[This essay, a Guest Post by Dhruva Gandhi, is the second in a four-part series excavating the role of the doctrine of arbitrariness in Indian constitutional litigation. The first part of the series is available here.]
Previously, along with Sahil Raveen (here and here) I have critiqued the use of manifest arbitrariness by Nariman J. in Hindustan Construction and Essar, stating how the use of the doctrine meant that the Supreme Court effectively usurped the mandate of Parliament and showing how the decision-making of the court itself was arbitrary. Commenting on the same doctrine, Naniwadekar seems to suggest that the scope of the doctrine should be confined to administrative law and we adopt a deeper understanding of the classification test. The doctrine ought not to be used to strike down statutes. However, the point remains that the doctrine of manifest arbitrariness is now a part of positive law and can be used to strike down statutes as well. Therefore, it is imperative that we find a way to make the doctrine workable that does not see the Supreme Court make policy decisions. With that objective, I look at Navtej Johar and show how the adoption of the doctrine of manifest arbitrariness by the Constitution Bench was actually an application of the concept of disparate impact or indirect discrimination. While I am not saying that ‘manifest arbitrariness’ must be equated with ‘disparate impact’, it could help us provide some meaning to the doctrine and limit its scope.
The Concept of Indirect Discrimination
‘Indirect Discrimination’ or ‘Disparate Impact’ has been adopted in different forms in various jurisdictions the world over. In the United States, the concept of indirect discrimination was initially spelled out in Griggs v Duke Power Co. as:
The Act proscribes not only overt discrimination but also practices that are fair in form, but discriminatory in operation. The touchstone is business necessity. If an employment practice which operates to exclude Negroes cannot be shown to be related to job performance, the practice is prohibited. (emphasis supplied)
Cutting across the Atlantic, several decades later the Equality Act, 2010 in the United Kingdom defined indirect discrimination as,
A person (A) discriminates against another (B) if A applies to B a provision, criterion or practice which is discriminatory in relation to a relevant protected characteristic of B’s.
(2)For the purposes of subsection (1), a provision, criterion or practice is discriminatory in relation to a relevant protected characteristic of B’s if—
(a)A applies, or would apply, it to persons with whom B does not share the characteristic,
(b)it puts, or would put, persons with whom B shares the characteristic at a particular disadvantage when compared with persons with whom B does not share it,
(c)it puts, or would put, B at that disadvantage, and
(d)A cannot show it to be a proportionate means of achieving a legitimate aim.
Definitions such as these can be multiplied and there do appear to be certain nuances that differentiate them. For the moment, I do not wish to delve into those. A common understanding of these definitions seems to be that when a facially neutral measure has or may have a disproportionately adverse impact on persons who belong to certain protected/vulnerable groups, the measure would be indirectly discriminatory or disparate in its impact.
History of Disparate Impact in India
While there has not been much of a discussion of indirect discrimination in India, there have been a few instances when this concept has been applied. For instance, when examining the constitutional validity of a section in the Hindu Marriage Act, 1955 that provided for a decree of restitution of conjugal rights, the Andhra Pradesh High Court said that such a section irretrievably alters the life of a wife while not having any such impact on the husband. It is the wife who has to beget a child and the practical consequences of such a decree would cripple her future plans. More recently, the Delhi High Court examined the validity of a practice in the Northern Railways where medical insurance was denied to those family members of an employee whose names had been struck off by that employee. The court said that even a facially neutral decision can have a disproportionate impact on a constitutionally protected class. In this case, the disproportionate impact was on women and children. Barring these and a few other instances though, there has been no explicit recognition of ‘disparate impact’ in India by the Supreme Court. Bhatia argues that Anuj Garg v Hotel Association of India could be a precedent to say that the effect of a measure must be observed to see its discriminatory nature, but as he himself concedes this case involved a directly discriminatory provision.
‘Manifest Arbitrariness’: A Finding of Disparate Impact.
With this background of indirect discrimination and its application in India, I now wish to turn to the case which adopted the doctrine of manifest arbitrariness as a doctrine that can be used to strike down a statute- Navtej Johar v Union of India. While it was first discussed by Nariman J. in Sharaya Bano v Union of India, there appears to be a lack of clarity on whether his opinion on that point was supported by a majority. Instead, writing previously on this blog, Abhinav Chandrachud has shown how the judges in Navtej adopted this doctrine as a tool that can be used to invalidate statutes.
