The set of cases that we have discussed in this series is not comprehensive. To the best of my knowledge, though, it is largely representative of the major lines of argument followed by the High Courts and the Supreme Court in their analysis of Articles 15(1) and 15(3). Readers interested in other accounts of sex discrimination under the Constitution, which use a slightly different set of cases, and advance a different analytical framework, may consult Ratna Kapur and Brenda Cosman’s article, On Equality, Women and the Constitution (available here), and Kalpana Kannabiran’s chapter on sex discrimination in her book, Tools of Justice.
Let us return to the questions that we raised in the introductory essay.
The first question was whether Article 15(1) invalidates all sex-based classifications, unless they are saved by Article 15(3), or some other constitutional provision. This issue has not been confronted squarely by most of the judgments, which have preferred to get around the issue by holding that a particular classification is not founded on grounds only of sex, but on other grounds (such as “physical structure”, “maternal functions” etc.). This includes cases where the impugned law or executive action expressly extends or withholds certain benefits/employment opportunities solely to men or women. Therefore, what seems to be the thinking behind these cases is that differential allocation of benefits and burdens to men and women may be justified in certain cases (i.e., in other words, it does not constitute “discrimination”). A significant amount of confusion could be avoided if these cases were analysed on the rubric of the meaning of “discrimination”, as opposed to whether they were based on grounds “only of” sex.
As a related question, we also saw that in the Air Hostess cases, the Supreme Court rejected discrimination claims on the basis that the parties had consented to a settlement. This conclusion is unsatisfactory, since it presumes that discrimination is a state of mind, and/or the right to non-discrimination can be waived. For reasons discussed before, both presumptions are unjustified.
The second question was whether the word “grounds” qualifies “The State shall not…”, and thus necessitate an enquiry into the reasons behind the discriminator’s (in this case, the State’s) actions – or does it qualify “race, sex, caste” etc (i.e., is “grounds” used in the technical sense of referring to the personal characteristics listed out in the operative part of Article 15(1))? The distinction between the two meanings is not identical to – although it tends to track – the distinction between direct and indirect discrimination. A reason-based reading of grounds will tend to exclude indirect discrimination, while an effects-based reading will tend to include it. We saw that there was a split in the High Courts on this issue, which is seemingly resolved by Anuj Garg in favour of the effects-based interpretation. I have argued that for various textual and philosophical reasons, this is a correct interpretation.
The third question was whether the word “only” limits discrimination claims to those that can demonstrate that discrimination was based solely on sex. By and large, the cases seem to answer this question in the affirmative. Notice, however, that the question itself reflects a bias in favour of the reason-based interpretation of Article 15(1). On an effects-based interpretation, what is required is to show that a people possessing a protected personal characteristic have received an unfair allocation of benefits or burdens, even if the law itself might stipulate bases other than that characteristic.
A fourth and related question was the scope of the word “sex“. We saw that in Anuj Garg, a long line of cases holding that classifications based on stereotypes about the sexual and social roles of women in society constituted impermissible discrimination. A combination of the effects test and the anti-stereotyping principle advanced in Anuj Garg suggests an analytical framework for adjudicating an Article 15(1) claim:
A. Does a particular classification allocate unequal benefits and burdens to men or women, independent of the reasons for which it might have been passed? This may happen in two ways:
A1: The statute or policy itself creates two groups, men and women (direct discrimination), and denies or withholds benefits or burdens to either of the groups
A2: The statute or policy is facially neutral, but disproportionately impacts one of the two groups
B. If yes, then is there any justification for the classification?
B1. No justification that relies upon the stereotypes about the social or sexual role of men or women in society can be accepted.
This, I would suggest, is the upshot of the jurisprudence leading up to and culminating in the Anuj Garg judgment, and also demonstrates why the judgments in Nargesh Mirza and Yashaswinee Merchant are incorrect.
The fifth question was with respect to the interpretation of Article 15(3), which allows special provisions to be made for women. We have seen that often, paternalistic measures are sought to be justified on the basis that they are made to benefit women. Sometimes, this argument has been accepted by the Courts, and sometimes it hasn’t. The anti-stereotyping principle may be pressed into service here as well: no classification that relies upon stereotypes can be justified on the ground that it is a benign classification for the benefit of women. A second issue is with respect to the scope of Article 15(3): cases have held that 15(3) measures cannot swallow up the guarantee of 15(1) entirely. Consequently, 15(3) provisions must themselves be justified on the basis that they are meant to remedy historical discrimination, and that their scope is limited.
While the judgments that we have discussed are divided and conflicted, I would suggest that these five tentative answers provide us with a provisional conceptual model for understanding sex discrimination under the Indian Constitution, and for resolving many of the confusions and contradictions that have plagued the jurisprudence thus far. The very different visions advanced through the course of the decades, however, means that – as for other significant areas of Part III – the future course that the jurisprudence might take remains radically open.