowledge or apprehension that the police were on the watch for the movements of the suspect, might induce a psychological inhibition against his movements but, as already pointed out, we are unable to accept the argument that for this reason there is an impairment of the “‘free” movement guaranteed by sub-cl. (d). We are not persuaded that Counsel is right in the suggestion that this would have any effect even on the mind of the suspect, and even if in any particular case it had the effect of diverting or impeding his movement, we are clear that the freedom guaranteed by Art. 19 (1) (d)has reference to something tangible and physical rather and not to the imponderable effect on the mind of a person which might guide his action in the matter of his movement or locomotion.”

The Court then noted that under the scheme of the Constitution, Article 21 contained the “residue” of the rights that were left out from Article 19. Consequently, the core of the Court’s reasoning was that the “right to movement” was solely contained within Article 19(1)(d); having once found that Article 19(1)(d) was not attracted, the question of Article 21, and  did not even arise.

However, in Gobind vs State of MP, the case which birthed the fundamental right to privacy, it was indeed held that privacy was an aspect of Article 21, on which issue Kharak Singh was not binding. There are independent reasons why Gobind was correctly decided: the “residue theory” in Kharak Singh was no longer good law; in Maneka Gandhi vs Union of India, the previous eleven-judge bench decision in R.C. Cooper had been relied upon specifically to abandon the interpretation of the Constitution that placed fundamental rights in separate, hermetically sealed containers. After Maneka Gandhi, therefore, the basis of Kharak Singh had, in any case, been obliterated.

Consequently, neither M.P. Sharma nor Kharak Singh are binding upon the question of whether there exists a fundamental right to privacy under the Constitution. The question turns upon interpreting the relationship between R.C. Cooper, Maneka Gandhi and Kharak Singh, and independently assessing whether, in light of the “inter-relationship of rights” theory that existed at Gobind’s time, whether it was correct in extracting a right to privacy out of a structural reading of the Constitution.

For various reasons, I believe that Gobind was correctly decided. For now, though, a Constitution Bench should decide the case without the constraining shackles of either M.P. Sharma or Kharak Singh.