The provision in contention in Navtej– Section 377 of the Indian Penal Code, 1860 was a facially neutral provision. It did not directly discriminate on the basis of sexual orientation. Instead, it potentially covered within its ambit, consensual and non-consensual homosexual intercourse, consensual and non-consensual heterosexual intercourse other than penal-vaginal intercourse and sexual intercourse with animals. Therefore, Section 377 was discriminatory in its operation. It had a disproportionate adverse impact on a vulnerable group- consenting homosexual adults. It is in this context that the four opinions in this case need to be studied.
The Opinion of Misra J: Parallels with Ladele
Misra C.J. (speaking for himself and Khanwilkar J.) said,
237. Per contra, the presence of this Section in its present form has resulted in a distasteful and objectionable collateral effect whereby even ‘consensual acts’, which are neither harmful to children nor women and are performed by a certain class of people (LGBTs) owing to some inherent characteristics defined by their identity and individuality, have been woefully targeted. This discrimination and unequal treatment meter out to the LGBT community as a separate class of citizens is unconstitutional for being violative of Article 14 of the Constitution…
.. 239. In view of the law laid down in Shayara Bano (supra) and given the fact that Section 377 criminalises even consensual sexual acts between adults, it fails to make a distinction between consensual and non-consensual sexual acts between competent adults. Further, Section 377 IPC fails to take into account that consensual sexual acts between adults in private spaces are neither harmful nor contagious to the society. On the contrary, Section 377 trenches a discordant note in respect of the liberty of persons belonging to the LGBT community by subjecting them to societal pariah and dereliction. Needless to say, the Section also interferes with consensual acts of competent adults in private space. Sexual acts cannot be viewed from the lens of social morality or that of traditional precepts wherein sexual acts were considered only for the purpose of procreation. This being the case, Section 377 IPC, so long as it criminalises consensual sexual acts of whatever nature between competent adults, is manifestly arbitrary.
The opinion of Misra CJ. on the point of manifest arbitrariness was not merely an instance of the court commenting on the policy decision of the legislature and disagreeing with it. Instead, the opinion first notes how the effect of the provision is important because it covers consensual sexual activity. Second, this effect was on a vulnerable group of individuals who shared a common immutable characteristic. Third, this effect had a disproportionately adverse impact- it subjected homosexual individuals to social stigma and even interfered with their privacy. Lastly, the opinion shows how the discriminatory effect came to be- by treating alike those who ought to have been differently. Therefore, it is the disparate impact (see the words, ‘this being the case’) of a neutral measure that made it manifestly arbitrary.
Misra J.’s reasoning is similar to that of English Court of Appeal in Ladele v Islington BC. Ladele, the applicant in that case was a marriage registrar employed by the Islington Borough Council who refused to perform civil partnership proceedings because she believed they were contrary to the will of God. She was then subjected to disciplinary proceedings where she pleaded indirect discrimination on the grounds of religion. While the proceedings were held to be justified, the claim of indirect discrimination was accepted in the following terms,
There is no doubt but that Islington’s policy decisions to designate all the registrars civil partnership registrars, and then to require all registrars to perform civil partnerships, put a person such as Ms Ladel, who believed that civil partnerships were contrary to the will of God, ‘at a particular disadvantage when compared to other persons’, namely those who did not have that belief.
Therefore, it was the failure to distinguish between civil registrars who had a religious belief and those who did not that lead to indirect discrimination as per Neuberger LJ, much like Misra J.
Opinion of Chandrachud J.: Parallels with Mandla v Lee
Similarly, consider the opinion of Chandrachud J.,
Section 377 is based upon a moral notion that intercourse which is lustful is to be frowned upon. It finds the sole purpose of intercourse in procreation. In doing so, it imposes criminal sanctions upon basic human urges, by targeting some of them as against the order of nature….It would have human beings accept a way of life in which sexual conduct without procreation is an aberration and worse still, penal. It would ask of a section of our citizens that while love, they may, the physical manifestation of their love is criminal. This is manifest arbitrariness writ large….
… Even behaviour that may be considered wrong or unnatural cannot be criminalised without sufficient justification given that the penal consequences that follow. Section 377 becomes a blanket offence that covers supposedly all types of non-procreative ‘natural’ sexual activity without any consideration given to the notions of consent and harm.
