The cases that we have discussed so far have followed a common argumentative pattern. It will be helpful to provide a step-by-step conceptual schema:
(1) All these cases are cases of direct discrimination. This means that the distribution of benefits and burdens directly tracks the line that divides the sexes (for the purposes of this argument, I will not get into the question of whether this line is physically real or socially constructed). In other words, the questioned law on its face denies to one sex a benefit that it allows to the other. To put it still another way, there is a complete overlap between the constituency adversely affected by the law, and the protected ground (sex). If we were to draw a Venn diagram, the smaller circle B (i.e., those affected by the law) would fall entirely within the larger circle A (all men, or all women). In Walter Alfred Baid, for example, the larger circle would be all men. The smaller circle would be those ineligible for the post of Senior Nursing Tutor. This circle would only include men, and would therefore lie entirely within the larger circle.
On the other hand, indirect discrimination occurs when a facially neutral law disproportionately burdens one of the sexes. As we discussed in the previous essay, for example, a workplace regulation requiring compulsory attendance for a certain number of days in a year, will disproportionately burden pregnant women employees (and of course, only women can become pregnant).
(2) In all these cases, the State’s justification was to rely upon the word “only”, and argue that the ground for the distinction was not only sex, but sex along with a host of other factors (capacity to manage property, “decorum”, ability to deal with male prison inmates etc.) In one set of cases, this understanding of “only” was accepted, and the State’s contentions were upheld. In another line of cases, these arguments were rejected.
It is at this point, however, that we need to make a further distinction. Consider, first, the latter category of cases: i.e., the ones that struck down the law or regulation in question, on the issue of sex-discrimination. Recall that in the first essay, we also discussed two potential meanings of the word “grounds“. On the first, “grounds” could refer to the State’s reasons for action. In other words, “the State shall not discriminate on grounds only of…” translates to “the State’s reason for discrimination shall not be only…”. On the second meaning, the word “grounds” does not qualify “the State shall not discriminate”, but instead qualifies “sex, race, religion…” etc. That is to say, “sex”, “race”, “religion” etc. are together known as “grounds”, and Article 15 prohibits the State from discriminating on those grounds. Here, the focus is not on the State’s reasons for acting, but on the effect of its actions.
With this in mind, consider first Walter Alfred Baid. In striking down the prohibition upon employing men as senior nursing tutors, the Court noted:
“… it is difficult to accept the position that a discrimination based on sex is nevertheless not a discrimination based on sex “alone” because it is based on “other considerations” even though these other considerations have their genesis in the sex itself… While it is true that there are patent physical disparities between the two sexes, yet it is not possible to justify a conclusion by any process of reasoning, without impinging on the guarantee against discrimination on the ground of sex, that all women or all men, as the case may be, would be unfit or, to put it differently, no man or woman, as the case may be, would be suitable for a particular class of work.”
Here, the Court was endorsing the first meaning of grounds. By holding that “sex plus” was not a valid justification because the “plus” had its “genesis in the sex itself“, the spotlight was placed squarely on the basis of the law. By comparison, not consider the Orissa High Court’s decision in Radha Charan Patnaik, striking down the rule that barred married women from applying as civil judges:
“Marriage does not operate as a disqualification for appointment as a District Judge in the case of men, whereas in the case of married women, by Rule 6 (2) they are being excluded from appointment.”
Here, the Court did not argue that the marriage requirement itself had its genesis in sex, which it obviously did: the thinking behind the law was clearly that marriage placed certain obligations (of domesticity) upon the woman, and only the woman. What the Court did do was simply to note that men and women who were, in all respects, in precisely the same position were treated differently, and consequently struck down the regulation. This is a case where the judicial focus was on the effect of the rule, and not its basis.
Interestingly, if we examine the cases that went the other way – i.e., those that upheld the State’s claim – we see a very similar split. In Mahadeb Jiew, while upholding the old Order XV, Rule 1 of the CPC, which distinguished between resident propertyless male paintiffs (no security for costs), and resident propertyless female plaintiffs (security for costs), the Court observed:
“Order 25 Rule 1 (3) of the Code is not. a law that can be said to discriminate against women on the ground of sex alone. The discrimination is made on the combined grounds of sex & property & as such is outside the bar of Article 15 of the Constitution, & the law providing it is not repugnant to Article 15 & is therefore permissible law within the competence & power of the lawmaking authorities. While therefore on a comparison between the case of a property-less woman pltf. with the case of a property-less man pltf. the difference can only be explained on the ground of sex only the fact remains that the impugned law itself does not proceed to make the discrimination on the ground of sex alone.“
Effectively, the Court holds in this case that even though it is clear that the reasoning behind the law’s differentiation is based only sex, but the law itself is not. The law simply differentiates between a male, propertyless plaintiff (Category A) and a female propertyless plaintiff (Category B). Consequently, the grounds stipulated in the law are sex (I) and property (II), and therefore the discrimination is not on grounds of sex alone. This is clearly an effects-based understanding of grounds.
On the other hand, in R.S. Singh, where the Court rejected a woman’s claim that barring her from becoming a jail superintendent in a men’s prison was discriminatory, the Court observed:
“… can it be said that the impugned order of the Governor making the women ineligible to posts in men’s jails other than those of the clerks and matrons is discriminatory on the grounds of sex alone? We do not think so. It needs no great imagination to visualise the awkward and even the hazardous position of a woman acting as a warder or other jail official who has to personally ensure and maintain discipline over habitual male criminals. Necessarily the inmates of these jails have a majority of hardened and ribald criminals guilty of heinous crimes of violence and sex… A woman performing these duties in a men’s jail would be even in a more hazardous predicament.”
Notice that the Court did not say that there were multiple grounds here, those of sex and jail superintendence. What the Court did say was that the reason why women were not allowed to become jail superintendents was not just because they were women, but because they would not be able to effectively and safely regulate the conduct of the jail inmates. This, as we can see, adopts the reason-based model of Article 15.
This analysis reveals a surprising conclusion. Indian sex discrimination cases do not divide along one clear axis, as the last two essays might have suggested. Rather, the division is along two intersecting axes. On one axis, the decisions divide upon whether directly discriminatory State acts violate Article 15(1), in cases where the government adduces additional reasons for making that classification. On the other axis, the decisions divide upon whether “grounds” refers to the reasons behind the State act, or to its effects. The distinction, it should be noted, is not a pedantic one, and has the potential to significantly affect the verdicts of important sex discrimination cases; this is evident from a comparison between American doctrine, which follows the reasons-based test, and the approach in UK/Canada/South Africa, which is effects-based.
(While a full analysis of which of the two approaches – reasons or effects – is beyond the scope of this essay, interested readers may consult Tarunabh Khaitan’s A Theory of Discrimination Law (OUP 2015), pp. 162 – 165 for a detailed analysis. I will note, in passing, Khaitan’s citation of Patrick Shin’s crucial point: “the argument in favour of liability for unconscious discrimination truly is an argument for moving our conception of discrimination away from the paradigm of individual human action and more toward a conception that defines it as something like the expression of our persistent structures of inequality” Ff. 54, p. 165).