In Chandrachud J.s’ opinion too, it is the ‘discriminatory effect’ of Section 377 that is tied to a finding of manifest arbitrariness. The words ‘section of our citizens’ in this paragraph are important. When seen in light of the definition that the Equality Act, 2010 in the United Kingdom offers for instance, they show how persons sharing a protected characteristic- ‘homosexuality’ as their sexual orientation- are disadvantaged as compared to those who do not. Chandrachud J.’s reasoning is also similar to the approach adopted by the Andhra Pradesh High Court where the effect on women was observed.
A parallel instance in the United Kingdom furthers this point. In Mandla v Lee, the House of Lords held that a school had unlawfully discriminated against a Sikh student by excluding him from school when he refused to take off his turban. The school had a rule which required boys to come bare-headed to school. While the rule was apparently neutral and applied equally to all students, it was recognised as wanting conformity to a Christian way of dressing and thus, creating barriers for other religions. Chandrachud J. similarly culls out the barriers created to homosexual intercourse by necessitating conformity to a notion that finds the sole purpose of intercourse in procreation.
Even the concluding portions of paragraph extracted above might take a different colour in this light. A differentiating factor between direct and indirect discrimination in the United Kingdom, for one, is that indirect discrimination can be justified whereas direct discrimination cannot. Even in the United States, a measure with a disparate impact is proscribed under Title VII unless it is shown to have some business necessity. Similarly, Chandrachud J.’s opinion says that a provision cannot have a discriminatory impact unless there is a ‘sufficient justification’.
Admittedly, there is an alternate broader interpretation that can be conferred upon this opinion. It can be read to mean that the Supreme Court can sit in review over the merits of criminalisation and de-criminalise a measure unless sufficient justification comes forth. This broader interpretation though, strikes at the separation of powers between the judiciary and the legislature. Analysing the concluding portion of the excerpt in light of an understanding of disparate impact offers a narrower interpretation that is in conformity with the separation of powers. This interpretation is also buttressed by the fact that Chandrachud J. himself finds Section 377 to be an instance of indirect discrimination under Article 15 as well. It may seem odd that he finds the same statute to be an instance of indirect discrimination under Article 15, but I will comment on that a little later.
Malhotra J: Some semblance of Disparate Impact
Moving then to the opinion of Malhotra J.:
Section 377 insofar as it criminalises consensual sexual acts between adults in privates, is not based on any sound or rational principle, since the basis of criminalisation is the ‘sexual orientation’ of a person, over which one has ‘little or no choice’.
Further, the phrase ‘carnal intercourse against the order of nature’ in Section 377 as a determining principle in a penal provision, is too open-ended, giving way to the scope for misuse against members of the LGBT community.
Thus, apart from not satisfying the twin-test under Article 14, Section 377 is also manifestly arbitrary, and hence, violative of Article 14 of the Constitution.
At the outset, this opinion does not seem to link manifest arbitrariness with disparate impact as neatly as the opinions of Misra CJ. and Chandrachud J. did. However, Malhotra J. has identified that in its operation, Section 377 impacts a vulnerable group defined by an immutable personal characteristic and that this impact is adverse in nature (criminalisation of consensual sexual activity and harassment caused by misuse of criminal law). Therefore, this opinion too looks at the discriminatory operation of an otherwise neutral measure and this discriminatory operation is one reason for the provision to be manifestly arbitrary. To this extent, it shares a common denominator with the opinions of Misra CJ. and Chandrachud J. The second reason, namely, the absence of a sound and rational principle is not in common with those opinions. However, the material fact is that four out of the five judges found Section 377 to be manifestly arbitrary because it had a disparate impact.
Nariman J. found the provision to be manifestly arbitrary because it was based on a capricious and irrational principle, namely, the fact that gay persons suffer from a mental disorder and ought to be penalised. To this extent, in my opinion, Nariman J. would be in the minority.
According to me, this reading of Navtej Johar seems to suggest that a statute is manifestly arbitrary when it has a disparate impact. Reading it in this way would have some direct, tangible benefits. First, it would militate against what Timothy Endicott terms arbitrary decision making by judges citing the doctrine of arbitrariness and to that extent, preserves the rule of law. Second, it might sanction the incorporation of the concept of disparate impact or indirect discrimination in India. While some have argued that Navtej was a precedent for indirect discrimination even otherwise (here and here), I disagree with these opinions. These opinions seem to rely on the observations made by Chandrachud J. in the context of Article 15(1). However, this was only the opinion of one judge and not the ratio. Therefore, there is some merit to unearthing the use of manifest arbitrariness.
Nevertheless, this understanding of ‘manifest arbitrariness’ in Navtej unpacks a set of issues as well. The first is that the application of ‘manifest arbitrariness’ in Navtej is not in sync with its application in Joseph Shine, another Constitution Bench decision. This criticism is fairly true, but one could say that the different opinions in Joseph Shine only applied the test laid down by Nariman J in Sharaya Bano. None of them tried to find a common minimum denominator to Navtej’s application of manifest arbitrariness, the first decision which technically incorporated this doctrine.
A second objection could be that the judges in Navtej applied or at least cited the test quoted by Nariman J. in Sharaya Bano and that disparate impact does not appear to be the sequitur of that opinion. Nariman J., in Sharaya Bano collapsed legislation and delegated legislation for the purposes of Article 14 and said that the doctrine of manifest arbitrariness would apply to both. After that, he said:
Manifest arbitrariness, therefore, must be something done by the legislature capriciously, irrationally and/or without adequate determining principle. Also, when something is done which is excessive and disproportionate, such legislation would be manifestly arbitrary.
The words ‘excessive and disproportionate’ appear to refer to the impact of a measure and to that extent cover the disproportionate, adverse effect which constitutes disparate impact. The absence of an ‘adequate determinative principle’ is the absence of a justification necessary to sustain a measure of indirect discrimination. Therefore, it is possible for judges in Navtej to apply this doctrine to arrive at a finding of disparate impact.
In fact, even the Delhi High Court in Harsh Mander v Union of India did something similar. Delivered a month before Navtej and dealing with the constitutionality of the Bombay Prevention of Begging Act, 1949, the court applied manifest arbitrariness in the following manner:
It is our view that law does not make any distinction between types of begging i.e. voluntary or involuntary as has been urged by the petitioners. The absence of any such distinction exposes the statute to a judicial evaluation on the ground of being arbitrary.
On the contrary, the comprehensiveness of the definition of begging in Section 2(1) (i), appears to indicate a legislative intent to cover a broad area, including in its sweep, all types of begging. It has also been pointed out that the respondents are using homelessness and begging synonymously and are in fact detaining the homeless as if they were begging and implementing the penal provision of the act qua them.
This in our view is manifestly arbitrary.
… As a result of the detention of the bread earner of the family, the entire family may be reduced to financial deprivation and penury. Such can never be the object, spirit and intendment of a welfare state by way of what is touted as a social benefit legislation.
The first reason to find the law arbitrary was in response to a contention of the petitioner within the classification framework. It is the second reason that is more interesting. The petitioners contended that 74% of those arrested belonged to the informal labour sector and 45% of those arrested were homeless. It is in response to this contention that the court says that the State was treating homelessness and begging synonymously. What the court thus found was a disproportionate impact on a vulnerable group- the homeless. The disadvantage caused to them was that they were detained and even exposed to criminal sanction. Not to mention, their families were pushed into penury as a consequence. Therefore, the finding of manifest arbitrariness was contingent on a finding of disproportionate impact.
The disproportionate impact becomes clearer when we compare this to DH v Czech Republic, heard by the ECtHR, where a claim of indirect discrimination was upheld. This was a case where ‘psychology tests’ were administered to determine whether or not children should go to special schools, which were in practice undemanding and inferior. Statistics showed that these tests had a disproportionate impact on Roma students, who were over-represented in special schools. Similarly, before the Delhi High Court it was the homeless who were over-represented before the Beggars Court.
The third and more structural objection to this placement of ‘indirect discrimination’ in Navtej is that the court applies manifest arbitrariness in the context of Article 14. Chandrachud J. even says that discrimination under Article 15 could be indirect. This raises interesting questions about the nexus between Articles 14 and 15. Does Article 15(1) only cover direct discrimination? Would all cases of disparate impact only be covered by Article 14? Is one way to reconcile Chandrachud J.’s opinion that Article 14 covers indirect discrimination for markers other than the grounds covered by Article 15? However, what then is so unique about direct discrimination? Is it harm? Is it intention? Similarly, if we are to say that Article 15 covers direct and indirect discrimination for the grounds and Article 14 covers indirect discrimination for other grounds, what is the significance of Article 15(1)? Why do we even have grounds?
While I cannot deal with all these questions at present and may do so at a subsequent stage, my point remains that this reading of Navtej and Harsh Mander might help us add substance to the doctrine of ‘manifest arbitrariness’.